The purpose of this study was to develop a tool for detecting connective tissue diseases using subjective symptoms and background data in order to establish appropriate criteria to advise patients in a community pharmacy setting.
The medical records of patients admitted to the medical ward of the Internal Medicine Department at Fujita University Hospital, Japan over a 5-year period between December 1995 and January 2001, were retrospectively reviewed. Fifty-eight patients with a fever of over 38. 5t and/or exanthema were found to exhibit drug allergies, and thus were included in this study. Of these, 22 patients were previously diagnosed to have connective tissue diseases. A logistic regression analysis was performed to examine the factors associated with the onset of connective tissue diseases.
A univariate logistic regression analysis for several variables showed the factors associated with connective tissue diseases to be age (per year relative risk) (Odds Ratio [OR] : 1.01, 95% confidence interval [95% CI] : 1.00-1.07), ranged onset age (OR : 4.12, 95% CI : 1.48-15.03), gender (OR : 5.07, 95% CI : 1. 27-20.22), anemia (OR : 4.57, 95% CI : 1.18-17.72), fever (OR : 0.32, 95% CI : 0. 11-0.97), and anti-nuclear antibody (OR : 5.14, 95% CI : 1.28-20.68). A multivariate analysis of the former four factors, which may be determined for each patient through interviews in a community pharmacy setting, revealed 3 factors ; namely, the range of the onset age (OR : 6.59, P=0.0006), gender (OR : 5.27, P=0.034) and anemia (OR : 5.21, p=0.041), to be the most important factors for predicting the onset of connective tissue disease. An equation to evaluate the onset of connective tissue diseases was created using the model coefficients for these factors. This equation was thereafter applied to these 58 patients. Predicting diseases with this equation revealed a sensitivity of 81.8%, a specificity of 77.8%, a predictive value for a positive test of 69.2%, and an OR of 15. 8.
A logistic regression analysis of several variables revealed 3 factors which were considered to increase the probability of connective tissue diseases onset in patients with fever and/or exanthema. Screening for connective tissue diseases using these factors may thus be a useful tool for evaluating diseases in a community pharmacy setting.
