BACKGROUND: Emergency calls are frequently made to ambulance services for older people who have fallen, but ambulance crews often leave patients at the scene without any ongoing care. We evaluated a new clinical protocol which allowed paramedics to assess older people who had fallen and, if appropriate, refer them to community-based falls services. OBJECTIVES: To compare outcomes, processes and costs of care between intervention and control groups; and to understand factors which facilitate or hinder use. DESIGN: Cluster randomised controlled trial. PARTICIPANTS: Participating paramedics at three ambulance services in England and Wales were based at stations randomised to intervention or control arms. Participants were aged 65 years and over, attended by a study paramedic for a fall-related emergency service call, and resident in the trial catchment areas. INTERVENTIONS: Intervention paramedics received a clinical protocol with referral pathway, training and support to change practice. Control paramedics continued practice as normal. OUTCOMES: The primary outcome comprised subsequent emergency health-care contacts (emergency admissions, emergency department attendances, emergency service calls) or death at 1 month and 6 months. Secondary outcomes included pathway of care, ambulance service operational indicators, self-reported outcomes and costs of care. Those assessing outcomes remained blinded to group allocation. RESULTS: Across sites, 3073 eligible patients attended by 105 paramedics from 14 ambulance stations were randomly allocated to the intervention group, and 2841 eligible patients attended by 110 paramedics from 11 stations were randomly allocated to the control group. After excluding dissenting and unmatched patients, 2391 intervention group patients and 2264 control group patients were included in primary outcome analyses. We did not find an effect on our overall primary outcome at 1 month or 6 months. However, further emergency service calls were reduced at both 1 month and 6 months; a smaller proportion of patients had made further emergency service calls at 1 month (18.5% vs. 21.8%) and the rate per patient-day at risk at 6 months was lower in the intervention group (0.013 vs. 0.017). Rate of conveyance to emergency department at index incident was similar between groups. Eight per cent of trial eligible patients in the intervention arm were referred to falls services by attending paramedics, compared with 1% in the control arm. The proportion of patients left at scene without further care was lower in the intervention group than in the control group (22.6% vs. 30.3%). We found no differences in duration of episode of care or job cycle. No adverse events were reported. Mean cost of the intervention was £17.30 per patient. There were no significant differences in mean resource utilisation, utilities at 1 month or 6 months or quality-adjusted life-years. In total, 58 patients, 25 paramedics and 31 stakeholders participated in focus groups or interviews. Patients were very satisfied with assessments carried out by paramedics. Paramedics reported that the intervention had increased their confidence to leave patients at home, but barriers to referral included patients' social situations and autonomy. CONCLUSIONS: Findings indicate that this new pathway may be introduced by ambulance services at modest cost, without risk of harm and with some reductions in further emergency calls. However, we did not find evidence of improved health outcomes or reductions in overall NHS emergency workload. Further research is necessary to understand issues in implementation, the costs and benefits of e-trials and the performance of the modified Falls Efficacy Scale. TRIAL REGISTRATION: Current Controlled Trials ISRCTN60481756 and PROSPERO CRD42013006418. FUNDING: This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 21, No. 13. See the NIHR Journals Library website for further project information.
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)