Predicting severity in spotted fever in India: A retrospective cohort study

<h2>Abstract</h2><h3>Problem considered</h3> Spotted fever, a neglected rickettsial illness in India, can present with life-threatening complications, yet early predictors of severity remain poorly defined. <h3>Methods</h3> We conducted a retrospective cohort study of 117 adult patients with spotted fever admitted to a tertiary care centre in Karnataka, India (2017–2024). Spotted fever was diagnosed based on a positive Weil-Felix test (OX-2 and/or OX-19) with exclusion of other common tropical infections. Severe disease was defined as the presence of oxygen requirement, vasopressor use, dialysis, or altered sensorium. Clinical and laboratory variables were compared between severe and non-severe groups, and multivariable logistic regression identified independent predictors. <h3>Results</h3> Severe disease occurred in 31.6% (37/117). On multivariable analysis, <b>shorter duration of illness at presentation to the hospital</b> was independently associated with severe disease <b>[aOR 0.70 (95%CI: 0.57–0.86)]</b>, while <b>higher serum creatinine</b> also predicted severity <b>[aOR 1.29 (95%CI: 1.00–1.66)]</b>. <b>Lower serum albumin</b> showed a strong association with severe disease <b>[aOR 0.16 (95%CI: 0.07–0.37)]</b>. Although platelet count demonstrated an inverse trend, the association did not reach statistical significance <b>[aOR 0.92 (95%CI: 0.83–1.03)]</b>. The model showed good calibration (Hosmer–Lemeshow p=0.128) and an overall accuracy of 77.7%. <h3>Conclusion</h3> In spotted fever, simple and accessible markers, shorter illness duration, renal dysfunction, and hypoalbuminemia, can help predict severity and guide early referral, especially in resource-limited settings. These findings merit validation in prospective cohorts.