Introduction

Periodic spirometry testing is performed in medical screening and surveillance programs for workers with various occupational exposures and for cigarette smokers. The recent American College of Occupational and Environmental Medicine (ACOEM) evidence-based statement, Spirometry in the Occupational Setting, comprehensively reviewed the issues involved in conducting and interpreting standardized spirometry tests in occupational medicine.(1) However, interpreting change over time was only briefly discussed in that statement and little other guidance on assessing longitudinal change in lung function is available for health professionals. As a result, many practitioners do not evaluate change in lung function over time, but instead repeatedly determine whether each year's test results fall within the normal range. Other practitioners evaluate change over time or "trending" but are unaware of the pitfalls that can distort their evaluations.

Although health professionals must determine whether evaluating lung function change over time effectively screens for a specific outcome disease,(2-4) the ACOEM Occupational and Environmental Lung Disorder Committee recognized the need to provide guidance in the selection and use of simple measures of change over time. The Committee developed this separate ACOEM position statement to: 1) explain the need for longitudinal analysis of pulmonary function when evaluating employee respiratory health; 2) describe the pitfalls to be avoided when collecting serial measurements for longitudinal analysis; and 3) recommend simple criteria to use for flagging abnormal change in pulmonary function over time. The statement's key points are summarized in Table 1. Real-life examples illustrate the pitfalls to be avoided and the application of longitudinal methods for evaluating pulmonary function.

Table 1: Evaluating Change over Time Key Points   Why Look at Change Over Time? Traditionally, lung function has been compared with average (i.e., predicted) values measured in asymptomatic nonsmokers similar to the worker Change over time compares each worker with him/herself and probably detects earlier lung function loss, especially if the worker's baseline is above average Consistent with OSHA intent of evaluating current and previous results How to Evaluate? Longitudinal Normal Limit ( LNL ) is based upon baseline results for a specific worker Compare the worker's Current results to his/her LNL Is BASELINE result above average (>100% Predicted)? - YES --> Method 1* for simple calculation analyzing % Predicted values - NO --> Method 2* for calculations analyzing actual (measured) values   Cautions - To Interpret Serial Tests: Spirometry must be high quality Timing of tests must be consistent Technician training and quality assurance program essential Results interpreted by well qualified health professionals * Note: If a screening program wants to adopt only one method to compute LNLs for all workers, ACOEM recommends choosing Method 2 .

 

Why Examine Change Over Time?

Spirometry in Medical Screening and Surveillance Programs
Spirometry is performed periodically in screening and surveillance programs for a variety of occupational exposures. Occupational Safety and Health Administration (OSHA) regulations require periodic spirometry testing for certain workers exposed to asbestos,(5) coke oven emissions,(6) cadmium,(7) cotton dust,(8) benzene,(9) and formaldehyde.(10) Many companies mandate medical surveillance with periodic spirometry testing for additional exposures as well as incorporating spirometry into their respirator medical clearance screening programs. In each case, health professionals must evaluate current and previous test results to determine whether an employee is at increased risk of impairment from further occupational exposure, or if any limitations should be placed on the employee's activities or use of personal protective equipment. However, the details of the evaluation of current and previous results are usually not specified. The OSHA Cotton Dust Standard is only slightly more explicit than the other OSHA regulations listed above, stating that "a determination [shall be] made by the physician as to whether there has been a significant change [between the current examination results and those of previous examinations]."(8)

Comparing Observed with Cross-sectional Predicted Values
Traditionally, an individual's measured lung function has been compared with a "predicted" value, i.e., the average expected for an asymptomatic non-smoker of the subject's age, height, race/ethnicity, and sex. Many sources of predicted values have been derived from studies of asymptomatic non-smoking populations, and some applications require the use of specific sets of prediction equations. Selection of reference values has been reviewed elsewhere.(1,11-15) The comparison of observed with predicted values is usually summarized in a numerical index, the percent of predicted (% Pred). Lower Limits of Normal should be determined for the prediction equations in use, and the individual's measured results are then interpreted relative to the normal range as normal or abnormal, and possibly impaired. Since Lower Limits of Normal generally decrease with age, the use of 80% Pred as a Lower Limit of Normal for all age groups is no longer recommended.(11) Definitions of the Lower Limit of Normal, choice of measurements for evaluation, and definitions of airways obstruction have evolved over time,(1,11) but the approach of comparing an individual with the average "predicted" from an asymptomatic population has been widely used for decades.

Need for Longitudinal Lung Function Evaluation in Occupational Settings
In the clinical setting, patients with lung disease are often tested to determine the severity of their disease.(3,16) In contrast, in the occupational setting, many healthy workers are tested periodically, not because they have abnormal lung function, but to monitor their response to potentially harmful occupational exposures. Because of their health, working populations usually have higher levels of pulmonary function than clinic populations, and many workers have lung function that is above average, i.e., > 100% Pred. Such individuals may lose their lung function at an excessive rate but still remain in the normal range throughout their working lifetime and into retirement. Remaining in the normal range does not indicate respiratory health, since their function may drop from the top to the bottom of the normal range, but these individuals must lose large fractions of their lung function before they will fall below the normal range. For these workers, the widespread practice of repeatedly comparing serial test results with the traditional normal range may not detect serious pulmonary function deterioration. Longitudinal evaluation that compares current measured values with previously measured values, "using the subject as his/her own control," is needed especially for this group, as summarized in Table 2. (11,17,18)

Table 2: Why Examine Change Over Time? OSHA- and industry-mandated medical surveillance programs require health professionals to assess respiratory health using previous and current examination results. Traditional evaluation of pulmonary function determines whether test results are in the normal range, which is based on asymptomatic non-smokers. Unlike clinic patients, many workers have above average lung function, i.e., >100% Pred. Such lung function can deteriorate dramatically, falling from the top to the bottom of the normal range, without dropping below the normal range. This loss of function will not be detected by simply determining whether each year's test results fall within the traditional normal range. Health professionals must determine whether monitoring change over time is an effective screening test for the outcome disease of interest.

 

Evaluating Longitudinal Change to Screen for Specific Diseases
While this statement provides a method for evaluating change over time, health professionals must decide whether screening for excessive loss of function is appropriate for specific outcome diseases of interest. Monitoring pulmonary function longitudinally may be more justified for some exposures, e.g., smoking-related chronic obstructive pulmonary disease (COPD), than for others.(2-4) The sensitivity, specificity, and positive and negative predictive values of excessive loss of pulmonary function relative to the outcome disease of interest should be investigated. Screening for excessive loss of function is recommended if the prevalence and severity of the outcome disease are significant and if the effectiveness of the intervention or treatment balances the financial and non-financial costs of the intervention. (19)

Pitfalls in Collecting Serial Measurements

Although "using the subject as own his/her own control" may detect pulmonary function declines that are missed by comparisons with predicted values, practitioners who analyze longitudinal spirometry data are often unaware of the pitfalls that can invalidate their conclusions. Since both technical and biological factors affect spirometry results at each test session, practitioners should attempt to hold these factors constant if longitudinal analysis is anticipated. (20-25) Failure to control these factors produces extraneous variability which may be interpreted as an excessive loss or gain of lung function. Therefore, users of spirometry data should appreciate the effects of technical and biological factors on measurements and be prepared to evaluate test quality and reject inadequate tests before evaluating change over time. Sources of technical and biological variability are summarized in Table 3.

Table 3: Pitfalls in Collecting Serial Measurements   Standardize and Document the Testing Protocol, Equipment Used, and All Changes in Protocol or Equipment. Technician Training and Periodic QA Audits of Spirograms Equipment - Minimize unnecessary equipment changes - Minimize changes in spirometer configuration - Insure spirometer accuracy Laboratory testing of spirometer submitted by manufacturer Calibration or calibration checks at least daily when in use On-going scrutiny of spirograms and patterns of test results - Retain calibration records indefinitely Biological variability - Standardize time of day and season of testing to evaluate long-term change - Postpone testing for 3 days if subject feels ill, for 3 weeks after severe respiratory or ear infection, for 1 hr after smoking, use of bronchodilator, or a heavy meal, and until medically approved after surgery.

 

Technical Variability

Standardization and Documentation of Testing Technique and Equipment
Spirometry testing procedures, type of spirometer, and spirometer maintenance and quality assurance (QA) checks should be standardized across location and time, particularly if longitudinal analysis of lung function measurements is anticipated. Current American Thoracic Society (ATS) recommendations were summarized in the recent ACOEM spirometry statement. (1,26) Testing procedures and equipment used should be fully documented, and the documentation should be updated whenever changes occur. Standardization and documentation are particularly important if testing is contracted out to multiple vendors over time. In fact, frequent changes in vendor and/or poor vendor quality control of testing may preclude any meaningful longitudinal evaluation of results. Equipment malfunction and errors in testing technique can cause measurements to be falsely elevated or reduced.(1, 26-32) Some technical errors cause increased variability that is random, though many problems cause results to be biased. When a series of erroneous measurements is examined, healthy workers may appear to "decline," while others' deteriorating function may be masked by the noise in the measurements. Figures 1and 2, discussed below, illustrate the difficulty in evaluating change over time using technically flawed measurements. A summary of technical errors that raise and lower test results is available on the Internet. (31)

 

 

 

 

 

 

 

 

 

 

 

Testing Technique
Health professionals should develop a written testing protocol and insure that technicians understand and follow the specified procedures. The many details involved in conducting tests and maintaining equipment may be easily misunderstood, resulting in non-standardized testing procedures. The details of the testing procedures should be spelled out in the written protocol, e.g., the definition of end-of-test [recording to a forced vital capacity (FVC) plateau vs. recording for a specified number of seconds], testing posture [standing vs. sitting], minimum number of acceptable maneuvers to be recorded, criteria for rejecting a maneuver, i.e., what makes a maneuver "unacceptable," and whether to print out curves during a test for coaching if there is no real-time graphical display. Changes in the testing procedures over time should be documented. Figure 1 illustrates the effect of poor coaching which elicited only sub-maximal inspirations from an employee who had recovered from pleurisy but appeared not to have returned to his baseline level of pulmonary function. When an experienced technician urged the employee to inhale maximally, his FVC and forced expiratory volume in one second (FEV₁) results increased by 0.8 and 0.5 L, respectively, returning to their baseline levels. The variability introduced by inconsistent testing technique, such as that shown in Figure 1, probably precludes meaningful evaluation of change over time.(18)

Spirometry training courses such as those approved by the National Institute of Occupational Safety and Health (NIOSH) are recommended, and NIOSH is developing a course-approval web page and reorganizing its program to insure better standardization among courses.(1, 26) A single vendor should provide training for all technicians at a location, if feasible, and training should be followed by supervised on-the-job testing experience (26) and QA review of spirograms for technical quality.(26, 33, 34) Periodic Refresher courses are recommended (1) and QA reviews of spirograms should be continued indefinitely, perhaps conducted at least on a quarterly basis.

Equipment
When longitudinal evaluation is anticipated, equipment variability should be minimized across locations and time. Variability may be increased if different spirometers are used, if calibrations or calibration checks are not performed correctly and consistently, or if spirometer temperatures vary widely.(35) Recommendations to minimize equipment variability are presented below.

1. Minimize unnecessary equipment changes
   Unnecessary equipment changes should be avoided if longitudinal analysis of results is anticipated, though excessively variable spirometers should be replaced by instruments with greater precision. The ATS recommends that spirometers should be accurate to within +/-3% of the volume introduced into a spirometer, so a spirometer meets minimum criteria for accuracy if it records between 2.91-3.09 liters when a 3.00 liter volume is introduced. But since variability exists both within and between spirometers, a 3-liter subject could record 3.09 liters on one spirometer and 2.91 liters on a different spirometer, even though both spirometers met minimum accuracy requirements.(18) Some variation between spirometers may be due to their different mechanisms for determining volume or their use of variable disposable sensors. Some flow-type spirometers measure slightly different volumes when air passes through the sensor at different speeds, while volume-type spirometers are less affected by the speed of air entering the spirometer. Some spirometer sensors may also be subject to changes in calibration over time. Table 4 gives one example of varying volumes recorded by a flow-type spirometer when a 3-liter syringe was injected at different speeds during a calibration check, as described below. Though all of the values are within the acceptable range of 2.91-3.09 L, this spirometer clearly records lower volumes when airflow is slower.

Table 4: 3 L Injected into a Flow-Type Spirometer at Various Speeds
Injection Speed (L/s)	Recorded Volume (L) *
0.52 0.98 3.33 5.45	2.94 2.98 3.01 3.08
* Recorded volumes from 2.91-3.09 meet minimum standards of accuracy.

 

2. Minimize changes in spirometer configuration
   Most spirometers permit users to customize various aspects of data-saving and reporting during testing. Often, there is a choice of how many maneuvers' results should be saved and reported [users should choose "all data" and "all curves", which spirometry measurements to save and report [users should choose FEV₁, FVC, or forced expiratory volume in six seconds (FEV₆), FEV₁/FVC or FEV₁/FEV₆, PEF and forced expiratory time (FET) if available, unless other requirements apply], and which values are selected from the maneuvers attempted [users should choose maximum FEV₁, maximum FVC or FEV₆, maximum PEF, and not the "best curve" FEV₁ and FVC values].(26) It should be noted that many regulations do not permit measurement of the FEV₆ in place of the FVC. Any changes in spirometer configuration over time or across locations should be documented. Changing the spirometer's configuration may change the data that are saved and reported, which will adversely affect longitudinal analysis of lung function.

3. Spirometer accuracy
   several steps help to insure that spirometers function accurately.(1) First, the ATS recommends minimum acceptable levels of accuracy and precision for spirometers.(26) Second, an independent testing laboratory injects 24 standard waveforms into spirometers that are submitted by manufacturers for evaluation, and analyzes the spirometer responses.(26) If a spirometer passes the laboratory testing, a letter is issued stating that the spirometer completed testing following the 1994 ATS Spirometry Update protocol for evaluating diagnostic spirometers. Users should request a copy of this letter, specifically citing the 1994 ATS testing protocol, from their spirometer manufacturers.(1)

However, passing laboratory testing does not guarantee continued functioning, so the third step in insuring that the spirometer works properly is to regularly check the calibration of the spirometer before it is used for testing.(1) These checks are performed at least daily when the spirometer is in use and more frequently if many subjects are tested. Calibration checks performed at the end of the testing session confirm the status of the spirometer during the preceding tests. Calibration checks are decision-making prompts: if the spirometer fails a properly performed calibration check, the spirometer is out-of-calibration and should not be used for subject testing. Though the ATS recommends checking the calibration every four hours and whenever temperature changes occur,(26) the frequency also depends on how many tests the health professional can afford to discard and repeat if a calibrated spirometer loses its calibration.

Spirometer calibration is either set or verified during the "calibration" routine; users should consult their manufacturer to determine which procedure is performed for their spirometer. If calibration is verified, the 3 L should be injected once at a moderate speed for a volume spirometer and 3 times, at slow, medium, and fast speeds (e.g., over 1 s, 3 s, and 6 s) for flow-type spirometers.(1, 26) If calibration is set, the 3-L volume should be injected at the speed specified by the manufacturer; after the calibration is set, flow-type spirometer accuracy should be verified at 3 speeds of injection using a manufacturer-recommended protocol. For flow-type spirometers with disposable sensors, it is prudent to perform a calibration check using the sensor that the subject will use,(37) but if this is not feasible, sensors used for calibration should at least be drawn from the same batch as those used for subject testing. Technicians should avoid the incorrect practice of using one sensor for calibration checks over extended periods of time while changing the subjects' sensors.

The calibration syringe must be accurate: syringes can be calibrated annually and checked for leaks periodically by trying to empty the syringe with the outlet blocked.(26) Store the calibration syringe near the spirometer in the testing environment, and perform calibrations and calibration checks in that environment. It is unacceptable to perform calibrations or checks in a warm environment to guarantee that the spirometer passes the calibration, and then move the spirometer into a colder environment, e.g., an unheated mobile testing van, for subject testing. If the testing environment can be maintained at 23 degrees C (73 degrees F) or above, testing errors due to cold temperatures will be minimized;(35) the ATS sets a minimum spirometer temperature at 17 degrees C.(11)

Volume-type spirometers should also be checked for leaks, daily and whenever breathing hoses are changed. The current ATS acceptable leak level is 0.01 l/min, though a slightly larger leak may be tolerated under revised ATS recommendations that are under development. If a chart recorder is used, the chart speed should be checked quarterly, along with linearity of the volume measurement.(26)

Finally, attention has recently been drawn to the fact that serious problems can develop during testing even after the spirometer passes its calibration checks.(30, 38) Particularly with flow-type spirometers, problems can develop due to faulty zeroing or contamination of the sensor, causing anomalous results and spirograms with unusual shapes. Therefore, even after calibration checks indicate that a spirometer is acceptably accurate, users should evaluate visual patterns in spirograms and be watchful for unlikely patterns of elevated results during testing.(30, 31, 38) Such vigilance is particularly important when longitudinal analysis is planned since falsely elevated baselines will exaggerate the loss of function over time in many individuals, while falsely elevated follow-up test results will have the opposite effect.

4. Save calibration records indefinitely
   Calibration records support the accuracy of employee spirometry tests conducted on the calibration date, and should be saved indefinitely.(39) When contracting out to vendor(s), users should obtain and save records from all calibrations or calibration checks performed while testing is conducted at their facility. If problems with test results are discovered later, calibration records may provide the solution to the problems. Figure 2 presents 3 years of spirometry surveillance results, showing that 5 out of 6 employees experienced significant declines in their FVC in April, 2002. The mean FVC for the 6 employees declined by 1.1 L from the previous test 18 months earlier, but then returned to baseline levels on further testing 2 months later. Calibration records later revealed that the spirometer used in April, 2002 was calibrated incorrectly, causing subject volumes to be grossly under-recorded, and producing the apparent FVC declines. As with errors in testing technique ( Figure 1), such increased variability probably precludes meaningful evaluation of change over time.(18)

 

 

 

 

 

 

 

 

 

 

Biological Variability

As airway caliber changes, spirometry measurements demonstrate diurnal (within a day) and seasonal (within a year) variability, so that time of day and year should be standardized when collecting serial measurements for long-term longitudinal analysis. Though diurnal variability, in particular, gives important information when short-term changes are evaluated, e.g., due to asthma, these factors should be controlled when long-term change in function is the outcome of interest. Many medical surveillance programs conduct examinations on the employee's birthday, so that seasonal variability is controlled.

Other factors may also affect test results and should be queried before conducting a spirometry test.(32) NIOSH recommends that testing be postponed for three weeks if the subject has had a recent severe respiratory infection. The test should be postponed for one hour if the subject has had a large meal, smoked a cigarette, or used a bronchodilator within the last hour. The one hour postponement can sometimes be achieved by performing the spirometry test later in a physical examination. If it is not feasible to postpone a test, these factors should at least be documented on the report of test results.

What Is a 'Significant Change' Over Time?

Because measurement variability strongly affects estimates of change in lung function over time, the expected rate of change is not as well defined as the cross-sectional "predicted" value. Definitions of "significant change" should minimize false negatives and false positives; deteriorating lung function should be detected early enough to permit the rate of loss to be slowed and the remaining function to be preserved, but at the same time, workers should not be labeled as having "significant loss" if they are not developing impairment. Definitions of "significant change" should be simple to apply even when practitioners do not have access to sophisticated statistical programs. The current ACOEM recommendations for evaluating change over time are summarized in Table 5.

Table 5: What Change Over Time Is 'Significant'?   Quantifying Change over Time Method 1  for Baselines >100 % Pred* : Calculate Decrease in FEV1 % Pred or FVC % Pred from Baseline to Follow-up Method 2  for Baselines <= 100 % Pred* : Calculate Decrease in Measured FEV1 or FVC from Baseline to Follow-up Fit "Slope" through Periodic FEV1s or FVCs over  &gt; 4-6 Years What Change Is " Significant"? Method 1*: Follow-up FEV1% Pred or FVC % Pred falls below Longitudinal Lower Limit of Normal (LNL) = [Baseline % Pred x 0.85] (See Table 6) Method 2*: Follow-up Measured FEV1 or FVC falls below Longitudinal Lower Limit of Normal (LNL) = [0.85 x Baseline Measured Value - (Baseline Pred - Follow-up Pred)] (See Table 7) Slope Steeper than 90-100 ml/yr over 4-6 or More Years What If Change Appears to be "Significant"? Re-test to Confirm Low Value Provide Medical Evaluation, even if Test Results Remain in the Traditional Normal Range. * Note: If a screening program wants to adopt only one method to compute LNLs for all workers, ACOEM recommends choosing Method 2 .

 

Length of Follow-up and Frequency of Testing
Estimates of individual rate of change become more precise as follow-up time increases, and only large losses of function can be reliably detected over short time periods, e.g., <2 years. To estimate longer term trends in an individual's FEV₁ or FVC, spirometric measurements should be made over at least 4-6 years using standardized equipment and testing techniques.(1,20-24,40) Precision is less affected by measurement frequency than by length of follow-up,(1,20,21,39) but periodic measurements are needed to detect workers experiencing rapid declines in pulmonary function and to detect systematic differences between examinations over time.(1,21,23,40)

ACOEM recommends that spirometry should be conducted every 1-2 years when indicated because of workplace exposures, unless otherwise specified by applicable regulations or recommendations.(1) The frequency of testing may vary with age and length of exposure as in the National Fire Protection Association (NFPA) examination protocol, which recommends spirometry testing every 3 years for fire fighters under age 30, every 2 years for ages 30-39, and annually for ages 40 and above.(41)

Evaluating and Defining 'Significant Change'
Loss of FEV₁ or FVC over time can be estimated simply by evaluating the difference between measurements at two points in time, or by fitting a least squares "slope" through an individual's periodic measurements.(1) Though epidemiologic studies often use complex statistical methods, this statement focuses on two simple approaches to use when evaluating individual workers: Method 1 (for BASELINE results > 100% Pred) evaluates change in FEV1 % Pred or FVC % Pred over time; and Method 2 (for BASELINE results <= 100% Pred) evaluates change in measured FEV1 or FVC over time. Method 1 is important because it provides a simple and more sensitive definition of abnormality for employees with above average baseline lung function. However, if a medical program wishes to adopt only one method for all workers, ACOEM recommends choosing Method 2, as recommended in the previous ACOEM Statement.(1)

Method 1. for BASELINES >100% Pred: Evaluate Change in % Pred
Method 1 provides a simple Longitudinal Lower Limit of Normal (LNL) for FEV1% Pred and FVC% Pred for individuals whose Baseline results exceed 100% Pred. The LNL should identify workers with accelerated lung function decline even though they remain in the traditional normal range. An employee is expected to remain above the LNL as he/she ages. Using the current estimate of 15% year-to-year measurement variability,(11) the Baseline % Pred is multiplied by 0.85 to obtain the LNL. Note that the same set of reference values (prediction equations) must be used for baseline and all follow-up tests.

Method 1: Longitudinal Lower Limit of Normal (LNL) for
Follow-up FEV₁ (or FVC) % Predicted = [0.85 x Baseline % Predicted)].

Table 6 and Figure 3 present FVC results for a 66-inch tall Caucasian woman, tested periodically from age 30-50 years. Her baseline FVC was 4.39 L, or 109% Pred, based on the National Health and Nutrition Examination Survey (NHANES) prediction equations.(12) At age 50, her FVC was 84% Pred. When each test was simply compared with the traditional normal range, all of her measured FVCs were above the traditional lower limit of normal and she appeared to be "normal."

However, evaluating her results relative to her own baseline value leads to a different conclusion. For her baseline FVC of 109% Pred, the LNL is [0.85 x 109%] = 93% Pred. As shown in Table 6, each of her tests remained above 93% Pred until age 50, when her result fell below the LNL. If this low value is confirmed by a re-test, she should be medically evaluated, even though her results remain within the traditional normal range. (17,42)

This example illustrates the insensitivity of repeatedly comparing periodic test results with the traditional normal range, particularly for employees with above average levels of pulmonary function. When baseline values exceed 100% Pred, lung function must decline dramatically before test results will fall below the traditional normal range. However, longitudinal evaluation using a LNL will be more sensitive to possible accelerated lung function decline.

 

Table 6: Is This a Significant Decline in % Predicted? A 66-inch tall Caucasian woman was tested periodically from age 30 - 50 years (Figure 3). Is her FVC loss "significant" at age 50? Since Baseline FVC > 100% Pred , determine a Longitudinal Lower Limit of Normal ( LNL ) for % Pred. The subject should remain above this LNL as she ages. Age FVC (L) FVC % Pred Lower Limit of Traditional Normal Range 30 35 40 45 50 4.39 4.22 3.82 4.12 3.17 109 106 97 106 84 3.29 3.26 3.21 3.14 3.05   Use Method 1 to calculate the LNL for Follow-up % Pred: Longitudinal Lower Limit of Normal (LNL) = Baseline % Pred x 0.85 = 109% x 0.85 = 93% Pred.   FVC remains > 93% Pred until age 50, when it falls to 84% Pred. If retest confirms this low value, medical review is recommended, even though her FVC remains in the traditional normal range.

 

 

 

 

 

 

 

 

 

 

 

 

 

Method 2. for BASELINES <= 100% Pred: Evaluate Change in Measured Values
ACOEM recommends Method 2 to calculate a Longitudinal Lower Limit of Normal (LNL) particularly for employees with Baseline results <= 100% Pred.(43) However, some medical programs may want to adopt only one LNL method for all workers. In that case, Method 2 can also be applied to workers with Baselines > 100% Pred, since both methods give the same results for this group. If only one method will be used for all workers, ACOEM recommends choosing Method 2 to compute the LNL.(1)

A "significant" decline should exceed both: 1) year-to-year measurement variability, currently estimated at 15%; (11) and 2) the expected age-related decline, which can be calculated as the difference between the baseline and follow-up predicted values.(1, 39, 42) These factors are used below to determine the LNL for follow-up test results. An employee is expected to remain above the LNL as he/she ages. Note that the same set of reference values (prediction equations) must be used for the baseline and all follow-up tests.

Method 2: Longitudinal Lower Limit of Normal (LNL)
for Follow-up Measured FEV₁ (or FVC) =
[0.85 x Baseline Measured Value - (Baseline Predicted - Follow-up Predicted)].

Table 7 and Figure 4 present FEV₁ results for a 65-inch  tall Caucasian woman, tested annually from age 67¨C73 years. Her baseline FEV₁ was 2.42 L, or 97% Pred, based on the NHANES prediction equations.(12) Though a LNL was computed for each test date, only the age 69 results are evaluated here. As illustrated in Table 7, the LNL at age 69 is 2.00 L, i.e., the FEV1 could drop as low as 2.00 L at age 69, due to measurement variability and aging alone. Since the age 69 test result is below 2.00 L, the FEV1 decline may be "significant" and should be medically evaluated if the low result is confirmed by a re-test.(17,41) Figure 4 shows that the FEV₁ remained below the LNL for all subsequent tests, though it did not fall below the traditional normal range until several years later, at age 73. This subject's deteriorating lung function was identified by longitudinal evaluation 4 years earlier than it would have been detected by comparisons with the traditional normal range.

Table 7: Is This Decline Significant? A "significant" loss of FEV1 should exceed year-to-year measurement variability and expected loss due to aging. A 65-inch tall 67- year old Caucasian woman with Baseline FEV <=  100% Pred was tested annually; biennial results are shown below. Longitudinal Lower Limits of Normal (LNL) were calculated for each test ( Figure 4). Has her FEV1 declined "significantly" by age 69? Age FEV1 (L) Pred FEV1 FEV1% Pred Longitudinal Normal Limit (LNL) 67 69 71 73 2.42 1.93 1.82 1.61 2.49 2.43 2.37 2.30 97 79 77 70 2.06 2.00 1.94 1.87   Using Method 2, allow 15% measurement variability: [0.85 x Baseline FEV1 ] = 0.85 x 2.42 L = 2.06 L Calculate expected aging effect: [Baseline Pred FEV1 - Pred FEV1 at Age 69] = 2.49 L ¨C 2.43 L = 0.06 L Calculate Longitudinal Normal Limit (LNL) for follow-up FEV1 = [0.85 x Baseline FEV1 - Expected Aging Effect] = 2.06 - 0.06 = 2.00 L   Her FEV1 at age 69 should be >=; 2.00 L (LNL). Since her FEV1 (1.93 L) is < LNL, she should be retested to confirm this result. If confirmed, medical review is recommended, even though the FEV1 remains in the traditional normal range until age 73, as shown in Figure 4.

 

 

 

 

 

 

 

 

 

 

 

 

Initial Identification vs. Progression
Once a worker is identified as having impaired lung function, ATS recommends a less conservative definition for evaluating progression of disease, since both the measured volumes and the percents of predicted are smaller than for the healthy individuals discussed above. In the statement on "Idiopathic Pulmonary Fibrosis: Diagnosis and Treatment," ATS and the European Respiratory Society (ERS) recommend interpreting a loss of 10% or more of the measured baseline VC (or at least 0.20 L) as a "failure to respond to therapy," i.e., a significant decline, if the change is accompanied by parallel changes in single-breath diffusing capacity or oxygen saturation 6 months after the baseline test.(44)

In addition, an increase from the measured baseline VC of 10% or more (or at least 0.20 L) is interpreted as a significant improvement if the change is accompanied by parallel changes in single-breath diffusing capacity or oxygen saturation and is maintained for two consecutive visits within a 3-6 month period.

Changes smaller than +/- 10 % of measured baseline VC (or < 0.20 L) maintained for two consecutive visits within a 3-6 month period indicate stable pulmonary function.(44)

Fitting a Least Squares 'Slope' through Periodic Measurements
Calculating a best-fit line of lung function measurements on test date requires more computational capability than calculating differences, but can be programmed or computed on a calculator. Based on reviews of the longitudinal spirometry literature, the previous ACOEM spirometry statement recommended that an FEV₁ or FVC decrease of 90-100 ml/ year, calculated over at least 4-6 years, should trigger further medical evaluation of pulmonary function.(1,21,24,45) Though this area remains one of current investigation, neither longitudinal predicted values nor 5th percentile LLNs have yet been recommended for the evaluation of individual rates of change over time in occupational or clinical settings.(1,21,46)

Conclusion

As summarized in Table 1, longitudinal evaluation of pulmonary function should be considered, particularly in the occupational setting, because many workers have above average levels of pulmonary function (i.e., >100% Pred). Such high levels of lung function can deteriorate substantially, falling from the top to the bottom of the normal range, without dropping below the normal range. This loss of function may not be detected by the common practice of simply determining whether each year's results fall within the traditional normal range.

To address this problem, ACOEM recommends simple methods for comparing an employee's periodic spirometry results with a Longitudinal Lower Limit of Normal (LNL) specific for that employee. Starting with an individual's baseline lung function level, the LNL describes the lowest results that might be expected for his/her lung function during follow-up, due to normal aging and measurement variability. Test results falling below the LNL may indicate significant deterioration of pulmonary function. However, to make such evaluations possible, spirometry data must be collected carefully, following standardized protocols. The rate of false positives will be high if test variability is not minimized through QA protocols, standardized testing procedures, and the continuity of well-maintained equipment.

Based on current recommendations, ACOEM recommends two methods to compute a worker's LNL. Method 1 (for employees with BASELINE results > 100% Pred) computes a simple LNL for the follow-up FEV1 % Pred or FVC % Pred using [Baseline % Pred x 0.85]. Each serial test can be compared to the LNL to determine whether the worker's pulmonary function has deteriorated significantly relative to his/her own baseline result. This approach is shown in detail in Table 6and Figure 3.

Method 2 (for employees with BASELINE results <= 100% Pred) computes a LNL for the measured FEV₁ or FVC using [0.85 x Baseline Measured Value - (Baseline Predicted - Follow-up Predicted)]. Each serial test can be compared to the LNL to determine whether the worker's pulmonary function has deteriorated significantly relative to his/her measured baseline value. This approach is shown in detail in Table 7and Figure 4. (In addition, if a medical program wishes to adopt only one method for all workers, ACOEM recommends choosing Method 2 to calculate the LNL. Both methods give the same results if a worker's Baseline exceeds 100% Pred, but only Method 2 should be used for those with lower Baseline values.)

If a test result falls below the longitudinal LLN calculated using either method, it should be confirmed by a re-test. Once confirmed, medical evaluation is recommended, even if the test results remain in the traditional normal range.

Finally, if multiple measurements are available over 4-6 or more years, a slope of lung function measurements over time can be calculated. ACOEM recommends that slopes that are steeper than 90-100 ml/yr should be flagged as significant losses of function, even if the worker's test results remain in the normal range.

Acknowledgement: This ACOEM guideline was developed by Mary C. Townsend, DrPH , and members of the ACOEM Occupational and Environmental Lung Disorder Committee under the auspices of the Council on Scientific Affairs. The guideline was peer reviewed by the Committee and Council and by John L. Hankinson, PhD, and William L. Eschenbacher, MD. It was approved by the ACOEM Board of Directors on July 26, 2003.

References

1. American College of Occupational and Environmental Medicine. Spirometry in the occupational setting. J Occup Enivron Med. 2000;42:(3)228-245.

2. Burrows B, Knudson RJ, Camilli AE, Lyle SK, Lebowitz MD. The "horse-racing effect" and predicting decline in forced expiratory volume in one second from screening spirometry. Am Rev Respir Dis. 1987;135(4):788-93.

3. Ferguson GT, Enright PL, Buist AS, Higgins MW. Office spirometry for lung health assessment in adults: a consensus statement from the National Lung Health Education Program. Chest. 2000;117(4):1146-1161.

4. American Thoracic Society. Occupational contribution to the burden of airway disease. Am J Respir Crit Care Med. 2003;167(5):787-97.

5. U.S. Code of Federal Regulations. Title 29, Part 1910.1001, Asbestos, revised 1994.

6. U.S. Code of Federal Regulations. Title 29, Part 1910.1029, Coke Oven Emissions, 1976.

7. U.S. Code of Federal Regulations. Title 29, Part 1910.1001, Cadmium, 1992.

8. U.S. Code of Federal Regulations. Title 29, Part 1910.1043, Cotton Dust, revised 1985.

9. U.S. Code of Federal Regulations. Title 29, Part 1910.1028, Benzene, 1987.

10. U.S. Code of Federal Regulations. Title 29, Part 1910.1048, Formaldehyde, 1987.

11. American Thoracic Society. Lung function testing: selection of reference values and interpretative strategies. Am Rev Resp Dis. 1991;144(5):1202-18.

12. Hankinson JL, Odencrantz JR, Fedan KB. Spirometric reference values from a sample of the general U.S. population. Am J Resp Crit Care Med. 1999;159(1):179-87.

13. Knudson RJ, Lebowitz MD, Holberg CJ, Burrows B. Changes in the normal maximal expiratory flow-volume curve with growth and aging. Am Rev Resp Dis. 1983;127(6):725-34.

14. Crapo RO, Morris AH, Gardner RM. Reference spirometric values using techniques and equipment that meet ATS recommendations. Am Rev Resp Dis. 1981;123(6):659-64.

15. Knudson RJ, Slatin RC, Lebowitz MD, Burrows B. The maximal expiratory flow-volume curve. Normal standards, variability, and effects of age. Am Rev Resp Dis. 1976;113(5):587-600.

16. Pauwels RA, Buist AS, Calverley PM, Hurd SS. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease. NHLBI/WHO Global Initiative for Chronic Obstructive Lung Disease (GOLD) Workshop summary. Am J Respir Crit Care Med. 2001;163(5):1256-76.

17. Hankinson JL, Wagner GR. Medical screening using periodic spirometry for detection of chronic lung disease. Occup Med: State Art Rev. 1993;8(2):353-61.

18. Hankinson JL. Pulmonary function testing in the screening of workers: guidelines for instrumentation, performance, and interpretation. J Occup Med. 1986;28(10):1081-92.

19. Harber P, Lockey JE. Pulmonary function testing in pulmonary prevention. Occ Med: State Art Rev. 1991;6(1):69-79.

20. Berry G. Longitudinal observations. Their usefulness and limitations with special reference to the forced expiratory volume. Bull Physiopathol Respir (Nancy). 1974;10(5):643-56.

21. Burrows B, Lebowitz MD, Camilli AE, Knudson RJ. Longitudinal changes in forced expiratory volume in one second in adults. Methodologic considerations and findings in healthy nonsmokers. Am Rev Respir Dis. 1986;133(6):974-80.

22. Clement J, Van de Woestijne KP. Rapidly decreasing forced expiratory volume in one second or vital capacity and development of chronic airflow obstruction. Am Rev Respir Dis. 1982;125(5):553-8.

23. Vollmer WM. Reconciling cross-sectional with longitudinal observations on annual decline. Occup Med: State Art Rev. 1993;8(2):339-51.

24. Buist AS, Vollmer WM. The use of lung function tests in identifying factors that affect lung growth and aging. Stat Med. 1988;7(1-2):11-8.

25. Becklake MR, White N. Sources of variation in spirometric measurements. Identifying the signal and dealing with noise. Occup Med. 1993;8(2):241-64.

26. American Thoracic Society. Standardization of spirometry, 1994 update. Am J Resp Crit Care Med. 1995;152(3):1107-36.

27. Graham WG, O'Grady RV, Dubue B. Pulmonary function loss in Vermont granite workers. A long-term follow-up and critical reappraisal. Am Rev Respir Dis. 1981;123(1):25-8.

28. Townsend MC. The effects of leaks in spirometers on measurements of pulmonary function. The implications for epidemiologic studies. J Occup Med. 1984;26(11):835-41.

29. Townsend MC, Morgan J, Durkin D, DuChene AG, Lamb S. Quality control aspects of pulmonary function testing in the Multiple Risk Factor Intervention Trial. Control Clin Trials. 1986;7(3 Suppl):179S-92S.

30. Townsend MC, Hankinson JL, Lindesmith L, Slivka WA, Ayres GT. Is my lung function really that good? Spirometer problems that inflate test results (Abstract). Am J Respir Crit Care Med. 2002;165:A200.

31. Townsend MC. Technique and Equipment Pitfalls in Spirometry Testing: Serious Threats to Your Respiratory Surveillance Program. NORA Medical Surveillance Workshop. November 8, 2001. www.cdc.gov/niosh/sbw/osh_prof/townsendhandout.html.

32. National Institute for Occupational Safety and Health, Division of Respiratory Disease Studies: NIOSH Spirometry Training Guide, 1997.

33. Enright PL, Johnson LR, Connett JE, Voelker H, Buist AS. Spirometry in the Lung Health Study. 1. Methods and quality control. Am Rev Resp Dis. 1991;143(6):1215-23.

34. Hankinson JL, Bang KM. Acceptability and reproducibility criteria of the American Thoracic Society as observed in a sample of the general population. Am Rev Resp Dis. 1991;143(3):516-21.

35. Hankinson JL, Castellan RM, Kinsley KB, Keimig DG. Effect of spirometer temperature on measurement of FEV1 shift changes. J Occup Med. 1986;28(12):1222-5.

36. Harber P, Peterson K, Lehman A. Change in spirometer: impact on longitudinal data (Abstract). Am Rev Resp Dis. 1996;153(1):A97.

37. Hankinson JL. Spirometry Training Program CD. Valdosta, Ga: Hankinson Consulting, Inc.; 2001.

38. Townsend MC, Hankinson JL, Lindesmith L, Slivka, WA, Stiver G, Ayres GT. Is my lung function really that good? Flow-type spirometer problems that elevate test results. Chest, 2004; In Press.

39. Hankinson JL. Instrumentation for spirometry. Occup Med: State Art Rev. 1993;8(2):397-407.

40. Wang ML, Gunel E, Petsonk EL. Design strategies for longitudinal spirometry studies: study duration and measurement frequency. Am J Respir Crit Care Med. 2000;162(6):2134-8.

41. National Fire Protection Association. Medical Requirements for Fire Fighters, NFPA 1582, 2000 edition. National Fire Protection Association, Quincy, MA, 2000.

42. USDHHS, PHS, CDC, NIOSH. Criteria for a Recommended Standard: Occupational Exposure to Respirable Coal Mine Dust. September 1995.

43. Townsend MC. Personal Communication. 2003.

44. American Thoracic Society/European Respiratory Society. Idiopathic pulmonary fibrosis: diagnosis and treatment. International Consensus Statement. Am J Respir Crit Care Med. 2000;161(2):646-64.

45. Camilli AE, Burrows B, Knudson RJ, Lyle SK, Lebowitz MD. Longitudinal changes in forced expiratory volume in one second in adults. Effects of smoking and smoking cessation. Am Rev Respir Dis. 1987;135(4):794-9.

46. Wang ML, Petsonk EL. Repeated measures of FEV1 over 6 to 12 months: what change is abnormal? (Abstract). Am J Respir Crit Care Med. 2003;167:A839