The effectiveness and cost-effectiveness of telephone triage of patients requesting same day consultations in general practice: study protocol for a cluster randomised controlled trial comparing nurse-led and GP-led management systems (ESTEEM)

Background: Recent years have seen an increase in primary care workload, especially following the introduction of a new General Medical Services contract in 2004. Telephone triage and telephone consultation with patients seeking health care represent initiatives aimed at improving access to care. Some evidence suggests that such approaches may be feasible but conclusions regarding GP workload, cost, and patients' experience of care, safety, and health status are equivocal. The ESTEEM trial aims to assess the clinical-and cost-effectiveness of nurse-led computer-supported telephone triage and GP-led telephone triage, compared to usual care, for patients requesting same-day consultations in general practice. Methods/design: ESTEEM is a pragmatic, multi-centre cluster randomised clinical trial with patients randomised at practice level to usual care, computer decision-supported nurse triage, or GP-led triage. Following triage of 350-550 patients per practice we anticipate estimating and comparing total primary care workload (volume and time), the economic cost to the NHS, and patient experience of care, safety, and health status in the 4-week period following the index same-day consultation request across the three trial conditions. We will recruit all patients seeking a non-emergency same-day appointment in primary care. Patients aged 12.0-15.9 years and temporary residents will be excluded from the study. The primary outcome is the number of healthcare contacts taking place in the 4-week period following (and including) the index same-day consultation request. A range of secondary outcomes will be examined including patient flow, primary care NHS resource use, patients' experience of care, safety, and health status. The estimated sample size required is 3,751 patients (11,253 total) in each of the three trial conditions, to detect a mean difference of 0.36 consultations per patient in the four week follow-up period between either intervention group and usual care 90% power, 5% alpha, and an estimated intracluster correlation coefficient ICC of 0.05. The primary analysis will be based on the intention-to-treat principle and take the form of a random effects regression analysis taking account of the hierarchical nature of the study design. Statistical models will allow for adjustment for practice level minimisation variables and patient-level baseline covariates shown to differ at baseline.