WO2001075439A2

WO2001075439A2 - Breath test apparatus and methods

Info

Publication number: WO2001075439A2
Application number: PCT/IL2001/000308
Authority: WO
Inventors: Ilan Ben-Oren; Ephraim Carlebach; Julian Daich; Boaz Giron; Daniel Katzman
Original assignee: Oridion Medical 1987 Ltd
Current assignee: Oridion Medical 1987 Ltd
Priority date: 2000-04-04
Filing date: 2001-04-04
Publication date: 2001-10-11
Anticipated expiration: 2002-10-04
Also published as: EP1282814A2; WO2001075439A3; JP2003529766A; AU4679801A; JP4860879B2

Abstract

Breath test methods and apparatus for increasing accuracy and reducing the time taken to achieve diagnostically meaningful results. In order to determine, in the shortest time possible, when an increase in isotopic ratio of the exhaled breath is clinically significant, methods are described for the use of a variable and multiple threshold level; for reducing the time taken to determine an accurate baseline level; and for avoiding the effects of oral activity when making measurements. To increase measurement accuracy, methods are described, using the results of the breath tests themselves, of continuous and automatic self-calibration to correct for drifts in the gas spectrometer absorption curves.

Description

BREATH TEST APPARATUS AND METHODS

FIELD OF THE INVENTION

The present invention relates to the field of breath test instrumentation and methods of use, especially in relation to their accuracy, reliability and the speed with which results are provided.

BACKGROUND OF THE INVENTION

Gas analyzers are used for many measurement and monitoring functions in science, industry and medicine. In particular, gas spectrometry is becoming widely used in diagnostic instrumentation based on the use of breath tests for detecting a number of medical conditions present in patients. Descriptions of much breath test methodology and instrumentation are disclosed in PCT Publication No. 099/12471, entitled "Breath Test Analyzer" by D. Katzman and E. Carlebach, some of the inventors in the present application. Methods of constructing and operating gas analyzers such as are used in breath test instrumentation are disclosed in PCT Publication No. 099/14576, entitled "Isotopic Gas Analyzer" by I. Ben-Oren, L. Coleman, E. Carlebach, B. Giron and G. Levitsky, some of whom are inventors in the present application. Applications of some breath tests for detecting specific medical conditions are contained in patents issued to one of the inventors of the present application, namely U.S. Patent No. 5,962,335 to D. Katzman on "Breath Test for Detection of Drug Metabolism", and U.S. Patent No. 5,944,670 to D. Katzman on "Breath test for the Detection of Bacterial Infection" and in U.S. Patent No. 6,067,989 to D. Katzman on "Breath test for the Diagnosis of Helicobacter Pylori Infection in the Gastrointestinal Tract". Each of the above documents is hereby incorporated by reference in its entirety.

Such breath tests are based on the ingestion of a marker substrate, which is cleaved by the specific bacteria or enzymic action being sought, or as a result of the metabolic function being tested, to produce marked by-products. These by-products are absorbed in the blood stream, and are exhaled in the patient's breath, where they are detected by means ofthe gas analyzer.

One well known method of marking such substrates is by substituting one of its component atoms with an isotopically enriched atom. Such substrates and their by-products are commonly called isotopically labeled. One atom commonly used in such test procedures is the non-radioactive carbon- 13 atom, present in a ratio of about 1.1% of naturally occurring carbon. Using ¹³C as the tracer, the cleavage product produced in many such tests is ¹³C0₂, which is absorbed in the bloodstream and exhaled in the patient's breath. The breath sample is analyzed, before and after taking this marker substrate, typically in a mass spectrometer or a non-dispersive infra-red spectrometer. Detected changes in the ratio of ¹³CO₂ to ¹²C0₂ may be used to provide information about the presence of the specific bacteria or enzymic action being sought, or as a measure of the metabolic function being tested.

Since the amount of C0₂ arising from the process under test may be a very small proportion of the total C0₂ production from all of the bodies' metabolic processes, the breath test instrumentation must be capable of detecting very small changes in the naturally occurring percentage of ¹³CO₂ in the patient's breath. Typically, the instrument should be capable of detecting changes of a few parts per million in the level of ¹³C0 in the patient's exhaled breath, where the whole ¹³C0₂ content in the patient's exhaled breath is only ofthe order ofa few hundred ppm. For this reason, the sensitivity, selectivity and stability ofthe gas analyzers used in such tests must be of the highest possible level to enable accurate and speedy results to be obtained.

Furthermore, since the instrument is intended to operate in a point-of-care environment, where there is generally no continuous technician presence, the instrument must have good self-diagnostic capabilities, to define whether it is in good operating condition and fit for use. For similar reasons, it should also have a level of self calibration capability, to correct any drift in calibration level revealed in such self-diagnostic tests or otherwise.

The use of the instrument in a point-of-care environment adds additional importance to the speed with which an accurate diagnosis can be given to the patient following the test. Consequently, to increase patient compliance, the methods used in the breath test for analyzing the results of the measurements in terms of meaningful diagnostic information should be designed to provide as conclusive and reliable a result in as short a time as possible. Furthermore, the execution of the test in the physician's office is greatly facilitated by the use of simple patient and substrate preparation procedures.

The disclosures of all publications mentioned in this section and in the other sections of the specification, are hereby incorporated by reference, each in its entirety.

SUMMARY OF THE INVENTION

The present invention seeks to provide new methods and devices for ensuring the accuracy, speed and reliability of breath tests. A number of separate aspects of the invention are disclosed herein, including but not limited to subjects related to:

(i) methods of patient preparation and substrate ingestion;

(ii) methods of analysis and calculation of the results of breath tests to provide accurate diagnoses in the minimum possible time; and

(iii) self-diagnostic facilities and calibration of breath test instruments.

According to one aspect of the present invention, there is provided in accordance with preferred embodiments of the present invention, methods of executing breath tests and of speeding up the attainment of meaningful results of such breath tests, related to the procedures adopted for patient preparation before administration of the breath test. These methods are rendered possible only because ofthe method of virtually continuous sampling and analyzing of breaths, as described in this application, and in the documents described in the background section. In addition, methods of preparation and administration ofthe substrate for ingestion before the breath test are described.

It is to be understood that, throughout this specification and as claimed, the use of terms to describe sampling and/or analyzing, such as "virtually continuous sampling" or "virtually continuous analyzing" or equivalent descriptive expressions, such as "substantially continuously", are meant to refer to methods of sampling or analyzing capable of being performed repeatedly and repetitively at a rate which is sufficiently high that a number of samplings and/or analyses are performed within the time taken for useful clinical information to be determined from the physiological effects under investigation by the breath test. This rate is thus highly dependent on the type of breath test involved. In the case of a breath test such as that for the detection of Helicobacter pylori, for instance, where a meaningful clinical result may already be obtained in a matter of a few minutes, "virtually continuous sampling" could be taken to mean a rate as fast as almost every exhaled breath of the subject. On the other hand, with breath tests such as that for liver function, in which it could be several hours before a meaningful result is obtained, the condition of "virtually continuous sampling" or equivalent terms, may be fulfilled by means of a breath sample collection and/or analysis every half hour, for instance.

It is this feature of virtually continuous sampling or analysis which provides the present invention with many of its advantages over prior art methods of sampling and analyzing individual bags of breath. It is difficult, from a practical point of view, if not almost impossible, to perform such prior art methods "virtually continuously", and it is this feature which thereby enables the present invention to provide clinically significant results both sooner and with a higher level of reliability than by prior art methods.

According to other aspects of the present invention, there are also provided in accordance with more preferred embodiments of the present invention, methods for analysis of the results of breath tests to provide accurate diagnoses within times significantly shorter than those possible by use of prior art methods. These methods include the use of a method for detecting the presence of oral activity in the subject, arising from the direct interaction of the labeled substrate with bacteria in the oral cavity, unrelated to the physiological state being tested for, or the bacterial infection being sought. It is important to detect such oral activity, and to delay the analysis of the collected breaths until after its subsidence. Otherwise the breath test's ability to detect by-products ofthe labeled substrate exhaled in the subject's breath after traversing a metabolic path through the subject's blood stream and lungs, would be severely degraded.

Further novel preferred methods are disclosed for calculating the change in isotopic ratio over the baseline isotopic ratio, which enable more reliable test results to be obtained in situations where there may be interference or excessive noise in the measurement. A further method is described for combating the effects of drift in the breath test instrumentation, which may limit the ability to accurately compare currently collected samples with a baseline sample collected earlier. According to this preferred embodiment of the present invention, the sample collected at each sampling point is compared with the sample collected at the previous sampling point, rather than with a baseline sample or an external reference gas.

Further preferred embodiments are disclosed in which the changes in isotopic ratio detected are analyzed using a newly proposed parameter, called the Relative Change in Isotopic Ratio, or RCIR, which compares the fractional change in the currently obtained ratio, normalized to a variety of isotopic ratios, each of which has its own specific advantages. A method is also disclosed of using alternating definitions for the RCIR parameter, according to the progress of the test results, in order to reduce the effects of physiological or instrumental noise in the test results. A method for more accurate detection of the baseline level is also disclosed, whereby multiple baseline measurements are made to eliminate the possible negative effects of a single rogue measurement point.

The operational function in a breath test is to determine when a change in the isotopic ratio of a component of breath samples of the subject is clinically significant with respect to the effect being sought. The criterion for this determination, as used in much ofthe prior art, is whether or not the isotopic ratio has exceeded a predefined threshold level, at, or within the allotted time for the test. According to another preferred embodiment of the present invention, in order to achieve the highest sensitivity and specificity in the shortest possible measurement time, the breath test analyzer does not use fixed criteria for determining whether the change in the isotopic ratio of a patient's breath is clinically significant. Instead, the criterion is varied during the course of the test, according to a number of factors manifested during the test, including, for instance, the elapsed time ofthe test, the noise level ofthe instrument performing the test, and the physiological results ofthe test itself.

Furthermore, although in many of the prior art procedures, the measurement used for the change in the isotopic ratio has been the level of the ratio over a baseline level, according to further preferred embodiments of the present invention, the measurement could be the change over a previous measurement point other than a baseline level, or the rate of change of the isotopic level, or any other suitable property which can be used to plot the course ofthe change.

As an example ofthe execution of such a variable criterion, the crossing of a threshold level by the isotopic ratio is used to illustrate the advantages of these preferred embodiments of the present invention. A calculation method is disclosed for the more accurate use of the threshold level, above which, according to the methods ofthe prior art, a test result is assumed to be positive, or below which it is assumed to be negative. The new methods according to more preferred embodiments of the present invention, make use of a dynamically variable threshold, whose value changes according to the progress of the breath test. This dynamically variable threshold may optionally and preferably be dependent on the elapsed time of he test, on the physiological significance of the results, on the scatter or quality ofthe results themselves, or on the noise level or drift of the instrument being used, or a combination of any of these factors. In addition, further preferred embodiments using multiple thresholds are disclosed.

According to another aspect of the present invention, there are also provided, in accordance with other preferred embodiments of the present invention, methods for self-diagnostic analysis of a breath test instrument, and for system checking of the instrument. According to these preferred embodiments, novel methods are disclosed for calibration of the instrument according to the data being collected, either automatically, or by means of operator intervention. Such methods generally are based on the assumption that if the absorption curve of the gas analyzer is accurately known, then the isotopic ratio measured in the gas being detected, according to the supposedly correct absorption curve, will show no dependence on changes in the concentration ofthe sample being measured. Any such dependence found is reduced by means of an iterative correction method, which adjusts the parameters of the absorption curve in such a manner as to reduce any such correlation.

Another preferred method disclosed according to this aspect of the present invention is operative for correcting any inaccuracy in the capnographic measurement performed at the entrance to a breath tester, by means of comparison with an accurate measurement performed by the self-calibrating gas analyzer, as summarized above. The capnographic measurement is used in order to determine which parts of the breath waveform are collected for analysis by the gas analyzer in the breath tester.

There is thus provided in accordance with a preferred embodiment of the present invention, a method of performing a breath test, consisting ofthe steps of using a predetermined criterion for determining when a change in a measurement of an isotopic ratio of at least one breath sample of a subject is clinically significant, and allowing the criterion to change during the breath test. The change in the measurement may preferably be the deviation of the isotopic ratio from a measurement of at least one previous sample of the subject, or alternatively, it may be the rate of change of the isotopic ratio, and the measurement ofthe at least one previous sample ofthe subject may preferably be a baseline measurement.

In accordance with other preferred embodiments of the present invention, the criterion may be a function ofthe elapsed time ofthe test, or ofthe noise level of the instrument performing the test, or it may be a function ofthe physiological results ofthe test.

There is further provided in accordance with yet another preferred embodiment ofthe present invention, a breath test method, consisting ofthe steps of performing a first measurement of the isotopic ratio of at least a first breath sample of a subject, performing a second measurement ofthe isotopic ratio of at least a second breath sample of the subject, and determining when the second measurement shows sufficient deviation from the first measurement that a clinically significant result of the breath test may be concluded, wherein the level of the sufficient deviation is allowed to undergo variation during the breath test. The first measurement ofthe isotopic ratio ofthe above mentioned at least a first breath sample of the subject, may be a baseline measurement. Furthermore, the level of sufficient deviation mentioned above may be a function of the elapsed time from ingestion of an identifying substrate by a subject, or of the physiological results ofthe analysis of at least one ofthe samples, or ofthe nature ofthe results obtained in the breath test. This nature ofthe results may preferably be a function of the standard deviation of the spread of results in the breath test, or a function of the noise level present in the results, or of the instrumental drift present in the results.

In accordance with still another preferred embodiment of the present invention, in the above described method, the level of sufficient deviation may cover a range of values between an upper threshold and a lower threshold, which may even converge as the test proceeds.

In any of the above-described embodiments, the first measurement of the isotopic ratio of the at least a first breath sample of a subject is performed only after significant subsidence of oral activity.

There is further provided in accordance with still another preferred embodiment of the present invention, a breath test method for determining the presence of a clinically significant state in a subject consisting of the steps of performing measurements ofthe changes from a baseline of an isotopic ratio in a plurality of samples of exhaled breath of the subject, following the effective cessation of oral activity, determining a polynomial which approximates the functional plot of the measurements with time, calculating a weighted standard deviation of the measurements from the polynomial, wherein for measurements over the baseline by more than a predetermined amount, a predefined fractional part of the measurement is taken, while for measurements not over the baseline by more than the predetermined amount, the measurement is taken in its entirety, and finally, determining whether the weighted standard deviation exceeds a predetermined level.

In accordance with a further preferred embodiment of the present invention, there is also provided a breath test method, consisting of the steps of performing a measurement ofthe isotopic ratio of at least a first breath sample of a subject, and determining when the measurement shows sufficient deviation from a baseline measurement that a clinically significant result of the breath test may be concluded, wherein the deviation consists of an upper and a lower threshold band of uncertainty, and wherein the extent of this band is dependent on at least one ofthe parameters selected from the group consisting ofthe elapsed time of the breath test, the standard deviation of the physiological spread of results, the dynamics of the physiological change in isotopic ratio, the number of points measured in the breath test, the environmental conditions present during the breath test, and the noise and/or drift levels of the instrument executing the breath test.

There is also provided in accordance with yet a further preferred embodiment ofthe present invention, a method for determining the reliability of a breath test, consisting of the steps of obtaining results from the breath test, defining a reliability parameter by combining at least one of the criteria selected from the group consisting of the instrument noise and/or drift level, the standard deviation ofthe physiological spread of results, the dynamics ofthe physiological change in isotopic ratio, and the time elapsed since ingestion of a labeled substrate, and finally, using the reliability parameter to assess the results of the breath test according to a predetermined reliability criterion. The reliability parameter may also be used in order to determine when to terminate the test.

Furthermore, in accordance with yet another preferred embodiment of the present invention, there is provided a method of calibrating a breath test instrument without the need for externally supplied calibration means, consisting of the steps of continuously measuring isotopic ratios of a gas species in the samples in a plurality of subjects, and searching for correlation between the isotopic ratios of the gas species and the concentration of the gas species in the samples. Neither operator involvement, nor active subject involvement are necessary for this calibration method. There is also provided in accordance with a further preferred embodiment of the present invention, a method of calibrating a breath test instrument without the need for externally supplied calibration means, consisting of the steps of substantially continuously measuring isotopic ratios of a gas species in the samples in a plurality of subjects, and searching for correlation between the isotopic ratios of the gas species and an environmental condition present at the time ofthe breath tests.

In accordance with yet another preferred embodiment of the present invention, there is provided a method of calibrating a breath test instrument, by analyzing results obtained on breath samples of a plurality of subjects not showing change of any significance in the isotopic ratios of a specific gas species measured in the samples, for correlation between the isotopic ratios and the concentration ofthe gas species in the samples.

There is further provided in accordance with yet another preferred embodiment of the present invention, a method of calibrating a breath test instrument, by analyzing results obtained on a plurality of collected breath samples from one subject for correlation between the isotopic ratios of a specific gas species measured in the samples and the concentrations of the gas species in the samples.

In accordance with still another preferred embodiment of the present invention, there is provided a method of calibrating a breath test instrument, consisting of the steps of (a) collecting a breath sample containing a specific gaseous species, (b) measuring the concentration of the specific gaseous species in the sample, (c) determining the isotopic ratio ofthe specific gaseous species in the sample, (d) diluting the sample such that the concentration of the specific gaseous species changes, (e) determining the isotopic ratio again, (f) repeating steps (d) to (f) to obtain measurements on a number of different concentrations of the sample, (g) looking for correlation between isotopic ratios and concentrations of the different concentrations of the sample, and finally (h) adjusting the calibration ofthe breath test instrument to reduce any correlation found. There is further provided in accordance with still another preferred embodiment of the present invention, a method of correcting a change in the calibration of a gas analyzer for determining the isotopic ratio between a first component and a second component of a gaseous sample, consisting ofthe steps of measuring the concentration of the first component by means of optical transmission measurements, calculating the concentration of the second component from the measured concentration ofthe first component, by assuming a predetermined ratio between the components, and correcting transmission measurements made on the second component such that a concentration derived therefrom is essentially equal to the concentration calculated in the previous step from the measured concentration of the first component. The components of the gas samples may preferably be isotopic components.

In accordance with a further preferred embodiment of the present invention, there is also provided a method of retroactively correcting the results of a breath test from the effects of incorrect calibration, consisting ofthe steps of performing a calibration procedure according to one of the above described methods to determine the existence of correlation between measured isotopic ratios and the concentration of gas species in the breath samples, correcting the calibration of the instrument by means of corrected parameters of the gas absorption curves to eliminate the correlation, and recalculating the data of prior breath tests using the absorption curves with corrected parameters.

There is even further provided in accordance with a preferred embodiment of the present invention, a method of calibration of a capnographic probe, operative for measuring input breath waveforms in a breath test instrument, consisting of the steps of estimating the integrated concentration of the accumulated breaths collected according to the measured capnograph waveforms, measuring the concentration of a sample of i e accumulated breaths in the gas analyzer of the breath test instrument, and correcting the calibration of the capnographic probe such that it provides the same concentration as that measured by the gas analyzer. Furthermore, in accordance with yet more preferred embodiments of the present invention, there is provided a breath test instrument which monitors changes in an isotopic ratio of a gas in exhaled breath samples of a subject virtually continuously, and determines that the test has a clinically significant outcome in accordance with the ongoing results of the test. The breath test instrument may preferably provide a signal, such as a visible or audible signal, for indicating that a clinically significant outcome of a breath test has been determined.

In accordance with yet another preferred embodiment of the present invention, the breath test instrument described above may be such that the outcome of the test is substantially independent of dynamic physiological effects occurring in the subject as a result of background conditions. These background conditions may be the result of treatment with a drug therapy or of food intake in a period prior to the performance ofthe breath test. As a result, in using the above breath test instrument, the need for a pre-test fast by the subject may be obviated.

Furthermore, in accordance with yet another preferred embodiment of the present invention, using that breath test instrument, the outcome ofthe test on the subject undergoing treatment with a gastro-intestinal drug therapy, is obtained more reliably or sooner or both, than using corresponding breath tests which do not monitor the changes in an isotopic ratio substantially continuously.

In accordance with still another preferred embodiment of the present invention, the ongoing results ofthe test enable a positive result to be determined even when the isotopic ratio does not clearly exceed a predetermined threshold level, or enable a negative result to be determined even when the isotopic ratio exceeds a predetermined threshold level.

There is further provided in accordance with still another preferred embodiment of the present invention, a method of determining whether the correct isotopically labeled substance kit is being used for a specific breath test, consisting of the steps of adding a marker element to the substance, the marker element being selected to have an immediate and short term effect on the breath test, and providing breath test instrumentation consisting of a detector for the

I marker element. Furthermore, the breath test instrumentation may also incorporate an enabling mechanism that allows the instrument to perform analysis of the results of the breath test samples only after detection of that marker element.

In accordance with a further preferred embodiment of the present invention, there is also provided a method for determining when the effects of oral activity have subsided during execution of a breath test, consisting of the steps of determining a characteristic time required to detect the physiological effect of interest in the breath test, monitoring change in an isotopic ratio in samples of breath collected from a subject following the ingestion of an isotopically labeled substrate, detecting the presence of a meaningful peak over a predefined minimum threshold level occurring in the isotopic ratio, within a time shorter than the characteristic time.

There is provided in accordance with yet a further preferred embodiment of the present invention, a method, in a breath test procedure, of determining a baseline level for an isotopic ratio of a gaseous species in exhaled breath of a subject before ingestion of an isotopically labeled substrate, consisting of the steps of performing a measurement of a first baseline point, assessing the reliability of the measurement, and performing a second measurement of at least one additional baseline point if the reliability of measurement ofthe first baseline point is determined to be inadequate.

There is even further provided in accordance with a preferred embodiment of the present invention, a method, in a breath test procedure, of determining a baseline level for an isotopic ratio of a gaseous species in exhaled breath of a subject, before ingestion of an isotopically labeled substrate, consisting of the step of measuring at least first and second baseline points. Preferably, the mean of the two points is used as the baseline value, if the first two ofthe at least two baseline points fall within a predetermined range of each other. Alternatively and preferably, a third baseline point is measured if the first two of the at least two baseline points do not fall within a predetermined range of each other. In addition, if the first two of the at least two baseline points do not fall within a predetermined range of each other, the point more distant from the third baseline point may be discarded.

Furthermore, in accordance with yet another preferred embodiment of the present invention, there is provided a method of determining change in isotopic ratio in a plurality of at least a first, a second and a third gaseous sample collected at different points in time, wherein the change in isotopic ratio is determined by measuring the isotopic ratio of the second sample in relation to the first sample, and in relation to the third sample.

There is also provided in accordance with a further preferred embodiment of the present invention, a method of reducing the effect of changes in the operating conditions of a gas analyzer on isotopic ratios measured in a series of at least three gaseous samples, by measuring the isotopic ratio of at least one sample in relation to a sample collected before and a sample collected after the at least one sample.

In accordance with yet another preferred embodiment of the present invention, there is provided a method of determining change in the isotopic ratio between a first and a second gaseous sample, consisting ofthe steps of measuring the isotopic ratio of the first sample, measuring the isotopic ratio of the second sample, determining the difference between the isotopic ratios, dividing the difference by one of the ratios, and adding the change to a previous change determined between a prior first and second sample.

There is further provided in accordance with yet another preferred embodiment ofthe present invention, a method of determining change in isotopic ratio in a plurality of at least a first, a second and a third gaseous sample collected at different points in time, wherein the change in isotopic ratio is determined by measuring the isotopic ratio of the second sample in relation to the first sample, and in relation to the third sample, each of the changes in isotopic ratio being determined by the above-mentioned method. In accordance with still another preferred embodiment of the present invention, there is provided a method of determining change in the isotopic ratio between a first and a second gaseous sample, consisting of the steps of (a) measuring the isotopic ratio ofthe first sample, (b) measuring an isotopic ratio of a reference sample, (c) computing a first difference between the first two isotopic ratios, (d) measuring the isotopic ratio of the second sample, (e) remeasuring an isotopic ratio of the reference sample, (f) computing a second difference between the second two isotopic ratios, and (g) subtracting one of the first and the second differences from the other.

There is further provided in accordance with still another preferred embodiment of the present invention, a method of determining change in the isotopic ratio between a first and a second gaseous sample, consisting ofthe steps of (a) measuring the isotopic ratio of the first sample, (b) measuring a first isotopic ratio of a reference sample, (c) computing a first difference between the isotopic ratio ofthe first sample and the first isotopic ratio of a reference sample, (d) normalizing the first difference relative to the first isotopic ratio of the reference sample, (e) measuring the isotopic ratio of the second sample, (f) measuring a second isotopic ratio of the reference sample, (g) computing a second difference between the isotopic ratio of the second sample and the second isotopic ratio of the reference sample, (h) normalizing the second difference relative to the second isotopic ratio of the reference sample, and (i) determining the change in the isotope ratio by subtracting one of the normalized differences from the other.

In accordance with a further preferred embodiment of the present invention, there is also provided a method of determining in a breath test, change of an isotopic ratio in a plurality of breath samples of a subject, consisting of the steps of (a) collecting a reference sample of breath, (b) determining the isotopic ratio of a first one of the plurality of breath samples by comparison with that of the reference breath sample, (c) determining the isotopic ratio of a second one of the plurality of breath samples by comparison with that of the reference breath sample, and (d) computing the change in the determined isotopic ratios between the first one and the second one ofthe plurality of breath samples.

The term "calibration check" is generally used in this specification and claimed, to refer to a measurement of the absolute calibration of the isotopic ratios measured by the breath tester, referred to a zero base line level, by the use of calibration checking gases with known isotopic concentrations or ratios, input to the instrument from externally supplied containers.

The term "system check" is generally used throughout this specification and claimed, to describe methods for determining correct functioning of multiple aspects of the measurement system, including primarily calibration of the gas analyzer, but also possibly including such functions as the radiation source stability, the input capnograph calibration, the gas handling system, the intermediate chamber system for collecting and diluting accumulated breath samples, and the detector operation.

Since a calibration check is part of a system check, overlapping use ofthe terms may have been made on occasion, according to the context under discussion. The use ofthe term "calibration" ofthe instrument, on the other hand, is generally used in this specification and claimed, to describe a process whereby the parameters of the absorption curves used for the infra-red absorption measurements ofthe gases are corrected so that they compensate for drift or other environmentally induced changes occurring in the instrument. Changes in the absorption curves are indeed generally the major cause for changes in the calibration of the instrument. According to this nomenclature, a calibration procedure as used in this application, as opposed to a calibration checking procedure, does not utilize externally supplied gases with known isotopic concentrations or ratios, but typically relies on checks for internal inconsistency in the results obtained in actual measurements performed by the breath tester. The usual inconsistency revealed is an unjustified correlation of measured values of isotopic ratio with gas concentration, as will be further expounded hereinunder. BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will be understood and appreciated more fully from the following detailed description, taken in conjunction with the drawings in which:

Fig. 1 is a schematic block diagram of the constituent parts of a breath tester as disclosed in PCT Publication No. W099/14576, incorporating an intermediate chamber system for accumulating and manipulating breath samples;

Fig. 2 is an isometric view of a NDIR molecular correlation spectrometer, of the type used in the breath tester shown in Fig. 1;

Fig. 3 is a schematic flow diagram of the main steps of a preferred embodiment ofthe calibration procedure operating in the breath tester;

Figs. 4 A to 4E show plots of various typical breath test results which can be correctly interpreted using methods of virtually continuous sampling and analyzing according to preferred embodiments of the present invention, but which could have been misinterpreted using prior art methods of collecting and analyzing discrete bags of sample breath;

Fig. 5 is a graphic plot illustrating the use of the method of comparing pairs of successively collected sample with each other, rather than with baseline or reference samples; and

Fig. 6 is a graphic plot of the threshold values used for determining when the change in isotopic ratio can be considered as providing a definitive result, as a function ofthe time elapsed from ingestion ofthe isotope labeled substrate. DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

Overall Breath Test System Construction and Operation

Reference is now made to Figs. 1 and 2, which are schematic illustrations of parts of a prior art breath test instrument, of a type in which can be incorporated many ofthe methods and devices ofthe various embodiments of the present invention. The illustrations are taken from PCT Publication No. W099/14576, mentioned in the background section of this application. Figs. 1 and 2 are presented solely for the purpose of illustrating and clarifying certain aspects ofthe present invention, and it is not to be construed that the methods and devices of the present invention are limited to applications in breath testers ofthe type illustrated in Figs. 1 and 2. The component parts and operation of the breath tester shown in Figs. 1 and 2 are described in terms of their use for performing ¹³C0₂ breath tests.

Reference is now made to Fig. 1, which is a schematic block diagram of the constituent parts of a breath tester incorporating an intermediate chamber system for accumulating and manipulating breath samples, in order to bring them to the desired concentration for analysis. The subject 1 undergoing the breath test breathes or blows into a nasal or oral cannula 2. The breath samples are input into a breath sensor module 3, which is an input capnographic probe whose function is to monitor the waveforms of individual sample breaths, and to determine which parts of each breath to accumulate for analysis, and which parts to discard. The intermediate chamber gas handling system 4, which includes a system of sensors and solenoid valves, is operative to direct parts of the sample breaths either into the sample accumulation chamber 5, or if uneeded, out into the room. As soon as enough gas has been collected, and at the desired concentration, the sample gas is transferred from the accumulation chamber 5 to the NDIR molecular correlation spectrσmetric measurement cell 6, for measurement ofthe isotopic concentrations in the gas. A computer-based control system receives and processes the results of the absorption measurements, calculates the isotopic ratios of the samples, and generally controls the complete operation ofthe intermediate chamber system.

Reference is now made to Fig. 2, which is an isometric view of a prior art NDIR molecular correlation spectrometer, of the type used in the breath tester shown in Fig. 1. The analysis chambers are built into an aluminum block 10. There are two chambers for each isotopic analysis, one sample chamber and one reference chamber. The minority isotope chambers 11 for the ¹³C0₂ , the ends of which are visible in the end plate 12 of the analyzer block 10, are much longer than those 13 of the majority isotope ¹²C0₂. A thin shutter 14 is used for switching the measurement between the sample and reference channels. In the spectrometer embodiment shown in Fig. 2, the isotopically specific sources 15, and the absorption chambers 11, 13, are directed such that the output beams from all four channels are directed by means of a light cone 15 into a single detector 16.

Self-diagnostics and Calibration

1. Self- diagnostics

Since the instrument is intended to operate in a point-of-care environment, where there is generally no continuous technician presence, the instrument must have good self-diagnostic capabilities, which define whether it is in good operating condition and fit for use. There are five main levels of activity associated with the operation of the diagnostic system, two levels of diagnostic activity, and three levels of consequential or corrective action, as follows:

(a) Accumulation of a historic database of functional parameters of the instrument operation, such as noise level, drift, correlation of unrelated results, and the like.

(b) Identification of the existence of problem and estimation of its severity. A problem is identified either because of a parameter falling outside the limits of the instrument specification, or because of a systematic change in comparison with the past performance of that instrument.

Once the existence of a problem has been established at level (b), it is dealt with as per levels (c) to (e). The level reached depends on the severity ofthe problem revealed and its impact on the measurements performed. The levels, in increasing order of severity are:

(c) Automatic application of a correction to measurements being made.

(d) A warning output that instrument maintenance or calibration is required.

(e) Complete disablement ofthe instrument.

As an example of the different operational significance of each of these levels, the effects of noise present in the measurements is used to illustrate the consequences of each ofthe above five levels.

(a) The measurements are constantly monitored and the results for a specific time period backwards are stored in a database. The noise level of the results, both in terms of scatter of the actual measurements, and in terms of various operational parameters of the instrument, such as lamp or detector noise are recorded.

(b) The noise level is checked, both for departure from the norm, or from past behavior. If the random noise is below a predefined critical level, the diagnostic method does not define the noise as being sufficiently problematic to prevent the attainment of an accurate measurement. Excessive correlation noise, on the other hand, as will be discussed in the section on instrument calibration below, always directs the instrument into one or other ofthe recalibration modes.

Identification of what constitutes a critical noise level is dependent on the type of measurement being performed. A measurement which is giving a definite clinical indication of the patient's state of health, showing a strongly positive or strongly negative result, is capable of tolerating a higher level of random noise than a measurement giving a result very close to the threshold level. For results close to the threshold level, a noisy signal could result in a false positive or false negative result, and a much lower critical noise level is therefore required. In this way, the reliability of the breath test measurement is determined as a function of the conditions prevalent during the execution ofthe breath test itself.

The instrument diagnostic system can preferably be constructed to output a measurement reliability parameter, which is a combination of many or all of the operational parameters affecting the measurement reliability, as described hereinabove. It could, for instance, be a predefined combination of the closeness of the measured breath test result to the threshold, the noise level encountered during the measurements, and the level of the result itself. The measurement reliability parameter thus operatively defines what constitutes an excessively high noise level, according to the result being obtained at the time the definition is being made.

This parameter, can also preferably be output with the results of the test, in order to give the doctor additional information as to what level of confidence can be attributed to that particular test result.

It should be pointed out that, throughout this specification, the use of the terms "positive" and "negative" to describe the results of breath tests or patients, is taken to mean that the patient shows respectively an elevated or non-elevated DoB (delta over baseline) isotopic ratio. It is appreciated that whether such an elevated DoB is indicative ofa state of normal health, or the reverse, is dependent on the particular test being undertaken. For the breath test for H. pylori, for instance, it is an elevated DoB that may be associated with the presence of the bacteria. On the other hand, in the breath test for liver function, for instance, a DoB which stays low may be indicative of a diseased state ofthe liver.

(c) If the noise level is high, but not so high that no meaningful measurements are possible, the diagnostic system preferably applies a compensation procedure to the measurement. A commonly applied compensation procedure is achieved, as an example, by increasing the averaging measurement time ofthe sample currently undergoing analysis in the gas analyzer.

Another compensation procedure for excessive noise, either instrumental, or a physiological result ofthe test, is the dependence ofthe width ofthe band of threshold levels for the definition of a positive result, on the noise level present, as is discussed in more detail hereinbelow. This compensation procedure has a direct bearing on the measurement reliability parameter output by the instrument. Yet another compensation procedure for excessive noise is the criterion used for ending the test. Should the noise level be such that a definitive decision concerning the outcome ofthe test is masked by noise fluctuations, a decision can preferably be taken to lengthen the test in order to try to achieve a more definitive result above the noise level.

(d) This level warning is preferably actuated either as soon as a level (c) situation is encountered, or at a higher level of noise severity, depending on the success of the compensation mechanisms in the instrument in achieving an acceptable measurement, with a good level of confidence. At this level, an output may preferably be issued by the diagnostic system, warning the user that instrument maintenance or calibration is required, so that the source of the noise can be determined and eliminated.

(e) Once the noise level becomes excessive, and compensation procedures do not enable the achievement of an accurate measurement, a level (e) status is reached. At this stage, the diagnostic system preferably disables the instrument, since there then exists the danger ofthe generation of false results.

2. System checks

The breath tester, according to more preferred embodiments ofthe present invention, is capable of performing independent checks of all of its major system functions by performing a pseudo-breath test on supplied samples of a calibrating gas. In particular, the calibration of the instrument is checked. There are a number of ways in which this may be preferably performed, as described in the co-pending PCT Application PCT/ILOO/00338 for "Gas Analyzer Calibration Checking Device" by some of the same applicants as of the present invention, published as International Publication No. WO 00/74553, herewith incorporated by reference in its entirety. The samples can be two physically separate samples of gas mixtures, supplied, for instance, in calibrating cylinders, each gas mixture having a known total C0₂ concentration, and a known 13cθ₂/^CO₂ isotopic ratio. The use of two separate calibration checking gases provides information about the absolute gain of the instrument, such that the positions of the two absorption curves are known. This information can then be used to confirm the true position of the ^C0₂ absorption curve, which, in the instrument calibration procedures to be described below, is assumed to be constant.

Alternatively and preferably, a single gas with a known gas mixture may be used and the intermediate chamber of the instrument used to generate separate samples, each having a different concentration. Even a gas with unknown properties may be used, and the intermediate chamber of the instrument used to generate separate samples, each having the same isotopic ratio but a different concentration.

As an alternative to performing the calibration check by analyzing external sources of gas with known or fixed isotopic ratios, the pseudo-breath test calibration check can preferably be accomplished by using a breath simulator device, which generates pseudo-breath samples with different isotopic ratios from one sample of gas. A suitable breath simulator device is described in the above-mentioned PCT Publication No. WO 00/74553. The parameters of the pseudo-breath sample are similar to those encountered in the normal operation of a real breath test, e.g., similar flow rate, similar "respiration" rate, similar C0₂ percentage, and similar ¹³CO₂/¹²C0₂ ratio.

The checks performed are preferably of instrument calibration, hardware, software, pneumatics and mechanics. There may be two levels associated with each system check - validation and correction. The former confirms that the system is functioning as specified, and the latter corrects readings in accordance with the results of the diagnostic system output. Alternatively and preferably, if the system check procedures identify the need for calibration, the calibration may be performed at a later time.

Furthermore, the checks may preferably be performed by means of an Internet connection with a central service center, either for on-line diagnostic assistance, or on a periodic basis for routine service checks and maintenance.

2.1 Processing system self-check

According to further preferred embodiments of the present invention, the system incorporates a self-checking facility, operating in a watchdog mode to ensure correct operation of the processing software and hardware. This facility preferably consists of a secondary microprocessor, with its own associated software distinct from the main instrument software. The main system microprocessor generates at regular intervals, a predefined synthesized output sequence. The secondary microprocessor analyses this sequence, and if any deviations from the predefined form are detected, the watchdog system issues a warning and closes down the main processor. The PC is then restarted under the control of the secondary processor, and the reason for the malfunction investigated.

2.2 Hardware self-check

The complete instrument preferably performs a self check of its hardware, and the software directly involved in operating the hardware. Some of these checks require the use of a known charge of gas in the reference and absorption chambers. Others check the functioning of components of the hardware which operate independently of the specific measurement being made, and therefore do not require the presence of a calibration gas sample.

Approximately sixty parameters may preferably be used to characterize the operation of the system. Some of them are monitored continuously during operation of the system, as a real time diagnostic facility. Most of them are monitored only between tests, or when the instrument switches from the stand-by mode to the ready mode. Of the sixty or so parameters, 16 are preferably defined as being critical parameters, and divergence from predetermined allowed values results in interruption in the use of the system. Amongst the critical parameters are:

(a) Light source stability as a function of time

(b) Reference cell absorption check

(c) Operation ofthe shutter between reference and sample channels

(d) Input capnograph operation

(e) Intermediate chamber operation - pneumatics and electronics

(f) Feasibility check on values of δ obtained (e.g. negative values, or values with large changes between subsequent breaths actuate the critical parameter flag.) The component parts of the system with which these parameters are associated are shown in Figs. 1 and 2, and in more detail in PCT Publication No. W099/14576, from which Figs. 1 and 2 are adapted.

To illustrate the use of the critical parameters, the test for light source stability is described in more detail. On initial switch on, the lamp intensity I, as conveyed by an optical fiber from the lamp to a detector, is monitored in the reference channel, in order to determine the lamp stability on warm up. As soon as the time differential of the intensity, dl/dt falls below a predefined level, the lamp is considered to be stable, and an enable signal is output to the instrument control. If the desired stability level is not reached, the instrument waits for a predetermined time for stability to be attained. After the elapse of this time period, a request for lamp maintenance is displayed, and the instrument is not enabled for operation. Similarly, if during operation, dl/dt rises above the predefined level, a disable signal is given to the instrument.

3. System calibration

According to further preferred embodiments of the present invention, the breath test instrument is capable of performing four levels of calibrations, which are operative to ensure that measured differences in isotopic ratios are accurate on an absolute level. These calibration procedures preferably operate by amending the absorption curve parameters used in the gas analyzer for converting optical transmissions into gas concentrations. In this way, they compensate for drifts in the absorption curves, whether due to environmental changes, or to instrumental component changes. Changes in these absorption curves are the most common cause of incorrect calibration in such breath tester instrumentation. These calibrations operate at the software level, within the routines concerned with converting series of optical absorption measurements into isotopic ratio differences. In this respect, they are to be distinguished from the calibration checking procedures described extensively hereinabove, which check the absolute accuracy of isotopic ratios measured, by the use of gases with known isotopic ratios.

The four preferred levels of calibration are denoted:

3.1 Soft calibration

3.2 Self-calibration

3.3 Patient calibration

3.4 Service calibration

The first three of these calibration procedures involve no operator or patient intervention, and operate automatically and continually without being requested. Furthermore, the first three of these test procedures, and even one embodiment of the service calibration procedure, are unlike any prior art gas analyzer calibration procedures, in that they use the subject's own breaths, both in order to determine whether calibration is necessary, and in order to perform the recalibration procedure itself. Procedures are known wherein a sample bag of breath provided by the operator or nurse is used as the calibration gas sample, but such a sample is like any unknown, externally provided calibration gas, and certainly requires operator initiation and intervention.

In addition, the first of these calibration procedures is an ongoing process, operating continually in the background. As a result, trends in the instrument calibration can be better identified than is possible using any external calibration which relies on a procedure performed at a specific point in time, which may, by chance, fall at a moment when a temporary change or an atypical event occurs in the instrument.

A second preferred embodiment of the service calibration procedure indeed utilizes an externally provided gas sample or samples for its calibration procedure. All ofthe calibration procedures described, except the soft calibration, are based on the measurement of the relationship between the isotopic ratios measured in gas samples of different CO₂ concentration derived from samples of gas with the same isotopic ratio.

3.1 Soft calibration

This preferred method of calibration is software based, and operates continuously in the background of the system, without requiring patient or operator intervention or involvement. This procedure preferably continually monitors the results of breath sample analyses obtained from subjects with results close to the baseline. For instance, for the H. pylori breath test, this means subjects who proved negative. According to a preferred embodiment, the system software monitors the results of all of the patients tested over the last 2 to 3 days who showed negative response to the breath test. The measurement points used for this test are those obtained for the baseline measurement taken before the ingestion of the isotopic labeled substrate, and those obtained after the cessation of any oral activity which arises from possible interactions of the labeled substrate with bacteria present within the oral cavity.

Each of these negative patients provides breath samples, each generally having a somewhat different and random level of C0₂ concentration, such that each absorption measurement is performed at a slightly different point on the absorption curve. Preferably, breath samples with CO₂ concentrations of from 2.3% to 2.7% are used, so that deviation ofthe absorption curve is checked over a range of values instead of at one point only.

This is in contrast to the routine measurement procedure, where the use of a constant concentration within each test is an important feature in the reduction of the sensitivity of measurement accuracy to the state of the instrument calibration. So long as all of the absorption measurements in one test are performed at one concentration level, any drifts in the isotopic absorption curves affect, to first order, all of the measurements equally, such that any lack of calibration thus becomes a second order effect. There may indeed be an error in the ratio measurement ¹³C0₂/¹²C0₂ because of change in the absorption curve, but the error appears equally in all of the measurements, and thus does not affect, to first order, the changes in ratio detected.

In the soft calibration procedure, all of the samples tested from a single negative patient should, within the noise limits of the measurement, have the same isotopic ratio, despite their having different concentrations. If, however, the ratios measured are not the same, but show a correlation with the sample C0₂ concentration, this is symptomatic of a change in the absorption curve from its correctly calibrated value. The "soft calibration" method then applies a correction, preferably to the shape or position of the absorption curve, to bring the instrument back into calibration, which is indicated by a lack of correlation between the isotopic ratios and concentrations. The way in which the correction is performed is explained hereinbelow. After correction of the shape of the absorption curve, the data of all of the negative patients over the past 2-3 days is again checked for correlation between concentration and isotopic ratio, to confirm that the recalibration procedure was successful, which is indicated by a reduction in the aggregate correlation level for all of the data. Since this calibration procedure operates continually in the background, it maintains a constant state of recalibration of the instrument with respect to shifts of the absorption curve in the operating concentration range.

If, for a soft calibration procedure for a particular patient, no significant correlation between differing measured isotopic ratios and the concentrations is found, then the system is considered to be correctly calibrated, and no adjustment to the absorption curve parameters is made at that point.

In addition to correlation between isotopic ratio and concentration, the soft calibration process can preferably be programmed to inspect for correlation between isotopic ratio and any other function which could affect the calibration of the instrument. Among such functions are environmental conditions, such as the temperature present within the instrument, which has a noticeable effect on the absorption curves.

3.2 Self-calibration

This procedure, like the soft-calibration, also operates automatically without operator or patient involvement. Unlike the soft-calibration procedure, however, it preferably involves both the instrument hardware and the processing software. The procedure is preferably programmed to commence at the conclusion of a breath test, if two conditions are fulfilled:

(i) that the patient tested showed a negative result, and

(ii) that the total percentage of carbon dioxide in the patient's alveolar breath was reasonably high, preferably 4% or more. The inteπnediate chamber operation is adjusted to provide a single accumulated sample with a high C0₂ concentration, e.g. 3.5%, such as would be obtained by collection of alveolar breath only. This sample is measured for isotopic ratio, and is then diluted down by the intermediate chamber system to provide preferably two additional samples with lower concentrations, such as 3%, 2%, each of which too is measured for isotopic ratio.

In order to provide more points for making the calibration assessment, the self-calibration procedure also preferably uses the results of data obtained at around the preferred operating point of the absorption curve, taken from the patient's previous breaths. A total of five points is preferably used, three derived from the high concentration single sample and its diluted derivatives, and two more from previous negative breaths taken during the test. The object of this spread of sample concentrations is to cover as large a part of the concentration range ofthe absorption curve as possible.

The isotopic ratio is checked at each ofthe five concentrations. Since each of the samples originates either from the same accumulated breath sample, or from other breaths taken from the same negative patient close in time to the collection of the accumulated breath sample, the measured isotopic ratios should be identical. Any divergence is indicative of a drift in the absorption curve, as described above, and the recalibration procedure is thus initiated to eliminate this correlation of isotopic ratio and concentration.

Since this self-calibration procedure takes place after the completion ofthe breath test, such as for instance, when the next patient is being readied for his test, it does not require virtually any additional instrument dead time.

3.3 Patient calibration

In a situation where the instrument is not in continuous use, or has not been used for a period of time, or if the initial system check detects a need for an immediate calibration, the system automatically initiates the performance of a patient calibration procedure, according to a further preferred embodiment ofthe present invention. In this procedure, the first several breaths ofthe patient, before administration of the labeled substrate, are preferably collected and diluted down by means of the intermediate chamber, to provide a number of successive samples of different concentration. Each of these samples should have the same isotopic ratio, since they are all taken from a single patient and at the baseline level. The calibration procedure then adjusts the absorption curves, as previously, until the ratios obtained from the samples of different concentration are all essentially the same. A preferred criterion for determining whether recalibration is required is that, for example, the isotopic ratio should vary by less than 3δ (i.e. less than 30 ppm) for changes in CO₂ concentration of from 3% to 1.5%.

Since this procedure lengthens the time during which the patient has to supply breath samples, it is less desirable from the point of view of patient tolerance than the self or soft-calibration procedures, but apart from the slightly lengthened sampling time, it too does not involve any conscious patient or operator involvement. With any of the three above-mentioned calibration procedures, there exists the possibility, according to further preferred embodiments of the present invention, of correcting the results of tests performed in the past, using the newly found calibration data. If, for instance, the test results show a strong correlation between ratio measured and concentrations, and the calibration calculation procedure applies a correction to the absorption curve, it is possible to use this correction not only for future measurements, but also to correct past measurements.

If, the lack of correlation is revealed at the conclusion of a certain test, but the data available from that test is insufficient to perform a complete calibration, or if the breaths available from that test do not cover a full enough range, then the system can preferably recommend the performance of a patient calibration procedure in order to accumulate sufficient accurate data for performing a retroactive calculation of the results of that test. According to this preferred embodiment, the patient need only give a few more breaths at the conclusion of his test, and can then be released.

3.4 Service or operator calibration

This is similar to the patient calibration procedure, except that it is technician or operator initiated when the need to perform calibration becomes apparent, or is mandated by external causes, such as following service, or after expiry of the maximum inter-calibration period required to maintain the instrument in accurate condition. In this procedure, the gas used can preferably be either operator breath samples, by means of a method as previously described, or an external container of a calibration gas, such as is included within the periodic system calibration check kit described in the section on the system calibration check hereinabove.

3.5 Calibration Correction Method

Recalibration is preferably required when the physical parameters of the gas analyzer undergo change such that the absorption curves differ from those which existed when the instrument was last calibrated. The significance of this is that the function which correlates the absorption cell transmittance to the detected gas concentration has changed. In this situation, recalibration is preferably achieved by applying a correction to the absorption curves to bring them back to their correct form, such that a specific detected intensity is equivalent to a given gas concentration. This recalibration process is accomplished by means of another preferred embodiment of the present invention, called the calibration correction method, whose stages are now described. The description is first given for a full hardware-involved calibration, such as the self, patient or service calibrations, and then for the soft calibration, which is a software-only procedure.

3.5.1 Regular calibration procedure

Reference is now made to Fig. 3, which is a schematic flow diagram ofthe main steps of the calibration procedure according to another preferred embodiment of the present invention. The input data for the procedure are a series of ^l2C0₂ transmittances {T₁₂}i 400, and a series of ¹³CO₂ transmittances {Tι₃)i 401, known from measurements of different samples ofthe same gas, each sample having a different concentration C_!2. Since all of these measurements come from the same gas sample, the isotopic ratio for all of the concentrations should be constant.

For each value of T₁₃ , the value of the equivalent concentration Cι₃ is known by fitting the values Tι₃ to the given C₁₃ absorption curve T ₃ = F₁₃ (C₁₃), as shown at step 402.

It is now assumed that within the operating range of C0₂ concentrations used, the absorption curves for both isotopes T(c) can be described with good accuracy by a single exponent ofthe form:

T(c) = yo + A exp(-c/t) , where T(c) is the transmission as a function ofthe concentration c, and yo, A and t are the parameters which define the absorption curve. Furthermore, it is found empirically that the ¹³C0₂ absorption curve is significantly more stable than the ^l2C0₂ absorption curve, and that its parameters yi₃, A₁₃ and t_I3 can be considered to be essentially independent of changes in environmental conditions. The ^I3C0₂ absorption curve is therefore regarded as a fixed function.

The calibration procedure preferably consists ofthe following steps:

1. At 402, the measured { T_[3}i values are inserted into the known ¹³C0₂ absorption curve, to obtain a series of { C₁₃}_Ϊ values, where Cι₃ = F^"'₁₃(T₁₃).

2. A series of the unknown {C₁₂}i can then be created at step 404, by using the isotopic ratio R, C₁₂ = C₁₃ R. The value of R used can be approximated, without loss of effectiveness of the convergence of the calibration process used, by the natural isotopic ratio, R = (C[₃/ C₁₂)_natural.

3. At step 406, the new generated values of {Cι₂}j and the initial values of

{T_!2}i, are used to determine new parameters y₁₂, A₁₂ and t₁₂, which more accurately characterize the current status ofthe C₁₂ absorption curve. This can be preferably done by means of a best fit calculation, such as the "minimum of mean squared error" method, as is well known in the art.

4. A series of corrected transmittance values {T_12c}j 408 are obtained by insertion of the new generated values for {C ₂}[ into the new C₁₂ absorption curve. As expected from the method by which these corrected transmittance values were calculated, a constant isotopic concentration ratio R, is now obtained, as required.

5. At step 410, the differences between the initial input transmittance values { Tι₂}i and the new corrected transmittance values { T_12c}i are calculated, and for each concentration { C₁₂}i , a normalized error difference Δ Tι₂ = (Tne- π)

Tπ is obtained. A series of these values, {Δ T₁₂}i , is thus obtained

6. At step 412, a best fit polynomial, Δ Tπ = P(Cj₃) is generated, using the new values of {ΔT_]2}i and the known values of {Cι₃}j . The order of the polynomial depends on the number of concentrations used as input data, and is typically of order 3 to 5.

The result of the above calibration procedure is that a new absorption curve is obtained for calculating the C₁₂ concentrations. Using this polynomial, each newly measured value of Tπ is therefore corrected to a more accurate value. The new correction function Δ T_\2 = P(C₁₃), by virtue ofthe way in which it was derived, ensures that a zero delta value is obtained between samples of gas with different concentrations but with the same isotopic ratio.

According to another preferred embodiment of the calibration procedure, it is possible to use an abbreviated calculation method, wherein the value of the correction polynomial P(C₁₃) is changed directly, using the values of C₁₃ obtained from the ¹³C0₂ absorption curve, instead of generating a new Cπ absorption curve. According to this method, step 406 in Fig. 3 is by-passed, and the assumption is made that the ¹²C0₂ absorption curve too is fixed, like the ¹³C0₂ curve. The advantage of using the full calculation procedure, however, as described hereinabove, is that a clearer physical picture of what is changing can be obtained if the changes in the C₁₂ absorption curve are followed.

3.5.2 Soft Calibration procedure

The Soft calibration differs from a full calibration in that only the correction polynomial function Δ T₁₂ = P(C₁₃) is optimized. The parameters of the absorption curves for Cι₃ and C₁₂ remain untouched.

12

This preferred procedure requires the input of a historic series of C0₂

13 transmittances { T₁₂)i and C0₂ transmittances { T_I3}f from tests with negative results, from samples taken either before the ingestion of the labeled substrate (baseline results) or after the subsidence of any oral activity. The transmittances are grouped by the test from which they were obtained.

The procedure preferably includes steps similar to those used for the regular calibration described above, as follows: 1. Using the current Cj₃, Cπ absorption curves and the last known correction polynomial, the deltas and concentrations for the input transmittances are calculated, and the correlation between the {C ₃}j concentrations and the deltas is determined.

2. A series of { Cι_2c}i is created by inserting the values of { C_i3}j into the relationship Cπ = C_i3/R, where R is constant and can be set equal to the natural ratio.

3. The C i₂ absorption curve and {C₁₂c}i are used to find a series of corrected transmittance values {T C₁₂}i, where Tπc ⁼ Fπ(Cπc)- These are the transmittances needed in order to obtain a constant R ratio.

4. A series of normalized differences {ΔT₁₂}i between the input transmittances { T₁₂}i and corrected transmittances { T₁₂c)i are created, where:

5. Using (Δ Tπ}i and { C₁₃}; , a best fit polynomial Δ T₁₂ = P(C₁₃) is created, of order 3 to 5 depending on the number of concentrations used as input data.

6. Using the current C]₃, C_]2 absorption curves and the new correction polynomial, the new deltas and new concentrations are determined for the corrected input transmittances, and correlation between these new concentrations and new deltas is determined.

7. If, using the new P(C₁₃) correction polynomial, the correlation between the new concentrations and the new deltas is reduced, the old polynomial is replaced with new one.

If the soft calibration procedure is successful, the result is a lower value of correlation between deltas and concentrations. Since the soft calibration operates continuously, adding to the database every new set of negative data obtained, there is need to perform more than a single iterative calibration cycle. So long as the correlation is reduced, the use of the new correction polynomial ensures that the soft calibration is operating in the correct manner, and that the correlation errors continuously converge.

4. Input Capnograph Calibration

In addition to the above mentioned calibration tests of the accuracy of the overall instrument operation, according to another preferred embodiment of the present invention, a specific test for the calibration of the capnographic probe at the input to the instrument is also performed. The capnographic probe measures the input breath waveform so that those parts of the waveform which are to be collected or rejected can be correctly defined. Since a capnograph does not have the same high measurement accuracy as the breath tester, a procedure using the results of the breath test measurement, which are highly accurate, is used to calibrate the input capnograph.

The C0₂ capnographic probe at the entrance to the system provides a measure of the C0₂ concentration. The concentration of the content of the accumulated sample at the end of the filling process is estimated, preferably by integration of the capnographically measured concentrations of all of the breath waveform parts collected by the intermediate chamber system. The accuracy of this measurement is dependent on the form of the capnograph's absorption curve, which may have changed because of operating conditions. The concentration of the content of this accumulated sample is now measured in the gas analyzer sample chamber, where a highly accurate measure of the concentration is obtained. This is then used to correct the absorption curve of the capnograph for the actual environmental conditions existent in the system, by correcting the CO₂ probe calibration, so that the estimated bag concentration is made equal to the measured concentration. Patient Preparation and Test Procedure

1. Patient preparation

Prior to application of breath tests, during the patient history intake, it is advisable and is common practice that the physician should note details about any medications taken by the patient, which could interfere with the results ofthe test. In particular, the patient is typically asked according to the methods of the prior art, whether he has been taking any antibiotic or other therapeutic drug recently, since these drugs may affect the results of the breath test, depending on what specific breath test is being performed. Some ofthe prior art describes breath test methods which use two measurement points, based on a single bag of breath samples collected before ingestion, and a single bag thereafter, or at best, three measurement points, based on one sample bag before, and two sample bags collected at different times after substrate ingestion. Using these methods, in order to avoid the danger of false negative results, a time interval of a number of weeks is typically recommended between the cessation of the taking of antibiotic or other specific gastro-intestinal therapeutic drugs and the execution of the breath test. For example, operating recommendations given by Alimenterics Inc., of Morris Plains, NJ, the manufacturers of the LARA (Laser Assisted Ratio Analyzer) system for the detection of Helicobacter Pylori in the upper gastro-intestinal tract, suggest that the taking of antimicrobials, omeprazole (a proton pump inhibitor) and bismuth preparations within 4 weeks prior to performing their breath test, may lead to false negative results

The reason for this recommended abstinence period is that the drug may significantly affect the physiological dynamics of the appearance of the isotope labeled component in the patient's exhaled breath, due to suppression of the bacteria responsible for the mechanism giving rise to the elevated isotopic ratio. According to such prior art methods, this may result in a misdiagnosed result, particularly a false negative result because of the reduced reaction level, or because of the delayed physiological response dynamics, and hence arises the need to question the reliability of breath tests performed within a specified time of such drug therapy.

According to preferred embodiments of the method of the present invention, the use of multi-sample, on-line, virtually continuous monitoring ofthe isotopic ratio in the exhaled breath described in the present application, as opposed to the prior art methods of measuring one, or at most two discrete samples following ingestion of the labeled substrate, enables most changes in the patient response to be more easily detected. Unlike the prior art methods, according to the preferred methods ofthe present invention, the breath test for H. pylori can thus be performed with an acceptable rate of specificity and sensitivity, even when the patient is currently undergoing PPI therapy for the treatment of gastric problems, or antibiotic or other treatment for the eradication of the H. pylori infestation. At worst, the knowledge that the subject has undergone such therapy in the period immediately preceding the test, can be used by the physician to assign a somewhat lower "Level of Confidence" parameter to the results, but need not lead to any effective change in their significance.

Furthermore, according to most ofthe recommended procedures according to prior art breath tests, the patient is advised to fast for a period typically of several hours before the breath test, to eliminate the effects of changes in isotopic ratio arising from particular food intake. It is known, for instance, that diets high in maize content result in a higher baseline ¹³CO₂ isotopic ratio than otherwise. Because of the short time required to perform the breath test according to the present invention, there is preferably no need for the patient to fast prior to the test, since any changes in isotopic ratio resulting from particular food intake typically occur at a considerably slower rate than changes measured in the breath test due to H. pylori activity. This advantage may be enhanced by the ability of the present invention to monitor changes in the isotopic ratio measured virtually continuously, thus countering the effects of possible different dynamic response to the urea because of uncertainty as to the time from the patient's last food intake. In addition, there is evidence pointing to the fact that the ingestion of a meal results in the covering of part of the stomach lining, such that the H. pylori activity is reduced. Even if this is the situation, the ability ofthe present invention to virtually continuously monitor changes in the isotopic ratio enables more abstruse changes in isotopic ratios to be detected, and thus provides a higher level of confidence to the measurement than other prior art methods.

Reference is now made to Figs. 4A - 4E, which show situations which typically arise during the execution of breath tests, which, according to the prior art methods of discrete breath sample collection and analysis, may have been misdiagnosed as giving false positive or false negative results. The use of the virtually continuous methods of sampling and analysis according to further preferred embodiments ofthe present invention, enable these cases to be correctly diagnosed.

In Fig. 4A is shown a plot of an isotopic ratio which is increasing very slowly, but monotonically. This kind of response can arise when, for instance, the test is performed on a subject too soon after food intake. The absorption of the marked substrate from a full stomach is considerably slower than otherwise, and there is also a strong dilution effect from the other stomach contents. Consequently, even if the subject is definitely positive, the result may be a slow rise in the resulting isotopic ratio. The same effect may be seen in a subject with a poor level of gastric absorption, or in a subject undergoing drug therapy for treatment or eradication ofthe disease or bacteria being tested for.

According to the prior art methods of collecting a single or at most two sample bags at predefined times after ingestion of the marked substrate, at those times, tj and t₂, the isotopic ratio has not reached the upper threshold level, T/Η, the crossing of which would be determined as indicating a positive result. As is seen, even a considerable time after t₂, the threshold level is still barely crossed, or may not have been crossed at all, long after the termination of the breath test according to all of the usually accepted protocols. This subject would thus have been determined to be negative.

According to preferred methods of the present invention, however, the ability ofthe breath tester to virtually constantly collect and monitor a plurality of breath samples, enables the analysis software of the breath tester to detect the continuous rise in isotopic ratio, and such a subject would thus be more correctly diagnosed as being positive. The use of this method therefore preferably allows more reliable breath testing to be performed. Furthermore, it preferably enables the breath tests to be performed more reliably without the need of pre-test fasting, and on subjects undergoing drug therapy for the treatment or eradication of the clinical state or bacteria being tested for. Furthermore, it preferably allows a result to be obtained earlier than by the prior art, discrete sample bag methods.

Reference is now made to Fig. 4B, which shows an example of the plot of a breath test of a subject who has a condition which results in an unstable level of metabolized substrate, and hence of isotopic ratio of his exhaled breaths, but who does not show the clinical symptoms of the condition being sought for in the breath test. According to some prior art methods, if a sample bag were, by chance, to be collected for analysis at point ti in time, the subject would be diagnosed as positive. Use of the preferred methods according to the present invention, would however, result in a correct negative result, since no definite rising trend is detected.

Reference is now made to Fig. 4C, which shows a situation in which the isotopic ratio rises fairly rapidly to over the threshold level T/H, but then reaches a steady plateau level just above the threshold. Such a physiological outcome would be determined as being positive according to a single point prior art test performed at point t_b yet would be correctly interpreted as negative by the analysis methods used in the present invention.

Finally, reference is made to Fig. 4D, which shows a plot 450 ofthe result of a breath test, which would initially be interpreted as giving a positive result, whether by a prior art discrete bag collection method, at time t_ls or by the methods of the present invention using virtually continuous collection and analysis of isotopic ratios. However, on the same graph are plotted the values 452 ofthe carbon dioxide concentration ofthe samples measured at each point in time on the graph. It is observed that the concentrations show a strong correlation with the ratios measured at each point in time.

According to preferred methods ofthe present invention, the correlation of the concentrations with the isotopic ratios would be detected by one of the self diagnostic routines operating within the instrument, as arising from an incorrect calibration state of the gas analyzer, probably as a result of a shift in one of the absorption curves. A patient calibration procedure would then be performed, to correct the parameters of the absorption curves so as to reduce the correlation discovered, and the results ofthe test recalculated retroactively, using the original data with the newly calculated absorption curves. Fig. 4E shows the result of this recalculation procedure after the calibration. As is observed, the isotopic ratio is now seen to be low and undulating, and the result of the test is shown in fact to be negative.

2. Substrate preparation and administration

According to further preferred embodiments of the present invention, a procedure is practiced for the administration of the marker substrate. This is described in terms of the method used for the breath test for the detection of H. pylori, where urea is used as an isotopically labeled substrate. The current state of practice of this procedure is well documented in a number of recent published patent applications, such as WO 98/21579 to A. Becerro de Bengoa Nallejo, entitled "Method and kit for detecting Helicobacter pylori" and WO 96/14091 to C. Νystrom et al, entitled "Diagnostic preparation for detection of Helicobacter pylori" both hereby incorporated by reference, each in its entirety.

It is known in the art, such as in the above mentioned deBengoa Vallejo PCT Application, that in order to detect the activity of H. pylori in as efficient a manner as possible, the pH value of the stomach antrum should be maintained at its natural acidic level, in which environment the H. pylori continues its urease activity undisturbed. This can be preferably achieved by giving the patient before administration ofthe urea, a drink of approximately 200 ml of aqueous citric acid solution, with a pH of 2 to 2.5, instead of the water used in the earliest breath tests. In the article entitled "Citric Acid as the Test Meal for the ¹³C-Urea Breath Test" by D.Y. Graham et al., published in The American Journal of Gastroenterology, Vol. 94, pp. 1214-1217, May 1999, there is described the results of breath tests performed on patients after taking citric acid solutions. Results on positive patients were obtained significantly more quickly using 4 grams of citric acid than using only 1 gram or plain water. The use of citric acid has three more added known advantages - firstly citric acid delays gastric emptying, thereby keeping the full amount of urea in the stomach for a longer period, secondly, citric acid assists in counteracting the effects of on-going P.P.I. therapy, as given to a large number of candidates for the H. pylori detection breath test, and thirdly, citric acid diminishes the activity of oral bacteria, which are known to cause interference with the breath test.

The known procedure is, therefore, to give the patient a drink of approximately 200 ml of dilute citric acid, before or with the administration of the urea. Since the stability of urea in solution cannot always be guaranteed for long periods, the generally accepted procedure is to provide the urea in powder or tablet form, which is then dissolved in water, and given as a drink.

According to a preferred embodiment of the method of the present invention, the urea is provided in the form of a tablet, which is dissolved directly in the citric acid solution, which is then drunk, or taken by means of a straw. The use of a straw ensures that the urea has minimal contact with the oral cavity, such that the effect of oral bacteria is reduced. This procedure has a number of advantages. Firstly, the use of a pill rather than powder is simpler to package and use. Secondly, the use of a pill makes it clear that all of the urea has dissolved, and therefore, that the whole of the dose is active immediately on ingestion. Thirdly, for some tablet structures, the urea dissolves more readily in the citric acid solution than it does in water. For example, a tablet composed of 50% urea and 50% sodium chloride, with silicate binders and cellulose disintegration agents dissolve completely in a citric acid solution in almost half the time required for dissolution in water, 4 minutes as opposed to almost 8 minutes. 3. Kit identification

The breath tester instrument according to preferred embodiments of the present invention can be used for a number of different tests, some of which are described in the "Background" section of this application, and even more in the PCT Publication No. WO 99/12471, mentioned in the background section. Each test uses its own specific kit of isotopically labeled substrate, and possible accompanying solution components, such as the urea and citric acid used in the breath test for the diagnosis of Helicobacter Pylori in the upper GI tract. Since each test procedure may also have its own specific test protocol, in terms of elapsed time and detection levels for the gas being detected, it is important that a means be provided for ensuring that the correct kit is being used for the selected breath test, and vice versa. Furthermore, the quantity of substrate and accompanying solvent used can be made dependent on the age, weight, medical history, or even ethnic or geographic origin of the patient, and the breath test parameters are adjusted accordingly. Finally, the pharmaceutical lifetimes of some of the active materials in the kits may be limited, so that it is important to warn the user, or even to disable the instrument, if an attempt is made to use a kit with an expired usage date.

According to a further preferred embodiment of the present invention, the materials for each individual breath test are supplied in a kit together with the disposable oral/nasal cannula or other breath conveyance tube used in performing the test. In the co-pending U.S. Patent Application, No. 08/961013, "Fluid Analyzer with Tube Connector Verifier", by some of the inventors ofthe present application, and hereby incorporated by reference, there is disclosed a tube connection verifier, which is operative to ensure that the correct tube is being used for the test being performed by the analyzing instrument, and that the connector is attached correctly. According to this embodiment of the present invention, the connector of the oral/nasal cannula or equivalent, can be coded with an identification code which contains information about which materials are contained in the kit together with that cannula, their quantity, and their date of expiry. Means are provided on the connector of the breath tester instrument, to read the information thus provided when the oral/nasal cannula or equivalent is connected to the instrument. These means can include one or more of optical, electronic, magnetic or mechanical means, including bar code scanning, digital impulses or any similarly effective means. The communication can be either automatic when the connector of the cannula or equivalent is plugged into the breath tester, the data being automatically input to the instrument, or it can be actuated in an interrogation mode when the operator keys into the instrument the test details.

Alternatively and preferably, a tracer or marker material is added to the materials in the breath test kit, and means are provided in the instrument for detecting the marker. According to this preferred embodiment, any of the contents, quantity and expiry date of the breath test material can be automatically identified by the breath test instrument, even if use is not made of a cannula with the relevant coded material information, such as those provided in the kit.

According to a further preferred embodiment of the present invention, a marker added to the substrate in the breath test kit can be used to initiate the analysis ofthe breaths collected. According to this embodiment, a substance such as labeled glucose is added to the substrate, the substance being very rapidly absorbed by the stomach into the blood stream, and its metabolic by-products appearing very shortly thereafter in the subject's exhaled breath. The detection by the instrument ofthe labeled marker by-product from the glucose can be used as a signal that the substrate has been ingested, that its absorption in the stomach has commenced, and that it has followed the complete metabolic pathway of the physiological effect being investigated, but being immune to the particular disease, bacteria or physiological malfunction being sought, appears independently of the presence of that disease or malfunction. This signal is used to issue a command to the instrument control system to commence analysis of collected breath samples for the specific by-product of the test being performed. The use of this method is particularly advantageous with breath tests which extend over a long time, since the marker provides a signal as to when to expect the commencement o the appearance ofthe substrate by-products.

According to other preferred embodiments ofthe present invention, as an alternative to a solid such as glucose, a gas can be incorporated into the substrate, the gas being released on dissolution of the substrate in the gastric juices, and detected directly in the breath without the need to perform the complete circuit of absorption, metabolism and pulmonary exhalation. As an alternative, the gas can be produced from a parent material which generates the marker gas on contact with the gastric acids.

In all of these embodiments employing marker materials, whether incorporating a gas, or resulting in a direct or an indirect gaseous by-product, if the gas is identical to the gas to be detected in the specific breath test, it is important that the effects of the marker gas be short term, so as not to interfere with the detection ofthe true by-products ofthe breath test.

Analysis of breath test results

In the above-mentioned PCT Publication No. WO99/12471, entitled "Breath Test Analyzer", by some ofthe inventors in the present application, there is disclosed a method whereby the breath test is terminated at a time determined by the results of the test itself. This is similar to the method disclosed hereinabove, whereby the method of virtually continuous collection and analysis of breath samples enables the instrument to determine that a clinically significant outcome has been obtained in accordance with the ongoing results of the test, such that the outcome of the test can be obtained earlier than by means of a sampling method using only a single or two discrete sampling points, as in most of the prior art. According to a further embodiment of the present invention, the breath test instrument is equipped with signaling means for indicating to the operator that the test may be concluded, since a clinically significant result has been obtained. The signal may preferably and alternatively be visual, by means of one or more indicator lights, or audible, by means of tones, or by any other suitable real time indicating method or device. In Fig. 13 of the above-mentioned PCT Application published as PCT Publication No. WO 00/74553, herewith incorporated by reference in its entirety, are shown on the front panel of the breath test instrument 210, two alternative embodiments for signaling to the operator that a meaningful result has been obtained, one in the form of an indicator lamp 231 and the other a loudspeaker 233. According to a further embodiment ofthe present invention, different signals may be used for indicating different outcomes of the test, such as different colored light outputs, or different tones, for indicating whether the outcome ofthe test is positive or negative.

According to further preferred embodiments ofthe present invention, there are provided a number of methods used in calculating the results of the gas analyses, such that a decision about the results ofthe breath test are obtained at an earlier time, or with more certainty than by prior art methods.

(a) Oral activity determination

One of the advantages of the virtually continuous analyzing of samples, according to preferred embodiments of the present invention, is that it becomes possible to differentiate between the effects of oral bacterial activity, arising from the direct effect on the substrate of bacteria in the oral, nasal or laryngetic passages, and true gastric effects. When only a single, or at most, a two point measurement after substrate ingestion is made, as in most of the prior art methods, it may be difficult to determine with certainty whether an elevated isotopic C0₂ ratio is due to a rise in the isotopic C0₂ ratio from a gastric interaction, or whether it is the fall of the isotopic CO₂ ratio from the tail-end of oral activity. This may result in a percentage of false positive results.

With the effectively continuous monitoring of the isotopic ratio according to the present invention, a method of calculation can preferably be used which determines whether the isotopic ratio is on a rising or a falling trend, thus discriminating between a true positive gastric result, and the fall-off of oral activity. The method involves plotting the results from the commencement ofthe test, such that the detection of the characteristic rise and fall of oral activity is completely clear. According to a preferred embodiment of the present invention, a response is regarded as resulting from oral activity, and is therefore ignored, if a characteristic peak of the DoB is detected, in the form of a rising and falling value, exceeding a lower threshold value, and returning to below it, all within a time which is clearly less than the time taken to detect the effects of the true physiological effect being sought after in the breath test. For the breath test for H.Pylori, a typical time frame for the completion of any oral activity is of the order of 8 minutes from ingestion of the labeled substrate. Typical values of the oral activity peak are a rise to about lOδ, together with a consequent fall of at least 5δ from the peak value, all within a time of 4 to 8 minutes from the ingestion ofthe urea.

It should be pointed out that the term "oral activity" is used and claimed in this specification to include any physiological side effects which result in an increased isotopic ratio in the subject's exhaled breath which is either unrelated to the sought-after effect being investigated by the breath test, or arises without involving the metabolic path associated with the physiological state being investigated.

(b) Isotopic ratio change

In the order to determine the increase in the isotopic ratio ¹³C0₂/¹²C0₂ of carbon dioxide in the subject's exhaled breath, the generally accepted method is to measure a baseline level ofthe background isotope ratio in the subject's breath before administration of any substrate. The fractional increase in isotopic ratio above this baseline is expressed in terms of the known "Delta over Baseline" parameter, or DoB. In generally used prior art methods, the DoB is commonly expressed as a normalized parameter, delta δ, or more strictly, delta per mil, where the delta between the isotopic ratio R_\ of a sample 1 and a reference sample R_R is defined as: δ, = 1000 * ( R_X - R_R ) / R_R

The reference sample traditionally used is a geological rock standard known as Pee Dee Belemnite limestone, and the reference isotopic ratio R_Pdb is thus the isotopic ratio of carbon, ¹³C / ¹²C , as found in naturally occurring PDB limestone, and has the value 1.11273%).

The Delta over Baseline between measurements 1 and 2 is thus given by: DoB = δ] - δ₂ , where,

δ₂ = 1000 * ( R₂ - R_pdb ) / Rpdb -

Therefore, DoB = 1000 * ( Rj - R₂ ) / R_pdb , where : Ri is the isotopic ratio measured on sample 1 at time 1, and R₂ is the isotopic ratio measured on sample 2 at time 2.

In normal subjects, the isotopic ratio of baseline breath samples is essentially that of the carbon dioxide resulting from the metabolism of organic compounds originating in the vegetable-originated or animal-originated food consumed by the subject. Since these foodstuffs generally have an isotopic carbon ratio noticeably lower than that typical of naturally occurring carbon dioxide in the air, and also lower than that of PDB, the baseline isotopic ratio of exhaled breath in normal subjects is usually significantly less than R_pdb, by an amount which can range from somewhat over 15δ to about 27δ, depending on the subject. The DoB, according to the generally used definition, is therefore expressed as the fractional difference in isotopic ratio between two measurements, relative to a specific fixed ratio, which is generally somewhat elevated from the typical baseline ratio.

According to another preferred embodiment of the present invention, it is possible in some cases to use the fractional difference in isotopic ratio between any two measurements, relative to a specific fixed ratio, but without the need to have made a baseline measurement. According to this preferred embodiment, since the measurement of change in isotopic ratio is sufficiently sensitive, measurements taken following ingestion of the substrate may be sufficient to detect a sought-after change in isotopic ratio, without knowledge of the baseline level. It is then important to note that when using the various parameters mentioned in this section, and throughout this disclosure, for making calculations of the isotopic ratio change, the term "Delta over Baseline" is to be interpreted broadly to mean the difference in Delta over some previously measured value, without strict adherence to knowledge ofthe baseline level.

During measurement of isotopic ratios in breath samples, by whatever means, the conditions of measurement in the sample and reference cells can change from those for the correctly calibrated conditions. Some types of breath test, such as fat mal-absorption estimation, gastric emptying rate, or liver function tests, may extend over a considerable period of time, even running into hours. In such cases, even very slight drift of the instrument during that time may become very significant. Therefore, if there is a systematic error in the measurement of the ratios, due for instance, to incorrect calibration arising from a shift of the gas absorption curves, even though the error in measurement of the change in two ratios close to each other is very small, the value of DoB calculated contains the full systematic error, since each ratio is normalized to a fixed value, R_{P b}, whose value could be quite different from the ratios currently being measured.

In order to avoid this disadvantage, an alternatively defined δ' has been proposed in Publication No. WO 97/14029, of the PCT application by the Otsuka Pharmaceutical Co., where, for the ratios R between samples 0 and 1:

This definition of δ' has Ro in the denominator, instead of R_pdb. The difference between R₀, the baseline ratio, and Ri, the next point of measurement, is generally much smaller than that between Ro and R_{p b.} as explained above. The value of δ' is, therefore, much less susceptible to changes in measurement conditions resulting from a shift of the absorption curve, than the value of δ, since δ' is normalized with respect to a ratio RQ close to the ratio R_\ being measured. By using this δ', some prior art measurement methods attempt to overcome the problem of the need to compensate for drift in the absorption curves.

On the other hand, since the value of δ' is dependent on the value of Ro, the absolute results are dependent on the baseline of the specific subject measured, and can thus vary with such factors as the diet of the subject, or the time elapsed since his last meal, or even his geographic origin, which it is known, can have an effect on baseline level. The difference in baseline levels between different subjects can cover a range of about lOδ, as mentioned above. For this reason, use of a δ' dependent on Ro does not enable absolute numerical comparisons to be made between the results obtained from different subjects.

Reference is now made to Table 1 below, which shows several calculated values of the DoB normalized to Ro in column 2, compared to the traditional DoB normalized to R_{P b} in column 3. Column 1 is the true isotopic ratio, as measured by mass spectrometry. The parameter RCIR will be explained hereinbelow. It can be seen that as the isotopic ratio of the sample increases, the DoB normalized to R₀ diverges from the traditional DoB value normalized to R_pd - Though the level of divergence shown at very large isotopic ratios has little clinically significance for a particular test, in statistical studies requiring comparisons of the results of breath tests on a number of different subjects, or for comparison of one subject's results taken at time intervals of typically some weeks, during which time his baseline ratio could be have changed significantly, the difference between the DoB parameters could become relevant, and the classical DoB, referred to _pdb, is thus to be preferred.

TABLE 1 Isotopic ratio DoB normalized to Ro DoB normalized to R_pdb RCIR 1.097 0 0 0

1.101 3.436 3.355 3.425

1.127 27.656 27.000 26.992

1.165 62.139 60.666 59.458

1.200 93.532 91.314 88.166 In order to overcome the dependence ofthe measured breath test results on changes in the absorption curves, according to preferred embodiments of the present invention, the breath test instrument may incorporate various compensation procedures, such as the soft-calibration, self-calibration or patient-calibration procedures, as described hereinabove. When one of these calibration procedures is performed, using for instance, a sample of the breath of a patient not showing meaningful change in isotopic ratio (a "negative" patient), the assumption is made in the calibration method that the isotopic ratio of that breath can be approximated by R_pdb. The parameters of the absorption curve are adjusted by the iterative calibration procedure, and as expected, the ratio measured is indeed found to be that of R_pdb. The use of such an assumed approximation to R_pdb, even though it is known that the true ratio may be 20δ or more below the value of R_pdb, has only a small effect on the DoB values measured. Thus, for example, if the true ratio of the above mentioned negative patient sample is, in fact, R = 1.07% (a typical value for a negative patient), instead of the assumed R_pdb = 1.11273%), then the error in the DoB measured resulting from the use ofthe R_pd approximation, is only ofthe order of 3% ofthe value measured. This means that instead of 5δ, a reading of 5.15δ is obtained, this deviation being quite insignificant. The absolute values of DoB measured thus have minimal dependence on the baseline level ofthe subject.

Taking the example, at the other extreme, of a subject with an unusually high baseline ratio, even with a baseline ratio as much as 60δ above the assumed value of R_pdb, by virtue of the iterative calibration method used, the measured DoB values obtained are affected by less than lδ. Since the spread in baseline isotopic ratio between different subjects is typically considerably below this value of 60δ, the use ofthe preferred calibration methods of this invention, as described hereinabove, enable accurate breath test results to be obtained, referable to the generally accepted DoB parameter, and independent of the actual baseline of the patient tested. Even the small deviations in the measured values ofthe DoB's engendered by the R_pdb approximation used in the calibration methods described above, can be compensated for, preferably by executing a ratio measurement on a sample with a known isotopic ratio, such as by the execution of a calibration check ofthe instrument. According to another preferred embodiment of the present invention, the above-amended definition for δ' is used in an alternative parameter, known as the "Relative Change in the Isotopic Ratio" or RCIR. The parameter RCIR can be preferentially used, instead ofthe prior art DoB, for determining the increase in the isotopic ratio of exhaled breath. The RCIR parameter is defined by means ofthe expression:

RCIR_n = RCIR_fl.₁ + 1000 * ( R_n - R_n-1 ) / R_n-1 , where R_n is the ^I3C0₂/^I2CO₂ isotopic ratio for the measurement n. By definition, at the baseline measurement, RCIRo = 0.

According to the above definition of RCIR, the normalization is preferably performed with respect to the isotopic ratio at the previous point measured, R^. An alternative preferred definition can also be used for RCIR, namely:

RCIR_n (+) = RCIR,,.! + 1000 * ( R_n - R_n-1 ) / R_n , where the normalization is done with respect to the ratio at the current measurement point. Since, during the course of a breath test showing a positive result, R_n > R_n-ι , the results achieved using RCIR_n(+) are closer to those obtained relative to R_pdb, than results obtained using the previously defined RCIR_n .

According to a further preferred embodiment of the present invention, these two types of normalization for RCIR, are used alternately for calculating the results of each measurement, depending on whether the measured isotopic ratio is on the increase or decrease. When the isotopic ratio is rising, R_n-1< R_n, and the second definition, RCIR (+), is used. For a falling isotopic ratio, in which case R_n< R_n-1, the first definition, RCIR_n, is used. This method of calculation, using alternate RCIR parameters, is advantageous for smoothing the trend of the results when the ratio curve changes direction from increasing to decreasing, or vice versa, or when there is a high level of noise in the measured points, whether from instrumental or physiological sources. The method is also advantageous for compensating for a measured point which is particularly deviant from the general trend of the plotted curve.

A further advantage of the use of RCIR parameters alternately defined according to the trend in direction of the isotopic ratio measured, lies in the accuracy of the results obtained. The use of a single definition RCIR parameter results in either the accentuation or the de-emphasis of the change in the ratio, depending on the direction of the change. Thus, for instance, the use of the first RCIR_n parameter, being normalized to the previous reading R_π._b results in the exaggeration of an increasing ratio, since when on the increase, the R_n-ι in the denominator of RCIR is smaller than R_n. Similarly, the use of RCIR_n results in an apparent decrease in the rate of decrease of a falling ratio, since the denominator R_n-ι is now larger than the present ratio R_n. The opposite effect is obtained when using the RCIR(+) parameter, de-emphasizing an increase, and exaggerating a decrease.

As a consequence, if either one or the other of the RCIR parameters are used in the analysis of an executed breath test, an undulating ratio curve will result in an accumulated ratio error. On the other hand, according to the preferred embodiment of the present invention whereby alternate RCIR parameters are used, depending on the trend in the measurements, an undulating curve always reflects the true measured result, and, as a result, for instance, always returns to its original level if the ratio returns to its original value.

Instrumental drift which takes place during the course of a measurement, does not affect the parameter RCIR significantly differently from the affect on the prior art DoB. In the last column of Table I above, are shown values of the RCIR parameter calculated for each isotopic value given. As is seen, the RCIR values follow the values of the classical DoB, with small deviations at high DoB values.

According to another preferred embodiment of the present invention, the RCIR can be used in a method of measurement which largely overcomes a major problem of performing breath tests over comparatively long periods of time, such as those tests mentioned above, which can extend for well over an hour. In such situations, instrumental drifts are common, for example due to changes occurring in the absorption curves with changes in environmental conditions, in particular with change in temperature. In this situation, if a baseline reference is taken near the start of the test, there is no simple way of accurately comparing this baseline measurement with isotopic ratios obtained much later in the test, since the measurement conditions are generally likely to have changed, and the comparison is not therefore valid.

Furthermore, there are optical spectrometric gas analysis methods, including some used in breath tests, in which there is a need to bring the samples to be measured to the same major isotopic component concentration as the baseline sample, so that it becomes possible to directly relate optical transmissions (or absorptions) measured in the sample cells, to the concentrations of the component gases therein. This equality of concentration is achieved by diluting each sample collected, by means of an inert gas, down to a predetermined concentration. The concentration typically chosen is such that it is at the low end of the range of commonly achieved concentrations to be tested, such that a majority of the samples collected in practice, can be diluted down to that same predetermined concentration value.

In the above mentioned Publication No. WO 97/14029, of the PCT application by the Otsuka Pharmaceutical Co., in order to achieve equality in concentration between two samples, a method is described of comparing the concentration of the two sample bags, and diluting the higher concentration one to that of the lower. Unfortunately, it is impractical to apply the method of diluting to the lowest concentration to a large number of samples, each collected in a different sample bag or to an on-line measurement method in which the breath of the subject is effectively monitored quasi-continuously, as is described in the present invention. The comparison of such a large set of samples would require a complicated system of sample handing and temporary storage, since each sample needs to be ultimately compared with every other sample.

Furthermore, since stable measurement conditions generally cannot be maintained during the duration of typical breath tests, as explained above, there is another disadvantage in the method of Otsuka, because ofthe use ofthe ratio, Ro, to which the changes in isotopic ratio of the samples are referred, which may be remote from the ratio ofthe measured samples. Likewise, if when comparing the n-th sample to sample n-1, the ratio R_n-1 is used for normalizing, the trend of the results becomes unduly exaggerated, and any changes over emphasized.

The utilization of alternating RCIR parameters, according to the preferred embodiments described in this invention, largely solves the above-mentioned deficiencies in the Otsuka method, enabling individual pairs of samples, n-1 and n, to be brought to the same concentration, without reference to any other of the pairs. Thereafter, the n+1 sample is measured relative to the new measurement of the n sample, and so on.

Another method of achieving the correct dilution of the samples, commonly used in prior art carbon dioxide breath tests, is by means of a comparison of the ¹²CO₂ IR transmission with a reference sample of known concentration. Since for the reasons stated above, comparative absorption measurements may be inaccurate if made at widely differing times, the ability to achieve samples of equal concentration is also affected by those same instrumental stability problems that affect the absorption measurements themselves. Thus, according to methods used hereto, when the samples may have been collected and measured at considerably different times, the accuracy of a breath test is dependent in two separate ways on the ability to perform accurate comparative absorption measurements referred to a baseline sample; firstly, in the ability to dilute the samples accurately to the same concentration, and secondly, in the ability to perform the actual absorption measurement accurately.

There are a number of methods of overcoming this problem. One method is to collect a very large baseline sample, and to divide it into separate individual parts that will suffice to compare each subsequent sample collected as the test progresses, with a part of the original baseline sample, under the conditions prevalent when the subsequent sample is measured, such that the comparison is more accurate. Alternatively, separate samples of the initial baseline sample can be drawn off for each successive breath comparison. These methods are very cumbersome to execute, and generally not easily feasible, because ofthe practical problem of collecting a large initial baseline sample that will suffice, after division, for each comparative measurement.

An alternative and simpler procedure that has been proposed in U.S. Patent No. 5,146,294 to R. Grisar et al, is to use a storage container for supplying successive samples of a reference gas, which, between measurements of diluted breath samples, is transferred into the measurement chamber under conditions identical, as far as is possible, to those of the diluted sample breaths. The accuracy of this method would appear to be limited by the accuracy with which the system can be temperature stabilized, and by the accuracy with which the reference samples can be repeatedly measured at the same pressure conditions.

According to another preferred embodiment of the present invention, an alternative reference measurement method is to collect a single initial baseline sample of exhaled breath, and to repeatedly measure this same baseline sample immediately before and/or after measurement of each sample collected during the test, by transporting the baseline sample into and out of the measurement cell between each collected sample measurement. In this way, the baseline sample is measured under conditions similar to those of the collected samples. The execution of this method requires an accurate gas handling system to avoid contamination ofthe single baseline sample by loss, leakage or dilution.

Alternatively and preferably, and even more simply, this single baseline sample may be stored in its own reference cell, and compared with the sample gas in the measurement cell at each measurement of a new collected sample. This has the disadvantage of having to switch the measurement path between different cells in order to perform each measurement.

Calculation methods for performing all of the above procedures can preferably be established using various different definitions of the change in isotope ratio from a fixed point, normalized either to an absolute fixed ratio or to a variable ratio. The isotopic deviation of any measured ratio is then given by any ofthe following expressions, depending on which reference ratio is used:

(a) ( R_n- Ro_(n))* 1000/ Ro_(n) , or

(b) ( R_n- R_0(n))* 1000/ R_pdb where R_n and Ro_(n) are the measured isotopic ratio and the baseline reference respectively measured at the n^th measurement point.

When an absolute normalized measurement, such as that to R_pdb , cannot be achieved, for instance because of excessive instrumental drift, the relative isotopic deviation may then take the form:

(c) ( R_n- R_0(n))* 1000/ R_n

Since R_n > Ro_(n), expression (c) is closer to the standard PDB related expression (b), than (a) is.

When the reference ratio is not a baseline measurement, the change in isotopic ratio from this point can preferably be calculated relative to the previous result. Thus, the deviation is obtained by subtracting for the previous result, one ofthe following terms, depending on the definition used:

(d) ( R_n- R_re?n))* 1000/ R_«fϊn), or

(e) ( R_n- R_ref(n})* 1000/ Rp_db,

Where R_n and R_ref(n) are the measured ratio itself and reference ratio respectively measured at the n ^l measurement point.

When an absolute normalized measurement, such as that to R_pdb , cannot be achieved, for instance because of excessive instrumental drift, the relative isotopic deviation may then preferably take the form:

(f) ( R_n- R_ref_(n))* 1000/ R_n

Since R_n > R_0(n), expression (f) is closer to the standard PDB referred expression (e), than (d) is.

For the first point (baseline), the relative change in isotopic ratio can thus be expressed as : (Rθ- R_ref(0))* 1000/ R_ref(0). pdb or 0 while for the n measurement point, it is:

( n- Rref(n))* 1000 Rref.n), pdb orn " ( θ^_ Rref(0))* 1000/ R_ref(0), pdb or O if the measurement is performed relative to a baseline, or:

(Rn- Rref(n))* 1000 / Rfef(n), pdb or n " (R -P Rref(n-1))* 1 00 / R_ref(n-1), pdb or n-1 if the RCIR calculation method is used.

According to another preferred embodiment of the present invention, a method is proposed, using the RCIR parameters, which largely overcomes the above-mentioned problems of comparing collected samples with a single baseline sample for breath tests which extend over a long period of time. According to this preferred method, a pair of samples is collected at each measurement point, except the first, where only one sample need be collected, generally a baseline sample. At any successive measurement point, n, one of the pair of samples collected is compared with one of the samples from the (n-1) point, generally collected a comparatively short time previously, while the second is kept for comparison with one of the pair of samples to be collected at the next measurement point, (n+1). At the last measurement point, two samples are collected, but only one need be measured, as the test is terminated at that point.

Though the method of this embodiment has been described in terms ofthe "collection" of two separate samples at each measurement point, it is to be understood that in practice, there is no need to physically collect two separate samples. It is possible, for instance to collect a single sample, and to use half at each of the two measurement points concerned, or to collect a single sample and to measure it twice, once at each measurement point, or any other suitable variation ofthe method.

The time between measurement points is comparatively short compared with the total elapsed time of the complete breath test, and can typically range from considerably less than a minute, to over 30 minutes, depending on the type of test being performed. It is thus simpler to maintain the integrity of the sample and the stability ofthe measurement conditions for the comparatively short period between one measurement point and the next, than from the beginning of the breath test till its end.

In this way, a moving frame of reference is created, whereby comparisons with the previous measurement point are made under the closest possible measurement conditions, and each measurement point may preferably be referenced back to the previous point, regardless of how the measurement conditions have changed in the interim because of instrumental drift.

The computational procedure of comparison of pairs of samples according to this preferred embodiment of the present invention, can be understood by reference to Fig. 5, which shows a typical plot of how a measured isotopic ratio, R, could change as a function of elapsed time, because of instrumental drift, such as results from changes in the absorption curves. The graph shown is for illustrative purposes only. The plot shows the change in the ratio actually measured by the instrument for an idealized fixed ratio, as if the breath test were giving a negative result, with no change whatsoever in the true isotopic ratio. A similar explanation can be given for the more likely situation when the ratio really is changing, but for the sake of simplicity, a fixed ratio is used to explain the method of this embodiment.

Using the prior art methods of comparing the change in isotopic ratio to a baseline reference measurement taken at time to, the change in isotopic ratio measured at time ti would be ΔRi. At the next measurement point, taken at time t₂, the change in measured isotopic ratio, as a result of drift in the instrument, would be measured as ΔR₂, and the accumulated change from the baseline level ΔR[ + ΔR₂. Similarly, at the third measurement point t₃, the measured change in ratio from the previous point is ΔR₃. and the accumulated change from the baseline level ΔR( + ΔR₂ + ΔR₃. As is seen from the graph, the various values of ΔR_n can be significant, and the accumulated ΔR even more so, especially for breath tests which continue for a considerable time in comparison to the stability level ofthe instrument.

According to this preferred embodiment of the present invention, the collection of pairs of samples at each measurement point can reduce this error substantially. Thus, for instance, at time t_\, one of the pair of samples collected is used for comparing the ratio measurement at time ti with the baseline reference sample measured at time t₀. The second sample ofthe pair is kept intact until time t_2} and is then used as the reference sample against which one of the pairs of samples collected at time t₂ is compared. In this way, the change in measured ratio ΔR₂ due to instrumental drift is nullified, since a sample from time ti is compared in the instrument with a sample from time t₂ under essentially identical conditions, namely those extant in the instrument at time t₂. Similarly, comparison of the second sample kept from t₂ with one from time t₃ enables the apparent shift in ratio ΔR₃ to be effectively nullified. A similar argument applies for all successive measurement points in the breath test.

The above explanation is somewhat simplified in that it assumes that the ratio measurement of each sample takes infinitesimal time, and that both samples are measured simultaneously at each point in time, and thus under identical conditions. In practice, each ratio measurement takes a time Δt, and at point 2, for instance, the measurement ofthe ratio ofthe reference sample kept from time tj is performed at a time Δt earlier than that of one of the pair of samples collected at time t₂. During that time Δt, the drift ofthe instrument continues, and the result is that the ratio r₂ measured at time t₂, is different from the ratio measured at time (t₂+Δt) by an amount Δr. However, since the time Δt taken for the ratio measurements themselves is generally significantly shorter than the elapsed time between successive measurement points, the use of the sample pair measurement method, according to this preferred embodiment of the present invention, still results in considerably increased accuracy, and considerably increased immunity from instrumental drift.

Furthermore, the sample pair method according to this preferred embodiemnt of the present invention still results in a significant measurement improvement, even when account is taken ofthe time taken to dilute each sample down to its target concentration and to measure the concentrations reached, this process being performed so that the samples are compared at identical concentrations. The dilution and concentration measurement process may, in fact, take considerably longer than the time, Δt, taken to perform the ratio measurement itself. If the dilution and concentration measurement procedure is performed in the interim period between measurement points, both for the reference sample collected from the previous measurement point, and for the sample currently being measured, then when the time to make a ratio measurement arrives, the samples are already diluted to their target concentration, and are ready to be measured immediately, with a time difference between comparative measurements of no more than Δt.

This preferred method of collection and measurement of pairs of samples , and use of the RCIR parameter for calculation of the relative change in isotopic ratio of the samples, may be advantageous and applicable for all types of breath tests, both those which use bags for sample collection, which are then analyzed at a time and place not necessarily related to the time and place of collection of the samples, and also those which are performed in real time, with the subject connected to the breath test instrument such that his breath is capable of being monitored almost continuously by the breath test instrument.

(c) Baseline determination methods

In the known prior art, the baseline is determined by determining the isotopic ratio in a single measurement taken from a single breath sample, or group of samples, obtained before the ingestion of the labeled substrate. This method may produce inaccurate results if the single baseline point obtained is incorrect, due either to drift or a high noise level in the instrument, or to physiological "noise" in the breaths supplied by the patient, when for some clinical reason, successive breaths give significantly different isotopic ratios.

In order to overcome these sources of potential inaccuracy, according to another preferred embodiment of the present invention, a baseline determining method is operative to review the quality of the measured baseline point, and if necessary, to measure one or more additional baseline points before the patient's ingestion ofthe labeled substrate.

According to a first additional preferred embodiment related to baseline measurement, if the self-diagnostic procedures operative within the instrument indicate that the quality of the point measured is high, such as is determined for instance, by the presence of a low standard deviation of scatter of the separate results making up that first baseline measurement point, or by the achievement of a carbon dioxide concentration close to the target value, then the control system concludes that a single baseline measurement is sufficiently accurate.

According to a further preferred embodiment, a second baseline measurement is taken. If the two measurements are within a predetermined value of each other, then it is assumed that no interference, neither instrumental nor physiological, was operative in the baseline measurement, and a simple mean of the two values is preferably used. If, on the other hand, discrepancy is detected between the two values, or one of the points is suspect as being of poor quality, as determined for instance, by the criteria given above, a number of possibilities present themselves. The point of poor quality can preferably be discarded and only the good point used in defining the baseline level. Alternatively and preferably, the system can request the measurement of a third baseline point. The result of this third baseline point preferably determines which of the results are used. If it is apparent that the first two values are scattered because of a noise problem, then the method takes a simple mean of all three values. Alternatively, if it is apparent from the third measurement that one of the first two values is severely discrepant from the other two, one ofthe first two values can be rejected as being delinquent.

According to further preferred embodiments of the present invention, the baseline measurements are performed in such a manner as to speed up the progress of the test. Thus, for instance, the second point can preferably be collected before completion of the analysis of the first point. If it becomes clear before completion of the measurement of point number 2, that point number 1 is of poor quality and cannot be used, then the calculation routine requests the collection of a third baseline sample, before completion of the measurement of point No. 2.

According to a further preferred embodiment of the present invention, in order to speed the test up even more, the patient is given the labeled substrate to ingest even before the results of the second measurement is known, but after the collection of the breath sample for the second baseline measurement. In this case, if a check of the quality of the measurement, as determined for instance by the standard deviation of the separate points making up the measurement, or by the accuracy of the achieved C0₂ target concentration, reveals that one of the measurements is likely to be of poor accuracy, then that measurement can be discarded in the calculation of the baseline. If both are of good quality, then their average may be used.

(d) Threshold utilization methods

A breath test for the detection of a specific clinical state, in common with many other diagnostic tests, relies on the attainment ofa specific result or level as the indication of a positive result of the test. The provision of a definitive diagnosis about the absence of the sought-after clinical state, or, in other words, the definition of a negative result, is a much more difficult task.

The operational function in a breath test is to determine when a change in the isotopic ratio of a component of breath samples of the subject is clinically significant with respect to the effect being sought. The criterion for this determination, as used in much of the prior art, is whether or not the DoB has exceeded a predefined threshold level, at, or within the allotted time for the test. When the result is positive, the decision is simpler, even though some doubt may still remain when the threshold crossing is not decisive, such as when a slow upward drift of the DoB value is obtained. When, however, the DoB hovers between the baseline and the threshold, with random noise perhaps sending it over the threshold occasionally, it becomes a much more difficult task to make a definite diagnosis that the result is really negative. The interpretation of negative results is of special importance in the commonly used urea breath test for detecting the presence of H. pylori in the upper GI tract, because of the widespread prevalence of ulcers and similarly related conditions of the upper GI tract in a significant proportion ofthe population.

One solution to this problem is to allow the test to continue for a longer time, to detennine whether the result does or does not remain negative. This, however, involves patient inconvenience and the unnecessary occupation of expensive instrumentation, which could otherwise be used for the testing of further patients. It is therefore important to devise a means of defining the optimum breath test threshold level which enables definite results to be obtained, and especially negative results, in a minimum of time.

In order to accomplish this objective, and to achieve the highest sensitivity and specificity in the shortest possible measurement time, the breath test analyzer according to another preferred embodiment of the present invention, does not use fixed criteria for determining whether the change in the isotopic ratio of a patient's breath is clinically significant. Instead, the criterion is varied during the course of the test, according to a number of factors operating during the test, including, for instance, the elapsed time of the test, the noise level of the instrument performing the test, and the physiological results ofthe test.

Furthermore, although the traditionally used measurement ofthe change in the isotopic ratio has been the level ofthe ratio over a baseline level, according to other preferred embodiments of the present invention, the measurement could be the change over a previous measurement point other than a baseline level, or the rate of change of the isotopic level, or any other property which can be used to plot the course ofthe change.

In order to illustrate these novel measurement methods, a preferred embodiment of the present invention is now presented in which a threshold utilization method uses the real time results of the test to determine whether enough data has been accumulated in order to decide that the test is complete. This is accomplished by using a dynamic threshold level, whose value can be changed by the threshold utilization method during the course of a measurement, as a function of one or more of the following quantities:

(i) as a function of the nature of the data accumulated, i.e. whether the test is giving a well-defined result or not,

(ii) as a function ofthe level and trends o the results obtained, (iii) as a function ofthe standard deviation ofthe accumulated data, and (iv) as a function of the noise and drift level in the particular instrument being used at that particular time.

In addition, according to a further preferred embodiment, two thresholds may be utilized, an upper and a lower threshold, which converge as the significance of the data collected becomes clearer.

A distinction can be made between a method of determining the use of variable thresholds at the beginning of a test, after the subsidence of oral activity, and as the test proceeds. During the early period of a test, before the accumulation of a large quantity of data, in order for a definitive result to be decided, the points must fall either very definitely above the baseline, which implies a positive result, or must be very close to the baseline, which implies a negative result. As the test proceeds, and the physiological result of a large number of accumulated points of measurement are taken into account, then the threshold criteria to be used should be more statistically based on the data accumulated.

The result of these two approaches is shown by reference to Fig. 6, which illustrates a preferred method ofthe use of double and dynamic threshold criteria. In Fig. 6 is shown a plot of the threshold values used as a function of the time elapsed from ingestion of the isotope labeled substrate at time t₀. In the region 500 from the beginning of the test at time t₀, up to time t_\, breath samples are generally not taken into account in determining whether the upper threshold has been reached because of the possible existence of oral activity arising from the breakdown of the substrate by bacteria present in the patient's mouth. If, on the other hand, the oral activity detection system determines that no significant oral activity is present, then rising results may be taken into account in determining when the upper threshold has been crossed, even from time t₀. Results close to the baseline may generally always be taken into account, since they are obviously unaffected by the presence of oral activity, if any.

After the elapse of time t_1: when oral activity, if any, has subsided and all of the breaths collected are taken into account for calculation of the test results, two threshold levels are used, an upper threshold T_u 506, and a lower threshold Ti 508. Throughout the whole of the region 502, when very little data has been accumulated, widely different values of T_u and T_\ are used, 85 and 2δ in the preferred embodiment shown in Fig. 6. A result below 2δ is regarded as a negative result, while one above 8δ is regarded as a positive result. Results falling within the region between T_u and Tj are indefinite and require the accumulation of more data. According to further preferred embodiments, the threshold levels need not be constant in this region, but could commence widely apart at time t_l5 and slowly converge with time, as shown by the alternative upper threshold curve 507.

An important aspect of this preferred embodiment is that already by the time ti has been reached, which could be as little as 4 minutes from the ingestion of the urea, it is possible to make a definite positive or negative diagnosis, if the results obtained are sufficiently deviant from the expected baseline, 8δ and 2δ in the preferred embodiment described. A preferred criterion for a definite diagnosis is that if two points are obtained above the upper threshold, to the exclusion of any points below the lower threshold, then the diagnosis is positive. Similarly, the existence of two points below the lower threshold, to the exclusion of any points above the upper threshold, is a sufficient criterion to provide a negative diagnosis. A more stringent and preferred criterion for a negative diagnosis, bearing in mind the difficulties mentioned above in determining whether a result is truly negative, is the presence of any 3 or 4 successive points after the urea administration, falling either below the lower threshold or with an average below this threshold, and a standard deviation of less than lδ and with a slope of less than O.lδ per minute, and with no significant instrument drift or trend. Expressed qualitatively, this implies that for a measurement which shows the DoB plot to be fairly flat, and on an instrument known historically to be stable in operation, then a negative result can be determined sooner than by any prior art method. The speed with which a diagnosis can be provided thus distinguishes one aspect ofthe preferred embodiments of the breath test of the present invention from those of prior art instruments and methods, both for positive and for negative results

After elapse of a time t₂ , shown at 6 mins in the embodiment of Fig. 6, when a sufficient number of points have been accumulated, the standard deviation of the results should have dropped, and the physiological result of the test, if conclusive in either direction, should be evident enough that the thresholds may be closed up gradually, and even moved in the direction which provides the highest sensitivity and specificity for the test in progress. The time t₂ from which this threshold closure begins to operate, and the rate of closure itself of the thresholds 510, 512, and the rate of change of any level common to both of them, are determined by the data themselves, or are predetermined. Points with standard deviations which converge rapidly from the best-fit curve, result in speedy threshold closure and movement, and vice versa for slowly converging standard deviations. After a time t₃ , which too is determined by the degree of scatter ofthe points, the two thresholds converge into one threshold 514, at the traditionally used level of 5 δ in the embodiment shown, and from that time on, a diagnosis is made on the basis ofthe result falling above or below this threshold.

If, on the other hand, the data points remain scattered, and their average level close to 5δ, a fault message is displayed and the test concluded without a meaningful result.

In use, the threshold utilization method shown in Fig. 6 may preferably operate as follows. A subject with a rising DoB result who is not above 8δ after 6 minutes, is evaluated again after 8 minutes with a threshold of 7δ, and if not over this new threshold, again after 10 minutes with a 6δ threshold, until the traditional 5δ is reached. Likewise, a subject showing a non determined DoB trend, who is not below 2δ after 6 minutes is preferably evaluated again after 8 minutes with a threshold of 3δ, and if not below this new threshold, again after 10 minutes with a 4δ threshold, until the traditionally used 5δ is reached.

As the time ofthe test proceeds, and the data accumulated more accurately reflects the physiological reality of the breath test mechanism, whereby the DoB of positive subjects rises as time proceeds, then according to a further preferred embodiment of the present invention, the threshold used 515 is allowed to rise slowly with the continuation ofthe test time.

According to yet another preferred embodiment, the lower threshold can remain independently existent instead of coalescing with the upper threshold to form a single threshold level. This is shown in Fig. 6 by alternative and preferable lower threshold 516, which either remains constant at a certain upper level, shown at 4δ in the example illustrated, or slowly falls with time, as the differentiation between a negative result and a positive result becomes more pronounced.

The validity of the use of variable thresholds according to the above preferred embodiments has been verified by comparisons of the results obtained with a breath tester incorporating such threshold embodiments, with results obtained by means of a standard monitoring test method. For the case of the H. Pylori breath test, such a standard monitoring method is by the endoscopic collection of a biopsy from the stomach ofthe patient, followed by a histological examination ofthe tissue, or by means of a culture ofthe bacteria present.

According to a further preferred embodiment of the present invention, the threshold values used are made dependent on the self-diagnostic outputs of the system. For an instrument giving poor results, such as having a high noise level, or a systematic noise pattern, or a trend in the results, the threshold utilization method uses wide initial threshold values, namely a comparatively high upper threshold and a low lower threshold. For an instrument giving good results, such as having a low noise level or a low level of drift, the upper and the lower threshold levels and also the difference between them can all be lowered, and the number of points falling below the threshold required to define a negative result can then be decreased. In this way, it becomes possible to define a test result as being negative or positive earlier than with a fixed threshold.

According to a further preferred embodiment of the present invention, a threshold utilization method is used for determining whether sets of points measured can be considered to give a definitive test result. A best fit polynomial, preferably of second order, is constructed to functionally approximate the plot of the points obtained in one test. Only those points measured following the effective cessation of oral activity are used in the construction of this polynomial. A weighted standard deviation of those points from the calculated polynomial curve is then calculated. The weighting is performed such that for points less than, for example, 5δ above the baseline, the deviation from the polynomial curve is taken as is. For points of over 5δ above the baseline, since the effect of random noise on such a measurement is far less significant, the calculation method takes only 20%) of the deviation of the point from the polynomial as the effective error value for the purpose of calculating the weighted standard deviation. If this weighted standard deviation is greater than 1.5σ, the measurement is regarded as problematic because ofthe high scatter of results.

In addition to the standard deviation criterion, there are other preferred criteria on the basis of which the measurement may be rejected as being inconclusive. An example of one such preferred criterion is for instance if at least 2 out of 5 points measured fall within the threshold limits, such as from 3δ to 7δ above the baseline, while at the same time, the last measured point is less than lOδ above baseline, indicating that there is no strongly positive result. Even though the final point alone would seem to indicate a positive result, the calculation routine rejects the measurement, as extant at that point in time, because of the uncertainty introduced by the previously mentioned "2 out of 5" criterion.

In some parts of this specification, the procedures and computational methods have been described in terms of the urea breath test for the detection of Helicobacter Pylori in the upper gastro-intestinal tract. It is to be understood that this test is only an example of many diagnostic tests which can be performed by means of breath testing, and that the invention is not meant to be limited to the preferred explanatory examples brought from the H. pylori detection breath test.

It will be appreciated by persons skilled in the art that the present invention is not limited by what has been particularly shown and described hereinabove. Rather the scope of the present invention includes both combinations and subcombinations of various features described hereinabove as well as variations and modifications thereto which would occur to a person of skill in the art upon reading the above description and which are not in the prior art.

Claims

CLAIMSWe claim:

1. A method of performing a breath test, comprising the steps of: using a predetermined criterion for determining when a change in a measurement of an isotopic ratio of at least one breath sample of a subject is clinically significant; and allowing said criterion to change during said breath test.

2. The method according to claim 1 and wherein said change in said measurement comprises the deviation of said isotopic ratio from a measurement of at least one previous sample of said subject.

3. The method according to claim 2 and wherein said measurement of at least one previous sample of said subject is a baseline measurement.

4. The method according to claim 1 and wherein said change in said measurement comprises the rate of change of said isotopic ratio.

5. The method according to any of claims 1 to 4 and wherein said criterion is a function ofthe elapsed time ofthe test.

6. The method according to any of claims 1 to 4 and wherein said criterion is a function of the noise level ofthe instrument performing said test.

7. The method according to any of claims 1 to 4 and wherein said criterion is a function of the physiological results of said test.

8. A breath test method, comprising the steps of: performing a first measurement of the isotopic ratio of at least a first breath sample of a subject; performing a second measurement of the isotopic ratio of at least a second breath sample of said subject; and determining when said second measurement shows sufficient deviation from said first measurement that a clinically significant result ofthe breath test may be concluded; wherein the level of said sufficient deviation is allowed to undergo variation during said breath test.

9. The method according to claim 8 and wherein said first measurement of the isotopic ratio of at least a first breath sample of said subject is a baseline measurement.

10. The method according to either of claims 8 and 9, and wherein said level of sufficient deviation is at least a function of the elapsed time from ingestion of an identifying substrate by a subject.

11. The method according to either of claims 8 and 9 and wherein said level of sufficient deviation is at least a function of the physiological results of the analysis of at least one of said samples.

12. The method according to either of claims 8 and 9 and wherein said level of sufficient deviation is at least a function of the nature of the results obtained in said breath test.

13. The method according to claim 12 and wherein said nature ofthe results is at least a function ofthe standard deviation ofthe spread of results in said breath test.

14. The method according to claim 12 and wherein said nature ofthe results is at least a function ofthe noise level present in said results.

15. The method according to claim 12 and wherein said nature ofthe results is at least a function ofthe instrumental drift present in said results.

16. The method according to either of claim 8 and 9 and wherein said level of sufficient deviation covers a range of values between an upper threshold and a lower threshold.

17. The method according to claim 16 and wherein said upper threshold and said lower threshold converge as the test proceeds.

18. The method according to either of claim 8 and 9, and wherein said first measurement of the isotopic ratio of at least a first breath sample of a subject is performed only after significant subsidence of oral activity.

19. The method according to claim 16 and wherein the presence of at least two measurements above said upper threshold within a predetermined time is indicative of a positive result of said breath test.

20. The method according to claim 16 and wherein the presence of at least two successive measurements above said upper threshold is indicative of a positive result of said breath test.

21. The method according to claim 16 and wherein the presence of at least two measurements above said upper threshold in combination with the absence of measurements below said lower threshold is indicative of a positive result of said breath test.

22. The method according to claim 16 and wherein the presence of at least two measurements below said lower threshold within a predetermined time is indicative of a negative result of said breath test.

23. The method according to claim 16 and wherein the presence of at least two successive measurements below said lower threshold is indicative of a negative result of said breath test.

24. The method according to claim 16 and wherein the presence of at least two measurements below said lower threshold in combination with the absence of any measurements above said upper threshold is indicative of a negative result of said breath test.

25. The method according to claim 16 and wherein the presence of at least two successive measurements falling below the upper threshold, with less than a predetermined difference in the value of delta between said three measurements, and with less than a predetermined average slope between them, is indicative of a negative result of said breath test.

26. The method according to claim 16 and wherein the presence of at least two successive measurements falling below the lower threshold, with less than a predetermined average slope between them, is indicative of a negative result of said breath test.

27. A breath test method for deteπnining the presence of a clinically significant state in a subject comprising the steps of: performing measurements of the changes from a baseline of an isotopic ratio in a plurality of samples of exhaled breath of said subject, following the effective cessation of oral activity; determining a polynomial which approximates the functional plot of said measurements with time; calculating a weighted standard deviation of said measurements from said polynomial, wherein for measurements over said baseline by more than a predetermined amount, a predefined fractional part of the measurement is taken, while for measurements not over said baseline by more than said predetermined amount, the measurement is taken in its entirety; and determining whether said weighted standard deviation exceeds a predetermined level.

28. A breath test method, comprising the steps of: performing a measurement of the isotopic ratio of at least a first breath sample of a subject; and determining when said measurement shows sufficient deviation from a baseline measurement that a clinically significant result of the breath test may be concluded; wherein said deviation comprises an upper and a lower threshold band of uncertainty, and wherein the extent of this band is dependent on at least one of the parameters selected from the group consisting ofthe elapsed time of said breath test, the standard deviation of the physiological spread of results, the dynamics of the physiological change in isotopic ratio, the number of points measured in the breath test, the environmental conditions present during the breath test, and the noise and/or drift levels ofthe instrument executing the breath test.

29. A method for determining the reliability of a breath test, comprising the steps of: obtaining results from said breath test; defining a reliability parameter by combining at least one of the criteria selected from the group consisting ofthe instrument noise and/or drift level, the standard deviation of the physiological spread of results, the dynamics of the physiological change in isotopic ratio, and the time elapsed since ingestion of a labeled substrate; and using said reliability parameter to assess the results of said breath test according to a predetermined reliability criterion.

30. A method for determining the reliability of a breath test measurement according to claim 29, wherein said parameter is used in order to determine when to terminate said test.

31. The method according to either of claims 29 and 30, wherein said reliability parameter is output with the result ofthe breath test.

32. A method of calibrating a breath test instrument without the need for externally supplied calibration means, comprising the steps of: continuously measuring isotopic ratios of a gas species in said samples in a plurality of subjects; and searching for correlation between the isotopic ratios of said gas species and the concentration of said gas species in said samples.

33. The method of calibrating a breath test instrument according to claim 32 and also without the need for operator involvement.

34. The method of calibrating a breath test instrument according to claim 32 and also without the need for active subject involvement.

35. A method of calibrating a breath test instrument without the need for externally supplied calibration means, comprising the steps of: substantially continuously measuring isotopic ratios of a gas species in said samples in a plurality of subjects; and searching for correlation between the isotopic ratios of said gas species and an environmental condition present at the time of said breath tests.

36. The method according to claim 35, wherein said environmental condition is the ambient temperature.

37. A method of calibrating a breath test instrument, by analyzing results obtained on breath samples of a plurality of subjects not showing change of any significance in the isotopic ratios of a specific gas species measured in said samples, for correlation between said isotopic ratios and the concentration of said gas species in said samples.

38. A method of calibrating a breath test instrument, by analyzing results obtained on a plurality of collected breath samples from one subject for correlation between the isotopic ratios of a specific gas species measured in said samples and the concentrations of said gas species in said samples.

39. A method of calibrating a breath test instrument, comprising the steps of:

(a) collecting a breath sample containing a specific gaseous species;

(b) measuring the concentration of said specific gaseous species in said sample;

(c) determining the isotopic ratio of said specific gaseous species in said sample;

(d) diluting said sample such that said concentration of said specific gaseous species changes;

(e) determining said isotopic ratio again;

(f) repeating steps (d) to (f) to obtain measurements on a number of different concentrations of said sample;

(g) looking for correlation between isotopic ratios and concentrations of said different concentrations of said sample; and

(h) adjusting the calibration of said breath test instrument to reduce any correlation found.

40. A method of correcting a change in the calibration of a gas analyzer for determining the isotopic ratio between a first isotopic species of samples of a gas and a second isotopic species of said samples of a gas, comprising the steps of:

(a) assigning a concentration Cl to each measured value of transmission TI of said first isotope species for each sample, by assuming given values for the parameters ofthe absorption curve for said first isotope species;

(b) calculating the concentration, C2 of the second isotopic species for each sample measured, by assuming a predetermined ratio between the concentrations of said first isotopic species and said second isotopic species;

(c) by means of a best-fit calculation, generating new parameters of the absorption curve for said second isotope species, by using, for each sample, said generated value of C2 and a measured value of the transmission T2 of said second isotope species;

(d) generating a set of corrected transmissions T2_C of said second isotopic species, by insertion of said calculated values of C2 obtained in step (b) into said new absorption curve for said second isotope species generated in step (c);

(e) calculating the differences between said measured values of transmission T2 of said second isotope species, and said corrected transmissions T2c obtained in step (d);

(f) calculating a set of normalized error differences ΔT2 = (T2 - T2) / T2 for said second isotopic species;

(g) by means of a best- fit calculation, generating from said set of normalized error differences ΔT2, a polynomial of ΔT2 as a function of the values of the concentration Cl of said first isotopic species; and

(h) using said polynomial to obtain iteratively corrected values of transmissions T2c' of said second isotopic species, in place ofthe initially measured values T2.

41. A method of correcting a change in the calibration of a gas analyzer for determining the isotopic ratio between a first component and a second component of a gaseous sample, comprising the steps of: measuring the concentration of said first component by means of optical transmission measurements; calculating the concentration of said second component from said measured concentration of said first component, by assuming a predetermined ratio between said components; and correcting transmission measurements made on said second component such that a concentration derived therefrom is essentially equal to the concentration calculated in the previous step from said measured concentration of said first component.

42. The method of claim 41, wherein said components of said gas samples are isotopic components.

43. The method of claim 42, wherein said predetermined ratio is the isotopic ratio of carbon 13 to that of carbon 12, as found in the naturally occurring mineral Pee Dee Belemnite limestone.

44. The method of claim 42, wherein said components are components of collected breath samples, and said predetermined ratio is a ratio of said components typical of a population to be measured.

45. The method of claim 41, wherein said step of correcting said transmission measurements made on said second component is performed by calculating corrected absorption curves for said second component.

46. A method of retroactively correcting the results of a breath test from the effects of incorrect calibration, comprising the steps of: performing a calibration procedure according to the method of claim

37 to determine the existence of correlation between measured isotopic ratios and the concentration of gas species in the breath samples; correcting the calibration of the instrument by means of corrected parameters ofthe gas absorption curves to eliminate said correlation; and recalculating the data of prior breath tests using said absorption curves with corrected parameters.

47. A method of retroactively correcting the results of a breath test from the effects of incorrect calibration, comprising the steps of: performing a calibration procedure according to the method of claim

38 to determine the existence of correlation between measured isotopic ratios and the concentration of gas species in the breath samples; correcting the calibration of said instrument by means of corrected parameters of its gas absorption curves to eliminate said correlation; and recalculating the data of prior breath tests using said absorption curves with corrected parameters.

48. The method of claim 47, and comprising the additional step of collecting at least one additional sample from said subject before correcting the calibration of said instrument.

49. A method of calibration of a capnographic probe, operative for measuring input breath waveforms in a breath test instrument, comprising the steps of:

(a) estimating the integrated concentration ofthe accumulated breaths collected according to the measured capnograph waveforms;

(b) measuring the concentration of a sample of said accumulated breaths in the gas analyzer ofthe breath test instrument; and

(c) correcting the calibration of said capnographic probe such that it provides the same concentration as that measured by the gas analyzer.

50. A breath test instrument which monitors changes in an isotopic ratio of a gas in exhaled breath samples of a subject virtually continuously, and determines that said test has a clinically significant outcome in accordance with the ongoing results of said test.

51. A breath test instrument comprising a signal for indicating that a clinically significant outcome of a breath test has been determined.

52. A breath test instrument according to claim 51, and wherein said signal is a visible signal.

53. A breath test instrument according to claim 52, and wherein said signal is an audible signal.

54. A breath test instrument according to claim 53, and wherein said signal also indicates the nature of the clinically significant outcome of said breath test.

55. A breath test instrument according to claim 50, such that the outcome of said test is substantially independent of dynamic physiological effects occurring in said subject as a result of background conditions.

56. A breath test instrument according to claim 55, and wherein said background conditions are the result of treatment with a drug therapy.

57. A breath test instrument according to claim 55, and wherein said background conditions are the result of food intake in a period prior to the performance of the breath test.

58. A breath instrument according to claim 57, and wherein the need for a pre-test fast by the subject is obviated.

59. A breath test instrument according to claim 50, and wherein the outcome of said test on said subject undergoing treatment with a gastro-intestinal drug therapy, is obtained more reliably than using correspondmg breath tests which do not monitor said changes in an isotopic ratio substantially continuously.

60. A breath test instrument according to claim 50, and wherein the outcome of said test is obtained more reliably than would be obtained by corresponding breath test instruments which do not monitor said changes in an isotopic ratio substantially continuously.

61. A breath test instrument according to claim 50, and in which said outcome of said test is obtained sooner than would be obtained by corresponding breath test instruments which do not monitor said changes in an isotopic ratio substantially continuously.

62. A breath test instrument according to claim 50, and in which said ongoing results of said test enable a positive result to be determined even when said isotopic ratio does not clearly exceed a predetermined threshold level.

63. A breath test instrument according to claim 50, and in which said ongoing results of said test enable a negative result to be determined even when said isotopic ratio exceeds a predetermined tlireshold level.

64. A breath test instrument according to claim 63, and in which said negative result is determined even when said isotopic ratio exceeds said predetermined threshold level because of instrumental drift.

65. A breath test instrument according to claim 50, and in which said ongoing results of said test enable a negative result to be determined by the detection of correlation between said isotopic ratio and instrumental drift.

66. A method of determining whether the correct isotopically labeled substance kit is being used for a specific breath test, comprising the steps of: adding a marker element to said substance, said marker element being selected to have an immediate and short term effect on the breath test; and providing breath test instrumentation comprising a detector for said marker element.

67. The method of claim 66 and wherein said breath test instrumentation also comprises an enabling mechanism that allows the instrument to perform analysis of the results of the breath test samples only after detection of said marker element.

68. A method for determining when the effects of oral activity have subsided during execution of a breath test, comprising the steps of: determining a characteristic time required to detect the physiological effect of interest in the breath test; monitoring change in an isotopic ratio in samples of breath collected from a subject following the ingestion of an isotopically labeled substrate; detecting the presence of a meaningful peak over a predefined minimum threshold level occurring in said isotopic ratio, within a time shorter than said characteristic time.

69. The method according to claim 68, and wherein said meaningful peak subsides within a second predefined characteristic time from the ingestion of said isotopically labeled substrate.

70. The method according to claim 69, and wherein said second characteristic time interval is 8 minutes.

71. A method, in a breath test procedure, of determining a baseline level for an isotopic ratio of a gaseous species in exhaled breath of a subject before ingestion of an isotopically labeled substrate, comprising the steps of: performing a measurement of a first baseline point; assessing the reliability of said measurement; and performing a second measurement of at least one additional baseline point if the reliability of measurement of said first baseline point is determined to be inadequate.

72. The method according to claim 71, and wherein said reliability of said measurement of said first baseline point is determined to be inadequate if the standard deviation of scatter of the separate results making up said measurement exceeds a predetermined value.

73. The method according to claim 71, and wherein said reliability of said measurement of said first baseline point is determined to be inadequate if a sample with a concentration of said gaseous species within a predetermined level of the target value has not been collected.

74. A method, in a breath test procedure, of determining a baseline level for an isotopic ratio of a gaseous species in exhaled breath of a subject, before ingestion of an isotopically labeled substrate, comprising the step of measuring at least first and second baseline points.

75. The method according to claim 74, and wherein the mean of said two points is used as the baseline value, if the first two of said at least two baseline points fall within a predetermined range of each other.

76. The method according to claim 74, and wherein a third baseline point is measured if the first two of said at least two baseline points do not fall within a predetermined range of each other.

77. The method according to claim 76, and wherein if the first two of said at least two baseline points do not fall within a predetermined range of each other, the point more distant from said third baseline point is discarded.

78. The method according to claim 74, and wherein the breath test procedure is expedited by collection of said exhaled breath of a subject for determining said second baseline point before analysis of said first baseline point is completed.

79. The method according to claim 74, and wherein the breath test procedure is expedited by ingestion of said isotopically labeled substrate before the results ofthe analysis of said second baseline point are known.

80. A method of determining change in isotopic ratio in a plurality of at least a first, a second and a third gaseous sample collected at different points in time, wherein said change in isotopic ratio is determined by measuring the isotopic ratio of said second sample in relation to said first sample, and in relation to said third sample.

81. A method of reducing the effect of changes in the operating conditions of a gas analyzer on isotopic ratios measured in a series of at least three gaseous samples, by measuring the isotopic ratio of at least one sample in relation to a sample collected before and a sample collected after said at least one sample.

82. A method of determining change in the isotopic ratio between a first and a second gaseous sample, comprising the steps of: (a) measuring said isotopic ratio of said first sample;

(b) measuring said isotopic ratio of said second sample;

(c) determining the difference between said isotopic ratios;

(d) dividing said difference by one of said ratios; and

(e) adding said change to a previous change determined between a prior first and second sample.

83. The method according to claim 82, and wherein said difference is divided by said isotopic ratio of said first sample.

84. The method according to claim 82, and wherein said difference is divided by said isotopic ratio of said second sample.

85. The method according to claim 82, and wherein said difference is divided by said isotopic ratio of said first sample when said ratio of said first sample is higher than said ratio of said second sample, and said difference is divided by said isotopic ratio of said second sample when said ratio of said second sample is higher than said ratio of said first sample.

86. A method of determining change in isotopic ratio in a plurality of at least a first, a second and a third gaseous sample collected at different points in time, wherein said change in isotopic ratio is determined by measuring the isotopic ratio of said second sample in relation to said first sample, and in relation to said third sample, each of said change in isotopic ratio being determined by the method of claim 82.

87. A method of using a gas analyzer for determining change in the isotopic ratio of gaseous samples relative to a predetermined isotopic ratio, wherein said gas analyzer is calibrated according to the method of claim 40 using the same predetermined isotopic ratio.

88. The method of claim 87, wherein said predetermined ratio is the isotopic ratio of carbon 13 to that of carbon 12, as found in the naturally occurring mineral Pee Dee Belemnite limestone.

89. The method of claim 87, wherein said gaseous samples are collected breath samples, and said predetermined ratio is a ratio of breath components typical of a population to be measured.

90. A method of determining change in the isotopic ratio between a first and a second gaseous sample, comprising the steps of:

(a) measuring said isotopic ratio of said first sample;

(b) measuring an isotopic ratio of a reference sample;

(c) computing a first difference between said first two isotopic ratios;

(d) measuring said isotopic ratio of said second sample;

(e) remeasuring an isotopic ratio of said reference sample;

(f) computing a second difference between said second two isotopic ratios; and

(g) subtracting one of said first and said second differences from the other.

91. A method of determining change in the isotopic ratio between a first and a second gaseous sample, comprising the steps of:

(a) measuring said isotopic ratio of said first sample;

(b) measuring a first isotopic ratio of a reference sample;

(c) computing a first difference between said isotopic ratio of said first sample and said first isotopic ratio ofa reference sample;

(d) normalizing said first difference relative to said first isotopic ratio of said reference sample;

(e) measuring said isotopic ratio of said second sample; (f) measuring a second isotopic ratio of said reference sample;

(g) computing a second difference between said isotopic ratio of said second sample and said second isotopic ratio of said reference sample;

(h) normalizing said second difference relative to said second isotopic ratio of said reference sample; and

(i) determining said change in said isotope ratio by subtracting one of said normalized differences from the other.

92. The method according to claim 91 and wherein said samples are breath samples collected from a subject.

93. The method according to claim 92 and wherein said reference sample is a baseline sample.

94. The method according to claim 92 and wherein said reference sample is obtained from a reservoir of said reference sample gas.

95. A method of determining in a breath test, change of an isotopic ratio in a plurality of breath samples of a subject, comprising the steps of:

(a) collecting a reference sample of breath;

(b) determining the isotopic ratio of a first one of said plurality of breath samples by comparison with that of said reference breath sample;

(c) determining the isotopic ratio of a second one of said plurality of breath samples by comparison with that of said reference breath sample; and

(d) computing the change in said determined isotopic ratios between said first one and said second one of said plurality of breath samples.

96. The method according to claim 95 and wherein said reference sample is a baseline sample collected before the effects of any ingested identifying substrate are observed.

97. The method of claim 95, wherein said steps of comparing said isotopic ratio of said plurality of breath samples with that of said reference breath sample is performed by measurement of said reference breath sample in its own reference chamber.

98. The method of claim 95, wherein said steps of comparing said isotopic ratio of said plurality of breath samples with that of said reference breath sample is performed by introducing said reference breath sample into a sample measurement chamber alternately between separate ones of said plurality of breath samples.

99. The method of claim 95, wherein a separate portion of said reference sample is used for comparing said isotopic ratios of each of said plurality of breath samples.

100. An apparatus for executing the method of claim 95, and comprising a reference chamber for containing said reference sample during measurement.

101. An apparatus for executing the method of claim 95, and comprising ports for introducing said reference breath sample into a sample measurement chamber alternately between separate ones of said plurality of breath samples.