Attributable Patient Cost of Antimicrobial Resistance: A Prospective Parallel Cohort Study in Two Public Teaching Hospitals in Ghana

Abstract

Objective The aim of this study was to evaluate the attributable patient cost of antimicrobial resistance (AMR) in Ghana to provide empirical evidence to make a case for improved AMR preventive strategies in hospitals and the general population. Methods A prospective parallel cohort design in which participants were enrolled at the time of hospital admission and remained until 30 days after the diagnosis of bacteraemia or discharge from the hospital/death. Patients were matched on age group (± 5 years the age of AMR patients), treatment ward, sex, and bacteraemia type. The AMR cohort included all inpatients with a positive blood culture of Escherichia coli or Klebsiella spp., resistant to third-generation cephalosporins (3GC), or methicillin-resistant Staphylococcus aureus (MRSA). We matched the AMR cohort (n = 404) with two control arms, i.e., patients with the same bacterial infections susceptible to 3GC or S. aureus that was methicillin-susceptible (susceptible cohort; n = 152), and uninfected patients (uninfected cohort; n = 404). Settings were Korle-Bu and Komfo Anokye Teaching Hospitals, Ghana. The outcome measures were the length of hospital stay (LOS) and the associated patient costs. Outcomes were evaluated from the patient perspective. Results From a total of 5752 blood cultures screened, 1836 participants had growth in blood culture, of which, based on our inclusion criteria, 426 were enrolled into the AMR cohort; however, only 404 completed the follow-up and were matched with participants in the two control cohorts. Patients in the AMR cohort stayed approximately 5 more days (95% confidence interval [CI] 4.0–6.0) and 8 more days (95% CI 7.2–8.6) compared with the susceptible and uninfected cohorts, respectively. The mean extra patient cost due to AMR relative to the susceptible cohort was US$1300 (95% CI 1018–1370), of which about 30% resulted from productivity loss due to presenteeism and absenteeism from work. Overall, the estimated annual patient cost due to AMR translates to about US$1 million and US$1.4 million when compared with the susceptible and uninfected cohorts, respectively. Conclusion We have shown that AMR is associated with a significant excess LOS and patient costs in Ghana using prospective data from two public tertiary hospitals. This calls for infection prevention and control strategies aimed at mitigating the prevalence of AMR.