PURPOSE: A comparison of predicted and lower limit of normal (LLN) values for diffusion capacity of carbon monoxide (DLCO) by two commonly used reference equations with implications on normalcy and severity.
METHODS: We reviewed our PFT database from 2005 to 2008 and calculated percent predicted based on reference equation by Crapo and Miller for men and women. The actual values were also assigned normal or abnormal based on the lower limit of normal (LLN) values for each equationt.
RESULTS: The database had a total of 2249 (1082 male and 1167 female) patients. 283/ 1082 (26.1%) males and 354 /1167 (30%) females who had normal DLCO by using Miller’s reference equation for LLN had abnormal DLCO by Crapo’s equation for LLN.143/1082 (13.22%) of males who had DLCO >40 % by Miller had DLCO <40% by Crapo. Similarly, 76/1167 (6.51%) females who had DLCO >40% by Miller had DLCO <40% by Crapo.The DLCO predicted using reference equation by Crapo et al resulted in higher values for both males (mean difference 5.94, SD 2.29, p value < 0.001) and females (mean difference 3.13, SD 0.84, p value < 0.001). The values were also high for LLN for males (mean difference 7.08, SD 2.04, P value < 0.001)and females (mean difference 4.87, SD 0.84, P value < 0.001).
CONCLUSION: The predicted and lower limit of normal values for DLCO were significantly higher by Crapo et al’s reference equation as compared to Miller et al in both males and females. This resulted in lower percent predicted values affecting degree of severity and substantial misclassification as abnormal in both sexes by Crapo’s reference equation.
CLINICAL IMPLICATIONS: Pulmonary function laboratories should indicate which reference equation they used while giving results for the DLCO measurements. An abnormal value can thus be normal based on which prediction equation has been used. Uniform use of the reference equation can solve this problem.
DISCLOSURE: Madan Joshi, No Financial Disclosure Information; No Product/Research Disclosure Information