We sought to describe the referral pattern to and between the medical specialist outpatient clinics in the University Hospital of Southern Denmark. Our study showed that referrals were evenly distributed between hospitals 49.5% (233) and GPs 50.5% (238), with a very small percentage either attending another speciality or were simultaneous referral. Of the referrals from hospitals, one third (72) were from the Medical Department, indicating that the department has an inefficient referral process. Out of these, inter-departmental referrals accounted for 70% (51) and inter-speciality referrals accounted for 30% (21), meaning that some of the outpatient clinics in the Medical Department referred patients to themselves (self-referral) e.g. pulmonology outpatient clinic referring the patient to their own clinic.
There is considerable information on referral from GPs to secondary care internationally (13, 15–19, 23). Unlike in our study, those studies (18, 19, 23) showed that the majority of referrals came from GPs. Based on these data and findings that often GPs refer patients based on their symptomatic picture and rather than with a confirmed diagnosis (17, 24, 25), we expected that our findings would also comprise a considerably higher proportion of patients referred from GPs. Additionally the finding that the Medical Department had also considerably higher proportions of inter-departmental and inter-specialty referrals was also unexpected. We propose several reasons for increased proportion of inter-departmental referrals in our study, including increasing numbers of patients with multi-morbidities that require complex care from several specialists (7, 8) and the internal referral procedure at the Medical Department between the outpatient clinics. However, the unusually high rate of inter-departmental referrals should be further investigated. Furthermore, many previous studies of referral have tended to use self-reported data or survey formats (19, 20, 26–28) to describe referral patterns and examine factors influencing healthcare professionals’ referring habits (4, 29). There is a possibility that these previous findings were impacted by recall bias. Our study used register data, which provides more accurate findings.
Despite the growing awareness of issues related to referrals, there has been relatively little research internationally on referrals between specialists in hospital outpatient clinics or departments. The literature on referral is diverse and differences in contexts, study methods, and measurement of referral patterns make it difficult to compare results between studies. Similar to our study, a qualitative study by Burkey et al. (30) on inter-departmental referrals in hospital indicated that inter-departmental referrals are likely to increase among specialists. Previous research shows that unnecessary referrals are often due to an inefficient referral process, which results in patients undergoing unnecessary diagnostic procedures (10, 11).
The University Hospital Medical Department in our study is typical of hospitals internationally in facing the same challenges of increasing demand and limited resources. It receives a high proportion of inter-departmental and inter-speciality referrals. It is difficult to explain what causes the inter-speciality referrals. However, according to the Head of the Medical Department, the inter-speciality referral or self-referring is often associated with outpatient clinics that are organised into small specialised teams or clinics that have local functions at the hospital’s other facilities. As a result of the organisational system, the physicians occasionally send formal referrals to their own outpatient clinic when the patient requires other services from a different team, instead of booking the patient directly for the next appointment. Additionally, less experienced or junior physicians feel more compelled to refer patients through the formal channels because they are not familiar with the referral procedure.
Unnecessary referrals due to inefficient referral processes can, in the long-term, lead to a bottleneck in services and create more demand for a health service already stretched and under pressure (15). This may increase the cost of providing care, consuming healthcare resources that could have been used to provide other services. For example, the practice exacerbates the administrative workload for the healthcare professionals and the administrative employees such as the secretaries that are responsible for screening and sending out the referrals as well as booking the tests for the preliminary appointments. Cost for patients may include delayed treatment and waste of time (31) related to unnecessary patient appointments and repetitive tests (10, 11). Visiting the hospital several times a year for further assessment could subject the patient to stress and affect their mental well-being (14, 31).
A new way of managing referrals is required in the hospital system. Several strategies can be implemented to improve the referral process within the Medical Department and thereby reducing the high inter-department and inter-speciality referral rates. Measures such as IT-solutions (32), joint triage of referrals (12, 33–35) and co-location (36, 37) of the medical outpatient clinics to create a joint medical outpatient clinic would provide opportunities for inter-professional collaboration due to the close physical proximity. Close proximity may stimulate more informal consultation between the medical outpatient clinics and replace formal referrals. It would allow the specialists to work together and jointly share information, diagnose and treat common patients. IT-software can be used to get a better overview of the patient flow and management (38, 39). The University Hospital of Southern Denmark already uses IT-software for management and workflow in the in-patient ward. Similarly, the healthcare professionals in the outpatient clinics can use the IT-software to send communication to each other and book patients directly for appointments and tests and thereby improving the inter-departmental and inter-speciality referral process (38, 39). An added benefit to using IT-software is that it may be easier to locate colleagues and result in fewer interruptions. Co-location in combination with IT-software and joint triage of patient referrals would provide patients with convenient access to necessary diagnostics and allow for same-day treatment from several providers at the same time. This would reduce the ineffective use of physician and hospital services and allow the patient care in the hospital to be organised around their needs rather than the needs of the medical specialties involved. Additionally, it would ensure improved workflow and safe ongoing care.
Implications for practice and future research
We anticipate that our results will be useful for clinicians, managers, decision-makers, and researchers in understanding and improving referral processes. Our study sheds light on the inefficiency of the referral process within a hospital owing to the organisation structure and the administrative system. Our findings could stimulate similar practices and hospital systems nationally and internationally to conduct an assessment of their current referral and routine practice to minimise unnecessary referrals and improve cross-sectoral collaboration between specialists. The findings may also help guide strategies for quality improvement and serve to initiate discussions on how to best organise the care of patients between multiple outpatient clinics within a single hospital system.
Strength and Limitations
While the size of the study was comprehensive, the study was limited to a short period of time and a single health service in a region of Denmark, which may limit the generalisability of our results. Second, the study was limited to new referrals to the medical outpatient clinics at the hospital and therefore this may have impacted the low rate of co-attendance referrals. A strength of our study is that we used registry data: the Danish hospitals routinely collect data linked to the 10- digit personal identification number (CPR number) and hence provide highly valid and reliable data for research.