Is evaluation of centralised nurse-led management of common STIs in a cluster randomised trial feasible and acceptable?

Problem

Around half of Chlamydia trachomatis (CT) and Neisseria gonorrhoea (NG) testing in England is amongst asymptomatic individuals attending Primary Care (PC) for unrelated reasons. 5-10% of these tests are positive. Evidence suggests that General Practitioners (GPs) feel that they have neither the training nor resources to manage these infections effectively and would welcome support.

Approach

We piloted extension of the option of management using the centralised nurse-led telephone-based service used locally to manage tests taken in the National Chlamydia Screening Programme to all CT/NG infections diagnosed in PC in patients age 18 or over. We investigated acceptability of this to providers and patients and feasibility of evaluation in a cluster randomised trial. We recruited 11 practices. Eight were randomised to the intervention arm where practitioners could choose the centralised management model in eligible patients. Patients in control practices received usual care. Clinicians in participating practices alongside patients who both consented and declined consent to study participation were interviewed about their experience of the study. We investigated fidelity of study allocation and feasibility of participant follow up in medical records with linkage undertaken through a trusted third party.

Findings

Findings from the qualitative research are reported in detail elsewhere, here we report quantitative and overall study findings. 41% of practices invited agreed to take part and 60% of eligible patients were recruited. Fidelity of study allocation was generally maintained and linkage to participant records for follow up was successful. 1154 tests (1030 female) were submitted amongst eligible patients overall positivity was 2.6 (95% CI 1.8-3.7)% for CT and 0.8 (95%CI 0.4-1.5)% for NG. Overall CT positivity was 3.5 (95%CI 2.2-5.5)% amongst individuals eligible but not recruited to the study compared to 2.0 (95%CI 1.1-3.3)% amongst those recruited, imprecision in all these estimates reflecting small numbers. The main reason reported for failure to recruit eligible patients was insufficient time to undertake consent procedures. Patients found the intervention acceptable and were generally happy with the use of their medical records in evaluation as long as their anonymity was maintained. Despite patient consent being recorded, patients were sometimes unclear that they were participating in a research study.

Consequences

Patients and practitioners liked this intervention and evaluation in a cluster-randomised design appeared feasible. Requiring individual consent impeded recruitment and there was evidence it also introduced bias. In view of this, the fact that the intervention is a variant of routine care and the views of patients expressed in this study we conclude that a viable and valid future trial should not require individual patient consent providing the choice to opt out is provided and records-based follow up procedures maintain patient anonymity.