Risk-Factors of Crimean-Congo Haemorrhagic Fever Across Six Tertiary Hospitals in Afghanistan from 2022 to 2025: Reported Exposures, Healthcare-Worker Clustering, and Ribavirin Treatment-Timing Analyses

Introduction Crimean Congo haemorrhagic fever (CCHF) is endemic in Afghanistan, but multi-hospital data characterising reported exposure patterns, healthcare-worker clustering, and ribavirin-mortality associations after accounting for treatment timing are lacking. Methods A retrospective cohort study included 737 laboratory-confirmed CCHF patients admitted to six tertiary hospitals in Afghanistan between January 2022 and December 2025. Reported exposures were assessed by occupation and tick-contact history. Clustering of healthcare-worker (HCW) infections during high caseload was assessed by permutation testing. Ribavirin-mortality association was assessed using multivariable Cox regression adjusted for clinical severity, age, sex, hospital, and comorbidity, with sensitivity analyses addressing treatment timing immortal-time bias. Results Among 737 patients, the median age was 35 years and 78.3% were male. The hospitalization fatality ratio (HFR) was 27.7% (204/737). Of 574 patients with recorded occupation: 33.6% were livestock handlers, 12.9% slaughterhouse workers or butchers, 24.9% household meat preparers, and only 11.1% reported tick exposure. Nineteen patients were HCWs, and 18 became infected during periods of concurrent CCHF admissions, and clustering exceeded chance (p=0.012), consistent with healthcare-associated exposure during surges. Personal protective equipment (PPE) was used in 728 (98.6%) admissions, but its composition was incomplete: 532 (72.2%) comprised mask and gloves without a gown, and only 205 (27.8%) with mask, gloves, and gown. Isolation rooms were available for only 274 of 737 admissions (37.2%). Among 727 patients in the ribavirin analysis, 85.0% received treatment. Conventional analysis suggested lower mortality (adjusted HR 0.57) but affected by immortal-time bias. The association weakened for same-day treatment versus no treatment (HR 0.63, 95% CI 0.41-0.96; sensitive to model specification) and was absent in time-varying analysis (HR 0.93). Conclusion CCHF in this hospitalised Afghan cohort was commonly associated with reported animal and livestock-related exposure, while healthcare-worker infections clustered during outbreak surges. Isolation capacity and caseload may be important correlates of healthcare-associated risk alongside PPE availability. Ribavirin findings were sensitive to treatment-timing adjustment and cannot establish a causal survival benefit.