A Process Evaluation of an Antibiotic Stewardship Intervention for Urinary Tract Infections in Nursing Homes

Abstract

ObjectivesTo assess the internal and external validity of a cluster randomized controlled trial (cRCT) evaluating a decision tool with supportive interventions for the empirical treatment of urinary tract infections (UTIs) in nursing homes (NHs), and to identify facilitators and barriers in implementing this antibiotic stewardship intervention. DesignMixed-methods process evaluation study. Setting and ParticipantsPhysicians, nursing staff, client council members, and residents of Dutch NHs. MethodsWe used cRCT data of the ANNA study (Antibiotic Prescribing and Non-prescribing in Nursing Home Residents With Signs and Symptoms Ascribed to Urinary Tract Infection). In addition, we sent out an online evaluation questionnaire, conducted semistructured interviews with physicians and nursing staff, and consulted client council members. ResultsInternal validity was lowered: control group physicians participated in several non–study-related activities regarding UTI. External validity was good: almost all intervention components had a high fidelity (52%-74%) and were perceived as relevant (physicians: 7.2-8.6 of 10, nursing staff: 6.5-8.5 of 10) and feasible (physicians: 7.5 of 10, nursing staff 6.4 of 10), with feasibility for residents with dementia and urine incontinence needing attention. The most common reason for deviating from the advice generated by the decision tool was an unclear illness presentation. Identified facilitators to implementation were confidence in the intervention, repeated intervention encounter, and having “champions” in the NH. Barriers were limited involvement of nursing staff, unstable nursing teams, residents' and representatives’ belief that antibiotics should be prescribed, and a low antibiotic prescribing threshold within the NH culture. Conclusions and ImplicationsLowered internal validity may have reduced the study effect. Attention should be paid to the feasibility of the intervention in residents with dementia and urinary incontinence. Improvement opportunities for implementation were higher nursing staff involvement and repeated intervention offering.