Document rBMppk4VgJ1qVbXy0bpxYmR57

American Thoracic Society MEDICAL SECTION OF THE AMERICAN W NG ASSOCIATION 4 LUNG FUNCTION TESTING: SELECTION OF REFERENCE VALUES AND INTERPRETATIVE STRATEGIES %SI OFFICW STATEMENTOF THE AMERICAN "~ORACICSOCIETYWASAD~FTEBDY THEATS BOARD OF DIRECTORS, MARCH 1991. clinicalcase management, they have become a part of routine health examinations in respiratory, occupational, and sports medicine Although the statement deals primarily adults, the conceptual issues apply to cMdren as well. Terms and abbreviations follow and in public health screening. It is common the American College of Chest Physicians Contents practice for the results of lung function tests (ACCP)-ATSjoint committeeon pulmonary to be interpretedin relation to referencevalues, nomenclaturerecommendations(15). The Introduction and in terms of whether or not they are con- four sectionsdeal with conceptual issuesand Background sidered to bewithin the "normal" range (1-6). their scientific basis; the last section dealswi& Focus A wide selectionof published referencevalues p d c a l considerationsand r e c o m m e ~ of variation in Lung Function Testing and "lower limits of normal" is available (4). Conceptual issues pertinent to the interpretation of lung function tests Technical sources of variation Procedural sources of variation Computerized equipment adds a new dimension with preselected or menus of reference values and interpretation algorithms whose origin and justification may be unclear. Sources of Variation in Lung Function Testing Conceptual Issues Pertinent to the I Biologic sources of variation To maximize the clinicalvalue of lung func- Interpretation of Lung Statistical Considerations in the tion tests and to assist those managing clini- Function Tests Derivation of Prediction Equations tal lung function testing laboratories, the General comments American Thoracic Society (ATS)(7-12). the Characterizing the distribution and European Community for Coal and Steel determinants of lung function in (ECCS) (4). and the European Society for reterence populations ClinicalRespiratoryPhysiology (13) havepub- Evaluarmg prediction equations lished guidelines, focusing primarily on spi- Distributions and "lower limits Of normal" rometry as the most widely used lung func- Uses* and Selection Of Reference tion test. n e 1987'A statement in spirom- Values etry (8) outlined the stepsneceSSary to achieve General comments standardization: (I)equipment performance, Sources of reference equations validation, and quality control; (2) subject Determination of the "normal range" performance; (3) measurementproceduresto Smoking as an independent determine acceptability and reproducibility; ease can be properly interpreted. So- of Cross-sectional and longitudinal (4)reference values and interpretation. The technical and biologic variation and theesti- predictions fmt threehave been addressedin o f f i d state- mated magnitude of their effects are listed ~ for Of reference ments or position papers of the A S (7-12). in tables 1 and 2. Published reference equations This statement addresses the fourth. Interpretation of pulmonary function tests Limitationsof currently availableequations depends upon establishing :he variation of Interpretative Strategies Focus interest (the signal)and its relation to all oth- Conceptual issues concerning normality and the limits of normal Obstructive and restrictive ventilatory defects Bronchodilator response Interpretation of lung function in clinical practice Recommendations Overall Selecting reference values Recommendations for interpretation The chargeby the ATS was to prepare a comprehensive and practical document dea!ing with conceptual issues and their scientificbasis and providing guidelines for daily use in two areas: (I) selecting reference values and (2)interpretativestrategies. The statementwas to address the concerns of those who generate lung function reports and those who use lung function reports to assist in clinical case management. Epidemiologic and public health issues are not addressed though epi- er sourcesof variation (the noise) (5). Which sources of variation constitute signd and which noise will depend on the question be- ing asked. For instance,in a physiologicstudy of the effects of posture on FEV1,variation , caused by posture would constitute the sip ; nal and all other sourcesof within-individual variation, the noise. Similarly, in an epidedologic study of the effects of an occupation- . al exposureOR a work force,variation c a d I' by exposurewill constitute the signal, and Introduction demiologic studies provide the scientific ba- other sources of between population h a sis for many of the concepts used in inter- tion, the noise. In the clinical context, sisnd Background preting lung function results. The ATS has and noise will vary according to the c r d d During the last 3decades lung function tests published standardization procedures for question. For instance,when assessingthe out- have evolved from tools for physiologicstudy epidemiologic studies (14). The focus of this come of a treatment, the signal would be the to clinical tools widely used in assessing re- statement is spirometry, but referenceis made change after treatment, and the noise would spiratory status. In addition to their use in to other lung function tests when pertinent. be within-individual variation in the absence 1202 AM REV RESPlR DIS 1991; 1 4 4 : 1 ~ - 1 ~ ' &N 'THORACIC -SOCIETY 1203 TABLE 1 SOURCES OF VARIATION IN LUNG FUNCTION. Source Determinants Technical fBiDbgic Within individual &tween individual ' Between population Instrument, subject, posture, observer, procedure (including number of tests), software; temperature; altitude All of the above Diurnal (circadian) and seasonal effects, endocrinologic effects All of the above Personal factors, including size. age, sex, physical activity, muscularity, race, and other genetic characteristics and past and present health Environmental factors, including tobacco smoke (personal and environmental), occupation, residence (urban or rural). air pdlution (home, environmental), and socioeconomic status All of the above Selection factors which determine inclusion or exclusion of certain subjects from study populations * 8.ad on table in reference5 and reproducedwith permisswon. of treatment. When lung function tests are used as an aid in diagnosis, the signal is usu- If d l y the patient's results compared with the apccted result for subjects without disease ! britsimilarin the personal characteristicsthat danminelung function such assex, size, age, a d , possibly, race (table 1). Rchnical Sources of Variation INSTRUMENTATION - Detection of instrument problems is an inteaJ part of interpretation. Readers should eorwlt ATS recommendations on spirometrycmd k o ,which give practical limits of waptable instrument variability (7-9). In- CISION AND ACCURACY ering the variability of a test, a disust be made between precision and if the values ob- 1)no2 easy to establish, refers to how close themeasurements made by an instrument are tothek e " value. Becausemost instruments bbetter precision than accuracy,between- hQtuncmt variation usually contributesmore hbtalmeasurementvariability than withinPptrament variation. COMPUTESROFTWARE AND HARDWARE m,the use of computers in spirometry ryJtems has reduced technical variability; rwvathelesserrors associiltedwith computers Oaau.Even smalldifferedcesin the techniques -uted tocalculate flow c8n produce relatively hgc differences in derived flow measuremCnts (17, 18). It is irfiperative that spirometry systems using computers be validated ini%and each time changes aremade in soft- or hardware. One simple method of Wdatingcomputer computations is to comPIUC manual calculations of spirometric *e with computer-calculated values. The *e should be close, f 2 to 3q0, but they will not be identical.Computers canalsoprovide immediate feedback on the success of a subject's performance and improve overall test quality. Quality control algorithms that detect coughs, late peak flows, prematuretermination of effort, excessive extrapolated volumes using the back extrapolation technique, and excessivevariationbetween maneuvers can be programmed to provide immediate feedback to the technician. SPECIAL CONSIDERATIONS FOR %STING CHILDREN Equipment for testing children should have an accuracy for volume of f 50 ml to below 0.5 L. The output for the hard copy display should be scaled to the size of the signalwith a variable attenuation to a minimum of 30mm/L. There should be a visible real-time display to encourage both the child and the technician and to ensure that effort is sustained over a sufficient time. Equipment, including mouthpieces and noseclips, should be adjustable and comfortable for children with heights aslow as 120an.Childm should be tested in a laboratory where personnel are familiar with clinical testing of children and whereinterpretationscanbe made by persons familiar with pulmonar): function testing in children. Detailedrecommendationsfor podiatric testing have recently been issued by the European Society of Clinical Respiratory Physiology (13). Procedural Sources of Variation The largest single source of within-subject variability is improper performance of the test. Therefore, interpretations of spirometry should includea statement about test quality before any other interpretation is rendered. The ATS (7-12), the ECCS (4). the California Thoracic Society (2), the Intermountain Thoracic Society (19), the European Society for Clinical RespiratoryPhysiology(13). and several texts (1.3.20-23) have all recognized the importanceof procedure in reducing measurementvariability. Readers should consult thesereferencesfor detailed recommendations. TABLE 2 ESTIMATES OF THE PROPORTION OF MEASURED BETWEEN-INDIVIDUAL VARIATION IN FEV, OR FVC IN ADULTS ATTRIBUTABLE TO IDENTIFIED FACTORS' Factor Proportion of Variation Attributable SSX Age Height Weight Ethnic differences Technical Unexplainedt Total ::: 0.02 up to 0.30 up to 0.30 0.10 0.03 0.27 1.oo * Reproduced with permission from reference5. t Includes all other determinants of biologic variation dls- CusILBd in SOURCES OF VARIATION IN LUNG FUNCTIONTEST. INO whether envimnmental(e.g., smoking. actlve. and pemtve. ocarpationalexposures, maidentill pollution, tocloeconomk status) or host, (e.g., genaic, allergic, past and present rert pirahny health status). The Mer lwo are usually the focus of interestto the cilnical pulmonary function laboratory. Biologic Sources of Yariation W r r m - I m m u (~INTRAINDIWDUAL) VARIATION This section addresses short-term intraindividual variations in lung f u n c t h tFat do not originate with instrumentation and are not related to disease, environment, the intake of drugs, smoking, or failure of the subject to inspire or expire maximally during spirometric maneuvers. The main residual sourc& of variation are: (I)body position, (2)head position, (3) effort dependence of maximal flows, and ( 4 ) circadian rhythms. (I)Body position. Body position affects spirometric volumes, particularly FVC and VC. which are 7 to 8Vo lower in the supine than in the standing position and 1 to 2% lower in the sitting than in the standing position (24-27). Body position should be kept constant in comparison studies. The standing position may beparticularly advantageous for obese subjects (28). (2) Headposition. Systematic increases in maximal expiratory flows have been documented during neck hyperextension (29). These increases are believed to be related to elongation and stiffening of the trachea and range from minimalto 35% of baseline values for lung volumes above FRC (a0to 80% of VC). Corresponding changes in FEV, have not been documented. Conversely,neck flexion may decrease peak expiratory flow rate (29) and increase airway resistance (30). Avoiding hyperextension and flexion of the neck seemssufficient to eliminate this source of variability. The effect of neck position is usually less than that of body position, but it may be important for patients tested in bed. (3)u f o r t dependency of maximal flows. The imperative for standardization is one reason for the recommendations that the expiratory maneuver be performed with maximal effort. Nevertheless, FEV, may be 100 to d 1204 AMERICAN THOR*CK: socrr\lL' 200 ml lower when the effort is maximal com- or those with a skeletal deformity such as has sustainedin the past including those from pared with submaximal efforts because the kyphoscoliosis (19, 54). Lung function is de- the prenatal and immediate postnatal peri- airways are narrower with respect to the ex- creased at both extremes of weight (55, 56). ods (86-88). haled volume (31-34). Variable expiratoryef- Including measurements of chest circumfer- fort may thus be a confounding factor when ence only slightly improves the prediction of assessing small changes in maximal flows or lung function (57-60). Variations in airway BETWEEN-INDIVID(UINALTERINDMDU~) VARIATION: ENVIRONMENTAL F A C I D ~ timed volumes such as those resulting from bronchodilator response, therapy, or aging. and air-spacedimensions and geometry also contributeto interindividualvariation in lung The effectsof exposureto tobacco smoke,by far the most important environmentalfe When a flow-volumecurve is available, peak function (61,62). Accurate methods of mea- known to alter lung function, are well d m - expiratory flow may be an index of maximal suringairwayand air-space geometry are not mented elsewhere (89). In this section con. expiratory effort (31). In some subjects, widely available, and the contribution these sideration is given to other environmental&- repeated maximal efforts may trigger bron- measurementswill make to increasing predic- tors that account for between-individw chospasm, resulting in a progressivedecrease tion accuracy is unknown. After correcting differences in lung function. in FVC and FEV, (35). This may also account for body size, girls appear to have higher ex- (I) Geogmphicfactors. Altitudes as for a subject's inability to achievethe reprodu- piratory flows than do boys, whereas adult as 1,500 m do not appear to cause m w - cibility standard recommended by the ATS. men have larger volumes and flows than do able changes in lung volumes, though mq, It is of interest that failure to meet these women (6, 63, 64). surement of some flow rates may be a f f m reproducibilitystandards may itself be a mea- (2) Aging. An appropriate model for lung by changes in air density even at these altf- sure of less than perfect health (36, 37). function changes caused by aging during the tudes (90-92). FEV, and forced expirator, (4) Circadian rhythms. Variations in lung adult years includesa period after adult height flows are slightly increased at high a l t i t u k function tests with a period of approximate- is attained in which there is either an increase mainly because of the decreased density of ly 24 h are well documented (38-40). For max- (usual in young men) or little or no decrease air (93, 94). During acute exposures to imal expiratory flows, the lowest values are in function (usual in young women), after tude there may be slight reductions in VC, usually seen in the early morning (4 to 6A.M.), which the function decreasesat an accelerat- TLC, and FRC, most likely b and the largest values are seen around noon ing rate with increasing age (6, 6s) (see also creased thoracic fluid (95). Those (38). In healthy subjects, FEV, has been GROWTHsectionbelow). Theseaccelerated ag- high altitudes probably have larger shown to increaseby about 0.15 L in the mom- ing effects are typically found in longitudinal volumes than do residents at low &ti- ing and decrease by 0.05 L in the afternoon studies and not in studies based on cross- The reasons are unclear because of the e- (39); for peak expiratory flow rate (PEFR), sectional data The differencesbetween cross- founding effects of variables such the peak-to-trough amplitude is on the order sectional and longitudinal studies are ex- tion (Sa, 97). of 8% (40). Circadian variations have also plained by both statistical issues (66-69) and (2)Exposure to environmentaland been documented for airway resistance,spe- cohort effects (5, 6, 52, 55, 70). tional pollution. Exposure to airborne cific airway conductance, functional residu- (3)Race. Racehas been consistentlyshown tants such as ozone, nitrogen dioxidq -s al capacity, total lung capacity, and residual to be an important determinant of lung func- dioxide, and sulfuric acid may producemqt, volume (41-44). The mechanisms responsi- tion (20, 55, 58, 63,64,71-83). When com- surabletransientchangesin pulm ble for these diurnal variations in lung func- pared with Caucasians of European descent, tion tests in controlled human ex tion h a q n o t yet been elucidated (45, 46). values for most other races usually show Much large4-tliurnalchanges are seen in asth- smaller static and dynamic lung volumes and matic patients who often exhibit a severe lower forced expiratory flow rates but simi- "morning dip" in pulmonary function lar or higher FEV,/FVC ratios. In somepopu- creased responses. For example, short-- parameters with decreases of 50% or more lation groups diffusing capacity (transferfac- exposures (minutes) to high c o n m w in PEFR (40,41,47,48). As with healthy sub- tor) is also lower (71). Regression equations of SO, can trigger transient bronchoconstrio. jects, the largest values areusually seen around derived from white populations using stand- tion in exercising asthmatics (103). Reducqd noon, but this pattern may be substantially ing height as the measure of sizeusually over- lung function levels and an increased rated shifted by the timing of treatment (49). Ex- predict values measured in black subjects by decline in lung function have been assoCirt- aggerated circadian variations have also been about 12% for TU=,FEV,, and FVC and by ed with long-termexposuresto sulfuroxides, observed in patients with chronic bronchitis approximately7% for FRC and RV (20). Peo- inhalable particles, and photochemical ox- (50. 51). Seasonal variations of respiratory ple of mixed race usually have intermediate idants (100). function have also been recorded (49). values. These differences persist after al- Environmentalexposure to tobacco smoke lowances are made for age, stature, smoking, appears to affect the lung function of chil- BETWEEN-INDIVID(UINATLERINDMDUAL)air pollution, habitual activity, and altitude. dren (104, 105) and, possibly, adults(106108). VARIABILITY: HOST FACTORS The reason for the differences between the More recent observations also show an & f a The most important host factors responsible races is unclear. Differences may be due in on bronchial reactivity in children (109). The for interindividual variation in lung function part to differences in body build (58,63, 64, health effectsof other indoor pollutantshave are (1) sex and size, and (2) aging, which ac- 72-76). Blacks, on average, have a smaller not yet been conclusivelyestablished (110, Ill). count for approximately 30, 22, and 8%, trunk:leg ratio than do whites (77). The use Exposure to occupational pollutants, includ- respectively,of the variation in adults ( 5 ) (ta- of sitting height as an index of body size in ing dusts, chemicals, gas, etc, may induce ble 2). Other sources of interindividual vari- prediction equations reduces but does not ful- acute and chronic changes in lung function ation are (3) race and (4) past and present ly eliminatethe observed differencesbetween (112-114). health. Approximately 27qo of interindividual whitesand blacks(58,63,64,77) or the differ- (3)Socioeconomic status. Adverse effeb, variation remains unexplained (5) (table 2). ences between Europeans,Indians,and Asians. of low socioeconomic status on lung f- (0Size and sex. Size is usually measured Environmental differences, perhaps relating tion are well documented and detectable- as standing height (6, 52). Sitting height, not to nutrition, physical activity, community air in industrialized countries (85, 115,116). Lon as easy to measure as standing height, gener- pollution, and socioeconomic factors are also socioeconomicstatus is often associatedwith ally explains less of the variability (53), but thought to contribute to these differences unfavorable environmental conditions such it may be a useful predictor in certain circum- (78-85). as living in polluted urban-industrial areaS, stances (e.g., when dealing with a population (4) Past andpresent health. Lung function increased environmental and occupational of mixed ethnic origins, see below). Arm span at any one point in time reflects not only the exposures, increased indoor air pollutioa measurements provide a practical substitute present health of the individual but also the increased rates of respiratory illness, for standingheight in subjects unable to stand sum of all the insults and injuries the lung decreased accessto health care. Moreover,dd- 6 IIyE#uN THORACIC SOCIETY 1205 I feences in lung function attributed to genet- gest that changes in lung and chest recoil are 1 ic factors may be partly or even largely at- well balanced during growth. tributable to differences in socioeconomic (5)Lung volumes and Ventilatoryflows. I status (84). I GROWTH From childhood to adulthood the FEVJFVC ratio and the ratio of maximal expiratoryflow (derived from flow-volumecurves)to the FVC Growth affects the relationships between in- are almost constant. Girls generate larger ex- dices of body size and spirometric measure- piratory flows than do boys of the same age ments in children and adolescents. Some of and stature (120,127, 135, 144, 145).This is the determinants of lung volumes and ven- due in part to the fact that girls have a smaller tilatory flows are therefore briefly reviewed VC for the same TU7 than do boys, but it here may also reflect both the smaller muscle mass (I)Relationshipto height.The relationship and the smaller number of alveoli found in of ventilatory function to height from child- girls (146).Airway tone appears to decrease 1 hood through late adolescence to adulthood in girls but not in boys after a deep inspira- is not linear. Prediction equations for chil- tion (147).Finally, in children between 2and dren are usually based on power or exponen- 12 yr of age airway resistance is less in girls tialfunctions of height, both of which seem than in boys (148).These observations war- to fit the data equally well (63,64,117-120). rant using different prediction equations for (2)Age-dependence. Growth in standing boys and girls at all ages. height, measured in cross-sectionalor longitu- dinalpopulation studies, is not in phase with 1 h q growth during the adolescent growth ii Spurt(1u)-125).Growth in chest dimensions i lags behindthat of the legs (60,122,124,125). Statistical Considerations in the Derivation of Prediction Equations hboysstanding height and VC areoften not 1l b I I B ~ ~ buyI1~7 y~r of age (123).VC continues tokrcase after growth in height ceases and 1 mfilmtbe maximal until after 25 yr of age. however, seem to attain their maximal I nffresat about 16 yr of age (120,122,123). ects, FVC and FEV, seem to percentiles over time (126). I pmental rather than chronob$E Lge should be included in prediction for children and adolescents, but General Comments Reference equations provide a context for evaluating the pulmonary function values of an individual patient or subject in comparison to the distribution of measurements in a referencepopulation. The clinician's request for tests often contains the implicitquestion: Are these resultsbelow the "lower limit of nor- mal?" This section deals with statistical aspects and limitations of this concept. I tionsare not availableor practical. tory muscles. The opposing ef- Characterizing the Distribution and (1 reasing muscularity and obesity invoked to explain the observedin- Determinants of Lung Function in Reference Populations t Subjects with similar characteristics for the variables that affect lung function (sex, age, i - consequent increase in respiratory muJckforce (124,127,128). However, data on maximal inspiratory and expiratory pres- i - generated at different ages are incondusivc.NO differences were observed between pressures in adolescentsand adults @%adolescents there is evidenceof only increase in maximal respiratorypres- With growth of the lung and thorax height, race) canbegrouped together in a stratum or a cell. Comparing the performance of an individual subject with the values gener- ated from a reference population requires one to know something about the data in the ap- propriate cell, specifically: (I) the number in the cell, (2)measuresof central tendency such as the mean value, (3)estimates of dispersion such as variance or standard deviation (SD), and (4)information about the symmetry of the distribution. If the number of subjects in each cell is sufficient, lung function can be described by providing descriptors of the 0-134). The average maximal respiratory distribution such as mean and SD. Such tabu- ofboys are larger than those of girls lations are infrequently used for lung func- (l*133). Although there is a large variabili- I b'm maximalinspiratoryand expiratorypres-between individuals of the same sex, re- I +W force accounts for only a small por- tion because there are too many possible cells (consider all possible combinationsof age and height). Regressionequations are an econom- ical and efficient alternative method to de- blofthe differencesin ventilatory function scribe expected values as a function of sex, (9(1% 139. height, and age Regression techniquesassume El&c properties. From the neonatal that pulmonary function varies in a symmet- F O d to old age, the thoracic cage grows ric fashion about the mean value in each cell (136).Lung recoil increases from birth and that the variance about the mean is con- a d t h o o d and then decreases with aging stant from one cell to another. The closer the 05143).The relatively constant FRC1TI.C distribution of pulmonary function values *t (120)and the measurements of respira9 r y s t e r n mechanical properties (136)sug- comes to symmetry or, better still, to a Gaussian distribution within cells, the more it is possible to take advantage of the simplifications possible with Gaussian data. Evaluating Prediction Equations Linear regression is the most common but not the only model used to describe pulmonary function data in adults. Such equations perform less well at the edges of the data distribution and in those cells where there are few data. Estimates are likely to be misleading if they go beyond the range of the independent variables used to createthe equation. Regression analyses are often simplified by restricting the range of possiblevalues to cells (ranges of height and age) in which reasonable predictions are possible. One approach to regression analysis is to use separate simple regression equations for several different age groups (149,150).This approach may introduce conflicting estimates at the points of transition between equations. Complex equations may provide more biologicallyplausible models and reduce the average differences between observed and predicted values for every cell (cg., age and height)in comparison with simplelinear equations. The improved predictions, however, usually come at the cost of increased complexity of computation. The most commonly reported measures of how well regression equations fit *%data they describe are the square of the cotrelation coefficient (r2) and the standard error of the estimate(SEE). The proportion of variation in the observed data explained by the inde- pendent variables is measured by r'. The SEE isthe averageSDof thedata around the regression line SEEwill decrease and r' will increase as r e p i o n methods diminish the differences between predicted and observed pulmonary function values in the reference population. When the sameequations are used to describe a different population, SEE will invariably be larger, and r*will be smaller. In addition, since these statistics reflect average characteristics of the regression, r* and SEE may not reflect the ability of the equation to describe the tails of the distribution or the limits of "normal," and therefore are not sufficient criteria on which to choose the best equations to evaluate a clinical population. Dktributions and "Lower Limits of Normal" Distributions of FEVl and FVC in population studies are usually found to be close to Gaussian in the middle age range, but not at the extremes. Distributions of flow measure- ments and ratio measures (cg., FEVJFVC) are usually not symmetric(149).Transformation or age statification of the data may help produce symmetric distributions about the mean. Ideally, publications describing refer- ence populations should include not only the prediction equations but also a means of defining their lower limits. In the absence of explicit recommendations, a lower limit can be estimated from a regression model. For spirometry, values below the fifth percentile are taken as below the expected range (below 1206 the `-`lowerlimit of normal"), and those above the fifth percentile are taken as within the expected range (149,150). Percentilescan be calculated directly from the data if there are sufficient measurements within each category (56,149, 150). If individualobservations have a distribution close to Gaussian, the value of the fifth percentile can be roughly estimated as: Lower limit of normal = Predicted value -1.645 x SEE. Ideally,the SDof the residu- als should be constant for all cells. This is true for some equations for adults (149). In other studies, the estimated SD for the loga- rithm of FVC and FEV, among preadolescent children, and for height-adjusted FVC and FEV, among adults, appears to be constant for each sex and race (56). If SD is proportional to the predicted mean value, as it may sometimesbe in children (126), the fifth percentile can be estimated as a constant proportion of the predicted mean, i.e., a percent of predicted. A comparison of several prediction equations for spirometry has shown substantial agreement using the fifth percentile criterion but not using the -1.645 x SEE criterion (151). .L Sourees, Uses, and Selection of Reference Values General Comments Normal ventilatory function has come to mean the averagespirometric values of a representative sample of healthy subjects drawn from the general population. Various criteria for excludingstudy subjectshave been sug- gested based on (I) past and present medical history cq.,presence of respiratory symptoms such as cough, sputum production, and wheezing; presence of physiciandiagnosedrespiratory disease such as asthma, bronchitis, emphysema, or tuberculosis; hospitalization for lung or chest conditions; the presence of heart disease; employment exposures;and cigarette smoking); (2)physicalexamination;and (3)chest radiographic findings. The most important selection criteria are those based on a history of past diseaseand respiratorysymptoms. A reference population should, ideally, be representative of the general population from which the clienteleof the laboratory comes. Although a random sample of a population is ideal, one report found that once hospital patients were excluded, the method for selecting the study sample used to generate referencevalues had relatively little effect on either the mean value or the range of values obtained (152). Sources of Reference Equations In the 1960s. a number of referenceequations were published based on data gathered in specific population groups such as laboratory personnel, workers in a particular industry, school populations, subjects attending a specific clinic, volunteers, and general industrial workers (153-157). Some are derived from population-based data gathered in epidemiologic studies carried out for other purposes; in these studies reference equations are a byproduct (56,63. 126, 149,150). Others are based on data gathered specifically for the creation of reference equations (91, 158). Determination of the `?vormal Range" FIXED PERCENT OF PREDICTED VALUES The practice in many clinical laboratories has been to classify values of FVC and FEV, less than 80% of predicted as abnormal. This fNed value has no statistical basis in adults (91, 159-162). Although some studies have shown that for adults of average age and height, 80% of predicted FVC and EEV, is closeto the fifth percentile, use of a fwed value will result in shorter,older subjectsbeing more readily classified as "abnormal" (159, 162), whereas taller, younger adult subjects are more likely to be erroneously classified as "normal." The practice of using 80% of predicted as the lower limit of normal for FEF1,-,sr or the instantaneous flows will also cause important errors since, for these flows, the lower limits of normal are closer to 50% of predicted (149, 150). The practice of using a fiied percent of predicted as a lower limit of normal may be acceptable in children (163) (see section on DISTRIBUTIOANNSD LOWER LIMITS OF NORMAL). FEVi/FVC RATIO Defining a fiied FEVJFVC ratio as a lower limit of normal is not recommended in adults because F%V,/FVC is inversely related to age and height (91, 149, 150). The use of a fixed ratio will therefore result in an apparent increase in the prevalence of impairment associated with aging or with age-confounded factors such as cigarette smoking or occupational exposures. In addition, some athletes have values for FVC that are relatively larger than those for FEV,, resulting in a lower FEVJFVC. This may also be true of workers in somephysicallydemandingoccupations such as mining and deep-sea diving. PERCENTILES AS THE "LOWER LIMIT OF NORMAL" One statistically acceptable approach for establishing lower limits for any spirometric measure is to define the lowest 5% of the reference population as below the lower limit of normal (see section on DEXEUBUTIONANSD L ~ W E RL ~ h aOsF NORMAL). This implies a 5% false positive misclassification, a rate generally considered acceptable. Smoking as an Independent Variable Subjects who smoke cigarettes usually have lower values for spirometry and forced expiratory flows even if they meet the samehealth criteria for "normal" as nonsmokers (164). Smoking has both biologic and technical effectson DLCO(9, 165). A clear choice for the most appropriate method of adjusting spirometric indices for the effect of smoking is not readily evident from published data in which any of the following have been used: smoking status (current smoker or exsmoker), amount currently smoked, duration of smok- ing, and pack-years of smoking. N e g l m k the correlation of some of these factors (e&, pack-years) with age can introduce errors ig analyzingthe effect of smoking. In one study, ' the lifetime lossof FEV, for the averageme . smokerwas 7.4 ml/pack-year, and for them- erage female smoker it was 4.4 ml/pack-y- (164). Current smokingalso adds an acute& icit in FEV, of approximately 150ml over above the cumulativeeffect of lifetime smok- ing (164, 166). The distribution of a smoking variable in the reference population and its relation other health indicators will affect the regns. sion term calculated for smoking. For sure ple, in one study a twofold greater defiit io spirometricmeasurements in relationto pact- years was found in subjects with chronic cough compared with those without c h & cough (167). The mean spirometricvaluew not be the best index for determining l w function deficit caused by smoking since& effect on the susceptible minority tends tok overwhelmed by the unaffected majo&y (168). Whether the effects of smoking ~t similar across other independent varia& such as sex and age is unknown. Someof@ ' sex differencesin smoking-associatedpufipg. nary dysfunction may be related todiff- , in smoking behavior (169). The effect eb smoking also increases with age (las). effect of smoking on the de likely to be different from the ing on the adult lung. Finally, the effects o on pulmonary function smokers are found to have irreversible ventilatory decrements( 170).Most cross-sectionalstudiesin 01 jects have found older exsmokers to values intermediate betwee tinue to smoke and those smoked. Young exsmokers er spirometric values tha probably as a result of health selectioneffbet (134, 171). Whether the pulmonary f& of ex-smokers is better or worse than thatof current smokers probaby depends on the age of the subjects, how long they have smoked, and on why they abandoned smoking. Cross-sectional and Longitudinal Predictions Cross-sectional data are subject to a bits called "cohort" effect. A person who L 40 yr of age today is different from one w h o b came 40two decades ago because of a ty of host and environmental factors (6, The age-related lung function deficit p e from cross-sectionaldata tends to bethan that predicted from longitudinalp b nary function data in adults (67-70)and& dren (172-174). Prediction equations based on cross-sectional data are appropriate for determining the prevalence of pulmonw` function impairment in defined populations They are less well-suited to determine a(prelated events including the incidence or progression of impairment. Percentilesofad- `THORACIC SOCIETY 1207 j-te4.I lung function (similar to those used by pediatricians to assess growth) have been advocatedby several investigatorsfor assessment of both growth and decline of pulmo- nary function (56, 63, 126). A person would k apected to track along the same percendieashe or she ages if the loss (gain) in functionwas at a rate comparable to that of the refaence population. portant issues in the selection of reference values. However, neither is as important as the choice of a reference population that (I) provides an appropriate comparison for the subjects to be evaluated, and (2)is based on measurements made with instruments and methods comparable to those used in the lab- oratory for which reference values are being selected (2, 5). Criteria for Selection of Published Reference &uations Reference Values For the convenienceof readers, selected pub- Criteria for selecting reference values to be lished reference equations for adult whitesand used in the clinical or in the epidemiologic blacks and scaling factors for blacks current- context fall into three categories: methodo- ly in use are listed in tables 3 to 9. A compre- logic, epidemiologic, and statistical (5). hensive listing up to 1983 was published by (0Methodologic criteria. If possible, ref- the ECCS (4). The results of a survey of ref- mnce values should be based on data ob- erence equations used in North American pul- tained by trained operators using equipment monary teaching centersis shown in table 10. andtechniquesthat meet ATScriteria (7-12). Equations for children and adolescents are Incontrast with the use of the FVC in Ameri- detailed elsewhere (13, 63, 117-119, 131, 149, ca,predictionsof VC from Europe are usual- 176, 177). Laboratories should use the pub- ly based on inspiratory vital capacity (IVC) lished reference equations that most closely ofslow expiratory vital capacity (EVC). The describethe populations testedin their labora- IVCandEVC are, on average, somewhat larg- tories. This may also be assessed empirically atfianFVC in healthy subjects; in subjects by comparing the results for a group of 20 with airflow limitation, the differences are to 40 local reference subjects with those lDonpronounced (4, 175). provided by the intended referenceequations. 'Q)E$idernioIogic criteria.The population The local reference subjects should be ap- fnwlwhich the subjects are drawn should be propriately selected by age, ethnic group, and &&r with respect to age, height, sex, and sex, to match the clientele of the laboratory ahnic composition to the population to and should meet the selection criteria listed rturmthe predictionvaluesareto be applied. in section CRITERIA FOR SELECTION OF REF- pr#liction equations should use age, height, ERENCE VALUES. -,@, probably, ethnic group as indepenFor most clinical uses they on cross-sectionalstudies of Limitations of Currently Available &uations Referenceequations now availableincluderel- tical criteria. These are discussed atively few results for adolescents and the CAL CONSIDERATIONSIN THE DERI- elderly. Even fewer equations span the ages OF F~EDICTION EQUATIONS.Both bio- from grade school through adulthood and, plausibility and sim-plicity.in the model with few exceptions, they are discontinuous evelop prediction equations are im- for children and adults (55, 178). Older sub- jects reflect their lifetimeexperiences with respect to nutrition, health status, and other factors and are therefore subject to a cohort effect. Most equations in current use arebased on linear statistical models. All these aspects are subject to change. For this reason, reference equations should be reviewed regularly. Interpretative Strategies Conceptual Issues Concerning Normality and the Limits of Normal The word "normal" is used in a number of ways (5,6, 13, 179). In popular use it means ideal, conventional, or usual. It is used by statisticiansto describea specificdistribution about a central tendency and by biologists in ways that vary according to their focus of interest. Anatomists, for instance, use it to describe structural variations consistent with good function; physiologistsuse it to describe variationsthat preserve the "internal milieu," and cliniciansuse it to describevariation within the limits of "good health" and exclusive of "disease"(5). Issues of biologic "normality" are discussed in greater detail elsewhere, and interested readers are referred to those reviews (5, 6, 179-181). -- Becausemost laboratory tests are quantitative variables with overlap between measurements in healthy and diseased subjects, the idea of a range of values defining 6idogically "normal" is, in the view of its critics, misleading (5,6, 182). For instance, in interpreting laboratory test results where there is an overlap between healthy and diseased populations, the "normal" range should theoretically change with different disease processes and with the clinical questions being asked (181). It has also been pointed out that selecting a normal range "requires careful evaluation of benefit in terms of morbidity or mortality, inconvenience, and distress caused to TABLE 3 PREDICTED VALUES FOR FEV, AND FVC DERIVED FROM SELECTED STUDIES OF -3,. NONSMOKING CAUCASIAN MEN' --, !.!To Age Uange (v3 Number Studied F N , t for Ht 1.75 m, Age45 yr Regression Coefficient Ht Age RSD or SEE FVCt for Ht 1.75 m, Age 45 yr Regression Coeffkient Hl Age RSD or SEE 1971 (224) aunL#c, 1972 (225) Qlmla. 1977 (4) 090,lael (91) -, 1983 (149) Dodrery. 1985 (56) M-&4 15-79 21-64 15-91 25-84 25-74 517 870 189 125 86 624 3.63 3.74 3.59 3.9& 3.61 3.78 3.62 -0.032 3.59 -0.023 4.05 -0.031 4.14 -0.024 6.65 -0.029 Equation 0.55 NR 0.43 0.49 0.52 0.40 4.84 4.52 4.51 4.89* 4.64 4.72 5.83 -0.025 4.76 -0.014 6.11 -0.032 6.00 -0.021 6.44 -0.030 Equation 0.74 NR 0.56 0.64 0.64 0.47 nonlinear5 nonlinead Roa,rsse (228) -. 1986 (150) **1986 (158) 20-70 29-64 16-65 443 59 176 3.95 4.99 -0.021 0.44 3.83 4.94 -0.027 0.48 3.94 5.66 -0.023 0.41 f".l - -d dJbmviatlons: RSD residualSandard deviation; SEE -' standard errm 01 the emtimate; NR not repofled. 5.15 6.76 -0.015 0.53 5.06 7.24 -0.027 0.58 4.84 7.74 -0.021 0.51 rrrrr o b induded studies had to ( 1 ) include man and women; rpirometer (56. 91. 224); dry or wedge spirometer (2)adequately describe the methods wsd; (3)analyze spirometric values in terms of (158, 225); pneumotachograph (4. 149, 150, 2%). Equatii to predict FEV, or WC age and height. instruments using this table: of measurement - - -Predicted rrV, or FVC Predicted value? for Ht 1.75 m. Age 45 + Ht Coefficient x (HI 1.75) + Age Coefficient x (Age 45) - -:pnQlct.d value for Ht 1.75 m, Age 45. -mwtatansnitmjeot 1.mm. "pvt - - - - -= HT (1.541 4.06 x 1P'Ago 6.14 x 10- Ag#); W C Ht'(1.75 1.35 x 10- Age 1.01 x 10- Ag#) 1208 PREDICTED VALUES TABLE 4 SELECTED STUDIES OF First Author, Year (Re9 Morris, 1971 (224) Chemiack, 1972 (225) Quanjer. 1977 (4) crspo, 1981 (91) Knudson, 1883 (149) Dockery. 1985 (56) R m . 1986 (226) paa(etti, 1988 (150) Milk, 1- (158) Age Range (v3 20-84 15-79 n-64 15-84 20-87 25-74 20-70 214 4 18-82 Number Studied 471 452 514 126 204 1,830 427 313 193 Ht 1.65rn. Age 45 yr 2.72 2.87 2.71 2.92$ 2.79 2.79 2.87 2.84 2.91 Ht Age 3.50 -0.025 2.37 -0.019 3.17 -0.031 3.42 -0.028 3.09 -0.020 Equation nonlinear 3.17 -0.025 2.43 -0.020 2.68 -0.025 RSD or SEE 0.47 NR 0.35 0.33 0.39 0.40 0.31 0.29 0.33 Ht 1.65 m, Age 45yr 3.54 3.36 3.39 3.54* 3.36 3.41 3.72 3.78 3.59 Ht Age 4.53 -0.024 3.08 -0.015 4.64 -0.027 4.91 -0.022 4.27 -0.017 Equation nonlinead 4.54 -0.021 4.12 -0.015 4.14 -0.023 -srs * To be InclutMsludaa had to (1) hc*d.man and wanen; (2)ubqu- desdta the methods us&; (3) analyze spimmetric values in t e r n of age and heght. Instruments d I were' water splmneter (5691,224); dry or wedge spifOmater (158, 225t; pneurnot- (4, 140, 150, 226). Equation to predlct FEV, or FVC wrng this tabk - -Predictsd FEV, or FVC Predicted duet kc HI 1.85 m. Age 45 + lit Coefficent x (HI 185) t Age Coefficient x (Age - 45) - -tPmdicwvatueform l . m r n . ~ g o eyr. * s t a canied O(I et an . ~ ~ v d1r.400 m. - - - - - -FEV, W (1.352 4.08 x l&`AW 6.14 x 101 Ag&); FVC HP (1.463 1.36 x l(r Age 1.01 x 10- TABLE 5 .? PREDICTED VALUES FOR FEV, AND FVC DERIVED FROM SELECTED STUDIES OF Men Johemslen, 1- (227) Miiler,qm(228) oschen;llz? 1972 (81) Rosdter. 1974 (229) Lapp, 1974 (230) codgon.1976 (231) W i c k , 1976 (232}- 20-50 35-54 50.3 f 6.6 21-70 34.9 f 11.9 43.6 f 15.1 18-65 120 96 110 147 79 141 213 Hi 1.75 m Age45Yr 2.96s 3.05 2.94 3.04 3.53 3.12 3.11 2.87 -0.017 3.40 -0.024 2.99 -0.031 4.51 -0.027 3.54 -0.025 2.20 -0.024 4.23 -0.023 0.46 0.37 0.64 0.520 0.23 0.50 NR Ht 1.75 m 3.79 3.76 3.84 4.11 3.74 3.72 4.44 3.70 5.77 3.94 3.90 3.51 -0.024 -0.027 -0.019 -0.021 -0.017 -0.025 Women Johannsen, 1968 (227) Miller, 1970 (226) Cookson, 1976 (231) Patrick. 1976 (232) 20-50 *35-54 36.7 11.6 18-65 100 109 102 117 Ht 1.65 m Age 45 yr 2.25$ 2.19 2.35 2.10 2.18 -0.013 2.45 -0.018 2.40 -0.028 1.49 -0.014 0.34 0.31 0.41 NR Ht 1.65 rn Aw45yr 2.74$ 2.74 2.86 2.64 2.51 -0.015 ~a 13.15 -0.020 13.00 -0.019 4.42 3.17 -0.020 WR . * Instrumentsof ~a.sUmmenlusad were Water spirometer 327,229,231), a dry or bellows spirometer(228.no).and vanous others (el,232) Predictedvalues for men and w u n m am abrllc ed as shown in footnotes to tables 3 and 4 *t Predmed value for a e y r - o i d man 1 75 rn tall, and a 45-yroid m a n 1 ~5 m tail Corrected from ATPs to BTPS conditlons. assuming a spirometer temperature d 2 2 O C 5 includes caucasian subjects 4 subjects by further investigation and treatment, and the costs of making the wrong decision" (182). The "normal" range only gives information about the distribution of test results in the healthy population from which they were derived. It says nothing about the true positive rate, the false negative rate, or the predictive power of a positive test. To draw inferences about the presence of disease from a test, one should, ideally, know the prior probability that the patient has the disease and the distributions of test values for subjects with and without the disease in question. Although this ideal is rarely met, cliniciansmust use their understandingof the clinical situation to put an interpretation in proper perspective. Obstructive and Restrictive Ventilatory Defects DEFINITION OF AN OBSTRUCTIVE DEFECT An obstructive ventilatory defect may be defined as a disproportionate reductionof maximal airflow from the lung with respect to the maximal volume (VC) that can be displaced from the lung. It indicates airflow limitation and impliesairway narrowing during expiration. The earliest changeassociatedwith flow limitation in small airways is thought slowing in the terminal gram even when the ini gram is unaffected (1, is reflected in a proport tion in the instantaneous flow m e a s d r b - Id ter 75% oftheFVC has beenexhaled(FI3Fm) or in FEF2s-,s*than in FEV,.A b n o r m d b in these midrange flow measurementsd- a forced exhalation are,however, not s@K for smallairway disease and, though sugga- tive, should not be used to diagnose s d dr- way disease in individual patients (183). AS airway disease becomes more advancedand/ THORACIC SOCIETY 1209 First Author, Ouanjer, 1977 (4) Csspo, 1981 (91) KMldson. 1983 (149) mi19,% (150) MiHer, 1986 (158) TABLE 6 PREDICTED VALUES FOR FEV,/FVC% DERIVED FROM SELECTED STUDIES OF CAUCASIAN AND BLACK MEN AND WOMEN' Age Range 0 Number Studied FEVJFVC%t for Ht 1.75 m and Age 45 yr Regression Coefficients Ht Age RSD or SEE Number Studied FEV,/FVC%t for Ht 1.65 m and Age 45 yr Regression Coefficients Ht Age 21-64 15-91 25-85 6-64 18-65 189 125 .-86 263 176 -Caucasian Men 78.4 -80.9* -13.0 82.0 75.9 -5.3 80.5 -.13.1 -0.16 -0.15 -0.11 -0.23 -0.15 5.3 4.8 6.3 6.1 5.6 514 126 204 538 193 Caucasian Women 80.2 - 81.9t -20.2 82.6 -18.5 70.5 -4.31 82.3 -21.5 -0.24 -0.25 -0.19 -0.31 -0.15 RSD or SEE 6.4 5.3 7.6 5.8 6.8 J o h a n m , I968 (227) Oscherwitz. 1972 (81) W i e r . 1974 (229) Cookson. 1976 (231) 20-50 50.3 { * 6.6) 21-70 43.6 ( * 15.1) 120 110 147 141 Black Men 75.0 - 77.7 4.2 77.2 0.62 81.4 - -0.29 -0.32 -0.34 -0.25 8.6 10.2 7.26 10.7 102 Black Women 82.3 - -0.38 11.7 * Tlbk camprises studies cited in tables 3 to 5, which also reportedvalues for FEV,INC% analyzed in relation to height and age. For the instrumentsof measurement used, see footnotas to trbler 3 lo 5. Note: studiss of Caucasian subjects were confined to nonsmokers:studies of black subjects includedall smoking categories.Predictedvalues for FEVJNC are calculated as rham in kotnotes to tables 3 and 4. Only one study givw equations for black woman. *t PmdlUd value for a 45yr-old man 1.75 m tall, and a 45yr-OM woman 1.85 m tall. shrdas carried out at an altitude of 1,400 m. 8 ldudes Caucasian subjects. 1c&ficimt not signitkant. TABLE 7 PREDICTED VALUES FOR DIFFUSING CAPACITY (DLw) AND KCO (DL@A) -". '\ DERIVED FROM SELECTED STUDIES OF MEN AND WOMEN' FintAuRhoc. v- (Ref) Age Mean f SDor Range Number Studied OLcotfoT Htand Age Regression Coefficients Ht Age RSDW SEE DLflAtfW Htand Age Regression Coefficients Ht Age RSD or SEE );ia ; W i , 1963 (233) coles. 1965(20) Miller, 1983 (165) MI1,985 (240) Krmdson, 1987 (241) b,1990 (242) 20-75 19-72 18-67 25-79 39 f 12 17-79 20-69 15-91 43 f 16 19-64 25-84 20-70 Ht 1.75 m, AgeGyr -57 35.3 57.6 0.24 4.2 -127 30.3 32.5 0.20 5.1 -47 32.6 33.3 0.30 4.2 -70 29.3 16.4 0.20 3.8 64 33.3 28.5 -0.14 4.2 -64 29.9 10.4 0.20 3.9 -69 30.7 14.2 0.23 3.6 123 36.65 41.6 -0.22 4.0 74 31.4 16.4 -0.23 4.8 80 37.1I 44.1 -0.19 5.8 71 38.41 35.5 -0.27 4.6 194 33.6 36.7 -0.20 4.4 Ht 1.75 m. Age 45 Yr 4.96 4.83 '5.17* - -0.04 -- 0.04 -- 0.04 .5.60 -0.90 -0.03 NR 4.59 - -0.03 5.02 5.455 4.77 4.81I 5.61I ' -3.53 - -2.24 -0.12tt -2.35l-t -0.03 - 0.03 -0.03 -0.02 -0.04 Equation nonstandard1 0.92 0.81 0.73 1.07 0.85 0.83 0.84 0.73 0.71 0.80 Women mkt, 1963 (233) V M Game, 1972 (235) Salorinne, 1976 (238) w.hll, 1979 (243) 1981 (239) Miller, 1983 (165) holetti, 1985 (240) hudson. 1987 (241) 20-68 24-76 20-69 27-74 17-84 43 f 15 16-64 20-86 Ht 1.65 m, Me 45 yr 41 25.2 21.9 -0.16 3.6 72 20.3 16.8 -0.16 3.6 101 25.0 21.9 -0.12 2.8 113 30.1" 28.3 -0.19 4.1 122 27.46 25.6 -0.14 3.6 -130 23.7 16.0 -0.11 4.0 291 27.9b 15.7 0.07 4.3 99 28.21 18.7 -0.15 4.5 Ht 1.65 m, Age 45 Yr 5.55 5.61 5.27 5.66" 5.465 4.62 4.8511 5.371 - -0.17 - 3.96 - - --1.81 2.51 -2.78tt - 0.03 - 0.01 - 0.01 - 0.02 - 0.03 - 0.02 - 0.02 - 0.03 0.05 0.99 0.74 0.74 0.70 0.80 0.85 0.85 T W e &rs to 0 ~ an~d inc0ludespredictedvalues from publishedreportsinwhich the number ot subjects studied and their age were given and in which equations for DLCOwere described in I M n s of height and age according to ATS recommendations(9). All but one study (20) refer to nonsmokers h i d u a l volume or FRC was measured85 follows: singlebreath helium dilution (106,234,236-242). multiplebreath helium dilutbn (20, n3,243),open circuit N, washout (235).Prmliied values for DLW and DLNAare calculatedas shown in footnotes to table8 3 and 4. * pndldsd value for a 15-yr-oldman 1.75 m tall, and a 45-yr.old woman 1.65 m tall. R g u b 8djusted to 1 8TPS. Mmwcmentr made st M altitude of 1,400 m. 'CumcIkm for breathholding time M in the Epidemiology standardization Project (240, 241). Note that calculated DL is sensitive to the methods used to calculate breathhdd time. Form of me equation not thet recommended by the ATS. * * Results calculated for all smoking cat- and adjusted for smoking effect. tt m n t not signiticant. 1 \' A AMERICAN THORACIC First Author, Year (Re9 Men Goldman, 1959 (92) Cotes, 1965 (20) Wren. 1966 (155) Black, 1974 (244) Crapo, 1982 (245) Women Goldman, 1959 (92) Grimby, 1963 (246) Black, 1974 (244) TABLE 8 PREDICTED VALUES FOR TOTAL LUNG CAPACITY (TLC) AND RESIDUAL VOLUME (RV) DERIVED FROM SELECTED STUDIES OF MEN AND WOMEN' Age Mean Number or Range Studied TLCt for Ht and Age Regression Coefficients Ht Age RSD or SEE RVt for Ht and Age Regression Coefficients Ht Age 44 f 17 19-72 20-62 16-59 15-91 44 127 422 83 123 Ht 1.75 m, Age 45 yr 6.61 6.68 6.35 6.84 6.72 9.40 -0.015 8.67 7.80 7.80 - 7.95 0.003 0.65 0.91 0.87 0.68 0.79 Ht 1.75 m, Age 45 yr 2.04 1.62 2.15 1.87 2.70 Not reported 1.90 3.80 2.16 0.017 0.012 0.034 0.021 - ma SEE 0.39 0.a 0.67 0.37 38 f 16 18-72 16-59 50 58 110 Ht 1.65 m. Age 45 yr 5.18 5.05 5.20 7.90 -0.008 7.31 -0.016 6.40 - 7.46 -0.013 0.53 0.52 0.51 Ht 165 m, Age 45 Yr s ,1 1 1 i, or more proximal airways become involved, timed segments of the spirogram such as the FEV, will become reduced out of proportion to the reduction in VC. Bronchodilator Response ithelium, nerves, mediators, and bronchial table 11. These studies sh whether responsewas co DEFINITION OF A RESTRICTIVE DEFECT A restrictive ventilatory defect is characterized physiologically by a reduction in TLC. One may\$fer the presence of a restrictive ventilatory %efect when VC is reduced and FEVJFVC is normal or increased. Severeairflow limitation is another common cause of a reducedVC either because airflow is so slow the subject cannot continue to exhale long enough to complete emptyingor because airways collapse Occasionally,patients will have a small VC, a normal FEVJFVC, and a normal TLC. If there is a contradiction between VC and TLC in defining restriction the classification should be based on TU=. TABLE 9 FACTORS FOR ADJUSTING REFERENCE VALUES FOR CAUCASIANS WITH A VIEW TO THEIR BEING USED FOR BLACK AMERICANS' FEV, FVC FEV,IFVC TLC RV RVKLC Diffusing Capacity (transfer factor) TINA (BTPS) o.88t o.88t 0 0.88 0.93* 1.05 0.93 1.05 * Source: Rossiter and Weill with annotation (229).Allhwgh the average Caucasian admixture in studies of Black A m & cans varies. a reasonable average is 22% (247). t Also apply to women younger than 55 yr of age; in older subjects, the correction may be larger (approximately 0.80; *Dockery et a/. [!XI). A larger correction (approximately 0.88) was proposed by Lapp el d.(230). 20to 30% of responsivesubjectswill respond er than those in general population and bronchodilator responses is imperfect, and it is not possible to infer with certainty the presence of one from the other. Data on the percent changein FVC,FEV,, and FEF2s-,aa,after bronchodilatoradministration in general population studies as well as in patient populations are summarized in whether the patient is differe patientsor from previousvisits, may provide the most appropriate re data. There is no clear consensus on w stitutes reversibility in subjects with airflw obstruction(192). In part, this is because then TABLE 10 SURVEY OF SPIROMETRY REFERENCE EQUATIONS USED IN NORTH AMERICAN PULMONARY TRAINING CENTERS* FVC or VC FEV, FEV,/FVCt MF MF MF Morris et a/. (224) 65 65 65 65 58 80 Crapo et a / . (91) Knudson et a / . (149) Kory et a/. (153) Kory et a/. (249) 27 27 27 27 29 29 24 24 25 25 78 78 Cherniack et a/. (225) 3 3 4 4 Miller et a/. (180) 22 22 22 Other studied 11 11 8 8 11 9 * Basedon a questionnairesurvey of adult respiratorydisease trainingprograms in the United State0 and Canada Responses from 139 of 180 instautions are summarued (248) *t Thirty-nine centers predlcted FEV,IFVC by dividing predlcted FEV, by predicted FVC Studies clted only once zb -d I 1 9! 1 !I I' 1 I 1 eTHORACIC SOCIETY 1211 TABLE 11 RESPONSE TO BRONCHODILATOR: RESULTS FROM SELECTED POPULATION STUDIES Popldation -1- ubjects 8-75 yr of age population sample Cora Tucson, AZ (186) wbjects; random sample :#@treeareas in Alberta, (187) CY rlected normal subjects (W AgenUMode of Delivery Two inhalations of isoproterenol via metereddose inhaler 500 pg terbutaline administered via spacer Two inhalations from a Bronkometer" metered-dose inhaler FVC 10.7% (403 mi) - 5.1% (231 mi) FEV, 7.7% (315 ml) Females 9% (224 ml) Males 9% (338mi) 10.1% (365rnl) FEF25-75~ or FEF, 20% Comments 95th percentile for percentage change from baseline (absolute value In parentheses) - 48.3% 95th percentile for percentage change from baseline in asymptomatic never smokers with FEV, > 80% predicted (absolute value in parentheses) Upper 9596 confidence limits (*tailed) for percentage change from baseline 40pUients referred to p*nonarv function lab (189) W6pmtients with COPD rnplgcipating in the IPPB (190) ith airway 191) RESPONSE TO BRONCHODILATOR: RESULTS FROM SELECTED PATIENT STUDIES Placebo 250 pg isoproterenol air compressor nebulizer 200 pg salbutamol or 500 pg terbutaline via metered,dose inhaler 14.9% (340 ml) - 12.3% (178 ml) 15% 15% (330mi) 10% (160 ml) 45.1% - - Upper 95% confidence interval change after placebo inhalation. Absolute values in parentheses. Average increase as percent of initial FEV, (5% as percent of predicted normal FEV,) 95% m h d e n c e interval for absolute change; absolute rather than relative change preferred measure of bronchodllator response nsus on how a bronchodilator re- sured at some fraction of the VC may also part of health assessmenton -of a third ould be expressed. The three most be misleading in assessing bronchodilator re- party (eg., an insurance company or a gov- ethods are: percent of the initial sponseif expiratorytime changesarenot con- ernmental agency) where the clinician is not value, percent of the initial sidered and if flows are not measured at the in his or her usual patient advocacy role and line value, and absolutechange same volume below TLC. the subject or patient is, consequently, wary. he change in FEV, as a percent Current published criteria and the Work- In each of these situations, the question asked FEV, deserves further study as shop recommendations for determining bron- of the pulmonary function laboratory is quite reported to have advantages over chodilator response are given in table 12. different. Ideally, interpretations of pulmo- thods (193). When using the perge from the initial values as the most authorities would require at Interpretation of Lung Function Tests in Clinical Practice nary function tests should depend on the purpose of the tests and, when performed on patients with known disease, should be orient- Pulmonary function tests may be used to ad- ed to answering the specific question of the value as necessary to define a dress major issues in clinical case manage- clinician ordering the procedure. %ts inter- ment. These include describing dysfunction preted without clinical information will be and assessing itsseverity,kplaining it in terms limited in their clinical utility and the interpre- of diagnosis,establishingprognosis, planning tation will usually represent only a refined management, and assessingtrends over time, description of the data obtained. includingchangesafter treatment. Pulmonary The first step in interpreting a lung func- function tests may also be used to identify tion test is to evaluate the quality of the study. an abnormality in subjects without a known If there are reasons to suspect the quality of pulmonarydisorder,asin preoperativeassess- the test, avoid specificdiagnostic statements. ments, in routine health status evaluations, Dysfunction discovered under these circum- and in clinical screening. Finally, pulmonary stances should indicateonly the need for more ngle laboratory testing session. function tests are increasingly requested as definitive testing. 7s1 is a highly variable spiro- part becauseof its dependence ? m C , which increaseswith expiratorytime I *Ithobstruction. If FVC changes,postbron-tor FEF,s-7s, is not comparable with measured prebronchodilator. Volume ad- TABLE 12 RECOMMENDED CRITERIA FOR RESPONSE TO A BRONCHODILATOR IN ADULTS b e n t of FEFIS-751has been used to deal I -this issue (194, 195). At least two studies b assessed the utility of FEF2s-7s%T. he I wts were disappointing, with only 8% of Organization American College of Chest Physicians (197) FVC (46) 15-25 FEV, (%) 15-25 FEF2,,-7m (46) 15-25 Comments ~~ ~~ % of baseline in at least cs (195) and 7% of patients with two of three tests obstructive pulmonary disease (1%) identified by FEFIS-Isl criteria BS outsidethe expected range Testssuch FEVJVC ratio and flow rates mea- IntermountainThoracic Society (19) ATS (current document) 15 12 12 12 45 % of baseline % of baseline and an absolute change of 200 ml 1212 AMERICAN THORACIC PATTERNOSF DYSFUNCTION Certain patterns of physiologic abnormaii- ties can be recognized, and although they are seldom if ever pathognomonic for a specific disease entity, the types of clinical illnesses most likely to produce the observed set of physiologicdisturbances can be pointed out. Regardless of the extent of testing, the most important point with regard to pattern recognition is the need to be conservative with respect to suggesting a specificdiagnosis for the underlying disease process based only on pulmonary function abnormalities. Recognition of characteristic patterns of dysfunction depends a great deal the comprehensiveness of the lung function evaluation. However, even with only spirometric results, one can determine whether the pattern is compatible with obstruction with or without a reduction in VC. A reduced VC without evidenceof ex- piratory slowing is a nonspecific finding. There was controversy among Workshop participants about using the term "restrictive" when VC is low. The majority thought it was acceptable to interpret the finding as indicating a "restrictive type of ventilatory impairment," or a "restrictive ventilatory defect" while recognizing that it does not necessarily indicate restrictivelung disease. Othersargued the interpretation should be descriptiveonly, i.e., simply noted as "reduced vital capacity" or "nonobstructive defect," and call for fur- ther testing, includinglungvolumes,to clarify its nature. The VC, FEV,, and FEVJVC ratio are the basic parameters used to interpret spirometry. Although FVC is often used in place of VC it 13 Wrferable to use the largest VC, whether obtained on inspiration (IVC), slow expiration(EVC), or forced expiration(FVC), for clinical testing. TheFVCis usuallyreduced more than IVC or EVC in airflow obstruction. Limiting primary interpretationof spirograms to three variables avoids the problem of simultaneously examining a multitude of measurements to see if any abnormalities are present, a procedure that will lead to an inordinate number of "abnormal"tests amongthe healthiest groups in a population (198, 199). Even when the rate of abnormality for any single test is only 5%, the frequencyof at least one abnormal test was shown to be 10% in 251 healthy subjects when FEV,, FVC, and FEVJFVC ratio were examined and increased to 24% when a battery of 14 different measurements were analyzed (198). The FEV,/VC ratio is the most important measurement for distinguishing an obstructive impairment. Expiratory flow measurements other than the FEV, and FEV,/VC should be considered only after determining the presence and clinical severity of obstructive impairment using the basic values mentioned above. When FEV, and the FEV,/VC ratio are within the expected range, abnormalities in flow occurring late in the maximal expiratory flow-volume (MEFV) curve should not be graded as to severity, and, if mentioned, interpretation of their clinical sig- nificance should be guarded. In the presence of a borderline value for FEVJFVC, how- test values to independent indices of perfor- ever, they may help confirm the presence of mance such as ability to work and function airway obstruction. The same is true for av- in daily life, morbidity, and prognosis (208- erage flows such as FEFIs-,lb. Even when 212). For instance, in general. ability to Work used in this limited way, the wide variability and to function in daily life relates to ows of these tests in healthy subjectsmust be tak- pulmonary function level. FVC and/or FEv,, en into account in their interpretation. which also relate to maximal Vo2 and Work One should be cautious in interpreting ob- effort, are used in several published systems structivedysfunction when the FEV, and VC to rate impairment (208, 209). Pulmonary are both above predicted even when the function level is also associatedwith mor&- FEVJVC ratio is below the lower limit of nor- ty; those with lower function having moreR- mal since this pattern is sometimes seen in spiratory complaints(212). Lung functionb- healthy subjects, including athletes. Tests oth- el is also associated with prognosis, includ- er than spirometry, including lung volumes, ing a fatal outcome from heart as well as1- diffusing capacity, and blood gas determina- disease (213, 214) even in patients who have tions allowamplifyingstatementson the over- never smoked (215). In the Framingham study, all pattern of the dysfunction observed dur- vital capacity was a major independentp- ing spirometry. tor of cardiovascular morbidity and m o r n - LAJWER LIMITSOF "NORMAL" IN CLINICAL INTERPRETATION ty (213,214). In several occupational cohoFEV, and FEVJFVC were independent predictors of all cause or respiratory d e Lower limitsof normal areoften used in clin- mortality (216-218). In addition, a ma. ical practice without thoughtful reflection analysis of mortality in six surveysin variou, q,about their inherent variability (5, 180, 181, U.K. working populations showed that & 200-207) or their implications (5, 182). (See risk of dying of COPD was related to also sectionsDISTRIBUITOANND LOWER LIMITS level. In comparison to those whose FEV,a =OF NORMADLE, TERMP~ATOIOF TNHE NOW initial examination was within 1 SD of a=- RANGE,and CONCEPTUAISLSUES CONCERN- age, those whose FEV, was more than 2 INGNORMALITYANDTHE LIMITSOF NORMAL.) below average were 12 times more lik Although clinical interpretation is usually die of COPD, over 10 times more lik straightforward when a pulmonary function die of non-neoplastic respiratory disease, 1 (wresult is well above or below a "lower limit more than twice as likely to die of vascula of normal," this is not so when a measured disease over a 20-yr follow-up period value falls close to the "lower limit of nor- A reduced FEV, also carries a 4- to 5-fdd mal." Predicting the presence or absence of excess risk of lung cancer mortality (adjult, disease requires knowledge about the distri- ed for cigarettesmoking) (220,221). Alth+ bution of dysfunctionin various diseasestates there is good evidence that FEV, conelat$ and the prior probability of disease For ex- with the severity of symptoms and progn- ample, consider the meaning of a spiromet- in many circumstances (208, 211, 212, 21% ric study that shows FEV, values and other the correlationsdo not allow oneto accum4 expiratoryflow rates to bejust abovethe lower predict symptoms or prognosis for indi limit of normal. If the patient were a healthy a1 patients. male who sought medical assistance because In clinicalpractice,predicted values arc he was disqualified for life insurance on the used to grade severity. The severity of & basis of his spirometry, it would-be appropri- spirometric abnormality is usually based k ate to interpret his spirometry as within nor- the actual or percent predicted FEV, in tbe mal limits. If, in contrast, the same data were case of obstructive disorders or on VC m obtained from a smoker with complaints of nonobstructive disorders. An example of an intermittent coughing and occasional wheez- algorithm sometimes employed for grading ing, it would be appropriate to suggest that seventy when nothing is known about theclin- the study is consistent with mild obstructive ical question being asked is shown in table dysfunction, although it could also represent 13. It is intended only as an example and not a variant of normal. In both of these instances, as a standard. Its approach is based as much computer printouts, or robotic physician in- on clinical impression as on objective data terpretation that simplistically declare the Although clinical experience has always results to be "normal" or "abnormal" on the played a major role in assessing severity, it basis of whether the observed values fall to can be enhanced by more exact methods, and one side or the other of a single number, could physicians should probably view arbitrary give information that does not perform a use- severity scoring systems with caution. ful serviceto the patient. One suggestion for Comments on the severity or significana minimizingthe problems of overly simplistic of any abnormality depend on the circum- use of the lower limits of "normal" in the in- stances under which a test is obtained. Eror terpretation of lung function tests is use of examplethe assessmentof severityof terms such as "unusually low" rather than "ab- tion illustrated in table 13may be relevantto normal" for tests close to the lower limit of COPD, but it would not be applicable to a normal. patient with tracheal stenosiswhose obstruc- tion could be life-threatening and yet classi- ASSESSING SEVERITY fied as only mildly reduced by this scheme Severityscores are most appropriatelyderived The VC has some relationship to the U- from studies that relate pulmonary function tent of loss of functioning lung parenchyma N- TH?RAClC SOCIETY 1213 TABLE 13 EXAMPLE OF CRITERIA FOR ASSESSING THE SEVERITY OF ABNORMALITIES* A. Normal: The test is interpreted as "within normal limits" if both the VC and the FEV,NC ratio are in the normal range. 6. Obstructive abnormality: This is interpreted when the FEV,NC ratio is below the normal range. The severity of the abnormality might be graded as follows: "May be a physiological variant" "Mild" 'Moderate" 'Moderately severe" 'Severe' Very severe" . % Pred FEV, 2 100 % Pred FEV, < 100 and ;z 70 % Pred FEV, < 70 and 2 60 % Pred FEV, < 60 and I50 % Pred FEV, < 50 and 2 34 % Pred FEV, < 34 C. Restrictive abnormality: This is most reliably interpretedon the basis of TLC. If this is not available, one may interpret a reduction in the VC without a reduction of the FEV,NC ratio as a "restriction of the volume excursion of the lung." The severity of the abnormality might be graded as follows Based on the TLC 'MiW "Moderate" "Moderately severe" Based on spirometry "Mild" "Moderate" "Moderately severe" "Severe" Very severe" % Pred TLC < LLN but I70 % Pred TLC < 70 and I60 % Pred TLC < 60 % Pred VC < LLN but 2 70 % Pred VC < 70 and 2 60 % Pred VC < 60 and > 50 % Pred VC < 50 and 2 34 % Pred VC < 34 of abbreviation: LLN = lower limit of normal. Thh schema was contributed by Burrows and Lebowi. It has been in use in the lung function laboratory at the Health gc*nceeCenter in Tucson. Arlzona for clinical purposes. It is intendedonly as an axempk of a transparentschemafor assessing ruasty.other schema may be acceptableas we4 Morework is requiredbeforeany schemacan be adopted as a standard. Note: EIatOmentS W r d i n g severity should be accompanied by a disclaimer such as "as assessed by spirometrf or 'physiologic -men& of severity may differ from clinical assessments." TABLE 14 CHANGE IN SPIROMETRIC INDICES OVER TIME Percent Changes Required to be Significant FVC FEV, FEF,s-,, Within a day (222) i t i ,%$ k%! Normal subjects Patients with COPD Week to week (222) Normal subjects Patients with COPD Year to year (69) 35 3 11 3 11 320 2 15 25 3 13 > 12 220 3 15 2 13 I23 3 21 330 in many nonobstructive lung disorders. It is alsoof some use in assessing respiratory mus- ekinvolvement in certain neuromusculardis- easeS. Here again, however, the VC may be onfy slightly impaired in diffuse interstitial diseases of sufficient severity to lead to marked loss of diffusing capacity and severe blood gasabnormalities, and a relatively small decrement in VC may indicate the onset of a severe respiratory problem in patients with a rapidly progressiveneuromuscular disease. The FEVJVC ratio should not be used in isolation to determine the severity of an obshctivedisorder. Both the FEV, and VC may decline with progression of disease, and an mV,/VC of 0.V1.0 indicates more impairment than one of 2.0/4.0, though both yield a ratio of 50%. Systemsthat use FEVJFVC to grade the severity of obstruction must deal W i t h the effect of total expiratorytime on FVC and FEVJFVC (19). CHANGES IN SPIROMETRY OVER m E Reliance should be placed on FEV, and VC for examining changes over time as they are the only spirometric variables that will consistentlyand correctlyreflect the direction of the change in overall ventilatory function. Even using these simple tests, it is never easy to determine whether a change is "real" or only a result of test variability. All lung function measurementstend to be more variable when made weeks to months apart than when repeated at the same test sessionor even daily (222, 223). Changes should therefore be interpreted cautiously. It is more likely that a real changehas occurred when there are a series of tests that show a consistent trend. As shawn in table 14 significant changes, whether statistical or biologic, vary by parameter, time period, and the type of patient. For FVC and VC in healthy subjects, within-day changeof 5% or more, between-weekschanges of 11 to 12% or more, and yearly change of 15% or more weregenerally thought by the Workshop to be clinically important. The clinician seeing the patient can often interpret results of serialtests in a useful manner, not reproducibleby any simplealgorithm. For example, seeminglystabletests may prove very reassuring in a patient receiving therapy for a diseasethat is otherwise rapidly progressive. The same tests may be very disappointing if one is treating a disorder that is expected to improve dramatically with the therapy prescribed. Depending on the clinical situa- tion, statistically insignificant trends in function may be very meaningful to the clinician. The greatest errors occur when one attempts to interpret serial changes in subjects without disease because test variability will usually far exceed the true annual decline, and reliablerates of change for an individual subject cannot be calculated without prolonged follow-up (69). Thus, in subjects with "normal" lung function, changes in VC or FEV, over 1 yr should probably exceed 15% (table 14) before any confidencecan be given to the opinion that a meaningful year-to-year change has occurred. Recommendations Overall ~ C H N I C A LISSUES Although technical sources of m a t i o n in spirometryhave been fully dealt with'in other documents, it was considered important to reemphasize their key role, particularly in relation to the following points. 1. Laboratory directorsshould be constantly on guard to maintain the precision and accuracy of the measurements made in their laboratories and should be aware of the potential sources of technical variation. Quality control includes strict adherence to ATS guidelines for equipment perfor- mance and calibration. 2. Attention should be given to the spirome- ter temperature where the tests are performed. Temperature-related errors will be reduced when the spirometer temperature is between 17" and 40" C. 3. Computer calculations should be validated at the time equipment is purchased and after any changes are made in software or hardware. BIOLOGIC VARIATION AND STATISTICAL ISSUES 1. Laboratory directors should be aware of the biologic sourcesof within- and betweenindividual variation in order to optimize the application of lung function tests to a particular patient. A number of withinindividual sources of variation fall within the domain and control of the laboratory, whereas between-individual sources of variation are important in selecting appropriate reference values. 2. Environmentalsources of variation pertinent to a given patient are more likely to be known to the referring clinician than to i 1214 AMERICAN THORICK: s the laboratory director and should be used be extrapolated for ages or heights beyond should frame this question as precisely in evaluating the clinical pertinence of a those covered by the data that generated possible. Likewise, the laboratory d m given lung function report. Laboratory them. If, for example, one calculates a responsible for seeing that the testsancgr. directors should request this information predicted FEV, for an 85-yr-old person ried out should insist that the clinicalqua- from clinicians. from prediction equationsbased on a popu- tion be included in the requisition. 3. Those who generate and report lung func- lation younger than 65 yr of age, the re- 2. Interpreters of lung function tests should tion tests should be aware of the strengths port shouldcontain a cautionarystatement. be conservative in suggesting a specific&- and weaknesses of the statisticaltechniques 5. The choice of referencevalues should con- agnosisbased only on pulmonary function used to generate the prediction values used sider the ethnic origins of the clientele of abnormalities. for interpretation. Laboratorydirectorsand the laboratory. Although it is preferable to 3. Borderline "normal" values should be in. chest physicians should also be aware of use equations based on the ethnic origins terpreted with caution. Such interpretations the strengths and limitations of the statisti- of the subject being tested, this is not al- should, when possible, use clinical infor. cal concepts of normality. ways possible or practical. For instance, if mation in the decisions as to what is nor- Selecting Reference Values a laboratory only occasionallyserves sub- mal and what is abnormkl. jects of a particular ethnic group, it is ac- 4. The first step in interpretation is to ev& GENERACLONSIDERATIONS ceptable to adjust for ethnic differencesby ate and comment on the qualityof them 1. Because of unexplained differences be- using a scalingfactor as suggestedin table 9. 5. The number of test indices (eg., FVC, tween published reference values, no one set of reference values is likely to be ap- b W E R LIMITS OF N O W FEV,, etc) used in interpretation s h w be limited to avoid an excessive n u m b of plicable to all laboratories and all clientele 1. Normal ranges should be based on calcu- false positive results. - -under all circumstances.The choiceof ref- lated fifth percentiles. Estimates of fifth 6.The primary guidesfor spirometryin- erencevalues should be a matter of careful percentiles based on the SEE are accept- tation shouldbeVC (slowor foxed), consideration by laboratory directors. It able for indices with distributions that are and FEVi/VC. should not be left to the judgment of close to Gaussian. 7. Testsperformed on children manufacturers of automated equipment. 2. Lower limits of normal are variable and, preted by those familar wit 2. Laboratories should indicate the sourceof therefore, should not be considered as ar- function in children. reference values on their reports. 3. Ideally, reference values should be based bitrary limits that correctly classify all patients into normal and abnormal groups. CONCERNING AIRWAY OBSTRU on data obtained using equipment and pro- Patient values that lie close to lower limits 1. FEVJVC should be the pri ceduresthat conformtocurrentATS recom- should be interpreted with caution. distinguishing obstructive mendations. Theprediction equationslisted 3. The use of 80% of predicted for a lower structive patterns. in tables 3 and 4 and published since 1981 limit of normal for adult pulmonary func- 2. Instantaneous and mid flows conform to current ATSrecommendations. tion parameters is not recommended. This to confirm the presence of ai criterion works only for average persons tion in the presence of a EPIDEMIOLOGIC CONSIDERATIONS and for a limited number of parameters. 1. R e f e q c e values should not come from It creates major errors when applied to studies BYLsed on hospital patients. FEFa5-75, and the instantaneous flows. 2. Reference valuesfor most clinicalapplica- Fixed percent of predicted values may be tions should be based on cross-sectional acceptable in children. studies. 4. In adults, it is not acceptableto used a fued greater than average 3. Subjects used to generate reference values FEVJFVC ratio asa lower limit of normal. should be free of respiratory symptomsand disease. It is preferableto choose reference OTHER CONSIDERATIONS healthy individuals. 5. The severity of airway obstruction values for men and women from the same 1. It is preferable for North Americanlabora- be based on FEV, rather than FE population source tories to select reference value studies based 6.Abnormalities in instantaneous flowsand 4. Reference equations based on nonsmok- on North American populations and Eu- FEFa5-,$,. should not be graded as to ers should be used for most clinical appli- ropean laboratories studies based on Eu- severity when FEV, and FEV,NC am cations. The problems in making adjust- ropean populations because an important within the normal range ments for the biologic effects of smoking portion of the variation between populalead to the recommendation that such ad- tion studies remains unexplained. justments should not be part of routineclin- 2. To assist in the choiceof reference values, CONCERNING BRONCHODILATOR RESPONSE ical interpretation. Such adjustmentsmay, it may be useful to make an empirical as- 1. VC (forced or slow) and FEV, should k occasionally, be made to address specific sessment of how different equations relate the primary indices used to judge bron- questions. to measurements made in 20 to 40healthy chodilator response. Total expiratoryt h e 5. Altitude may be important in the selection subjects typical of the laboratory's clien- should be considered when using FVC to of referencevaluesfor flow ratesand DLCO. tele. If the distribution of these measure- assess bronchodilator response sinceF W ments is, on the whole, within the range increases in obstructed patients as ex&p STATISTICAL CONSIDERATIONS 1. Prediction equations for adults should predicted, the choice is probably suitable tory -time increases. 3. If this is not the case, the differences may 2. A 12% increase, calculated from includeage and height as independentvari- be due to the laboratory (apparatus, tech- prebronchodilator value, and a 2lNhn@- ables. Usually, separate equations are used nician, procedure) or it may be that the ref- crease in either FVC or FEV, are rrag0, for men and women. erence values are inappropriate for the able criteria for a positive bronchodhklf 2. Linear equations perform adequately for laboratory's clientele Both possibilities response in adults. adults though they may overpredict in should be considered. 3. FEFIS-,sl and the instantaneous flW young adults and underpredict in the elderly. Recommendations for Interpretation should be considered secondarilyin e&- ating bronchodilator response If used, thev 3. Prediction equations should come from OVERALL CLINICALINTERPRETATTON must be volume-adjusted or the effect OC studies that present lower limits of normal 1. Because interpretation of the lung func- changing FVC must be dealt with in or present information from which such tion tests of an individual patient is best interpretation. lower limits can be calculated. made in light of the clinical question asked 4. Ratiossuch as FEV,/VC shouldnot bed 4. Referenceequations should, in general, not of the tests, the clinician requestingthe test to judge bronchodilator response 1215 5. patients may respond to bronchodilator therapy even though a bronchodilator response is absent in a laboratory test. 14. Ferris BG Jr, ed. Epidemiology standardization project. Part 2. Am Rev Respir Dis 1978; 118:1-120. 15. Pulmonary termsand symbols: a report of the CONCERNING RESTRICTION 1. The diagnosis o f a restrictive lung abnormality is based on a reduced TLC.A re- duced VC in the presence of a normal ACCPATS Joint Committee on Pulmonary Nomenclature. Chest 1975; 67:583-93. 16. Jay F, ed. IEEE standard dictionary of electrical and electronic terms. 3rd ed. New York: In- stitute of Electrical and Electronics Engineers, Inc, FEV,/VC may be used to suggest but not 1984. diagnose the presence of restriction. 17. Nelson SB, Gardner RM, Crapo RO, Jensen 2. Theseverityof restriction should be based RL. Performance evaluation of contemporary onTIC.If VC isused to infer the presence spirometers. Chest 1990, 97:288-97. of restriction, severity may be based on VC. 18. Pistelli G, Carmignani G,Paoletti P, et al. Comparison of algorithms for determining the end This Statement was prepared by the par- point of the forced vital capacity maneuver. Chest ticipants of a Workshop on Lung Function 1987; 91:loO-5. Testing: Selection of ReferenceValues and In- terpretative Strategies. The workshop participantswere: MARGARET BECKLAMKE.D,., and R0e;eRIO.CRAPO,M.D., Co-Chairpersons,A. SONIA BUIST, M.D., BENJAMIN BURROWS, M.D., JACK L. CLAUSEN, M.D., ALUN L. COATIS, M.D., JOHN COTES, D.M., DOUGLAS 19. Morris AH, Kanner RE, Crapo RO, Gardner RM. Clinical pulmonary function testing: a manual of uniform laboratory procedures. 2nd ed. Salt Lake City, U T Intermountain Thoracic Society, 1984. 20. Cotes JE. Lung function: assessment and application in medicine. 4th ed. Oxford: Blackwell Scientific Publications, 1979. W. DOCKERYP,H.D., REED M. GARDNER, 21. Wilson AF, ed. Pulmonary function testing, PED., JOHN L. HANKINSOPNH, .D., JAMES indicationsand interpretations. Orlandq n;Grune HANLEY, PH.D., ROBERT L. JOHNSON, JR., & Stratton, 1985. M.D., MICRAELD. LEBOWITPZH, .D.,PAOLO 22. Burrows B, Knudson RJ, Quan SF, Kettel LJ. PAOLETITM, .D., RENE PESLINM, .D., GEORGE ~ X G A RM, .D., PHILIP H. QUANJER, M.D., s.~ ~ E L V Y NT ~ ~ K MMA.DN., ScumT.WEISS, Resphtory disorders:a pathophysiologicapproach. Chicago: Year Book Medical Publishers, 1983. 23. Enright PL, Hyatt RE. Office spirometry: a practicalguide to the selectionand use of spirome- .M.D., M A , MARY ELLEN B. WOHL, M.D. ters. Philadelphia: Lea & Febiger, 1987. 24. Allen SM, Hunt B,Green M. Fall in vital ca- References ' 1. BatesDV. Respiratory function in disease. 3rd pacity with p t u r e . Br J Dis Chest 1985;79267-71. 25. Pierson DJ, Dick NP,Petty TL. A comparison of spirometric valueswith subjectsin standing cd Philadelphia: WB Saunders. 1989. and sitting positions. Chest 1976; 7Q17-20. 2 Clausen JL, ed. Pulmonary function testing: 26. Blair E, Hickam JB. The effect of change in Ioidelinesand controversies. New York: Academ- body positionon lung volume and intrapulmonary ~ R C S1S98,2. gas mhhg in normal subjects. J Clin Invest 1955; 3. Milkr A, ed.Pulmonary function tests: aguide 34383-9. for the student and house officer. Orlando, F L 27. Townsend MC. Spirometric forced expirato- & Stratton, 1987. ry volumes measured in the standing versus the sit- . 4. Quanjer PhH, ed.Standardized lung function nport of the working party. Bull Eur Phys- I bthd Rmpir 1983; 19(Suppl SA-95). ting posture Am Rev Respir Dis 1984; 130:123-4. 28. GlindmeyerH. Predictable confusion. J OcCUPMed 1981; 23845-9. Becklake MR. Concepts of normality applied 29. Melissinos CG, Mead J. Maximum expirato- to themeasurement of lung function. Am J Med ry flow changes induced by longitudinal tension on twcheainn o d subjects. J Appl Physiol1977; AS. Evaluation of lung function: Con- 43531-44. of normality. In: Simmons DH. ed.Current 30. Liistro G, Stanescu D, Dooms G. Rodenstein onology. Vol4. New York: John Wiley and D, Venter C. Head position modifies upper airway SOnS, 1982; 141-65. resistance in men. J Appl Physioll988: 64A285-8. 7. American Thoracic Society. Snowbird work- 31. Krowka MJ, Enright PL, Rodarte JR, Hyatt h p on standardization of spirometry. Am Rev RE. Effect of effort on measurement of forced ex- k p k Dis 1979; 119831-8. piratory volume in one second. Am Rev Respir Dis 8. American Thoracic Society. Standardization of 1987; 136~829-33. tpkomctry: 1987 update. Am Rev Respir Dis 1987; 32. Ingram RH Jr, Schilder DP. Effect of thorac- 1%1285-98. ic gas compression on the flow-volume curve of 9. AmericanThoracic Society. Single-breath car- the forced vital capacity. Am Rev Respir Dis 1966; boa monoxide diffusing capacity (transfer factor). 9456-63. b m m e n d a t i o n s for a standard technique. Am 33. Sadoul P. Mesure de la capacitC vitale et des Rev Respir Dis 1987; 1361299-307. debits maximaux. In: Denolin H, Sadoul P, Orie 10. Gardner RM, Clausen JL, Epler G, Hankin- NGM, eds.L'exploration fonctionnellepulmonaire. son JL. Permutt S. Plummer AL. Pulmonary func- 2nd partie. Paris: Flammarion, 1971; 389. tion laboratory personnel qualifications. Am Rev 34. Van de Woestijne KP, Afschrift M. Airway k p i r Dis 1986; 134623-4. dynamicsduring forced expiration in patients with 11. Gardner RM, Clausen JL, Crapo RO, et al. chronic obstructive lung disease. In: Orie NGM, Quality assurance in pulmonary function labora- Van der Lcnde R, eds. Bronchitis 111. Assen: Royal tories. Am Rev Respir Dis 1986; 134:625-7. Van Gorcium, 1970; 195-206. 12. Gardner RM, Clausen JL, Cotton DJ, Crapo 35. Gimeno F, Berg WC, Sluiter HJ, 'Thmmeiing RQ Hankinson JL, Johnson RL. Computer guide- GJ. Spirometry-inducedbronchial obstruction, Am b e s for pulmonary laboratories. Am Rev Respir Rev Respir Dis 1972; 10968-74. 1986, 134628-9. 36. E m n E, Dockery DW, Speizer FE, Fay ME, 13. Quanjer PhH, Helms P, Bjure J, Gaultier C. Ferris BJ Jr. Theassociation between health status Standardizationof lung functiontestsin paediatrics. and the performance of excessively variable spirom- Eur Respir J 1989; 2(Suppl 4121s-264s). etry tests in a population-based study in six U.S. cities. Am Rev Respir Dis 1987; 1361371-6. 37. Becklake MR. Epidemiology of test failure. Br J Ind Med 1990; 47:73-4. 38. Hetzel MR.The pulmonary clock. Thorax 1981; 36:481-6. 39. Guberan E, Williams MK, Walford J, Smith MM. Circadian variation of FEV, in shift workers. Br J Ind Med 1969; 26121-5. 40. Hetzel MR, Clark TJH. Comparison of nor- mal and asthmatic circadian rhythms in peak expi- ratory flow rate Thorax 1980; 35:732-8. 41. Clark TJH, Hetzel MR. Diurnal variation of asthma. Br J Dis Chest 1977; 71237-92. 42. McDermott M. Diurnal and weekly cyclical changesin lung airwaysresistance. J Physiol W n d ) 1966; 18690. 43. De Millas H, Ulmer WT. Circadian rhythms of airways resistance in healthy subjects and patients with obstructiveainvaysdisease Pneumologie 1971; 1M237-52. 44. Kerr HD. Diurnal variation of respiratoryfunction independent of air quality. Arch Environ Health 1973; 26144-52. 45. Barnes PJ. Circadianvariation in airway function. Am J Med 1985; 79(Suppl 6A:5-9). 46. Barnes P, Fitzgerald G, Brown M, Dollery C. Nocturnal asthma and changes in circulating epi- nephrine, histamine and cortisol. N Engl J Med 1980; 303~263-7. 47. Hetzel MR. Clark TJH, Branthwaite MA. Asthma: analysis of sudden deaths and ventilatory arrests in hospital. Br Med J 1977; 1:808-11. 48. Bags LR, Hughes DTD. Diurnal variation in peak expiratory flow in asthmatics. Eur J Respir Dis 1980; 61:298-302. 49. Halberg F. Chronobiology and thelung: implications and applications. Bull Eur Physibpathol Respir 1987; 23529-31. 50. Lewinsohn HC, Capel LH, Smart J. Changes in forced expiratory volumes throughout the day. Br Med J 1% k462-4. 51. Minette A. Contribution to the chronobiolo~y of lung function: changes of baseline values of four lung function indices between 8h and 17h in patients with bronchitic complaints without asthmatic components. Bull Eur Physiopathol Respir 1987; 23541-3. 52. B$ AS, Vollmer WM.The use of lung func- tion tests in identifying factors that affect lung growth and a&. Stat Med 1988; 7:11-8. 53. F e m s BG Jr, Stoudt HW. Correlation of anthropometry and simple tests of pulmonary function. Arch Environ Health 1971; 22:672-6. 54. Hibbert ME, Lanigan A, Raven J, Phelan PD. Relation of armspan to height and the prediction of lung function. Thorax 1988; 43:657-9. 55. SchoenbergJB, Beck GJ, Bouhuys A. Growth and decay of pulmonary function in healthy blacks and whites. Respir Physiol 1978; 33:367-93. 56. Dockery DW, Ware JH, Ferris BG Jr, et 01. Distribution of forced expiratory volume in one second and forced vital capacity in healthy, white, adult never-smokersin six US.cities. Am Rev Respir Dis 1985; 131511-20. 57. Howatt WF, DeMuth GR.11. Configuration of the chest. Pediatrics 1965; 35:177-84. 58. Damon A. Negro-whitedifferences in pulmonary function (vital capacity, timed vital capacity, and expiratory flow rate). Hum Biol 1966; 38: 381-93. 59. Marcus EB, Buist AS, Curb JD, et 01. Correlates of FEV, and prevalenceof pulmonary condi- tions in Japanese-American men. Am Rev Respir Dis 1988; 138:1398-404. 60. DeGroodt EG, van Pelt W, Quanjer PhH, van Zomeren BC, Borsboom GJJM. Growth of lung and thorax dimensionsduring the pubertal growth spurt. Eur Respir J 1988; 1:102-8. 61. Green M, Mead 1, firner JM. Variability of 1216 AMERICAN T H W * maximum expiratory flow-volume curves. J Appl Physiol 1974; 37:67-74. 62. Mead J. Dysanapsis in normal lungs assessed by the relationship between maximal flow, static recoil, and vital capacity. Am Rev Respir Dis 1980; 121~339-42. 63. Schwartz JD, Katz SA, Fegley RW, Tockman MS. Analysis of spirometric data from a national sample of healthy6- to 24-year-olds(NHANES 11). Am Rev Respir Dis 1988; 138:1405-14. 64. Schwartz DJ, Katz SA, Fegley RW, Tockman MS. Sex and race differences in the development of lung function. Am Rev Respir Dis 1988; 138~1415-21. 65. Becklake MR, Permutt S. Evaluation of tests of lung function for "screening" for early detection of chronic obstructivelung disease. In: Macklem PT, Permutt S, eds. The lung in the transition between health and disease New York: Marcel Dekker, 1979; 345-88. 66. Vollmer WM, Johnson LR, McCamant LE, Buist AS. Methodological issues in the analysis of lung function data. J ChronicDis 1987; 40:1013-23. 67. Vollmer WM, Johnson LR, McCamant LE, Buist AS. Longitudinal versuscross-sectional estimation of lung function decline: further insights. SMt Med 1988; 7:685-%. 68. Louis TA, Robins J. Dockery DW, Sgiro A 111, WareJH. Explainingdiscrepanciesbetween longitudinal and cross-sectional models. J Chmhic Dis 1986; 39831-9. 69. Burrows B,LebowitzMD, CamilliAE, Knud- son RJ. Longitudinal changes in forced expiratory volume in one second in adults. Methodologic considerations and findings in healthy nonsmokers. Am Rev Respir Dis 1986; 133974-80. 70. Van Pelt W, Quanjer PhH, Borboom G. Cohort effects in ventilatory function and chronic respiratory symptoms in a 12 year follow-up study (abstract). Eur Respir J 1989; YSuppl 93665). 71. Weispan IM, ZeballosRH.bwer singlebreath carbon mohmide diffusingcapacity@Lco)in black subjects compared with Caucasians. Chest 1987; 92(Suppl:142S). 72. National Center for Health Statistics. Body weight, stature, and sitting height: white and Ne- gro youths 12-17 years. Washington, DC:Health Servicesand Mental Health Administration, 1973. (Vital and Health Statistics. Series 11, no. 126) (DHEW publication no. jHRASl74-1608). 73. National Center for Health Statistics. Body dimensions and proportions, white and Negro children 6-11years. Washington, DC: Health Resources Administration, 1974. (Vital and Health Statistics. Series 11, no. 143)(DHEW publication no. [HRAS] 75-1625). 74. National Center for Health Statistics. Forced vital capacity of children 6-11 years. Washington, DC: Public Health Service, 1978. (Vitaland Health Statistics. Series 11, no. 164)(DHEW publication no. (PHS178-1651). 75. National Center for Health Statistics. Skin folds, body girths, bicromial diameter and selected anthropometric indices of adults. Washington, DC: Public Health Service, 1970. (Vital and Health Statistics. Series 11, no. 35) (DHEW publication no. [PHS] 100-11-35). 76. National Center for Health Statistics. Selected body measurements of children 6-11 years. Washington, DC: Health Services and Mental Health Administration (Vital and Health Statistics. Series 11, no. 123) (DHEW publication no. [HSM]73-1605). 77. Van de Wal BW, Erasmus LD, Hechter R. Stem and standing heightsin Bantu and whiteSouth Africans: their significance in relation to pulmonary function values. S Afr Med J 1971; 45(Suppl: 568-70). 78. Woolcock AJ, Coleman MH, Blackburn CRB. Factors affecting normal values for ventilatorylung function. Am Rev Respir Dis 1972; 1M692-709. 79. Cotes JE, Anderson HR, Patrick JM. Lung functionand the responsetoexercisein New Guineans: role of genetic and environmentalfactors. Phi10s Trans R SOCLond [Biol] 1974; 268:349-61. 80. Miller GJ, SaundersMJ, Gilson RJ, et at! Lung function of healthy boys and girls in Jamaica in relation to ethnic composition, test exercise performance, and habitual physical activity. Thorax 1977; 32486-%. 81. OscherwitzM, Edlavitch SA, Baker TR, et 01. Differencesin pulmonary functions in various racial groups. Am J Epidemiol 172; %319-27. 82. Mueller WH, Yen F, Soto P, et ul. A multinational Andean genetic and health program. VIII. Lung function changes with migration between altitudes. Am J Phys Anthropol 1979; 51:183-95. 83. Brown P, Sadowsky D, Gajdusek DC. Ventilatory lung functions studies in Pacific Island Micronesians.Am J Epidemioll978; 108:259-65. 84. Myers J. Differentialethnic standardsfor lung function or one standard for all? S Afr Med J 1984; 65:768-72. 85. Steinberg M, Becklake MR. Socio-environ- mental factorsand lung function. SAfr Med J 1986; 70270-4. 86. Burrows B, nussig LM. As the twig is bent, the tree inclines(perhaps). Am Rev Respir Dis 1980, 122813-6. 87. Burrows B, Knudson RJ, Lebowitz MD. The relationship of childhood respiratoryillnessto adult obstructiveairway disease.Am Rev b p i r Dis 1977; 115:751-60. 88. Bertrand JM, Riley P, Popkin JA, Coates A. The long term pulmonary sequelae of prematurity: the role of familial airway hyperreactivityand the respiratory distress syndrome. N End J Med 1985; 312742-5. 89. U.S.Department of Health and Human Services. The health consequencesof smoking:chronic obstructive lung disease, a report of the Surgeon General. Rockville,MD. Public Health ServiceOffice on Smoking and Health, 1984. 90. Kryger M, Aldrich F, Reeves JT, Grover RF. Diagnosis of airflow obstruction at high altitude. Am Rev Respir Dis 1978; 117:1055-8. 91. Crapo RO, Morris AH, Gardner RM. Refer- encespirometricvalues using techniquesand equipment that meet ATS recommendations. Am Rev Respir Dis 1981; 123:659-64. 92. Goldman HI, BecklakeMR. Respiratory function tests. Normal values at median altitudes and the prediction of normal results. Am Rev Tuberc 1959; 79457-67. 93. Gautier H, Peslin R, Grassino A, er uf. Mechanical properties of the lungs during acclimatization to altitude. J Appl Physiol1982; 52:1407-15. 94. Mansell A, Powles A, Sutton J. Changes in pulmonary PV characteristics of human subjects at an altitude of 5366 m. J Appl Physiol 1980; 49: 79-83. 95. Jaeger JJ, Sylvester JT, Cymerman A, el 01. Evidence for increased intrathoracic fluid volume in man at high altitude. J Appl Physiol 1979; 47~670-6. 96. Frisancho AR. Functional adaptation to high altitude hypoxia. Science 1975; 187:313-9. 97. Heath D, Williams DR. The lung at high altitude. Invest Cell Pathol 1979; 2147-56. 98. Lippmann M. Health significanceof pulmonary function responses to irritants. J Air Pollut Cont Assoc 1988; 38:881-7. 99. Goldsmith JR. Effects on human health. In: Stern A, ed. Air pollution. Vol VI. New York: Academic Press, 1986; 391-463. 100. Committee on the Epidemiologyof Air Pollutants. Epidemiologyand air pollution. Washing- ton, DC: National Academy Press, 1985. 101. Lippmann M. Health effects of ozOnc`ac& i d review. J Air Pollut Cont Assoc 19W,3%72-9$. 102. Shy CM. Health effects of acid auosok Re- view, discussion and summary of e p i d e m i o l w studies. Environ Health Perspect 1989; 79187-90. 103. Linn WS, Theodore G, Venet E.et ol.Re- spiratory effects of sulfur dioxide in heavily am. cising asthmatics. Am Rev Respir Dis 1983: 127: w-278-83. 104. Department of Health, Education and fare. The health consequences of involuntarypnd, ing: a report from the Surgeon General. Wash. ton, DC: USGPO, 1986. (DHHS publication no. (CDCl87-8398). 105. Fielding JE, Phenow KJ. Health effects of involuntary smoking. N Engl J Mod 1988; 319: 1452-60. 106. Spitzer WO, Lawrence V, Dales R,duL Li& between passive smoking and disease: a bat & dence synthesis. Clin Invest Med 1990; 1 3 1 7 4 2 107. Lebowitz MD, Quackenboss JJ. Theeof environmental tobacco smoke on p u l m o q function. Int Arch Environ Health 1 9 8 9 ; 4 4 ( S ~ 147-52). 108. Masi M, Hanley JA, Ernst P, BecklakeM& Environmentalexposunto tobacco s m o l r c a r d ~ function in young adults. Am Rev Res* Dk 1% 13832969. 109. Martinez FD, Antognoni G, Maai F. a.L Parental smoking enhances bronchial ness in nine-year-old children. Am Rev 1988; 138518-23. 110. Samet JM. Marbury MC, S Health effects and sourcesof indoor air poU Part I. Am Rev Respir Dis 1987; 136: 111. Samet JM. MarburyMC,Spengkr effects and sources of indoor air poUuti 11. Am Rev Respir Dis 1988; 137:22142. 112. Cotes JE, Steel J. Work related I ers. Oxford: BlackwellScientificPu 436. 113. Becklake MR. Occupational dence of a causal association with c tive pulmonary disease. Am Rev Respu 140585-91. 114. Chan-Yeung M, Lam S. Occu ma. Am Rev Respir Dis 1986; 133 115. Lebowitz MD. The relationshi environmentalfactors to the prevalen tive lung diseases and other chronic con Chronic Dis 1977; 30599-611. 116. Viegi G, Paoletti P, Prediletto R,d d Rcvc alence of respiratory symptoms in an unpolhtOca area of northern Italy. Eur Respir J 1988; 1~11-8. 117. Polgar G, Weng TR. The functional dcnlopment of the respiratory system: from tbepaiod of gestationto adulthood. Am Rev RespU Dis 1974 120625-95. 118. Polgar G, Promodhat V. Pulmonary hoc- tion testing in children: techniques and staodrrdr Philadelphia: W B Saunders, 1971; 272. 119. Quanjer PH, Stocks J, Polgar G Wise Karlberg J, Borsboom G. C o m p i l a t i o n o f d w values for lung function measu Eur Respir J 1989; 2(Suppl 4 120. DeGroodt EG, Quanjer Zomeren BC. Changing relatio ure and lungvolumesduring pu 1986; 65~139-53. 121. ColeTJ. Linear and proportional models in the prediction of ventilatory fimcti6a. J R Stat SOC[A] 1975; 138:297-338. 122. Schrader PC, Quanjer PhH, van Zoamea BC, Wise ME. Changes in the FEV,-height d- tionship during pubertal growth. BuU Eur iopathol Respir 1984; 20381-8. 123. Seely JE, Guzman CA, Becklake MR H F and lung function at rest and during arerdLe adolescence. J Appl Physiol 1974; 3634-40. 1217 1w. Tanner JM. Growth at adolescence, with a geeralconsideration of the effects of hereditary &environmental factors upon growth and matu- nuon from birth to maturity. 2nd ed. Oxford: Wkwell, 1978. 125. Simon G, Reid L, Tanner JM, Goldstein H, Mjamin B. Growth of radiologically determined beart diameter, lung width, and lung length from $19 years, with standards for clinical use. Arch D ~ Cs hild 1972; 47~373-81. U6. Dockery DW, Berkey CS, Ware JH, Speizer E,FerrisBG Jr. Distribution of forced vital ca- prityand forced expiratory volume in one second m childnn 6 to 11 years of age. Am Rev Respir Dk 1983; 128405-12. VI. Rea H, Becklake MR, Gheuo H. Lung func- tionchangesas a reflectionof tissueaging in young adults.Bull Eur Physiopathol Respir 1982; 185-19. 128. Black LF, Hyatt RE. Maximal respiratory pressures: normal values and relationship to age and sex. Am Rev Respir Dis 1969; W6%-702. j' 129. Smyth RJ, Chapman KR, Rebuck AS. Maximalinspiratoryand expiratorypressuresin adoles- ants: normal values. Chest 1984; 86568-72. I UO. Wagner JS, Hibbert ME, Landau LI. Maximp1nspinuory pressures in children.Am Rev Respir Dk 1984, 1291873-5. Ul. Wilson SH, Cooke NT, Edwards RHT, Spiro SG Predicted normal values for maximal respira- tory pressures in Caucasian adults and children. lhorax 1984; 39535-8. U2. Cook CD, Mead J, Orzalesi MM. Static , rdumcpressure characteristics of the respiratory rgstemduringmaximalefforts. J Appl Physiol 1964, U3. Gaultier C, Zinman R. Maximal static pres(prrr in healthy children. Respir Physiol 1983; If(. Leech JA, Ghezzo H, Stevens D, Becklake MR.Respiratory pressures and function in young vahr Arn Rev Respir Dis 1983; 128:17-23. US. SchrPder PC, Quanjer PhH, Olievier ICW. muscle forceand ventilatoryfunctionin dolescents. Eur Respir J 1988; 1:368-75. I U6 sharp JT, Druz WS, Balagot RC, Bandelin %Daaon J. Total respiratory compliance in inandchildren. J Appl Physiol1970; 29775-9. ' U?. Fagan DG.Shape changesin staticV-P loops gs related to growth. Thorax Samanek M. Pulmonary dolescents.J Appl Physiol m. 40353-61. Bfuan D, Yernault JC, Paiva M, Englert M. mechanical lung properties in healthy chil- *m.h Scand J Respir Dis 1976; 57:139-47. Mansell AL, Bryan AC, Levison H. Relation- Of lung recoil to lung volume and maximum *ory flow in normal children. J Appl Physiol 1m;42817-23. &?toyer A, Yernault JC, Englert M, Baran 4 Mva M. Evolution of intrathoracic airway -during lunggrowth. J Appl Physioll978; us21-7. Turner JM, Mead J, Wohl ME. Elasticity of humanlungs in relation to age. J Appl Physioll968; 8:664-71. '43. Knudson RJ, Clark DF, Kennedy TC,Knud- *DL. Effect of aging alone on mechanical propof the normal adult human lung. J Appl Phys- -1977; 439054-62. la. Cotes JE, Dabbs JM, Hall AM, Heywood c, kurence KM. Sitting height, fat free mass and pody fat as reference variables for lung function Q bealthy British children: comparison with stat- Ann Hum Biol 1979; 6307-14. &5. M e r SR, Swan AV, Peat JK, Woolcock AJ, k k b u r n CRB. Maximum expiratory flow-volume in children: changes with growth and in- dividual variability. Bull Eur Physiopathol Respir 1977; 13:249-60. 146. Thurlbeck WM. Postnatal human lung growth. Thorax 1982; 37564-71. 147. Taussig LM, Cota K, Kaltenborn W. Different mechanical properties of the lung in boys and girls. Am Rev Respir Dis 1981; 123:640-3. 148. DuivermanEJ, Clement J, van de Woestijne KP, Neijens HP, van der Bergh ACM, Kerrebijn KF. Forced oscillation technique. Reference values for resistance and reactanceover a frequencyspectrum of 2-26 Hz in healthy children aged 2.3-12.5 years. Bull Eur PhysiopatholRespir 1985;21:171-8. 149. Knudson RJ, Lebowitz MD, Holberg CJ, Burrows B. Changes in the normal maximal expiratory flow-volume curve with growth and aging. Am Rev Respir Dis 1983; 127:725-34. 150. Paoletti P, Pistelli G, Fazzi P, et of. Reference values for vital capacity and flow-volume curves from a general population study. Bull Eur Physiopathol Respir 1986; 22:451-9. 151. Lebowitz MD, Holberg CJ. Comparisons of spirometric reference values and the proportions of abnormalsamong male smokersand those symp tomatic in a communitypopulation.Am Rev Respir Dis 1990, 141:1491-6. 152. Van Ganse W,Billiet L, Ferris BG. Medical criteria for the selectionof normal subjects. In: Arcangeli P, Cotes JG, Cournand A, et a/, eds. Normal values for respiratoryfunctionin man. Alghero: Panminerva Medica, 1970; 15-27. 153. Kory RC, Callahan R, Boren HG, Syner JC. The Veterans Administration-Army cooperative study of pulmonary function. I. Clinical spirometry of normal men. Am J Med 1961; 30243-58. 154. De &on JPM. Joosting PE, VisserBF. Les valeursnormales de la capacitk vitale et du volume expiratoire maximum seconde. Recherchechez les ouvrien d'une aci6ie aux Pays-Bas. Arch Mal Prof 1964; 25:17-30. 155. k r e n HG. Kory RC, Syner JC. The Veter- ans Administration-Armycooperativestudyof pulmonary function. 11. The lung volume and its sub- divisionsin normal men. Am J Med 1966,4196-114. 156. Amrein R, Keller R. Joos H, et 01. Valeurs thhriques nouvelles de l'exploration de la fonction ventilatoire de poumon. Bull Physiopathol Respir 1970; 6317-49. 157. JouasSa D. Normalisationdeskpreuvesfonctionnelles respiratories dans les pays de la Communaue Europdnnedu Charbonet del'Acier. Poumon Coeur 1% 161145-59. 158. Miller A, Thornton JC, Warshaw R, Bernstein J, Selikoff IJ, Teirstein AS. Mean and instantaneous expiratory flows, FVC and FEV,: prediction equations from a probability sampleof Michigan, a large industrial state. Bull Eur Physiopathol Respir 1986; 2239-97. 159. Miller MR, Pincock AC. Predicted values: how should we use them? Thorax 1988; 43:265-7. 160. Miller A. Predictionequations and "normal values." In: Miller A, ed. Pulmonary function tests in clinical and occupational lung disease.New York: Grune & Stratton, 1986; 197-213. 161. Oldham PD. Percent of predicted as the limit of normal in pulmonary function testing: a statistically valid approach. Thorax 1979; 34569. 162. SobolBJ, Weinheimer B. Assessment of ven- tilatory abnormality in the asymptomatic subject: an exercise in futility. Thorax 1966; 2145-9. 163. Quanjer PhH. Predicted values: how should we use them (letter). Thorax 1988; 43563-4. 164. Dockery DW, Speizer FE, Ferris BG Jr, Ware JH, Louis TA, SpiroA 111. Cumulative and reversible effects of lifetime smoking on simple tests of lung function in adults. Am Rev Respir Dis 1988; 137:286-92. 165. Miller A, Thornton JC, Warshaw R, Ander- son H, Teirstein AS, Selikoff IJ. Single breath & diffusing capacity in a representativesample of the population of Michigan, a large industrial state. Am Rev Respir Dis 1983; 127:270-7. 166. Beck GJ, Doyle CA, Schachter EN. Smok- ing and lung function. Am Rev Respir Dis 1981; 123:149-55. 167. Burrows B,Knudson RJ, Cline MG, Lebowitz MD. Quantitative relationships between cigarette smoking and ventilatory function. Am Rev Respir Dis 1977; 115:195-205. 168. Fletcher C, Pet0 R, Tinker C, Speizer FE. The natural history of chronic bronchitis and em- physema. London: Oxford University Press, 1976. 169. Enjeti S, Hazelwood B, Permutt S. Menkes H, Terry P. Pulmonary function in young smokers: male-female differences. Am Rev Respir Dis 1978; 118667-76. 170. Higenbottam T, Clark TJH, ShipleyMJ, Rose G. Lung function and symptomsof cigarette smokers reiated to tar yield and number of cigarettes smoked. Lancet 1980; 1:409-11. 171. Becklake MR, La110 U.The "healthy smoker"effect: a phenomenon of health selection.Respiration 1990, 57:137-44. 172. Strope GL, Helms RW. A longitudinal study of spirometry in young black and young white children. Am Rev Respir Dis 1984, 1301100-7. 173. Pattishall EN, Helms RW, Strope GL. Noncomparability of cross-sectional and longitudinal estimates of lung growth in children. Pediatr Pulmono1 1989; 7:22-8. 174. Hibbert ME, Lannigan A; Landau LI, Phelan PD. Lung function values from a longitudinal study of healthy children and adolescents. Pediatr Pulmonol 1989; 7:lOl-9. 175. Bencowitz HC. Inspiratory anb'expFatory vital capacity. Chest 1985; 85:834-5. 176. Taussig LM, Cherniack V. Wood R. Farrell PM, Mellins RB. Standardization of lung function testing in children. Recommendations of the GAP ConferenceCommittee, CysticFibrosisFoundation. J Pediatr 1980; 97:668-76. 177. Zapletal A, Samanek TP. Lung function in children and adolescents: Methods and reference values. Basel-Munchen: Karger, 1987. 178. Burrows B,Cline MG, Knudson RJ, Taussig LM, Lcbowtiz MD. A descriptive analysis of the growth and decline of the FVC and FEV,. Chest 1983; 83:717-24. 179. Galen RS. The normal range: a concept in transition. Arch Pathol Lab Med 1977; 101561-5. 180. Zweig MH, Beck JR, Collinsworth WL. Assessment of clinical sensitivity and specificity of laboratory tests. Proposed guideline. Philadelphia: National Committee for Clinical Laboratory Standards, 1987 (NCCLS Document GP 10-P vol. 7 no. 6). 181. Henry RJ. Reed AH. Normal values and the use of laboratory results for the detection of disease. In: Henry RJ, Cannon Dc,Winkelman JW, eds. Clinical chemistry: principles and techniques. New York: Harper & Row, 1974; 343-70. 182. Cochrane AL, Elwood PC. Laboratory data and diagnosis (letter). Lancet 1969; 1:420. 183. Flenley DC.Chronic obstructive pulmonary disease. Dis Mon 1988; 34537-99. 184. Mendella LA, Manfreda J, Warren CPW, Anthonisen NR. Steroid response in stable chronic obstructive pulmonary disease.Ann Intern Med 1982; 96~17-21. 185. Guyatt GH, Townsend M, Nogradi S, Pugs- ley SO, Keller JL, Newhouse MT. Acute response to bronchodilator, an imperfect guide for bronchodilator therapy in chronic airflow limitation. Arch Intern Med 1988; 148:1949-52. 186. Lorber DB, Kaltenborn W, Burrows B. Responses to isoproterenol in a general population sample. Am Rev Respir Dis 1978; 118:855-61. 187. Dales RE, Spitzer WO, Tousignant P. I I 3 1 b 1 t i E 1 4 I 1218 Schechter M, Suissa S. Clinical interpretation of airway responseto a bronchodilator: epidemiologic considerations. Am Rev Respir Dis 1988; 138: 317-20. 188. Watanabe S, Renzetti AD Jr, Begin R, Big confidence. Br Med J 1989; 1-140. 208. Becklake MR, Kalica AR. NHLBl workshop summary. Scientificissues in the assessment of respiratory impairment. Am Rev Respir Dis 1988; 137:1505-10. 229. Rossiter CE, Weill H. Ethnic d i f f e r e m h lung function: evidence for proportional dillb, ences. Int J Epidemiol 1974; 355-61. 230. Lap NL, Amandus HE, Hall R, Mom aWKC. Lung volumes and flow rates in black ler AH. Airway responsiveness to a bronchodila- 209. Evaluation of impairment/disabilitysecond- white subjects. Thorax 1974; 29185-8. tor aerosol. Am Rev Respir Dis 1974; 109:530-7. ary to respiratory disorders. Am Rev Respir Dis 231. Cookson JB, Blake GTW, Faranisi C. Nor. 189. Sourk RL, Nugent KM. Bronchodilator test- 1986; 133:1205-9. mal values for ventilatory function in Rho&& ing: confidence intervals derived from placebo in- 210. Kanner RE, Renzetti AD, StanishWM, Bark- Africans. Br J Dis Chest 1976; 70107-11. halations. Am Rev Respir Dis 1983; 128:153-7. man HW, Klauber MR. Predictors of survival in 232. Patrick JM, Femi-Pearse D. Reference- 190. Anthonisen NR, Wright EC, and the lPPB subjects with chronicairflow limitation. Am J Med for FEV, and FVC in Nigerian men and wo- Trial Group. Bronchodilator response in chronic 1983; 74249-55. a graphical summary.Med J Nigeria 19766:3&5. obstructive pulmonary disease. Am Rev Respir Dis 211. Ttaver GA, Cline MG, Burrows B. Predictors 233. Billiet L, Baisier W, Naedts JP. Effa & h 1986; 133:814-9. 191. lkeeddale PM, Alexander F, McHardyGJR. of mortality in chronic obstructive pulmonary disease. Am Rev Respir Dis 1979; 119895-902. taille, et du sexe de l'age sur la capacitk de djf& sion pulmonairede l'adulte normal. J Physid-1 Short term variability in FEV, and bronchodilator 212. Anthonisen NR, Wright EC, Hodgkin JE. 963; 55:199-200. responsivenessin patients with obstructive ventila- Prognosis in chronic obstructive pulmonary dis- 234. Teculescu DB, Stanescu DC. Lung di tory defects. Thorax 1987; 42:487-90. ease. Am Rev Respir Dis 1986; 133:14-20. capacity. Normal values in male smokersand- 192. Eliasson0,Degraff AC Jr. The use of criteria 213. Kannel WB, Lm EA. Vital capacity as a smokers using the breath-holding technique sepd for reversibility and obstruction to define patient predictor of cardiovascular disease: the Framin- J Respir Dis 1970; 51:137-49. groups for bronchodilator trials: influence of clin- gham study. Am Heart J 1983; 105:311-5. 235. Van Ganse WF, Ferris BG Jr. CotesJE. ical diagnosis, spirometric and anthropometric 214. Kannel WB, Lew EA, Hubert HB, Castelli arette smoking and pulmonary diffusing I values. Am Rev Respk Dis 1985; 132:858-64. WP. The d u e of measuring vital capacityfor prog- (transfer factor). Am Rev Respir Dis 19`Q 193. Brand PLP, Quanjer PhH, Postma DS. et nostic purposes. Trans Assoc Life Insur Med Dir 30-41. uL A comparison of different ways to expressbon- Am 1981; 6456-83. 236. Frans A, Stanescu DC.Veri chodilator response Part 2. Am Rev Respir Dis 215. Tockman MS, Comstock GW. Respiratory T, Brasseur L. Smoking and pulmo 1990; 141:A20. risk factors and mortality: longitudinal studies in capacity. Scand J Respir Dis 1975 194. Olsen CR, Hale FC. A method for interpret- Washingtoncounty, Maryland. Am Rev Respir Dis ing acute response to bronchodilators from the 198% 14o(S~ppl:56-63). in smokers with nor spirograms. Am Rev Respir Dis 1968; 98:301-2. 216. Foxman B,Higgins ITT,Oh MS. The effects Am Rev Respir Dis 195. Boggs PB, Bhat KD, Vekaius WA, Deb0 MS. of occupation and smoking on respiratory disease 238. SalorinneY.Si Theclinical significanceof volume adjusted maxi- mortality. Am Rev Respir Dis 1986; 134649-52. mal mid expiratory flow (Iso-volume FEF,,-,,%) 217. Annesi 1, Kauffmann F. Is respiratory mu- connective tissue diseases and in in assessingairway responsiveness toinhaled bron- cus hypersecretionreally an innocent disorder? Am eases. %and J Respir Dis 1976; chodilator in asthmatics. Ann Allergy 1982; 48: Rev Respir Dis 19W. 134688-93. 239. Crapo RO. Morris AH. S 13942. 218. Ortmeyer CE, Costello J. Morgan WKC. breath normal values for carbon monoxi 1%. Berger R, Smith D. Acute postbronchodila- Swecker S, Peterson M. The mortality of Ap- ing capacity. Am Rev tor changes in pulmonary function parameters in palachian coal miners 1963to 1971. Arch Environ 240. Paoletti P. V patients with chronic airways obstruction. Chest Health 1974; 2967-72. 1988; 93541-6. 219. Pet0 R, Speizer FE, Cochrane AL, et al. The 197. Co&ttee report. Criteria for the assessment relevancein adults of air-flow obstruction, but not of reversibility in airwaysobstruction: reportof the of mucoushyperseattion, to mortalityfrom chronic committee on emphysema, American College of lung disease: results from 20 years of prospective Chest Physicians. Chest 1974; 65:552-3. observation. Am Rev Respir Dis 1983; 128:491-500. 198. Veda1 S, Crapo RO. False positive rates of 220. Tockman MS, Anthonisen NR, Wright EC, from a healthynonsmoking pop multiple pulmonary function tests in healthy sub- Donithan MG. Airways obstruction and the risk of hematocrit. Am Rev Respir Dis 1 jects. BuU Eur Physiopathol Respir 1983; 19263-6. of lung cancer. Ann Intern Med 1987; 106512-8. 242. Roca J, Rodriguez-Roisin R, 199. Knudson RJ, BurrowsB, LebowitzMD. The 221. Kuller LH, Ockene J, Meilahn E, Svendsen gos F, Perez J. Clausen JL. Single- maximal expiratory flow-volumecurve: its use in KH. The relation of forced expiratory volume in monoxide diffusing capacity predi the detectionof ventilatoryabnormalities ina popu- one second (FEV,) to lung cancer mortality in the from a Mediterranean population. lation study. Am Rev Respir Dis 1976; 114:871-9. multiplerisk factor intervention trial (MRFIT). Am Dis 1990, 141:1026-32. 200. Ash KO, Clark SJ, Sandberg LB, Hunter E, J Epidemiol 1990; 132235-74. 243. Hall AM, Heywood C, Cotes JE. Lung- Woodward SC. The influences of sample distribu- 222. Lebowitz MD,Quackenboss J, Camilli AE, tion in healthy British women. Thorax -, tion and age on referenceintervals for adult males. Bronnimacn D, Holberg CJ, Boyer B. The epide- 34359-65. Am J Clin Pathol 1983; 79574-81. miological importance of intraindividual changes 244. Black LF, Offord K, Hyatt RE. -V 201. Herrera L. The precision of percentilesin es- in objective pulmonary responses. Eur J Epidemi- in the maximal expiratory flow volume in tablishing normal limits in medicine. J Lab Clin 01 1987; 3:390-8. asymptomaticsmokersand in nonsmokers Am- Med 1958; 52:34-42. 223. Cochrane GM, Prieto F, Clark TJH. In- Respir Dis 1974; 110282-92. 202. Shulz EK. Willard KI, Rich SS, et 01. Im- trasubject variability of maximal expiratory flow 245. Crapo RO, Morris AH, Clayton PI), Nhroa proved reference-interval estimation. Clin Chem volume curve Thorax 1977; 32:171-6. CR. Lung volumes in healthy nonsmoking ad& 1985; 31:1974-8. 224. Moms JF, Koski A, Johnson LC.Spiromet- Bull Eur Physiopathol Respir 1982; 18:41%25. 203. Dybkaer R. International Federation of Clin- ric standards for healthy nonsmoking adults. Am 246. Grimby G, Soderholm R Spkometrk sttub ical Chemistry (IFCC)1),2) The theory of reference Rev Respir Dis 1971; 1035767. in normal subjects. 111. Static lung volUma d values. Part 6. Presentationof observed values relat- 225. Chemiack RM, Raber MR Normal standards maximum voluntary ventilation in adults d%a ed to reference values. J Clin Chem Clin Biochem for ventilatoryfunction using an automated wedge note on physical fitness. Acta Med scarsd 1982; 20841-5. spirometer. Am Rev Respir Dis 1972; 1M38-44. 173~199-206. us4. Elveback LR, Taylor WF.Statisticalmethods 226. Roca J, Sanchis J, Augusti-Vidal A, et al. 247. Reed TE. Caucasian genes in of estimating percentiles. Ann NY Acad Sci 1969; Spirometric reference values from a Mediterrane- Negroes. Science 1%% 165762-8. 161~538-48. an population. Bull Eur Physiopathol Respir 1986; 248. Ghio AJ, Crapo RO, Elliott CG. Re 205. Linnet K. Tivestage transformation systems 22: 217-24. equations used to predict pulmonary fund00 a for normalization of reference distributionsevalu- 227. Johannsen ZM, Erasmus LD. Clinical institutions with respiratory disease trainhg PP ated. Clin Chem 1987; 33:381-6. spirometry in normal Bantu. Am Rev Respir Dis grams in the United States and Canada: a s W W 206. Solberg HE. The theory of reference values. 1%8; 9735-97. Chest 1990, 97:400-3. Part 5. Statistical treatment of collected reference 228. Miller GJ, Ashcroft MT, SwanAV, Beadnell 249. Gaensler EA, Smith AA. Attachment f0fm values. Determinationof referencelimits. IFCC sci- HMSG. Ethnic variation in forced expiratory vol- tomated single breath diffusing capacity messlor- entific Committee proposed recommendations. J ume and forced vital capacity of African and Indi- ment. Chest 1973: 63:136-45. Clin Chem Clin Biochem 1983; 21:749-60. an adults in Guyana. Am Rev Respir Dis 1970; 207. Gardner MJ, Altman DG,eds. Statisticswith 102979-81.