Establishment of a scoring system for diagnosing diabetic patients with smear-negative pulmonary ;tuberculosis / 实用医学杂志

ABSTRACT

Objective To establish a diagnostic scoring system for diabetic patients with smear-negative pulmonary tuberculosis , and then to improve the rapidity and accuracy of clinical diagnosis and save medical expenses. Methods A case-control study was applied. 150 diabetic patients with smear-positive pulmonary tuberculosis who on initial treatment were assigned to a study group; 150 patients with lung infection were recruited as a study group. The data on general status, symptoms, chest X-ray manifestation, and laboratory examinations was collected. Univariate logistic regression analysis was used to gain significant indexes for multiple logistic regression analysis. β-coefficients derived from the independent predictors in our logistic regression model was applied to develop a scoring system. Results Toxic symptoms of tuberculosis, cough, upper lung, cavity, multiple lung field , and PPD positive entered into the final multipie logistic regression model , and the scoring system was accordingly established. The patient with a score of more than 11 had higher probability of TB , while those with a score of smaller than 11 were not likely to have TB. Conclusion The scoring system can be used as a predictive tool in diagnosis of diabetes mellitus complicated by smear-negative pulmonary tuberculosis , helping diagnose active tuberculosis rapidly.