Bronwen Holloway: Acute febrile illness, antibiotic use, and the role of diagnostics to target treatment in India

Abstract

    Aim: This thesis examined the causes of acute febrile illness (AFI), the current use of antibiotics and diagnostics, and evaluated the diagnostic accuracy of C-Reactive Protein (CRP) to differentiate bacterial from non-bacterial causes of AFI in children and adult outpatients at R.D. Gardi Medical College hospital, in Ujjain, India.

  Methods: A prospective cross-sectional study of children and adult outpatients with fever ≥37.5°C, or history of fever in the past 48 hours, and no signs of severe illness. Patient history, physical examination, culture, rapid diagnostic tests, and follow-up after one week was performed for all patients. Whole blood and urine were collected from all patients, and symptom based nasopharyngeal throat swabs, stool, and skin/ear/joint/aspirate specimens. Fever was classified as bacterial or non-bacterial based on microbiology and laboratory results together with an expert panel review. Data on antibiotic use before, during, and after enrolment was described by Anatomical Therapeutic Chemical classification and AWaRe categories. Serum CRP levels were measured and the performance characteristics for CRP to differentiate between bacterial and non-bacterial AFI were calculated. The area under the receiver operating curve (AUC), sensitivity, specificity, positive and negative predictive values, and likelihood ratios were estimated using 10, 20, 40, 60 and 80 mg/L thresholds. A rapid ethnographic qualitative study on the utilization of diagnostics was conducted using unstructured observations, structured observations and 43 semi-structured interviews. Interview data were analyzed using inductive thematic analysis.

   Results: Of 1000 outpatients, 24.4% were categorized as bacterial; 71.8% non-bacterial; and 3.8% an undetermined cause of fever. Throughout the course of AFI, 41.0% of patients received one or more antibiotics. The leading contributors to total antibiotic volume were macrolides. ‘Watch’ antibiotics accounted for 72.3%, 52.7%, and 32.6% of encounters before, during and after the outpatient visit. The overall median CRP was low but higher in the group classified as bacterial compared to non-bacterial (3.6 mg/L vs. 2.7 mg/L, p<0.0001, respectively). The AUC was low at 0.60 (95% CI 0.56 - 0.65). Caregivers trusted and understood the importance of diagnostics, but their acceptance wavered depending on the severity of illness and preference to treat their child directly with medicines. Caregivers struggled to get tests done and return for follow-up due to costs, delays in testing, further complicated by travel time, distance and competing priorities.

 Conclusion: This thesis highlights the challenges in determining the cause of AFI. Over,under, and inappropriate use of antibiotics throughout the course of AFI are of major concern.The organization of diagnostic services, together with direct and indirect costs, hinder caregiversfrom utilizing diagnostics. CRP is too weak as a single indicator of bacterial infection to safelysupport physicians in making treatment decisions for febrile outpatients in India.