Febrile patients admitted to remote hospitals in Northeastern Kenya: seroprevalence, risk factors and a clinical prediction tool for Q-Fever. Issue 1 (December 2016)

• Record Type: Journal Article
• Authors: Njeru, J.; Henning, K.; Pletz, M.; Heller, R.; Forstner, C.; Kariuki, S.; Fèvre, E.; Neubauer, H.
• Is Part Of: BMC infectious diseases. Volume 16:Issue 1(2016)
• Page Start: 1
• Page End: 15
• Publication Date: 2016-12
• Subjects: Seroprevalence -- Epidemiology -- Coxiella burnetii -- Q fever -- Kenya; Communicable diseases -- Periodicals; Sexually Transmitted Diseases -- Periodicals; 616.905
• DOI: 10.1186/s12879-016-1569-0
• Languages: English
• ISSNs: 1471-2334

Abstract

Abstract Background Q fever in Kenya is poorly reported and its surveillance is highly neglected. Standard empiric treatment for febrile patients admitted to hospitals is antimalarials or penicillin-based antibiotics, which have no activity againstCoxiella burnetii . This study aimed to assess the seroprevalence and the predisposing risk factors for Q fever infection in febrile patients from a pastoralist population, and derive a model for clinical prediction of febrile patients with acute Q fever. Methods Epidemiological and clinical data were obtained from 1067 patients from Northeastern Kenya and their sera tested for IgG antibodies againstCoxiella burnetii antigens by enzyme-linked-immunosorbent assay (ELISA), indirect immunofluorescence assay (IFA) and quantitative real-time PCR (qPCR). Logit models were built for risk factor analysis, and diagnostic prediction score generated and validated in two separate cohorts of patients. Results Overall 204 (19.1 %, 95 % CI: 16.8–21.6) sera were positive for IgG antibodies against phase I and/or phase II antigens orCoxiella burnetii IS1111 by qPCR. Acute Q fever was established in 173 (16.2 %, 95 % CI: 14.1–18.7) patients. Q fever was not suspected by the treating clinicians in any of those patients, instead working diagnosis was fever of unknown origin or common tropical fevers. Exposure to cattle (adjusted odds ratio [aOR]: 2.09, 95 % CI: 1.73–5.98), goats (aOR: 3.74, 95 % CI: 2.52–9.40), and animal slaughter (aOR: 1.78, 95 % CI: 1.09–2.91) were significant risk factors. Consumption of unpasteurized cattle milk (aOR: 2.49, 95 % CI: 1.48–4.21) and locally fermented milk products (aOR: 1.66, 95 % CI: 1.19–4.37) were dietary factors associated with seropositivity. Based on regression coefficients, we calculated a diagnostic score with a sensitivity 93.1 % and specificity 76.1 % at cut off value of 2.90: fever >14 days (+3.6), abdominal pain (+0.8), respiratory tract infection (+1.0) and diarrhoea (−1.1). Conclusion Q fever is common in febrile Kenyan patients but underappreciated as a cause of community-acquired febrile illness. The utility of Q fever score and screening patients for the risky social-economic and dietary practices can provide a valuable tool to clinicians in identifying patients to strongly consider for detailed Q fever investigation and follow up on admission, and making therapeutic decisions.