INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In an attempt to decrease the variability in pulmonary function testing, the American Thoracic Society (ATS) has developed standardized procedures (1). One of the most important sections of these guidelines is the one outlining the criteria for acceptability and reproducibility. Acceptability is determined for each maneuver whereas reproducibility compares at least two maneuvers for consistency. The criteria for end of test (EOT), a mostly objective measure of maneuver acceptability, require that there be: (1) a minimum change in volume of 0.03 L over a 1 s period given a forced expiratory time (FET) of at least 6 s or; (2) a FET of reasonable duration (often 15 or 20 s); or (3) a point where there are clinical reasons not to continue forced exhalation. Recognizing that children, the elderly, and many patients often have difficulty in meeting EOT criteria (4), the ATS standards also state that exhalation times shorter than the recommended 6 s are often acceptable and that "early termination is not by itself a reason to eliminate a maneuver from further consideration. . . . The volume accumulated over a shorter period of time may be used as an approximate surrogate for FVC" (3). Although this statement allows for flexibility, it adds a subjective element to the determination of what constitutes an acceptable maneuver. This becomes particularly difficult when dealing with large epidemiological surveys or multicenter studies where clearly defined objective criteria are required.

The current recommendations for the EOT criteria were developed based on a report by Hankinson and Gardner (4) who used both actual volume time traces from patients as well as computer generated "ideal" waveforms. The computer generated waveforms were obtained using the exponential equation v(t) = FVC(1  e^t/TC), where v(t) = instantaneous volume in liters; FVC = forced vital capacity in liters; t = time in seconds; TC = time constant in seconds. FVC and TC were varied and chosen so as to produce a variety of waveforms representative of different lung volumes and flow rates. Based on these findings, the criteria for EOT were updated in 1987 (2) and again in 1994 (3). In reality, EOT criteria are often not met when testing young children, and it is, therefore, up to the operator to determine whether to retain any given effort as being "acceptable." During periodic reviews of data obtained in our pediatric pulmonary laboratory, it was found that while the strict criteria for acceptability are often not met, most of the tests do meet the criteria for reproducibility. One possible reason for the failure to meet the current criteria for acceptability in a pediatric setting is that these criteria were developed using adult subjects. To our knowledge, there have been no formal investigations as to whether the criteria for acceptability are in fact appropriate for use in a pediatric setting.

By using exponential curve fitting techniques, methodology was previously developed whereby actual volume time signals can be closely approximated by an exponential equation with a mean difference between measured and calculated FEV₁ of 0.0001 L (7). From this equation, it is then possible to determine the theoretical full FVC (FVC _FULL) for a given maneuver which would have occurred had the patient exhaled until the volume time (V/ T) curve reached its asymptote. This method could possibly be used to develop individually tailored guidelines for EOT for each maneuver since it would be based on the actual expiratory volume time signal rather than on predetermined limits. Since attainment of FVC _FULL as the new EOT criteria could be too stringent for all patients to meet for all maneuvers, 95% of FVC _FULL (FVC ₉₅) was designated as the new EOT.

The purpose of this report was to determine to what extent the current ATS recommendations for acceptability and reproducibility were met during routine testing in a pediatric pulmonary function laboratory. An additional goal was to determine new EOT criteria for each maneuver based on its exponential equation and resulting FVC₉₅ and to examine how well these new criteria are met in children.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Data used in the analysis were obtained from all patients referred for pulmonary function testing at the Children's Hospital of Eastern Ontario (CHEO) from January 1 to June 30, 1994. When a patient was referred on more than one occasion, only the most recent (last) visit was used in the analysis. Only tests with three recorded efforts were considered for analysis. Spirometry was performed using a Sensormedics 2800 Autobox (Sensormedics, Yorba Linda, CA). As is routinely done, all data (expired volume, box volume, time and flow data from three efforts) were first stored on hard disk and subsequently archived to floppy disks.

Evaluation of meeting ATS standards for EOT criteria (3) was carried out by determining the following: (1) what percentage of children had FET  6 s for at least two FVC maneuvers (FET  6); (2) what percentage of children met criteria for a minimum volume change of 0.03 L over 1 s on at least two FVC maneuvers (MINVOL); (3) what percentage of children met both of the above criteria on at least two FVC maneuvers (BOTH).

Reproducibility was evaluated by determining how many children met ATS reproducibility criteria for both FVC and FEV₁. This requires that at least one FVC be within 0.20 L of the largest FVC and that at least one FEV₁ be within 0.20 L of the largest FEV₁.

A previous report (8) found that the volume-time curve of 95% of patients was best described by a two-compartment model. The data used for curve fitting were obtained by sampling the curves at incremental changes in both volume and time for each effort. This was done because points selected based on time increments alone would emphasize the latter part of the expiratory effort, whereas points sampled exclusively on volume would emphasize the early part of the curve.

Using exponential curve fitting techniques, the waveforms for these FVC maneuvers were described using both a one-compartment model of the lung: v(t) = VOL(1  e^t/TC), and a two-compartment model: v(t) = VOL1(1  e^t/TC1) + VOL2(1  e^t/TC2). A two-compartment model was chosen over a one-compartment model to represent any given FVC waveform if it was found to result in a statistically significant partial F statistic (9) at the p < 0.01 level.

In order to validate the information from the fitted curves, data were examined from the best curve for each patient. Best curve was defined as that with the largest sum of actual FEV₁ (FEV_1meas) and actual FVC (FVC_meas) (1), and with a volume of back extrapolation  0.15 L or 5% FVC whichever is greater (3). From the exponential equation of the corresponding fitted curve, FEV₁ was calculated (FEV_1calc), and FVC was calculated at the point in time where the actual maneuver was terminated (FVC_calc). These calculated values were then compared with the corresponding measured values using paired t-tests.

Once the curve fitting techniques were validated, this procedure was used on each of the three efforts for each child to generate the fitted curves. From each fitted curve, the volume at which the curve reached its asymptote was determined as the potential maximal FVC (FVC_FULL) for that maneuver. The FVC₉₅ was then calculated as 0.95 × FVC_FULL.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

During the 6-mo period from January 1 to June 30, 1994, 512 pulmonary function tests were performed on 394 children with a variety of diseases, mainly asthma, but some with cystic fibrosis or other conditions. Of the 394 tests considered for analysis, seven were rejected for technical reasons (e.g., large hesitations, notation by the technician which questioned cooperation), and five were rejected because fewer than three trials were recorded. The data from the resulting 382 tests represent a broad range for both age (5 to 18 yr) and severity (FEV₁ ranging from 20 to 120% predicted).

The results of the evaluation based on the current ATS EOT and reproducibility criteria are presented in Table 1. These results were initially determined for all children together and then subsequently divided into two groups: first, according to age (those older than 7 yr and those younger) and, second, according to severity of airflow limitation (those whose FEV₁ < 65% predicted and those whose FEV₁  65% predicted). Only 10% of the younger children were able to attain the goal of FET  6 s for at least two breaths; whereas 28% of the older children managed to meet this criterion. When the results were examined on the basis of airflow limitation, more children with FEV₁ < 65% had FET  6 s (36%) than those with FEV₁  65% (23%). There was a similar pattern for attaining MINVOL although the percentages are somewhat higher. In contrast, more than 90% of all groups met the ATS criteria for reproducibility with at least one FVC within 0.2 L of the largest FVC and at least one FEV₁ within 0.2 L of the largest FEV₁. As can be seen in Table 2, 90% of the children met the new proposed EOT criteria based on the FVC₉₅, which is more in keeping with the reproducibility criteria.

The mean r² for the fitted exponential curves was 0.998 ± 0.002 (mean ± 1 SD). Two examples of actual and fitted exponential curves are presented in Figures 1A and B. These figures represent patients of the same age but with very different FEV₁s who both show excellent r². Figure 2 shows the actual and fitted curve for a child with one of the worst fits as judged by the r². Results of the paired t-test analysis of all 382 children showed that values for FEV_1calc and FVC_calc closely approximated those for FEV_1meas and FVC_meas although they were statistically different. The mean difference for FEV₁ was 0.015 ± 0.04 L (mean ± 1 SD). There were 16 patients whose difference exceeded 0.05 L or 3% (whichever is greatest); however, all were within 0.20 L. Similarly, FVC_calc differed from FVC_meas with a mean difference of 0.004 ± 0.03 L and here again all differences were within 0.20 L. 

View larger version (13K): [in this window] [in a new window]   View larger version (14K): [in this window] [in a new window]   Figure 1.   (A) Actual volume time curve (solid line) and fitted exponential curve (dotted line) for a 12 yr-old with FEV1 = 82% pred. The equation for the exponential curve was v(t) = 1.65*(1  e(t/ 0.4)) + 0.79*(1  e(t/0.95)) with r 2 = 0.99936. (B) Actual volume time curve (solid line) and fitted exponential curve (dotted line) for a 12 yr-old with FEV1 = 51% pred. The equation for the exponential curve was v(t) = 0.95*(1  e(t/0.4)) + 1.86*(1  e(t/2.45)) with r 2 = 0.9998.

View larger version (13K): [in this window] [in a new window]   Figure 2.   Actual volume time curve (solid line) and fitted exponential curve (dotted line) for a 15 yr-old with FEV1 = 100% pred. The equation for the exponential curve was v(t) = 0.8*(1  e(t/0.4)) + 2.8*(1  e (t/0.5)) with r 2 = 0.99197.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The finding that over 90% of the children met the proposed new acceptability criteria based on their individualized EOT criteria, FVC₉₅, in at least two FVC maneuvers is in sharp contrast to the 19% who met acceptability criteria based on strict application of the current ATS criteria. In addition, the results obtained with FVC₉₅ criteria are more consistent with the findings for reproducibility.

Although it is a widely held belief among those responsible for pediatric pulmonary function testing that the current criteria for EOT are not applicable in a pediatric setting, to our knowledge, the only study examining acceptability criteria that included a pediatric population is that of Hankinson and Bang (10). In their study of 6,486 subjects, aged 8-90 yr, they found 301 (4.6%) had fewer than two acceptable curves. Although they eliminated these patients from further consideration in their analysis of reproducibility, it is interesting to note that 261 (87%) of these patients were under 18 or over 55 yr of age. It is not possible to tell from their report how many of these test failures were less than 18 yr of age and were due to failure to meet EOT criteria.

The development of the EOT criteria was an attempt to quantify the otherwise qualitative assessment of maneuver acceptability. However, this report and others have shown, there are many situations in which strict adherence to the objective elements of the current EOT criteria are difficult and may lead to rejecting useful information (4). This has led over the years to revisions in the ATS standards as well as proposals for alternative procedures in determining EOT (11, 12). In fact, the standards of lung function testing issued by the European Respiratory Society (13) do not include EOT criteria as they felt that many patients could not exhale long enough to meet the criteria.

Similar to the techniques of Hankinson and Gardner (4), the volume time tracing was expressed as an exponential equation. However, in the present study this was done for each maneuver on each subject making the results very specific for each individual which could then take into account the possible effects of age and/or pulmonary disease on the volume-time curve.

Failure to reach FVC₉₅ can occur simply because of premature termination of expiration or because an adequate exponential curve fit cannot be found. It is important to note that none of the patients in this series were eliminated because of how well they fit the exponential model. Of the 38 patients who did not have two acceptable FVC maneuvers based on FVC₉₅, 25 (66%) were  7 yr old. Most of these children had normal convex shaped flow volume curves (14) (using plethysmograph versus flow at the mouth) with inconsistent peak flows followed by abrupt cessation of flow suggesting failure to expire completely. This highlights the difficulty in obtaining consistent efforts when testing many younger children. In the remaining patients who did not meet EOT criteria based on FVC₉₅, most showed a wide variability in effort that would again suggest a problem with cooperation.

The children in this study represent a cross-sectional analysis of patients referred for pediatric pulmonary function testing and, as such, this study does not undertake to determine how applicable exponential curve fitting is in certain specific disease conditions or in other age groups. The vast majority of children were referred for investigation or follow-up of asthma or management of cystic fibrosis. In addition, it did not include any patients with neuromuscular disease as their pulmonary function tests are performed using different equipment. In this series of patients, we could only clearly identify one patient in whom exponential curve fitting was probably not justified. This was a patient referred with the diagnosis of tracheal stenosis whose volume time trace did not have an exponential shape.

The present study has shown that FVC₉₅ is a better indicator of EOT in most children referred for pulmonary function testing than the current ATS criteria as it is individually determined and yields acceptability rates that are more consistent with those for reproducibility. Although one would assume that this methodology would be applicable to adult patients, further research is required to confirm this. In addition, it is important to examine specific disease groups as well as a group of normal subjects.

It is important to note that the FVC₉₅ is determined for each individual maneuver and yields information as to whether the patient has expired to 95% of the potential FVC for that trial. As with previous EOT criteria it does not, by itself, give information about maximal effort or reproducibility. These are very important but separate issues. This again highlights the need for qualified experienced personnel to determine whether additional efforts are required.

In conclusion, these data suggest that a new EOT be established for children based on the achievement of 95% of their theoretical FVC. This new EOT would allow for the application of strictly objective criteria for determining the acceptability of a given maneuver.