Proposal to update the review: Telephone consultation and triage: Effects on health care use and patient satisfaction
The proposed protocol was prepared for the commencement of the specified systematic review and meta-analysis using the Cochrane methods. The following is the brief introduction about the proposal.
The population ageing and the shortages of health professionals leads to increasing demands for health care and escalating burden on health care systems in countries world-wide including China, where the healthcare system supports more than 1.4 billion people and faces an unprecedented challenge. An exploration of alternative ways to manage a variety of patient conditions across the continuum from the identification of problems towards the desirable recovery is necessary to satisfy the diversified needs and expectations from both the health care providers and the receivers. The telephone consultation and triage is an expanding innovative service model to navigate patients and/ or their caregivers for rapid access to appropriate professional services while freeing up opportunities for face-to-face medical consultation. It has been operated for decades in many developed countries to strengthen the primary care and efficient referral to medical services and to reduce the burden on all stakeholders.
A Cochrane Review (Bunn et al., 2004) evaluated the effects of telephone consultation and triage on health care use and patient satisfaction. It was found that telephone consultation appeared to decrease the number of immediate visits to general practitioners. The authors stressed that further rigorous evaluation was needed to confirm the effectiveness, safety, and cost-effectiveness of the telephone consultation and triage services. However, all the included studies in this review were from the UK and the USA, the absence of studies from other countries may limit the generalisation of the evidence in the other contexts. To understand the latest evidence on telephone consultation and triage, we have conducted a preliminary search and found new studies since 2007, including some large-scale studies and studies from the other countries (e.g., China and Australia). With these newly emerging studies, it is possible to synthesize findings from those interventional studies in different health systems and to further substantiate the results derived from the previous review. It becomes imperative for us to update the review to bring into surface the best evidence on the effectiveness and safety of telephone consultation and triage.
A new outcome will be introduced to evaluate the effects of the telephone consultation and triage service on health care providers’ workload ─ defined as number and total duration of triage contacts as well as the subsequent face-to-face consultation (including the referral to other health care professional), as such outcomes may affect decision makers’ adoption of the service model. The primary aim of this review is thus modified “to assess the effects of telephone consultation and triage services on caller safety, service usage, patient satisfaction, and health care provides’ workload.
No changes will be made to types of study, participants / population, and intervention. With the above newly added outcome, the outcomes will be reorganized according to the recommendation of categorising outcomes. Primary outcomes: patient safety (i.e., deaths/mortality, visit to A&E and unplanned hospital admission within 7 days post-indexed call); service usage (ambulance use; visit to GPs’ service including surgeries; walk-in centre contacts; out-of-hours contacts;); patients’ satisfaction; d. health care providers’ workload (calls handled by different responders alone; home visits by GPs/deputizing services within normal hours; occurrence of different types of individual consultation). Secondary outcomes: visit to A&E within one month post-indexed call; follow-on (routine) general practitioner appointments within one month post-indexed call; patient quality of life; health care providers’ attitudes/satisfaction; cost to health care system; and cost to patients and/or their families.
Changes will also be introduced to the Analysis. The pilot search identified a few RCTs with noticeable differences in populations, interventions, comparisons or methods, and settings. When applicable, the random-effects meta-analysis (i.e., τ2 statistics) will be performed to estimate the between-study variance, alternatively, estimated standard deviation (τ) of underlying effects across studies. For absolute measures of effect (e.g. mean difference of time duration of the index and fallow-up calls per patient or patient satisfaction), the 95% confidential interval of the underlying effects will be calculated as from 2τ below to above the pooled estimate. For the relative measures (e.g. odds ratio of adverse events), the Mantel-Haenszel fixed-effect meta-analysis will be performed to estimate the standard errors of the effect measure estimates.
The revised Cochrane risk-of-bias tool for randomized trials (RoB 2) will be used to document the level of bias arising from the randomization process, bias due to the deviation from intended intervention, bias due to missing outcome data, bias in measurement of the outcome, and bias in selection of the reported results. The certainty/ quality of the evidence (i.e., high, moderate, low, and very low) or confidence in the estimate will be graded using the approach recommended by the GRADE Working Group.
The brief economic commentary framework will guide the economic analysis. The analysis of costs or cost-effectiveness will be carried out to compare the effects on callers (i.e., patient safety and callers’ satisfaction) and the costs of resources (i.e., service usage, healthcare providers’ workload) of telephone triage and consultation with that of usual care (i.e., appointment-based primary care). Monetary estimation will be provided when appropriate.
Databases being searched included MEDLINE, EMBASE, CINAHL, Cochrane Central Register of Controlled Trials (CENTRAL), PubMed, ProQuest (Health & Medical Complete), and OpenGrey. Additionally, we also did a web search using google scholar. Terms we used for search: to avoid the missing of any potential studies and ensure the inclusion of as many as possible of relevant in the review, we used broad search terms as followings: triage, consultation*, tele*, hotline*, helpline, telephone, phone, NHS direct, intervention*, controlled, experiment*. The tentative strategies used for the search of different databases were provided separately.
We did not find any non-Cochrane reviews with the exact same topic. However, there are some similar reviews but with differences in at least one PICO component. A total of five studies were identified during the preliminary search: Two cluster RCTs (#1 ESTEEM study; Richards et al., 2004) and three individual RCTs (Gamst-Jensen et al., 2019; Hui, Kong, & Wong, 2010; Vorster & Stott, 2011), and one time-series study.
