The Impact of COVID-19 on Chronic Care: A Qualitative Study Among Primary Care Practices in Belgium

Background. The COVID-19 pandemic affects the processes of routine care for chronic patients. A better understanding helps to increase resilience of the health system and prepare adequately for a second wave or are-ups of the pandemic. Methods. A qualitative study was conducted in 16 primary care practices: 6 solo working, 4 monodisciplinary and 7 multidisciplinary. 21 people (doctors, nurses, dieticians) were interviewed, using semi-structured video interviews. A thematic analysis was done using the domains of the Chronic Care Model (CCM). Results. Three themes emerged: changes in health care organization, risk stratication and self-management support. All participating practices reported drastic changes in organization with a collective shift towards COVID-19 care, and reduction of chronic care activities, less consultations, and staff responsible for self-management support put on hold. A transition to digital support did not occur. Few practitioners had a systematic approach to identify and contact high-risk patients for early follow-up. A practice with a pre-established structured team collaboration managed to continue most chronic care elements. Generally, practitioners expected no effects of the temporary disruption for patients, although they expressed concern about patients already poorly regulated. Conclusion: Our ndings show the delivery of chronic care as disrupted. It indicates that the establishment of the CCM can facilitate continuity of care in crisis times. Short term actions should be directed to facilitate identifying high-risk patients and to develop a practice organization plan to organize chronic care and use digital channels for support, especially to vulnerable patients, during a second wave or in are-ups.


Background
Resources at all levels have shifted away from chronic disease management and prevention during the outbreak, and the lock-down of many services has translated into reduced access, a decrease in referrals and reduced hospitalisations of patients with non-COVID-19 pathology 6 . Scattered reports suggest chronic patients have used alternate pathways or have postponed health care seeking 7 . In addition, patients have less options for communitybased support and care. This leads to a serious concern about the indirect health footprint of COVID-19, especially on chronic diseases with increased complications and accelerated progression due to delayed and diminished access to secondary care and to a disruption in follow-up at primary care level.
These concerns indicate the need for an analysis of chronic care adaptation during the COVID-19 pandemic. Primary care providers have been struggling how to organise chronic care amidst the peak of the outbreak, when infection risk was high and resources extremely tight 8 . Pressure of COVID-19 on primary care is well documented, but the associated adaptation for chronic care is less so 9 . Chronic care models are based upon productive and active interactions between a patient, their informal care givers and the health care team, facilitated by a strong health care organisation and community embedding. How is such a model adapted in a context of a pandemic, in which the danger for serious and widespread infections absorbs most resources, and drastically changes the physical and social context in which to deliver care and support? How do primary care providers adapt their chronic care models to emerging crisis situations? How is the workforce adapted and what is prioritized? And what can we learn from these responses for the resilience of chronic care models?
This paper addresses this gap by examining the primary health care response among primary care providers in Belgium. The study aims to examine how both content and delivery of chronic care is being affected by the pandemic. Better understanding can enable us to identify ways to increase the resilience of the health system and be better prepared for are-ups of the COVID-19 epidemic and other emergency situations.

Context and study population
This study was performed among primary care practices in Belgium. Belgium has registered an estimated 842,2 deaths per million inhabitants in the beginning of June 10 , and the all-causes excess mortality is the highest in Europe after Spain and England, almost double that of the USA. 5 weeks after the rst con rmed case in Belgium, the federal government in collaboration with the National institute of Public Health Sciensano issued an emergency plan for general practice with guidelines that stated to postpone all non-urgent care and to start triage centers 11 for COVID-19-suspected complaints. After emergent signals of an increasing burden on emergency units of non-COVID-19 patients with urgent problems having postponed needed care, the guidelines explicitly allowed the provision of chronic and psychiatric care 12 if urgent and to prevent worsening. 6 weeks thereafter, general practices were allowed to re-open for non-urgent care, provided they maintained strict COVID-19-prevention and hygiene measures. With the introduction of the emergency plan, teleconsultations became permitted and remunerable, a novelty that had been postponed since many years. Suddenly, physicians were now expected to triage patients with complaints by phone. But teleconsultations were also explicitly allowed to guarantee the continuity of care for patients with chronic diseases.
The study was an additional study embedded in a larger study on the scale-up of integrated care for diabetes and hypertension, SCUBY 13 . For the current study General practitioners (GPs) were recruited in a semi-rural area in the northern part of Belgium (Flanders). General practitioners work traditionally as self-employed providers in small practices, but over recent eras more practices are being transformed into small multidisciplinary group practices. Data collection occurred until data saturation, which was reachec after twelve interviews. Purposive selection was done to recruit an equal number of GPs from different types of practices: solo working, monodisciplinary group practices and multidisciplinary group practices (with at least one nurse or a dietician).

Data collection and analysis
Because of the COVID-19 related restrictions, interviews were held via a secured online video connection by 2 researchers. The interviews took place from over a 6 week period, starting in week 12 of the epidemic, three weeks after the initial peak. All interviews were recorded, transcribed verbatim and independently analyzed by the two rst authors. The interview guide contained questions about chronic care for diabetes and hypertension and an additional part with 12 questions about changes in care organization as a result from the COVID-19 (appendix 1).
A thematic analysis was done using the elements of the Chronic Care Model (CCM) 14 : processes and incentives to improve the care delivery system; self-management support; team function and practice systems; evidence-based guidelines and implementation support; and information systems to facilitate the development of disease registries, tracking systems, and reminders and to give feedback on performance. The COREQ checklist was lled to assure complete reporting (appendix 2) Results 16 primary care practices were selected: 5 solo working, 4 monodisciplinary and 7 multidisciplinary. Within these practices, 21 people were interviewed: 3 dieticians (all female), 2 nurses (all female) and 16 GPs (mixed female and male). An overview of the participants is provided in appendix 3. Three major themes emerged: a) changes in health care organization; b) risk strati cation; and c) self-management support.
Changes in the health care delivery system and team approach As consequence of the emergency plan and because of fear among patients, there was a general drop in consultations for chronic care. In all primary care practices, the initial response was a re-organization with a focus on securing access to and safety of acute care with much attention to COVID-19 suspects. This entailed telephonic consultation and collaboration with triage posts for patients with COVID-19 suspected symptoms and the reorganization of the practice in line with the hygienic guidelines to enable access for patients with acute non-COVID-19 related health problems. This absorbed most time and energy, leaving little room to consider anything else. 'In the beginning it was also very busy, so we just tried to do the most urgent.' (IV 7) The majority of primary care practices did not plan the (re-)organization of chronic care.
Nurses and dieticians were frequently put on temporary unemployment by the practice owner due to a loss of revenues following the drop in consultations and their services considered 'not essential'. However, practices with an established culture of dialogue took a more systemic approach with team meetings about organization and patients.
'Throughout the corona pandemic, so for seven weeks now, we have been meeting every afternoon for an hour about our patients, about the care, about the triage center, about having enough material, about cases, about yes, suicidal patients, about everything and more. Every day for an hour, so I think we are very alert for that and are ercely engaged in doing the best possible care in this di cult period.' (IV10) Collaboration and concertation with medical specialists was more di cult for non-acute matters, also because not all referral centers communicated clearly about their changes in schedule and way of working. Access for acute care was no problem.

Risk strati cation
Few GPs had made a selection of high-risk patients to proactively contact for early follow-up once possible. Most respondents recognized the value of such approach, but they mentioned barriers such as a lack of time and staff, ethical objections, and a limited capacity to use the Electronic Medical Record (EMR) system.
'I have a problem with people calling patients myself. There are colleagues who do that, but I have a bit of a problem with that. I have a regular audience, they will come.' (IV8) An important reported facilitator for pro-actively contacting patients was the availability of a list of the high-risk chronic patients extracted from the EMR, which was present in some larger group practices. GPs in solo practices indicated that they know their patients personally and that they would be able to identify high-risk patients by heart. When asked for examples of such patients, they mentioned those receiving home visits, of very old age, those not well-controlled, with recent change of medication, or patients reporting di culties. GPs would approach such highrisk patients for a face-to-face consultation at their home or at the practice.
'Firstly, we have coded everyone in our practice with chronic pathology: hypertension, diabetes, COPD, asthma. It has been very easy for us to draw lists.We also exported lists of patients with depression and oncological disorders. We started by calling the diabetes lists: if you get sick or if you feel anything contact us.' (IV13) In contrast to GPs, the dieticians interviewed stated the intention to contact all their clients for renewed appointments as soon as possible. This would also compensate for their unemployment during COVID-19.

Self-management support
The new option of teleconsultation provided primary care practices with a potential tool to monitor and support patients with chronic diseases from a distance. However, most respondents said to mainly use these teleconsultations to prescribe medications and to get a quick overall impression of the patient. 'I cannot follow diabetes from a distance. I need to take lab tests, measure blood pressure.' (IV8) 'I did ask if they wanted it by phone or skype. But there are actually very few who have responded to that.' (IV9) In addition, self-management support was usually provided by the nurse or dietician, but due to the lower revenues, these staff members were put on temporary unemployment.
'A nurse has not been able to work all the time, because everything a nurse does is not urgent or not essential or not life-threatening, or how should I put it.' (IV10)

Perceptions on the changes and effects on chronic patients
Respondents indicated that for the large majority of previously well-managed chronic patients the consequences of the COVID-19 outbreak and the associated re-organization of primary care would be limited. They argued that missing only one consultation is not problematic.
'Most of those who follow the quarterly check-ups and are stable, are not going to suddenly get worse.' (IV21) However, there were worries about the effects on some patients, speci cally those with socio-economic problems, whom they expected to experience more distress from COVID-19 and the lockdown. GPs mentioned that for these people, more unhealthy food and especially less physical exercise would probably be important causes of diabetes getting out of control.
'Because you know a lot of patients have had a lot less exercise than normal. They've only been able to nd their salvation in the fridge. So in terms of pounds and exercise, that's been dramatic for a lot of patients in the last few weeks.' (IV12) Most primary care practices were quite satis ed with the way their practice was organised and were proud of all the work they had done. Therefore, they did not plan on taking other measures next time, besides increasing their stock of protection material.
'I think that we as general practitioners and certainly we as a practice do that super well, and I think that from the side of the government some other things might have happened there.' (IV10)

Discussion
This study examined how primary care practices in a highly affected area in Belgium organized services for chronic conditions like diabetes and hypertension during the COVID-19 pandemic. Our ndings show that the delivery of chronic care was severely disrupted.
We learned that most examined practices were not able to adjust quickly to changing circumstances. To reach the most vulnerable and frail chronic patients and to keep them as healthy as possible, risk strati cation within a practice population is essential. The Kaiser Permanente`s triangle' disease management model, focusing on intensive care and support of people with complex chronic conditions 14 , may be used as an example. In the case of a new emergency, good record keeping and listing the patients in strata could help to keep in contact with these chronic patients. Pro-active measures may then be taken, using innovative techniques like telemedicine as a mitigation strategy as proposed by the World Health Organization. The potential of mobile and digital selfmanagement support channels for patients can be further promoted especially among the pre-identi ed vulnerable patients 19 .
The limitations or our study encompass the selection of our sample population and the lack of quantitative data of the care process and follow-up parameters. We sampled practices in one region in Belgium (Flanders), but a large variance of infection rates across the various regions exists as in the neighboring countries like the Netherlands 20 . The experiences of the chronic patients were also beyond the scope of the present study. The strength of our study is that we interviewed different members of a range of primary care practices.
Primary care should prepare for new are-ups or even a second wave in the medium term. Our qualitative study deepens the knowledge of how primary care practices vary in their organization of care in times of a pandemic. We suggest primary care practitioners to better identify the chronic patients in their patient population, and to proactively plan the steps to be taken in order to keep track of them, using a team-based approach. The study results generated two important pathways to achieve this: (1) a more systematic implementation of the CCM and (2) the establishment of a stable nancing structure supporting staff (nurses and dieticians) so that they can play a role in managing chronic patients in times of a crisis. This will help to quickly switch between acute and chronic services and will improve continuity of care.

Conclusions
This study shows that the COVID pandemic affected the continuity of chronic care drastically. Face-to-face consultations had to be ceased and focus shifted towards COVID care. In most practices there was no proactive reach out to patients with chronic diseases and multidisciplinary teamwork was pushed to the back burner. Otherwise, some good hope is present, as practices with reliable pre-existing structures did notably better. Important ways to improve are implementing the CCM through strati cation of the patients according to their needs and planning ahead in anticipation of are-ups or a second wave.

Consent to publish
The information about participants and the used quotes cannot be used to identify the participants and therefore do not compromise anonymity.

Availability of data and materials
The data that support the ndings of this study are available on request from the corresponding author KD. The data are not publicly available due to them containing information that could compromise participant privacy.