Compliance with advice and appropriateness of emergency presentation following contact with the HealthDirect telephone triage service

Introduction

Many health systems around the world, both public and private, have introduced telephone health services with the aim of providing health consumers with timely and convenient access to expert health information.¹ Increasingly, 24 hour, seven days a week telephone triage services staffed by experienced triage nurses are being introduced with the apparent aim of reducing the after hours burden of work for primary health care practitioners and improving the appropriateness of utilization of emergency services.^2–5

In some cases, a justification for the introduction of such services has been an expectation that telephone triage would reduce the overall burden of work in increasingly overcrowded emergency departments.⁶

A reduction in after-hours calls for internal medicine specialists and paediatricians, and home visits for general practitioners have been reported following the introduction of telephone triage services.^7–11

In the ED setting, modest improvement in the appropriateness of paediatric presentations has been reported¹² however, the capacity to influence overall workload appears to be reduced by self selection to attend despite triage advice to defer presentation.¹³ At least one study has failed to identify any reduction in ED workload despite reductions in general practitioner after-hours workload.³

In the Australasian context, the only definite impact on emergency workload reported thus far is a reduction in the volume of telephone advice calls handled by emergency departments due to the automatic diversion of these calls from the hospital to the telephone triage service rather than to the ED.⁶

The purpose of this study was to examine the impact of a telephone triage service upon a typical Australasian metropolitan ED. The study examined the compliance rates and appropriateness of presentation of patients presenting to an Australasian metropolitan ED following contact with a telephone triage service.

Methods

HealthDirect is a telephone triage service staffed by experienced triage nurses and has been providing 24 hour, seven days a week telephone advice and triage to metropolitan Perth, Western Australia since May 1999. The service is operated by McKesson Corporation on behalf of the Department of Health of Western Australia and uses the Centramax computerized clinical decision support system to assist triage decision-making. This clinical guideline system has been tailored to Western Australian conditions by an expert group of clinicians. The establishment and performance of HealthDirect has been described in detail elsewhere.⁶

Probabilistic data linking techniques were used by the Western Australian Data Linking Unit (DLU) to link 2000 calendar year HealthDirect contacts with 2000 calendar year Fremantle Hospital Emergency Department presentations.¹⁴ Based upon previous named data linkages conducted by the Unit, the DLU estimates the data linkage rate to be in excess of 98%.¹⁴ Fremantle Hospital Emergency Department is a central metropolitan teaching hospital, mixed adult and paediatric ED, and the trauma centre for the southern metropolitan Perth region. A linkage was only accepted if the HealthDirect contact was during the 24 h preceding ED triage data and time. Linkages included both patients triaged to seek emergency care and those triaged to alternative sources of medical care. The Australasian Triage Score (1 — resuscitation to 5 — non-urgent) triage distribution and admission rate of the HealthDirect contacts population were compared to the non-HealthDirect patient contact population in order to determine whether HealthDirect had any impact upon the appropriateness of ED presentation. It was estimated that a 12 month study period would capture approximately 500 HealthDirect contacts amongst 10 000 presentations within a 2 km radius, assuming a HealthDirect contact rate of 5%. The power of detecting a 10% difference in admission rate between HealthDirect and non-HealthDirect contacts assuming α of 0.05 approaches 1.00. Compliance with HealthDirect advice was estimated for HealthDirect contacts who gave a call location within one of four postcodes within a 2 km radius of Fremantle Hospital. The 2 km radius was chosen to minimize the impact of patients choosing to attend alternative emergency services within the Perth metropolitan area upon the compliance estimates.

Some 28 different dispositions were available to HealthDirect nursing staff varying in urgency (e.g. immediate care, see within 4 h, seek care within a week, home care only required) and type of service (e.g. emergency, general practitioner, dental care) at the time of the study. Those indicating that medical attention should be sought immediately (‘immediate care’) or within 4 h (‘prompt care’) were considered emergency dispositions for the purposes of this study. Those dispositions indicating that medical attention could be safely deferred until at least the next day were considered ‘non-urgent’ dispositions for the purposes of this study. The denominator for compliance estimates included all HealthDirect triaged contacts, regardless of whether the contact was subsequently linked to an ED presentation.

In calculating compliance estimates it was assumed that patients complying with advice to attend an ED would be equally likely to choose to attend Fremantle Hospital Emergency Department (rather than an alternative emergency service) as patients who attended an ED despite advice not to attend an ED. Statistical analysis was performed using the Statistical Package for the Social Sciences, version 10.¹⁵

Ethical approval for the study was given by the Confidentiality of Health Information Committee in the Health Department of Western Australia and Fremantle Hospital Human Research Ethics Committee.

Results

There were a total of 13 019 presentations to Fremantle Emergency Department from the four postcode area during the study period of which 842 (6.5%, 95% CI 6.0–6.9) were identified as having contacted HealthDirect within the 24 h period prior to presentation.

A higher proportion of HealthDirect contacts were less than 15 years of age (290, 34% 95% CI 31–38) when compared to non-HealthDirect contacts (1598, 13.1%, 95% CI 12.5–13.7, Pχ² < 0.001). HealthDirect contacts (481, 57%, 95% CI 54–60) were also more likely to be female than non-HealthDirect contacts (5610, 46%, 95% CI 45–47, Pχ² < 0.001).

Despite these age and gender differences, the Australian Triage Score (ATS) distribution for HealthDirect contacts was similar to the non-HealthDirect contacts triage distribution (Fig. 1). Similarly, the proportion of HealthDirect presentations formally admitted to hospital (317, 37.6%, 95% CI 34–41) was also similar to the non-HealthDirect admitted proportion (4683, 38.4%, 95% CI 38–39, Pχ² = 0.67).

Compliance analysis

There were 3996 calls to HealthDirect from residents living in the four postcode area. The estimated compliance rates with advice to seek emergency care are presented in Table 1. These data indicate a generally high level of compliance with HealthDirect advice. The triage distribution of patients triaged by HealthDirect to immediate or prompt care who complied with HealthDirect advice was very similar to the triage distribution of patients triaged to non-urgent care who presented to the ED despite a HealthDirect non-urgent disposition (Fig. 2).

Table 1. Estimated compliance with HealthDirect advice to seek emergency care

HealthDirect triage	Total	Complied	Percentage complied (95% CI)
Immediate or prompt	1579	963	61% (57–63)
Non urgent	2416	2204	91% (90–92)
Total	3996	3167	79% (78–80)

Similarly, the proportion of patients admitted who had been triaged by HealthDirect to immediate or prompt care and who complied with HealthDirect advice was very similar to the proportion of patients admitted who had been triaged to non-urgent care but who did not comply with HealthDirect advice to delay seeking medical attention (261, 38%, 95% CI 34–41 vs 56, 37%, 95% CI 30–44, Pχ² = 0.89).

Discussion

There has been a rapid uptake by health consumers of telephone health advice and triage services like HealthDirect, and consumers appear well satisfied with these types of services.^6,16–19 Telephone services provide a convenient means of accessing health advice for those with limited mobility, difficulty with access to transport, those who live in outer metropolitan, rural and remote locations or those who wish to maintain a degree of anonymity in seeking health advice.²⁰

There are sustainable arguments for the use of formalized triage guidelines by staff specifically trained in telephone triage, from quality of service and medicolegal perspectives, when compared to the ad hoc arrangements existing in many emergency departments where dedicated telephone advice services are not available.²¹ In addition, diversion of calls to telephone advice services reduces distraction and saves time for busy ED staff.^6,22

It appears however, that there is little published evidence in the Medline indexed literature, despite considerable debate, to support the hypothesis that telephone triage services can significantly reduce ED attendances.^23–25 The finding in this study that only 6.5% of ED presenters had made prior contact with HealthDirect imposes an upper limit upon the capacity of this telephone triage service to substantially influence the ED utilization or workload, particularly when it is noted that there were over three times as many presentations to the ED from the four postcode area as calls to HealthDirect from the same area during the study period. The higher proportion of women contacting HealthDirect prior to presentation identified in this study is consistent with the known HealthDirect caller casemix.⁶

An important additional finding was that, despite differences in the age and gender mix of the HealthDirect contact population, the similarity of the HealthDirect and non-HealthDirect triage distributions and admission rates indicates that HealthDirect contact did not significantly influence appropriateness of presentation to this ED.

The apparent compliance rate with advice of just under 80% is within the range reported in the international literature.^26–28

An unexpected finding was the similarity of the triage distributions and admission rates of the HealthDirect contacts referred for emergency care and those who chose to present despite advice to defer presentation. This finding suggests that HealthDirect contacts self select to attend the ED and base their decision to attend an ED on factors independent of HealthDirect advice.

This finding is consistent with previous studies that have found that calling ahead to an ED is not associated with more appropriate ED attendances and the previous finding that patients tend to self select to attend an ED despite telephone advice.^13,25,29

It may be observed that for many patients, the effort required to make a telephone call to a telephone triage service is less than the effort required to physically attend a triage station at an ED and that a range of additional factors, such as perceived symptom severity, access to transport and expectations about ED waiting times would influence the decision to attend an ED when compared to telephoning for advice.

Conclusions

This study indicates that HealthDirect contacts are neither more nor less appropriate in their utilization of this ED than non-HealthDirect presenters and appear to attend the ED independent of HealthDirect advice. Given the small proportion of total ED presenters that contact HealthDirect prior to presentation, and the apparent lack of influence of HealthDirect upon the decision-making of callers, it is concluded that HealthDirect has a limited capacity to influence ED utilization or workload.