Out-of-hours demand for GP care and emergency services: patients' choices and referrals by general practitioners and ambulance services

Following the UK and Denmark, over the last five years, Dutch provision of out-of-hours primary health care has shifted from practice-based services towards large-scale general practitioner (GP) cooperatives [1, 2]. There are currently more than 130 GP cooperatives in the Netherlands, generally with 40 to 120 full-time participating GPs, which cover over 90% of the entire Dutch population and serve between 50,000 and 500,000 people. Most GP cooperatives are known to lie in close proximity of the hospital. Although most GP cooperatives operate independently from the hospital, recently, some have decided to integrate with the local Accident & Emergency Department (AED), to form one out-of-hours emergency centre [3]. One of the motives for this reorganisation was to prevent patients from self-referring themselves directly to the AED without first consulting the GP cooperative. Some authors have pointed out, that many of these so called self-referrals present with minor problems that can also be treated by a GP [4, 5]. This has led Dutch health policy makers to believe that integration of all out-of-hours services using one triage system will offer a chance to improve the efficiency and quality of care at a lower cost. Likewise, patient organisations have pointed out that patients find it increasingly difficult to determine to whom they should turn with their out-of-hours demand: the GP cooperative, the AED or the ambulance services.

Before major reorganisations are to take place, comprehensive data on overall out-of-hours care utilization should be provided, based on well defined populations. Interestingly, so far, only few studies have attempted to do so [3, 6‐11].

In this paper we describe the out-of-hours demand for a Dutch population of 62,000 people. The objective of this study is (1) to determine the out-of-hours patterns of use of general practice and A&E services; (2) to compare AED visits by self-referrals, patients referred by the GP cooperative, and patients brought in by the ambulance services.

This study shows that the GP cooperative is the main provider of out-of-hours care for the population studied. Within the group of all patients who contacted the AED, self-referrals constituted a large group, although they only represented a small percentage of all out-of-hours demand. With nineteen percent fractures among the patients with injury and one third of non-injury patients being admitted to the hospital, a substantial part of self-referrals appeared to have made a reasonable choice to attend the AED. Finally, compared to the self-referrals, both the GP and the ambulance services appear to be an adequate filter to the AED services, referring patients with more fractures and resulting in more hospital admissions.

This study was based on a relatively small population and therefore the results may not be generalisable to other regions. Although the population studied appeared similar to that of the Netherlands in terms of age-sex structure (Table 5), people from the respective age-groups for 15–44 and ≥65 years may have been slightly under- and overrepresented, leading to differences in overall demand from these groups. Furthermore, the proportion of self-referrals within the group of all patients who visited the three AEDs combined was derived from the contacts of the population of Velsen only. It is therefore unknown whether this was also representative for the other populations that these AEDs were serving. Hence, the conclusion that further integration between the GP cooperative and AED may not be effective refers to the population of Velsen only. Another limitation of the study was the relatively high percentage of patients visiting the GP cooperative without calling in advance (10%). Many of these patients presented with an injury (54%), suggesting some similarity with AED self-referrals (the GP cooperative being located in a former AED) and perhaps resulting in an underestimation of the population's overall AED self-referral rate. Nevertheless, even if most of these patients would have attended the AED, they still represent a small part of the overall demand only.

The coding of presenting symptoms to the AED took place using hospital records rather than the data collection sheets that were used in the GP cooperative. Even though the same coding methods were used for both settings, it is not clear whether the order in which the complaints were written down was similar to the order in which they were mentioned. Furthermore, it is uncertain whether the ICPC is valid for problems presented at the AED. Therefore, coding differences between both setting may have occurred.

Finally, the Winter period could have yielded a higher out-of-hours demand, although this effect is likely to have been small [14, 15].

Although GPs had been concerned that the cooperative would face an increasing demand once the service became more widely known to the public, no such increase was observed, which is consistent with findings from a study in another Dutch city surveying five consecutive years of out-of-hours demand.

Our study findings appear to be similar to those found in a few other population based studies in the Netherlands [3, 11]. Nevertheless, one study in the city of Amsterdam found a substantially lower contact rate with the GP services (171/1000/yr) and higher contact rate with the AED (170/1000/yr). Even larger differences in demand for GP care were found between studies from the UK, Ireland, Denmark and Finland (ranging from 130/1000/yr in England to 533/1000/yr in Finland) [14, 16‐18]. International comparisons should be interpreted cautiously because of varying definitions of the out of hours period and differences in health service organisation. Salisbury et al. found that the variation in call rates between different British cooperatives could not be accounted for by local demographic features (age structure, deprivation, and rurality) [14]. Finally, although no literature review has been performed yet, self-referral rates to the hospital too appear to vary substantially across some European countries (ranging from 57/1000/yr in one British study to 190/1000/yr in Denmark) [8, 10, 19, 20] Perhaps the enormous variation is in part explained by differences in the effectiveness of the gatekeeper. Boerma and others have shown, that while as a rule patients need a GP referral to make use of hospital services, all gatekeeping systems make an exception for emergencies that can be presented directly to the AED [21]. This important leakage of gatekeeping systems may lead to variation in AED use, especially since the (perceived) availability of GP services varies from country to country [22‐24].

Various authors have expressed their concern about the overcrowding of AEDs as a result of high numbers of self-referrals [25, 26]. There has been much debate on how to redirect these allegedly 'inappropriate attenders' to the GP. However, without a clear definition of what constitutes an 'inappropriate attender', it seems not surprising that a wide variation (6–80%) was found in the literature [27], and that others have cast doubt on the usefulness of the term itself [28]. In our study, with nineteen percent fractures among the injuries and thirty-five percent admissions among the non-injuries, at least the self-referrals emerge as a self-selected group with a severity level that appears to be higher than patients calling the GP cooperative, but still lower than those who were referred by the GP [29]. Also, in part, the care seems complementary: while 80% of the self-referrals presented with an injury, GPs also referred significantly more injury patients (11.6%) to the AED than patients without an injury (6.6%). Overall, many patients may have made reasonable choices when deciding which service to contact [9, 30]. Nevertheless, compared to the self-referrals, the GP and ambulance services provide an effective patient filter to the AEDs.

If GP cooperatives and AEDs are to further integrate their services, more research is needed on out-of-hours demand. Patient characteristics and their motives to attend to an AED as a self-referral or call the emergency services rather than contacting a GP cooperative need further elucidation before radical organisational changes are to be carried through [9, 31‐33]. Also, more insight is needed into the triage activities of the regional ambulance services, both on the telephone and during field assessments, as they appear to overlap with both GP and AED services [34].

From the patients' perspective, having one national or regional emergency number and one out-of-hours emergency service for all problems presented may seem an obvious development. Although some studies have indicated that GPs working within the AED handle self-referrals equally safe with fewer use of resources [5, 35], it is unclear whether integration of all services would become more efficient in terms of professional care and costs, as this may be more dependent on the size of the population the cooperative covers than the way the GP cooperative is organised, i.e. separated or integrated [36]. Moreover, except for the major cities, self-referrals not only represent a relatively small group in the total out-of-hours demand, many of these patients may have made a reasonable choice for AED services [9, 30].

It is not unlikely that different, regional models of integration could evolve from local patient demand. In the Netherlands, a few integrated out-of-hours emergency centres are now operational [3], although many GP cooperatives and AEDs prefer to keep the provision as it is and focus on working together more closely.

GP cooperatives appear to deal with the large majority of all out-of-hours problems presented.

Within the total out-of-hours demand, self-referrals at the AED constitute a small group of patients who, in part, seem to have made a reasonable choice of service. Compared to the self-referrals, GPs and ambulance services appear to make stronger selections of injury and non-injury patients as is indicated by higher percentages of fractures and hospital admissions.