Self-reported antibiotic stewardship practices and perceived barriers and facilitators among healthcare workers in three tertiary hospitals in Sierra Leone: a convergent mixed- methods study

Abstract Background: This study aimed to describe self-reported antibiotic management practices and perceived barriers and facilitators among healthcare workers in three tertiary hospitals in Sierra Leone and to integrate survey and interview findings to identify priorities for hospital antimicrobial stewardship. Methods: This study employed a mixed-methods design, triangulating quantitative data from a cross-sectional survey with qualitative data from key informant interviews. Data were collected in clinical settings at three tertiary referral hospitals in Freetown, Sierra Leone: Connaught Hospital, Princess Christian Maternity Hospital, and Ola During Children's Hospital. The survey included 214 respondents, comprising doctors, nurses, and pharmacists, representing 55.7% of the 384 distributed questionnaires, along with 15 purposively selected key informants. Results: Clinical guidelines exist across the three facilities, but informed prescribing is influenced by structural and social factors beyond availability. Qualitative interviews revealed concerns about outdated guidelines and whether they applied to the local pathogen picture. Senior physicians whose practices predate stewardship policies favour an experience-based culture that can run counter to updated guidelines. Quantitative data showed that adherence varied across facilities and cadres, with junior prescribers mirroring patterns among seniors. Participants showed awareness of antimicrobial resistance (AMR) risks; however, this did not translate to their prescribing behaviour. Interview data identified a gap: prolonged turnaround times, limited microbiology capacity, treatment availability, and affordability meant that empirical treatment was clinically based. Quantitative analysis indicated that facilities without microbiology lab functions had higher rates of broad-spectrum empirical prescribing. Interviews suggested prescribers may ignore guidelines and prescribe broad-spectrum antibiotics as the “magic bullet” to address these challenges. Participants also described choosing free broad-spectrum agents in settings with unreliable supply chains, not for clinical necessity, but because they were the only available treatment. Antibiotic escalation was also linked in interviews to potential patient severity or deterioration and social expectations from patients and families. Conclusions: Healthcare workers across the three hospitals reported high awareness of AMR and stewardship-relevant challenges; however, antibiotic decisions were shaped by diagnostic scarcity, medicine availability, affordability, and professional norms. Stewardship efforts should prioritise visible local guidance, stronger microbiology and diagnostic turnaround, reliable antibiotic supply, and repeated multidisciplinary training and feedback.