What if something happens tonight? A qualitative study of primary care physicians’ perspectives on an alternative to hospital admittance

The increased proportion of older adults and people with chronic diseases in the general population has resulted in an increased demand for health services worldwide [1‐3]. Hospital emergency departments in several high-income countries are overcrowded due to the large proportion of non-urgent patients [4‐6]. Many countries are launching primary healthcare models as alternatives to hospitalisation, as well as aiming to increase coordination across healthcare levels to improve patient care, reduce costs and improve public health [7‐10]. Community hospitals, observation wards and hospitals-at-home are examples of health service models at the interface between primary and secondary care that provide acute and/or non-acute services and offer a variety of treatment and diagnostic services [11, 12].

In Norway, health services are primarily divided into two levels: primary care services, including general practice, out-of-hours emergency services offered in casualty departments where patients may be assessed by primary care physicians, home-based care, nursing homes and municipal acute wards (MAWs), and specialist services, including hospitals, outpatient specialist care and contract specialists [13]. The MAW model has been in use since 2012 as an alternative to hospitalisation for adult patients who need acute treatment and care but not specialist health services [13, 14]. MAWs are located in the local community, near where people live. The inclusion criteria for admission to a MAW are as follows: patients who would otherwise be admitted to the hospital; maximum expected length of stay of 72 h; acute deterioration of a known, chronic condition; and other clarified conditions where the risk of acute deterioration is low. Eligible patients must be aged above 18 years [15]. In 2019, there were 216 MAWs in Norway, with a total of 735 beds [16]. MAW beds are located either inside nursing homes, in casualty departments or in separate units at local medical centres. MAWs range from small units with 3 or fewer beds to the largest units with 15 beds or more [16]. Studies have indicated that the MAW model is a good alternative to hospitalisation and that patients prefer to be admitted to a MAW rather than the hospital [17‐20]. Moreover, a study indicated that implementation of MAWs led to a 1.9% reduction in hospitalisations for patients aged over 80 years [14].

In many countries, primary care physicians (PCPs) are assumed to be gatekeepers responsible for assessing patients within the catchment area of their practice and for referring them to specialist health care services [21]. Since 2001, when a list patient system was implemented, Norwegian inhabitants have the right to be listed with a specific PCP [22]. Outside of office hours, patients can also access PCPs in a casualty department, where the PCP on duty does not necessarily know the patient or have access to his or her medical records [23].

Previous studies have indicated that PCPs’ attitudes towards the development of health systems and the treatment of fragile patients may be barriers to the implementation of alternatives to hospitalisation [24‐26]. Hence, for the implementation of new health service models, it is important to gain knowledge about the aspects that PCPs find important when referring patients to different healthcare services. Studies focusing on PCPs’ perspectives on alternatives to hospitalisation are lacking. Consequently, the aim of this study was to explore PCPs’ perspectives on the MAW model as an alternative to hospitalisation.

This qualitative study utilised semi-structured interviews with PCPs. The study adheres to the Consolidated Criteria for Reporting Qualitative Research (COREQ) guidelines [27].

A total of 21 PCPs agreed to participate in the study: 12 males and nine females. The PCPs’ mean age was 39 years, their mean work experience as a physician was 11 years, and their mean work experience at a hospital was 2 years (Table 2).

The findings indicate that the PCPs had different views on which level of healthcare service a MAW represented. MAWs were described as a “mini-hospital”, a “peripheral hospital unit”, a “specialised nursing home” and a “primary healthcare service level”. Despite varying descriptions of MAWs, most of the PCPs had a clear idea of “the typical MAW patient”, frequently described as an elderly patient with an acute need for treatment with a higher level of health services. Two main themes were identified through the analysis. The first theme was “What if something happens tonight?” The subthemes were a) safety for all, b) competence and responsibility, and c) collaboration across health service levels. The second them was “User participation”.

This study contributes knowledge about PCPs’ perspectives on the MAW model as an alternative to hospitalisation. Several factors affected the participating PCPs’ judgement of whether the use of MAWs is a safe alternative to hospitalisation, both from their own perspectives and from those of MAW personnel, patients and patients’ relatives. Moreover, disagreements regarding admission criteria, concerns related to the competence of MAW personnel, the unclear distribution of medical responsibility and collaborative challenges between primary and hospital health service levels all impacted the PCPs’ decisions on where a patient should be treated. Involving patients and relatives in the decision about the treatment level was also assumed to be important but challenging.

The PCPs were worried about what would happen if a patient’s condition deteriorated in the evening or night when there was no physician on duty at the MAW or if the PCP had misjudged the situation. A recent study of 2744 admissions to a MAW showed that 23.6% of admitted patients were transferred to the hospital as a result of deterioration or the identification of a clinical condition that required hospital care [34], which underlines the PCPs’ worry. To compensate for the PCPs’ insecurity and to avoid later hospitalisation, the diagnostic loop was considered a good solution. In addition, ‘the loop’ was requested by physicians and nurses at the MAWs. In contrast, a recent study found that PCPs felt that the diagnostic loop was an inappropriate use of time and resources [31]. Managing diagnostic uncertainty has been reported as a recurring challenge for physicians [35]. Several studies have shown that more experienced PCPs tolerate more risks and diagnostic uncertainty than other physician groups, who rely more on examinations to make a diagnosis [9, 36, 37]. This was not in line with findings in a recent study.

The PCPs reported that many patients wanted to be admitted to a MAW, especially older adults who had been to a MAW before. Such positive patient experiences of MAWs have also been highlighted in other studies [18‐20]. However, a Cochrane review also showed that patients with acute illness were satisfied with receiving treatment in primary care after a short hospitalisation [38]. Hence, the issue of whether patients who are acutely admitted to either a MAW or a hospital are able to objectively evaluate the services separately should be examined. It may also be argued that a patient’s condition when treated in an MAW is less acute and therefore impacts his or her impression of the stay.

In cases of deterioration or misjudged conditions, the PCPs were not convinced that admission to a MAW was a safe alternative to hospitalisation. They reported that they would prefer that MAWs be staffed by hospital physicians or medical specialists, and they had the impression that MAW physicians were quite inexperienced. It may be argued that MAW physicians could also be described as PCPs due to their role in primary healthcare. Hence, this may indicate a lack of insight from the PCPs into the actual competence of MAW physicians. A comparison of PCPs and hospital physicians in hospitals-at-home in Denmark showed that the PCPs managed to avoid the hospitalisation of elderly patients with acute medical diseases to a greater extent than hospital physicians [37]. This finding indicates a gap between PCPs’ assessments of their own, MAW physicians’ and hospital physicians’ competence.

Moreover, the PCPs in this study felt insecure because nurses staffed the MAWs for 24 h, while physicians were present only during the daytime. Vatnøy and colleagues found that reliance on nurses in acute inpatient settings requires that they have expertise to identify problems and take action [39]. The nurses in the MAWs were all registered nurses with a bachelor’s degree, and many of them had a specialisation, e.g., in critical care, geriatrics or diabetes. In addition, many of the nurses had considerable experience working in hospitals before they started working at the MAW. Hence, the PCPs seemed to lack insight into the nurses’ educational backgrounds and experience. Johannessen and Steihaug found that a lack of medical expertise was perceived as stressful and unsafe for nurses working in MAWs [31]. A Cochrane review indicated that nurse-led health services provide similar or better health outcomes, such as reduced mortality or reduced symptom burden for patients, than physician-led health services [40]. Regardless, neither the PCPs in the current study nor the personnel at the MAWs deemed nurses’ competence to be sufficient to provide safe and quality health services. This finding may indicate that it is time to reconsider how MAWs or similar primary care alternatives to hospitalisation should be staffed in the future. In addition, this finding indicates a need to better inform physicians about nurses’ competence, and vice versa.

The finding on PCPs’ feelings of insecurity when admitting a patient to a MAW is supported by, e.g., Leonardsen et al. [29]. Nevertheless, another study indicated that there were no differences in mortality and morbidity after patient admission to a MAW than after admission to a hospital [17]. Therefore, research on these primary healthcare alternatives to hospitalisation is essential to ensure that the different health care levels are used as intended. The findings in our study show that hospital physicians were considered medical specialists, while MAW physicians were considered collaborative partners. This finding was also reported in another study of general practitioners’ perspectives on patient handovers in primary healthcare [41]. The Vanguard study in England indicated that collaboration is difficult because there are barriers between the different health care levels [42], which is supported by our findings. A lack of communication and collaboration between healthcare providers is one of the main factors for stress among PCPs in high-income countries [43]; hence, this situation demands action. A focus on better collaboration between primary healthcare services and hospitals is essential to ensure the exchange of knowledge and experiences related to various diseases and how they are handled [44, 45].

The PCPs in our study reported that patients’ and relatives’ participation in deciding the level of healthcare services was important but challenging. This finding is in line with findings of recent studies [30, 46]. Patients’ involvement in decisions about their own health is associated with improved healthcare outcomes and patient satisfaction [47‐49]. Such involvement requires that physicians ensure that the patient is cognitively oriented and sufficiently informed [50], which was also noted by the PCPs in our study. Since many older patients are cognitively impaired, it can be very difficult for them to be involved in accordance with a patient-centred care model [30, 51]. Conflict of the wishes of patients and their relatives conflict with what PCPs consider to be the best medical decision can result in PCPs taking a more biomedical view rather than a patient-centred care approach when deciding the medical treatment [52]. PCPs have an obligation to not only ensure the patient’s medical safety but also involve the patient in his or her own medical treatment [49, 53]. Healthcare professionals report that patients have unrealistic expectations of what is feasible [54], which is supported by our findings. Thus, there is a potential ethical conflict between shared decision-making, on the one hand, and patient safety, on the other [53].

The implementation of MAWs in Norway was based on a political initiative, and important stakeholders, such as PCPs, were accorded minimal involvement in the development of the model. In a constantly developing healthcare system, organisational change is more likely to succeed if healthcare professionals have an impact on the changes [55]. The participating PCPs’ uncertainty and concerns probably resulted in patients being referred to the hospital instead of the MAWs. Therefore, in the further development of the MAW model, it is important to consider these aspects to ensure quality services and safety for all stakeholders.

The PCPs in our study reported several concerns related to the security of patients, MAW personnel and themselves in the consideration of admission to a MAW as an alternative to hospitalisation. Their concerns had an impact on how MAW services and hospital services were utilised. Moreover, user participation was assumed to be challenging due to conflicts between patients’ and relatives’ desires and what the PCPs found professionally justifiable.

Our findings indicate that PCPs do not have an overview of the competence of MAW personnel and resources of MAWs. For new health service models to be implemented and utilised as intended, our findings indicate that including key personnel perspectives is essential. Hence, it is important to increase the flow of information about alternatives to hospitalisation. It is also important to control whether this information is actually received.