Interobserver agreement on signs and symptoms of patients with acute febrile illness.

PURPOSE

To assess the interobserver agreement on clinical history and physical examination when using a semi-structured questionnaire to evaluate patients with an acute febrile illness (AFI).

METHODS

A cross-sectional study was conducted with outpatients aged 12 years and over, presenting with an AFI defined as fever up to 7 days and no evident focus of infection. Clinical data were collected independently by two physicians using a semi-structured questionnaire. Interobserver agreement was estimated using kappa coefficients with a 95% confidence interval (CI).

RESULTS

A total of 140 patients (age range 13-73 years; 56.4% females) were enrolled. All symptoms showed weighted kappa values significantly greater than 0.6, indicating an at least substantial agreement. As most physical signs were infrequent and of mild intensity, they were recoded and analyzed as absent/present. Of the signs with prevalence ≥15%, exanthema, pallor, lymph node enlargement, and eye congestion showed agreements significantly greater than 0.6, while kappa confidence limits for pharyngeal erythema and dehydration included values classified as regular.

CONCLUSIONS

High agreement was observed for most of the clinical data assessed, and symptom grading was feasible. Some physical findings were rare and their inclusion in a structured form may not be justified in this setting. The questionnaire application showed good reliability for the most frequent signs and symptoms and may prove to be useful at gathering data for surveillance and research at sentinel sites.