Main findings
The study objective was to explore which cardiology-related referral indications and reasons for referral are appropriate for specialist care in PC+ and which are more appropriate for specialist care in the hospital setting. Results showed that significant predictors of the advice ‘follow-up in hospital care setting’ are gender (male), age (per ten years), the referral indications ‘Stable Angina Pectoris’ or ‘Dyspnoea’ and the reasons for referral ‘To confirm disease’ or ‘Screening of unclear pathology’.
Interpretation of the main findings
Patients with one or more of the significant predictors have a higher probability of being referred to hospital care after a consultation at the PC+ centre. For example, a patient with an age of 30 years with the referral indication ‘Stable Angina Pectoris’ has a probability of 22.4% of getting the advice ‘Follow-up in hospital care setting’, and when in this case the reason for referral is ‘To confirm disease’ the probability is 51.1%, and if this patient is 70 years (instead of 30 years) it is even 72.2%. The reason for referral ‘To reassure the patient’ is a significant predictor of the advice ‘Follow-up in primary care setting’. The probability of these patients getting the advice ‘Follow-up in hospital care setting’ is only 1%.
The aim of PC+ is to substitute specialist care in the hospital setting with specialist care in the primary care setting. To achieve this it is crucial that PC+ is not designed as an intermediate station and thus it is important to select the appropriate patient groups for PC+. However, the results of this observational explorative study based on quantitative data of the PC+ centre need further interpretation. The results provide only insight which patients are more likely to be referred from PC+ to hospital care. This insight can be used as an input and can provide guidance for the healthcare professionals to improve the referral patterns (e.g. list of inclusion criteria for PC+). GPs and cardiologists should deliberate and collaborate to improve the referring procedures in order to enhance the connectivity between the different healthcare providers and the alignment of care. In addition to the characteristics taken into account in this study, more factors will probably influence which patient groups are appropriate for PC+, such as clinical information and specialist characteristics (e.g. level of expertise, experience and confidence). Hence, further research is needed to find out more precisely which patients are appropriate for PC+ and which patients should be excluded from PC+.
Reflection with existing literature
Previous studies have shown that outreach clinics could lead to increased healthcare costs [15, 27]. This is also a pitfall of PC+. Focusing on the appropriate patient groups and preventing PC+ from becoming an intermediate station increase its chances of success. Additionally, previous research showed that GPs experienced difficulty and uncertainty in referring eligible patients to PC+ [18]. Medical specialists sometimes saw patients in PC+ who should have been referred directly to hospital care or who could have been treated by the GP him−/herself [18]. Selecting the appropriate patient groups seemed to be essential for achieving efficient substitution. Moreover, PC+ is related to the concept integrated care as described by Kodner and Spreeuwenberg [16]. Intensifying the communication and collaboration between GPs and medical specialists in order to connect and align multiple services, providers and settings is one of its core features. Additionally, it is aimed at enhancing population health, quality of care as experienced by patients and reducing the number of unnecessary and inappropriate referrals in order to achieve efficiency. Hence, to improve the referral procedures it is recommended that GPs and cardiologists deliberate and collaborate. The patient groups that have a significantly higher probability of receiving the advice ‘Follow-up in hospital care setting’ should not be referred first to PC+. Additionally, they should deliberate about the patients who are referred with the reason ‘To reassure the patient’. Previous research indicated that perceived medical need is the strongest predictor but not the only predictor of GPs’ behaviour: perceived pressure from patients is also a strong independent predictor of GPs’ behaviour that affects the referral behaviour [28, 29]. It is plausible that some of the patients with the reason for referral ‘To reassure the patient’ were not referred to PC+ based on a medical need but on perceived pressure from the patients. With reference to efficiency, the results indicate that GPs should refer patients with a medical need and as few patients as possible with only a ‘To reassure the patient’ need. The strengthened collaboration and intensified communication generates knowledge transfer between the GPs and cardiologists. In the long term it is expected that this will induce a learning-effect for the GPs, which will probably result in less inappropriate referrals.
Strengths and limitations
This study used quantitative data extracted from the electronic medical record system of a cardiology PC+ centre. The study is based on a large sample size, namely on almost all referrals to the centre over one year. The results provide insight into which patients are less appropriate for PC+. Eventually this will lead to an increased number of appropriate referrals and fewer inappropriate referrals, in order to contribute to containing the rising healthcare costs.
A limitation of this study concerns the generalizability of the findings. The findings are context-bound, and generalizability to other countries (e.g. countries where GPs do not have a gatekeeping role), specialties (as this study is focused on cardiology) and/or other PC+ models is limited.
Future research
The results of this study are based on traditional statistical analyses providing insight into the appropriate patient groups for cardiology PC+ centres. This research is limited to quantitative data, and complementary (qualitative) research may be helpful for interpreting the findings in more detail.
The results showed a gender difference in the probability of being referred to specialist care in the hospital setting after a consultation at PC+; male patients have a significantly higher probability. This gender difference should be further investigated because previous studies demonstrated that the risk of heart disease in women has been underestimated. The awareness of the cardiovascular disease health risk in women among healthcare providers is relatively low and women are often misunderstood in terms of their symptoms [30, 31].
The results of this study can provide guidance for further research. To be better able to select the appropriate patient groups for PC+ it is recommend to investigate the influence of other factors and characteristics, and to perform complementary qualitative research to interpret and clarify the quantitative results. Future research could, for example, focus on the specialist characteristics (e.g. level of expertise, experience and confidence) and it should also include clinical outcomes (e.g. including follow-up data after a consult at the PC+ centre).
Additionally, regarding the cost-effectiveness of this PC+ intervention it is supposed that the introduction of PC+ will lead to reduced healthcare costs by lower costs in PC+ than in hospital care due to lower overhead costs, fewer referrals to hospital care, and less use of additional hospital-based services. Future research will take into account the effects on the health of the population, patients’ experiences of care and healthcare costs and will specifically focus on the cost-effectiveness of PC+ [17].