Document JvBqJMQGmvBdwBb0vjjKK56DK

-Reference Equations Used to Predict Pulmonary Function* Survey at Institutionswith Respiratory Disease Training Programs in the Unitedstates and Canada A n d m I. Ghio, M .D.;Robert 0 Crapo, M .D.,EC.C.E; and C . Gregory Elliott, M.D.,E C.C.P J 2 Adult respiratory disease training programs in the United States and Canada were surveyed to determine which referenceequationswere used to predict wrmal pulmonary function and how ethnic differences were approached. Replies from 139 of the 180 (77.2 percent) institutions surveyed were received and evaluated. Surprisingly few studies account for most of the equations in use: three studies account for 85 percent of the spirometric equations, two for 83 percent of the lung volume equations and five for 84 percent of the &sing capacity equati there are no debite data, the fwm of many suggests that equipment default settings may selection process. Of those responding to the ence question, !S percent of institutions applied no tion for ethnic differences. There was no consistent to the method of correction among those who did. Many linear and nonlinear regression techniques have been used to generate equations to predict "normal values" since Hutchinson' first provided a United States and Canada with training programs in adult respiratory disease to determine which equa- tions are used-and what attention is paid to ethnic quantitative index of pulmonary function based on differences. measurements of vital capacity in more than 2,000 healthy men in 1846. The increasing importance of pulmonary function testing in diagnosing and managing lung disease and assessing impairment has required more accurate definitions of normal. Innovations in technology and equipment, standardization of procedures, and changing concepts of normality have stimulated further studies of healthy subjects and there is now a large number of different regression equations from which to choose. In response to a frequently asked question about which equations are actually being used, we surveyed institutions in the METHODS Letters were mailed to the directors of training programs in ad& respiratory disease listed by the American Thoracic Society. Mb requested specific citations for the prediction equations used for spirometry, lung volumes, carbon monoxide diffusing capacity, and airway resistance. We also asked for a description of how they d d t with ethnic daerences in pulmonary function. A follow up request was sent to institutions which had not responded to the first inquiry. RESULTS One hundred and eighty institutionswere surveyed; 96 replies were received after the first mailing, 50 after the second. Seven replies were excluded because *From the Pulmonarv Division, Department of Internal Medicine, LDS Hospital, and Division of Respiratory, Critical Care, and Occupational (Pulmonary) Medicine, De tment of Internal Medicine, University of Utah School of M s i n e , Salt Lake City. Manuscript received August 5; revision accepted July 17. Re rint requests: Dr.Crapo, ttrlnwnary Dioision, LDS Hospitd, Sa[ Lake City 84143 they lacked adequate detail. We were able to analyze 139 replies (77.2 percent of those surveyed). The geographic distribution of inquiries and responses is shown in Table 1. All 139 institutions provided reference equations Table 1--Geographic DistributionOf 1?&7UkifXaI nd Responses Region United States Northeast (ME,NH, VT,MA, RI, Cr,NY, PA, NJ, DE) Midwest(WI, IL, MI, IN, OH, MN, ND, SD, NB, KS, MO, IA) West (AK, HI, WA, OR, CA, MT, ID, Nc: UT, CO, AZ,NM) South(W, VA, MD, DC, NC, SC, GA, FL, ICY, TN, MS, AL, OK, AR, LA, TX) Puerto Rico Canada Total Inquiries 68 33 2.5 42 1 12 181 Responses 48 30 22 28 0 11 139 I i ! 1 I 4 I 8I 1 78 55 -nine predicted FEV,/FVC by dividing predicted FEV, by predicted FVC. 1 We chose to deal only with standard indices and have listed only equations ore than one institution; equations cited e are included as "other studies." The re- es are summarized in Table 2. Six institutions reference equations from different studies to FVC and FEV,; six others selected prediction different studies for men and women. ations for FEV,/FVC and FEF25-75% , sources different from those used for FVC in six and 16 institutions, respectively. e centers predicted normal FEV,/FVC by ng predicted FEV, by predicted FVC. ne hundred thirty-five (97.2 percent) institutions furnished reference equations for assessing lung vol- umes (Table 3).Six of the nine institutions using Boren et all4 to predict lung volumes for men used the Goldman equations for women. The 138 responses for carbon monoxide diffusing capacity are summarized in Table 4. Two centers predict Dco based on unpublished studies conducted at their institutions. Sixty-six institutions did not comment on airway resistanceand 16specificallysaid they did not measure airway resistance. Of the 57 which supplied informa- tion on normal values for airway resistance, 48 used reEerence equations to predict normal values while nine used a numerical range derived from ``clinical experience." The equations used to predict airway mistance are summarized in Table 5. Of the 139 replies analyzed, 95 (68.4 percent) responded to thehuestion on ethnic differences. Fifty centers did not apply any ethnic correction and 45 did. Only five centers used population-specific reference equations to deal with ethnic differences.10.s28 The others reduced predicted values for whites by a fixed percentage: 29 institutions by 15 percent; four by "10 to 15percent;" and seven by 10 percent. The ethnic groups to which corrections were applied usually depended upon the geographic location of the institution and usually included the predominant minorities. Our survey showed ethnic adjustments were made for blacks in all regions of the United States but not in Canada. Adjustments for American Indians were made only in Utah and Alberta, Canada. Corrections for Hispanics were reported only from Texas, for Asian Indians only from New York, and for Asians only from California. DISCUSSION Despite the number of reference studies and equations available in the literature, surprisingly few equa- tions are widely used. Three studies accounted for 85 percent of the equationsused for standard spirometric indices. It was not uncommon for institutions to use a different reference equation for each spirometric parameter. While we are not aware of studies on the effects of this practice, it seems likely to increase the uncertainty that prediction equations will match the . .Table 4-Summary of ZnatiiurioM Citing R e j k m afor thdadwn ofcarbon Monoxide DijJbingCapacity CddmanandBecklake** clp,et a l ~ b n et ai" h e s et a~ dies TLC RV FRC --1 MFMFMF 84 90 85 91 74 81 2.82828282929 99 9 3 3 3 31111 333333 8 11 7 10 B 11 B u n a ~est all6 C r a p et all' Miller et a P Gaensler and WrightL* cotes= Bates et dL6 SalOrinneU McCrath and Thompson= Other studies MF 25 25 27 26 25 26 23 23 16 15 32 33 12 15 16 dnical population. In general, we recommend refer- ence equations for spirometric indices be selected from a single study which has been matched to the patient population. This should be relatively easy for spirometry, but is more difficult for lung volumes, Dco, and airway resistance because there are few studies available which include complete pulmonary function tests. Two studies account for 83 percent of the lung volume equations (Table3)and five for 84 percent of the Dco CIgble 4). The greater number of DCO equations in use may reflect the larger interlaboratory differences for Dco than spirometry or lung volumes, though a preliminary study suggests academiccenters do not consistently choose reference equations which match the Dco values produced in their laboratories.28 While all reference equations should be demonstrated to show a reasonable match with the individual labo- ratory's instruments and clientele, this is especially important in Dco where the variability has been demonstrated to be so great. We believe laboratories should assure that prediction equationsfor DCOmatch the data produced in their facilities by comparing measurements made on 20 to 30healthy subjects with predicted values from several different equations.30 Of those who responded to the question on the adjustments made for ethnic differences, roughly half made no adjustment at all while the majority of those who did simply reduced white values by a fixed percentage which varied from institution to institu- tion. These results confirm our impression that there is no clear consensus on whether ethnic adjustments are necessary, which ethnic groups require some sort of adjustment, and how adjustment should be ap- proached if it is necessary. Ethnic differences in pulmonary function parameters are best documented in blacks and but studies of ethnic differ- ences in lung function are frequently confounded when the different ethnic groups are not studied with the same equipment and techniques.. Interlaboratory variability in lung function measurements due to technical factors is large in comparison to the magni- tude of ethnic differences and could mask or accen- tuate ethnic differences. However, comparisons among blacks, Asians and whites have generally shown whites to have the largest lung volumes, blacks to have the smallest, and Asians to have intermediate lung vol- umes, after adjustment for body size.m326.n.31People of mixed race have been found to have intermediate values. Flow rates have usually been found to be about the same in all three ethnic groups. Studies differ on whether or not proportional adjustments of white equations are an acceptable method of dealing with the difTeren~es.28A.~t~the moment, it is acceptable to use either proportional differences or population spe- cific equations to deal with ethnic dserences. It is 402 *Nine inst3utions used a numerid range for n o d than a prediction equation. not, however, acceptable to ignore them. Reference equations horn studies done are not used frequently in the United Canada. The summary equations reported standardization project of the European for Coal and Steel2 were rarely cited in The pattern of the prediction equations changed from that reported from a si conducted in 1968.= Equations from larger using current standardized methods and n nology are now more frequently cited. The replies to our survey suggest the avai automated systems-and manufacturers' d tings-may also influence the choice of Forty-seven institutions responding to our returned manufacturer-supplied compute ing the equations used. Three institutions used software provided by the manufacturer not know what equations it was based responses raise the possibility that the del the publication of good reference equations an use in practice is, in part, a function of the ti takes for such information to find its way into mu= facturers' computer software. 1 f `f REFERENCES ft 1 Hutchinson J. On capacity of lungs and on respiratory functions with view of establishing a precise and easy method of detecting disease by spirometer. Tr Med-Chir SOC1846:29:137 2 Quanjer Ph H, ed. Standardized lung function testing: raporC of the working party for standardizationof lung function tests d the european community for mal and steel. Bull Europ Phyrki path01Respir 1983; lQ(Suppl5):1-92 3 Training programs in respiratory disease. Am Rev Respir Db 1987; 136:497-518 4 Moms JF, Koski A, Johnson LC. Spirometric standamh fW healthy nonsmokingadults. Am Rev Respir Dis 1971; 10357-67 5 Morns JF, Temple W,Koski A. Normal values for the ratlo d one second forced expiratory volume to forced vital cappdt)r Am Rev Respir Dis 1973; 108:1oOQ(x3 6 Crapo RO, Morris AH, Gardner RM. Reference spirevalues using techniques and equipment that meet Al'S ~COUI- mendations. Am Rev Respir Dis 1981; 123:659-64 7 Knudson RJ, Lebowitz MD, Holberg CJ, Burrows B. in the normal maxi& expiratoryflowvolume curve with gmwtb and aging. Am Rev Respir Dis 1983; 127:725-34 8 Kory RC, Callahan R, Boren HG, Syner JC. The V e t e m Administration-Army cooperative study of pulmonary functioll: clinical spirometry in normal men. Am J Medicine 1961; 58 t ! \ i 4 \ i ii i i t i Normal standards for ventilatory wedge spirometer. Am Rev Respir JC, Warshaw R, Bernstein J, Selikoff IJ, ean and instantaneous expiratory flows, FVC FEV1: prediction equations fhm a probability sample of I MR. Respiratory function tests: normal i es and the prediction of normal results. ,Clayton PD, Nixon CR. Lung volumes in healthy nonsmoking adults. Bull Eur Physiopathol Resp 1982; I 18419-25 14 Boren HC, Kory RC, Syner JC. The Veterans Administrationh m y cooperative study ofpulmonary function: the lung volume t and its subdivisionsin normal men. Am J Medicine lh;4 1 : s 114 I 111 Bates D\! Macklem PT, Christie RV Respiratory function in disease, 2nd ed. Philadelphia:WB Saunden, 1971 16 Burrows B, Kasik JE, Niden AH, Barclay WR. Clinical usefulness of the single-breath pulmonary &sing capacity test. Am Rev Respir Dis 1961; &1:789-806 17 CrapoRO, Moms AH. Standardizedsingle breath normal values for carbon monoxide diffusing capacity. Am Rev Respir Dis 1Q811;23:185-89 l8 Miller A, Thornton JC, Warshaw R, Anderson H, Tierstein AS, SeW IJ. Single breath &sing capacity in a representative -pie of the population of Michigan, a large industrial state. Am Rev Respir Dis 1983; 127270-77 19 Gaensler EA, Wright GW. Evaluationof respiratory impairment. Arch Environ Health 1966; 12:14&89 20 CotesJ. Lung function: assessment and application in medicine. 1st ed.London: Blackwell Scientific Publications, 1965 %SdoIrinne Y. Single-breath pulmonary diffusing capacity. S a n d b p i r Dis 1976;96(suppl):1-84 I I II I I 22 McGrath MW,Thompson MC. The effect of age, body, size, and lung volume change on alveolar capillary permeability in man. J Physiol1959; 146572432 23 DuBois AB, Bottelho SY, Comroe JH Jr. A new method for measuring airway resistance in man using a body plethysmograph: Values in normal subjects and in patients with respiratory disease.J Clin Invest 1956;35.327-35 24 Briscoe WA, DuBois AG: The relationship between airway resistance, airway conductance,and long volume in subjects of Merent age and body size. J Clin Invest 1959;37:1279-85 25 Clausen JL, Zarins LP, eds. Pulmonary function testingguidelines and controversies. New York: Academic Press, 1982:165-85 26 Corey PN, Ashley MJ, Chan-Yeung M. Racial differences in lung function: search for proportional relationships. J Occup Med 1979;21:395-98 27 Woolcock AJ, Colman MH, Blackburn CRB. Factors decting normal values for ventilatory lung function. Am Rev Respir Dis 1972; 106:6!32-709 28 Schoenberg JB, Beck CJ, Bouhuys A. Growth and decay of pulmonary function in blacks and whites. Resp Physiol 1978; 33367-93 29 Clausen JL, Crapo R, Gardner R. Interlaboratory comparisons of pulmonary function testing (Abstract). Am Rev Respir Dis 1984; (Suppl:A37) 30 Crapo RO, Gardner RM. DLCO reference equations: a perspective. Am Rev Respir Dis 1986; 134:856 31 Rossiter CE, Weill H. Ethnic differences in lung function: evidence for proportional differences. Int J Epidemiol 1974; 3:55-61 32 Leiner CC. Abramowitz S, Small MJ. Pulmonary function testing in laboratories asscciated with residency h-aining programs in pulmonary diseases. Am Rev Respir Dis 1969;100:240- 44 33 Baldwin E, Cournand A, Richards DW Jr. Pulmonary ins&- ciency: physiological classification, clinical methods of analysis, standard values in normal subjects. Medicine 1948;27243-78