Introduction: Although the Spanish version of the Zung's Self-Rating Depression Scale (SDS) is widely used, there are no studies about its validity as a diagnostic test in primary health care patients.
Methods: In a first phase, a sample of 350 consecutive primary care patients was assessed with the SDS. In a second phase, a subsample composed by all the positive test results and 1/3 of the negatives selected at random, was assessed with the modules of current Major Depressive Episode and Dysthymia of the Structured Clinical Interview for DSM-IV. Specific methods to avoid verification bias were used. Prevalence, sensitivity and specificity, predictive values, Receiver Operating Characteristic (ROC) curve, and Stratum Specific Likelihood Ratios (SSLR) were calculated.
Results: Prevalence estimations of major depression and dysthymia were 14,7% (IC95%: 10,7-18,7%) and 4,6% (IC95%: 2,4%-6,8%) respectively. Sensitivity and specificity to detect both diagnoses were 0,95 (IC95%: 0.87-1) and 0,74 (IC95%: 0,68-0,79). Area under ROC curve was 0,93. SSLR for scoring < 50 led to a post-test probability of 0.01. In the stratum with scoring > 69 the SSLR generated a post-test probability of 0.96. Less conclusive results were obtained by intermediate strata.
Conclusions: The SDS is effective in primary care patients and shows operating characteristics comparable to other depression assessment scales. SSLR provides practical information to estimate the probability of suffer a depressive disorder in individual patients.