Telephone triage in out of hours call centres

Telephone triage, both for in hours and out of hours consultations, has increased dramatically in recent years, and in many respects this is welcome. Telephone consulting can improve access to care for many patients,1 and out of hours care provided by call centres in particular can improve the efficiency of healthcare provision.2 Several unanswered questions remain, however, with respect to the quality and safety of such clinical encounters because of the relative paucity of evidence on this mode of consulting.3

The linked study by Derkx and colleagues (doi: 10.1136/bmj.a1264) highlights the potential shortcomings of telephone based consultations in the context of out of hours triage of patients in the Netherlands.4 Strengths of the study include a carefully considered sampling strategy of call centres and the use of standardised clinical encounters using simulated patients.

Out of hours consultations are risky because the patients and professionals usually do not know each other and the situation is often a clinical emergency. Predictably, the study found that the quality of these consultations was consistently poor for all cases and for all centres. The quality of the history taking, decision making, and providing advice phases of the consultation varied greatly. More surprisingly, general practitioners were little better than less trained “triagists.” Triage management outcomes were appropriate in only 58% of calls, and urgency was underestimated in 41% of cases.

For patients with serious conditions, such as meningitis or malaria, incorrect risk stratification and the subsequent delay in treatment could have grave consequences. Several high profile cases of serious adverse outcomes associated with telephone provision of out of hours care have been reported in the United Kingdom.5 However, the findings from the limited number of robust studies are mixed. They identify safety concerns in relation to the process of care but indicate that telephone consulting is safe overall in terms of clinical outcomes.6 Although we need to clarify the link between the quality of telephone consultation (questions asked, evaluation of the answers, and the care advice given), process measures, and actual clinical outcomes in patients, the most credible explanation for this discrepancy is that because serious adverse outcomes are relatively rare, most of the studies have lacked the power to detect harm to patients.7

Internationally, call centres employ “triagists” with variable amounts of clinical knowledge, skills, and experience. They can be trained lay people, nurses, or doctors, working with or without protocols. The organisation of these call centres is also diverse; for example, some have a tiered approach, with an initial screen by personnel who are paid less and then, if necessary, second level triage by a doctor to reduce the need for face to face clinical encounters further.8 The best approach to use is unclear, and it may depend on the quality of management protocols used by triagists. We do not know the degree to which protocols vary between call centres and countries—because they are invariably confidential—or to what extent variation from the guidelines is tolerated or even recorded. How then can we know that these protocols are up to date, evidence based, and followed? Why are they not public like other clinical guidelines? Moreover, why are they not available online to empower patients who may wish to use them?

So what is the future for out of hours telephone consulting? We need to define the contexts in which telephone consulting is most appropriate and develop safeguards to minimise the risk of inadvertent harm to patients. We can then closely monitor performance using methodologically robust approaches.9 10 It would be useful if evaluations of call centres in other countries were also published, including details of the methods used to triage patients (for example, the protocols used) and the approaches used to evaluate the care provided.

International cross cultural descriptive and evaluative approaches, such as recording consultations and comparing management with best care standards and guidelines would also be welcome to help develop interventions that can be locally implemented and evaluated. Underpinning all this is a need for policy makers, funders, and academics to recognise the importance of treating new models of care—such as out of hours call centres—as interventions in their own right and to evaluate them before rather than after they are introduced. Until we have such upfront commitment to evaluation we will not know whether developments such as the one reported by Derkx and colleagues are beneficial or harmful.