Frequency and Causes of Combined Obstruction and Restriction Identified in Pulmonary Function Tests in Adults Enrique Diaz-Guzman MD, Kevin McCarthy RCPT, Alan Siu, and James K Stoller MD MSc FAARC BACKGROUND: The frequency of combined obstruction and restriction identified in pulmonary function tests has not been well described. Moreover, although the causes of combined-obstruction- and-restriction patterns are known, the frequency of the various etiologies has received little at- tention. METHODS: We retrospectively reviewed medical records and surveyed pulmonologists. RESULTS: 43,212 PFT sessions were evaluated, which yielded 130 patients who satisfied our criteria for spirometry evidence of combined obstruction and restriction. Their demographic fea- ,tures were: mean ؎ SD age 54 ؎ 14 y, 51% male, mean ؎ SD body mass index 28.8 ؎ 6.7 kg/m2 mean ؎ SD height 174 ؎ 9 cm (men) and 162 ؎ 7 cm (women). The causes of combined obstruction (%38 ,49 ؍ and restriction were classified as either a pulmonary parenchymal disorder (Group A, n .(%48 ,63 ؍ or a combination of pulmonary parenchymal and non-pulmonary diseases (Group B, n In 18 patients (14%) no clear etiology of combined obstruction and restriction could be determined. The most common pulmonary disease was chronic obstructive pulmonary disease (45/130, 35%), and the most common non-parenchymal disease was congestive heart failure (27/130, 21%). We electronically sent a survey to 55 pulmonary physicians, of whom 30 (55%) responded. The re- spondents estimated that combined obstruction and restriction occurs in approximately 20% of all the pulmonary function tests performed in their practices and that pulmonary parenchymal dis- eases were responsible for 35% of all instances of combined obstruction and restriction. CONCLU- SIONS: Combined obstruction and restriction occurs infrequently and is more commonly caused by a combination of pulmonary parenchymal and non-pulmonary disorders. Pulmonologists’ impres- sions regarding the frequency and causes are generally discordant with the observed frequencies. Key words: pulmonary function tests; spirometry; airway obstruction; lung disease, interstitial. [Respir Care 2010;55(3):310–316. © 2010 Daedalus Enterprises] Introduction sify patterns of chest disease that may then facilitate di- agnosis and therapy decisions. Spirometry is commonly By providing an objective physiologic assessment, pul- used to identify patients with obstructive airway disease, monary function tests (PFTs) allow the clinician to clas- defined by a decreased forced expiratory volume in the first second (FEV1) and a decreased ratio of FEV1 to forced vital capacity (FVC).1 At the same time, limitations of At the time of this study Enrique Diaz-Guzman MD was affiliated with spirometry include its inability to establish the presence of the Department of Pulmonary and Critical Care Medicine, Respiratory a concomitant restrictive disorder or a pattern of combined Institute, Cleveland Clinic, Cleveland, Ohio; he is now with the Division (also called mixed) obstruction and restriction. of Pulmonary, Critical Care, and Sleep Medicine, University of Kentucky College of Medicine, Lexington, Kentucky. Kevin McCarthy RCPT, Alan Siu, and James K Stoller MD MSc FAARC are affiliated with the De- partment of Pulmonary, Allergy, and Critical Care Medicine, Respiratory Institute, The Cleveland Clinic, Cleveland, Ohio. Correspondence: James K Stoller MD MSc FAARC, Department of Pul- Dr Stoller has disclosed relationships with Boehringer-Ingelheim, Talecris, monary, Allergy, and Critical Care Medicine, Respiratory Institute, A90, CSL Behring, Asmatx, Grifols, Baxter, Philips-Respironics. The other The Cleveland Clinic, 9500 Euclid Avenue, Cleveland OH 44195. E- authors have disclosed no conflicts of interest. mail: [email protected]. 310 RESPIRATORY CARE • MARCH 2010 VOL 55 NO 3 FREQUENCY AND CAUSES OF COMBINED OBSTRUCTION AND RESTRICTION In this context and based on American Thoracic Soci- AS) after reviewing the text of the electronic medical record ety/European Respiratory Society recommendations that and radiology and pathology reports. lung volumes should be measured to establish restriction We developed a Web-based survey at SurveyMonkey. when spirometry indicates a decreased FVC,2,3 combined com, and via e-mail asked all the pulmonary/critical-care measurement of spirometry and static lung volumes is com- attending physicians and fellows in the Department of Pul- mon in clinical practice, though the actual frequency of monary, Allergy, and Critical Care Medicine of The Cleve- these tests and the rate of occurrence of combined obstruc- land Clinic, and some pulmonologists who had trained at tion and restriction have received little attention. Further- The Cleveland Clinic but were practicing at other institu- more, although the differential diagnosis of combined ob- tions, to participate in the survey. struction and restriction is known, information about pulmonary physicians’ familiarity with this differential di- Results agnosis and about the frequency of combined obstruction and restriction is also sparse. During the study period, 43,212 PFT sessions were per- To address this gap, we undertook the current study to formed. Of those, 7,506 (17%) included simultaneous spi- determine the frequency and causes of combined obstruc- rometry and body plethysmography. 2,203 (29%) patients tion and restriction, in the pulmonary function laboratory satisfied our criteria for a restrictive pattern, and 151 (2%) of a tertiary-care center. We also conducted a survey to satisfied our criteria for combined obstruction and restric- assess pulmonologists’ perceptions about the frequency tion. Figure 1 depicts the distribution of PFTs and corre- and etiologies of combined obstruction and restriction, and sponding number of patients (as some patients underwent we compared the respondents’ impressions about the fre- more than one PFT session during the study period). quencies of the causes to the actual frequencies. Altogether, 4,767 patients underwent simultaneous spi- rometry and body plethysmography. Among these, 130 pa- tients (2.7%) met our criteria for combined obstruction and Methods restriction. The remaining 24,723 patients did not com- plete lung-volume measurements at the same visit as spi- The study protocol was approved by the institutional rometry. Among those 24,273 patients, 11% (2,670 pa- review board of The Cleveland Clinic. tients) had a low FEV1/FVC and a low FVC, of whom 319 We retrospectively reviewed PFT results from consec- underwent lung-volume measurements at a later date; 30 utive adult patients referred to the pulmonary function (9.4%) of those patients showed restriction (ie, TLC was laboratory of The Cleveland Clinic between January 1, below the lower limit of normal). 2000, and December 31, 2003. For patients who under- Study subjects’ demographic features were as follows: went more than one PFT session during that interval we mean Ϯ SD age 54 Ϯ 14 y, 51% male, 87% white, 12% considered only the results from the first PFT session. African-American, mean Ϯ SD body mass index Spirometry (MasterLab Pro, Jaeger, Wu¨rzburg, Germany) 28.8 Ϯ 6.7 kg/m2, mean Ϯ SD height 174 Ϯ 9 cm (men) and plethysmographic lung-volume measurements (Mas- and 162 Ϯ 7 cm (women). Fifty-two percent (68/130) had terScreen Body PFT, Jaeger, Wu¨rzburg, Germany) were a history of cigarette smoking. performed in accordance with the recommendations of the The causes of combined obstruction and restriction were American Thoracic Society/European Respiratory Soci- classified as either a pulmonary parenchymal disorder ety.4,5 We used the post-bronchodilator values in the anal- (Group A) or a combination of parenchymal and non- ysis. The percent-of-predicted values were determined with pulmonary diseases (Group B). Approximately half of all reference equations for spirometry,6 lung volumes,7 and instances of combined obstruction and restriction were at- 8 diffusing capacity of the lung for carbon monoxide (DLCO). tributed to a combination of parenchymal and non-pulmo- A pattern of obstructive airway disease was defined as nary diseases (Table 1). The most common parenchymal an FEV1/FVC below the 5th percentile of the predicted disease was chronic obstructive pulmonary disease value.4,9 A restrictive pattern was defined as a plethysmo- (COPD), and the most common non-parenchymal disease graphically measured total lung capacity (TLC) below the was congestive heart failure. In the group with combined 5th percentile of the predicted value.6,7 A pattern of com- conditions, obesity (mean body mass index 42.1 Ϯ 4.9 kg/ bined obstruction and restriction satisfied both the latter m2) and a history of thoracic surgery were the most com- criteria. mon features. In 14% (18/130) of the subjects the cause of In the majority of subjects the causes of abnormal PFT the combined obstruction and restriction could not be de- results were determined by the physician who ordered the termined from the medical record. PFT and were documented in the medical record. For the Groups A and B had similar FVC, FEV1, FEV1/FVC, remainder of the subjects the causes of the PFT abnormal- and TLC values. Compared to patients in Group B, pa- ities were determined by two of the investigators (EDG, tients with pulmonary parenchymal disorders had lower RESPIRATORY CARE • MARCH 2010 VOL 55 NO 3 311 FREQUENCY AND CAUSES OF COMBINED OBSTRUCTION AND RESTRICTION ϭ ϭ Fig. 1. Distribution of pulmonary function test results. The n values represent