Diagnostic criteria clusters for diagnosis of chronic obstructive lung disease in primary care

Objective: to determine diagnostic strength of clusters of criteria based on medical history and physical examination findings commonly used in daily practice for diagnosis of chronic obstructive lung disease (COPD).Method: Study was conducted with 195 patient admitted to Izmir Basin Sitesi Family Health Center and out patient clinic of Dokuz Eylul University Faculty of Medicine Department of Pulmonary Medicine who were above 40 years of age with cough and met the inclusion criteria. All patients were subjected to the questionnaire, physical examination and spirometric evaluation. After the spirometric evaluation, sensitivities, specificities, positive and negative likelihood ratios (LR) and post test probabilities of all variables were calculated for COPD positive and COPD negative groups. Fischer?s Exact Test SPSS 11.0 was used for hypothesis testing. Criteria with p values lower than 0.05 were divided into three groups namely weak, moderate and strong based on their impact on post test probability. Post test probabilities of different variable groups were calculated by using chain LR method.Results: Smoking more than 20 package-years was an important determinant of COPD. Package-years show positive correlation with COPD. Wheezing, intercostal retractions and increment of tactile fremitus were found to be strong criteria where as sputum, exertion dyspnea and shortness of breath increasing with time were found to be weak criteria. If all of the nine weak criteria or three of the five moderate criteria, or two of the three strong criteria are positive, post test probabilities were high enough to suggest diagnosis.Conclusion: Primary care physicians can come closer to the diagnosis of COPD by using their signs and symptoms and clusters of criteria that facilitates their clinical decision making.Key words: COPD, diagnostic test, likelihood ratio, post test probability.