A multifaceted intervention to REduce Antimicrobial prescribing in Care Homes (REACH): a non-randomised feasibility study and process evaluation

Problem

There are concerns about the level of prescribing of in care homes for older people. A Canadian study found an intervention using a decision-making algorithm and small group interactive training was useful in reducing the use of antimicrobials in Canadian care homes. The aim of this study was to update and adapt the Canadian intervention for use in the United Kingdom and to test the feasibility of delivering it in care homes prior to planning a cluster randomised trial.

Approach

The original Canadian decision-making algorithm was updated and adapted by using best current evidence (literature review), an expert consensus panel, focus groups with care home staff and relatives of residents, interviews with GPs, and internal review by the study team. The small group interactive training programme was adapted informed by current best practice. The aim of the decision-making algorithm and the associated training was to help care home staff decide when to contact a GP when they suspected a resident had an infection (e.g. Urinary tract infection (UTI), Respiratory tract infection (RTI) or a Skin infection (SI)). Six care homes participated in the study: three in Northern Ireland (Belfast) and three in England (West Midlands), with two nursing and one residential home in each area. Staff in the homes were provided with training prior to a six-month period of implementation. Our primary outcomes were process related and we carried out an extensive mixed methods process evaluation underpinned by Normalization Process Theory (NPT). NPT informed the development of pre and post intervention focus group and interview discussion guides, observational field notes throughout the intervention, and the analysis of the resulting data. We also aimed to test the feasibility of collecting effectiveness data. Our primary quantitative outcome is defined daily doses (DDDs) of antimicrobials dispensed to the home: obtained from pharmacy records. In addition, we collected information on basic demographics, use of the decision-making algorithm, use of hospital services, contacts with health and social care professionals, adverse events and health economic data.

Findings

The implementation period ended in late 2017. Regarding feasibility, we successfully recruited six care homes, developed and implemented the decision-making algorithm, delivered training to care home staff, and collected outcome and process related data. Analysis of these data is ongoing. Our process evaluation outcomes will be mapped against the four components of NPT giving an indication of the extent that the intervention could be embedded into practice. Early results indicate a meaningful reduction in antibiotic following the implementation of the intervention.

Consequences

We have a model of best practice for considering medical intervention for suspected infections in care home residents and the data needed to support an application for a future cluster randomised trial.