The COVID-19 pandemic has stress-tested health systems around the world and in doing so has accentuated pre-existing health inequities.[[1]] This includes disparities in health outcomes and access to health services for rural populations. In the United States, for example, COVID-19 case fatality rates are higher in rural areas, and rural hospitals have struggled to deliver the advanced respiratory support needed by many COVID-19 patients.[[2,3]]

Although New Zealand has to date been spared the worst health and socioeconomic impacts, there was no certainty that this would be the case in the pandemic’s early stages. New Zealand’s rural hospitals, along with the health system as a whole, needed to plan for the scenarios that were at the time unfolding in Europe and North America. The Ministry of Health (MoH) reports COVID-19 case data at a district health board (DHB) level, which does not permit accurate urban–rural comparisons. Rural clinicians’ impressions that rural hospitals (especially those in tourist areas) were dealing with a disproportionate number of the very early COVID-19 cases requiring hospital admission is supported by the available data and news releases.[[4,5]] The first COVID-19-related death in New Zealand occurred in a rural hospital.[[6]] In some regions, the level 4 lockdown saw an exodus from the cities to rural areas, compounding the pressure on rural health services.[[7]] Iwi in Northland and the East Coast set up road blocks to keep COVID-19 out of their areas.[[8]] These factors combined to heighten the anxiety and sense of urgency felt by those working in the rural health sector.

Even before factoring in access to advanced respiratory care, rural New Zealand is a high-risk population with respect to the COVID-19 pandemic. New Zealand’s rural towns have on average the lowest socioeconomic status, highest proportion of Māori, oldest age structure and highest levels of dependency of any of New Zealand’s geographic categories.[[9,10]] There is evidence of poorer health outcomes for residents of rural towns, an effect that is accentuated for Māori.[[11]] Māori retain a strong historical memory of the disproportionate burden their rurally based communities bore during previous pandemics, and in 2020 the projected COVID-19 infection fatality rate for Māori was 50% higher than for non-Māori.[[12,13]]

New Zealand’s rural hospitals are estimated to serve at least 10% of New Zealand’s total population.[[14]]

Rural hospitals deliver a range of inpatient, outpatient and community services. These services are often integrated and do not align neatly with the concepts of “primary” and “secondary” care that are used to organise urban services.[[15–18]] International definitions of rural hospitals are varied and highly country dependent.[[19]]

In 2008 the Medical Council of New Zealand (MCNZ) recognised the scope of rural hospital medicine (RHM) and the Royal New Zealand College of General Practitioners’ (RNZCGP) Division of RHM (DRHMNZ) was established along with a vocational RHM training programme. Although there is no formally recognised MoH definition of rural hospitals in New Zealand, the defining features of rural hospitals accepted by the MCNZ and DRHMNZ are their geographic distance from specialist services, their acute in-patient bed capacity and their predominantly generalist workforce.[[20]] Rural hospitals work closely with their relevant clinical referral facilities or base hospitals. The current DRHMNZ list of 24 rural hospitals, used for purposes of its training programme, is shown in Figure 1.[[20,21]] New Zealand’s rural hospitals have neither the specialist anaesthetists/intensivists nor the facilities necessary to manage ventilated patients beyond brief periods in an emergency situation prior to transfer to a base hospital. Rural hospitals continue to face chronic medical staffing shortages and remain heavily reliant on locums.[[22]] The extent to which New Zealand rural hospitals improve access to healthcare, improve health outcomes and improve health equity for rural communities, particularly for Māori and Pacific peoples, is currently unknown.

Figure 1: New Zealand’s rural hospitals.[[20,21]]

By applying pressure to a poorly understood part of the health system, the pandemic has provided an opportunity to further our understanding of rural healthcare delivery in New Zealand. The aim of this study was to explore rural hospital doctors’ experiences of the COVID-19 pandemic, with particular emphasis on the rural hospital–base hospital interface.

Methods

Participants

Rural hospital doctors working clinically at the frontline during the pandemic were invited by email to participate in the study. Recruitment of participants was facilitated by the Rural Hospital Clinical Leaders Forum. Sampling was purposive with the aim of recruiting one representative from each rural hospital across the country: 21 of the 24 rural hospitals recognised by the RNZCGP DRHMNZ were invited to participate (logistical factors prevented an approach to all 24 rural hospitals). Those rural hospitals that had not provided a response were then directly contacted by email with a further invitation for one of their medical staff to participate. Access was facilitated through four of the authors (GN, KB, SW and RM) being known to participants both as peers and through regional and national rural hospital medicine networks.

Data collection

Semi-structured individual interviews were conducted via videoconference (Zoom) by a member of the research team (GN, TS or SW) between August and October 2020.[[23]] The interview schedule explored each participant’s view of the base hospital–rural hospital interface during the pandemic, and included questions regarding inter-hospital transfers as well as processes for ensuring equity of access to services. Each interview lasted between 30–40 minutes and was recorded and transcribed using Zoom’s inbuilt automatic transcription service. The researchers took notes during the interviews to record participants’ responses in detail. At the completion of each interview, the interviewer listened to the recording to check the accuracy of the participant’s responses in the transcription, and to ensure the documented responses were a comprehensive account of the interview.

Analysis

A thematic analysis was conducted using a framework-guided rapid analysis method.[[24]] KB, GN and TS developed a structured template by which the data were classified according to the study topic guide questions. The interview data (including corrected transcripts and the interviewer’s summary) were converted to a standard summary format. All interview summaries and templates were individually reviewed by each team member. The research team then met (via videoconference and once in person) to review summary responses and templates and to refine emerging themes with reference to the original recorded responses. Similar themes were grouped together and relationships between themes explored.

Researcher positionality

The research subject and setting was part of the real-life experience of all members of the research team except TS. Although “insider status” can be a research strength, it is important that the research remains rigorous.[[25]] In this study, rigour was promoted by (1) acknowledging the insider status of the researchers in participant information and consent forms and (2) a whole-team approach to analysis, which included the participation of TS, who is not a rural hospital doctor.

Ethics

Ethics approval was obtained from the University of Otago Human Ethics Committee D20/150.

Results

Seventeen interviews were conducted with rural hospital doctors representing 17 rural hospitals (four failed to respond). Ten participants were based in North Island rural hospitals and seven in South Island rural hospitals. The majority of participants were vocationally registered in RHM and two were senior RHM registrars. Participant characteristics are shown in Table 1. Participants were designated a number (P1–P17) and were referred to throughout the study by this coding (eg, P5).

Table 1: Characteristics of participating doctors.

Participants’ accounts covered five themes: initial reaction to COVID-19 pandemic, local leadership, the rural hospital–DHB relationship, understanding the rural hospital context and access to advanced care.

Initial reaction to the COVID-19 pandemic

Participants’ accounts as the pandemic began portrayed a sense of being alone, waiting and worrying. The usual communication channels and everyday processes linking rural hospitals to their relevant base hospitals had been disrupted. The role rural hospitals would take in the pandemic was uncertain. With DHBs assumed to be preoccupied with preparations within their base hospitals, participants reported being left to themselves, without clear directives:

“[F]or those first two and a half weeks we essentially were rudderless. Yeah it was quite a big moment… Holy shit we are on our own.” – P2

At the same time, rural hospitals were receiving an enormous barrage of information, which was difficult to get on top of, let alone make sense of:

“[We were] feeling very battered around by rapidly changing information coming from a number of different places that was frequently changing and often mutually contradictory.” – P4

Rural hospital teams rapidly realised that they would have to take responsibility for their own pandemic planning:

“So we felt we were left to ourselves… we had to blunder our way through that.” – P3

Local leadership

Going into the pandemic, hospitals either had, or did not have, established local leadership. Established local leadership facilitated a rural hospital’s ability to make an effective local response. Small, flexible and well-connected teams could respond and adapt quickly. Participants described how their hospital teams divided up responsibilities and actively looked after each other:

“[We] know each other well enough to know ‘this person is going to be really good at this, that person is going to be really good at that’. We identified support: ‘second- in- command’ and support people that were particularly there to safety- net those clinical leads or site managers who were going to be bombarded.” – P5

Established local leadership also meant that hospitals had the mandate and thus the confidence to adapt policies and other advice coming from outside to ensure these were locally relevant. The local leadership thus acted as an information filter.

In rural hospitals where established local leadership was absent, an effective and timely local response was initially more challenging. However, participants described how their existing small, well-connected teams helped individuals step up into a clinical leadership role and also enabled the local leadership team to be confident in making their own decisions for their community:

“From a motivation point of view, we've always had a number of individuals that stand out as the local ‘Clinical Directors’ even though that's not a recognised official role.” – P16

“We came together as a senior medical officer group and started to make our own decisions that were relevant for our community and our hospital. That was when we started to gain strength and confidence in our response.” – P5

The rural hospital–DHB relationship

District health boards’ support for their rural hospitals varied widely and largely reflected the status of the pre-existing relationship prior to the pandemic. Where there were established relationships (at both managerial and clinician level), the communication channels were better developed and the engagement was more effective:

“Relationships that existed prior… it makes things much easier, our Clinical Director had very well established links with… [DHB CEO]… that I think was essential in trying to get that visibility.” – P7

Rural hospitals that had previously faced major events had established clear processes with their DHB and drew on that experience to manage their response.

Participants who perceived a highly functional and supportive relationship with their DHB emphasised the importance of autonomy and of adapting policy and procedures to the local context:

“The DHB can tell you what their advice is, but in rural hospitals you have to make your own or devise, your own plan and setup. The DHB can’t tell you exactly what to do, but they would give recommendations and I generally found that they were supportive, but were giving us freedom.” – P9

However, many participants reported feeling unsupported by their DHBs, describing a continued pattern of poor communication:

“[It] just again highlighted this big mess of confusion and mixed messages and incoherence from the centre to the periphery.” – P13

This resulted in an inability to establish clear processes, including those for patient transfer for advanced care. Some participants even went so far as to see their hospital being completely forgotten about by the DHB. Others saw the opportunity to reconsider what their future relationship with the DHB should look like:  

“I think there has been a tendency to… have an umbilical cord, if you like. ‘You know what we need, when we need it’, and perhaps not so much to have the confidence to break that and to go out on our own. And I think what happened during this pandemic, was it really forced us to do that.” – P5

Understanding the rural hospital context

Participants thought DHBs had a poor understanding of rural hospital facilities and processes. In particular, facilities and processes were not set up in such a way that base hospital plans and protocols, especially those around escalating respiratory support, could simply be rolled-out with no adaptation to the local rural hospital context.

Participants from many rural hospitals also discussed the lack of “surge capacity” (defined as elective activities that can be temporarily halted in order to increase capacity for acute care), which was not well understood at base hospital level.

Local hospitals frequently responded with practical adaptations and innovation, often taking a “number 8 wire” approach:

“We managed in the end to work out something, a system whereby we could convert the positive pressure rooms into [negative] pressure rooms with some kiwi ingenuity and extractor fans and taping up things, but they all took a while to work out and to get it signed off by the right people.” – P7

The lack of understanding regarding the rural hospital context was also present in participant narratives that highlighted the absence of “fit-for-rural-hospital-purpose” guidance. The RNZCGP was providing guidance from early on for primary care and medical specialist colleges for secondary care. At the base hospital level, multiple hospital specialty guidelines (often conflicting) were gradually forthcoming, but there was no input from a rural hospital perspective:

“[If] they [specialities] want us to follow their best advice procedures, they need to be involving us in the discussion that helps apply to our setting, rather than just tell it. We’re not going to be a mini paediatric ward. We’re not going to be a mini respiratory ward following their processes and protocols.” – P4

Participants reflected that, to address this issue in future, rural hospitals needed to work together to raise their profile as a clinical specialty of equal standing to general practice and other specialist care.

Access to advanced care for COVID-19 patients

All participants raised concern regarding ongoing uncertainty around managing and transferring acutely unwell patients with COVID-19. There was often no explicit guidance and no mutually agreed arrangements around access to advanced care for rural hospital patients. Participants were not confident they could get patients to the right place in a timely way if the worst-case scenario had eventuated. Two particular issues were highlighted to be of concern: ambulance services and transfer issues.

Ambulance services, a key stakeholder for the escalation of care for rural patients, were subject to centralised, national processes and policies, which were frequently not aligned to local or even regional protocols. The inability to adapt these to the rural context was concerning:

“[St John] was initially just saying that they weren’t going to transport anyone where there was a respiratory problem... anyone who’s needing any kind of respiratory support… I know it was raised at a national level, because it was St John’s policy. I think one of the things that was very difficult was that this was unilaterally declared by St John to the DHB and to us all, so there was no balancing of risks, it didn't seem to us.” – P4

Transfer issues were seen as particularly problematic when participants viewed their rural hospital–DHB relationship to be poor. Here the DHB would give unilateral advice to the rural hospital, which the hospital had no choice but to follow. Such advice covered a number of situations, including admitting patients with suspected COVID-19 to the rural hospital and helicopter retrieval of acutely unwell patients with suspected COVID-19:

“[The] advice from [DHB base hospital] was that we were to admit, nobody that had suspected COVID… [instead] they were to be put in the ambulance and transferred two hours up the road to [name of base hospital]. There was no thought about what that meant in terms of transfer resources and how that would look for a nurse and two hours in the back of an ambulance in PPE.” – P16

“[We] were told we couldn't have a helicopter retrieval—the ambulance would not transfer anyone who was query COVID…” – P16

Overall, participants thought that rural patients did not have equitable access to specialist and advanced care prior to the pandemic and that the pandemic would simply exacerbate this existing inequity. Some participants seemed resigned to the continuation of this business-as-usual situation:

“[We] definitively dodged a bullet... but... that’s what we do all the time…” – P8

Others were more optimistic, recognising the good intent from DHBs, while also acknowledging further progress is needed to better work together in future across the sector:

“We were extremely worried about it [equitable access for advanced care] and I think [name of DHB] were not insensitive to those issues. And I think that’s one of the things that drove them being quite proactive about including us because they saw our communities in the rural hospitals being vulnerable to be fair, I think there was a real commitment at the DHB level to try and fight against that. I don’t know if they necessarily understood the complexities of that as well as we did and whether they actually put in place effective measures.” – P4

Discussion

This study has found that the experience of planning for the pandemic highlighted the challenges rural hospitals face in operating at the margins of the healthcare system. In the early pandemic phase, participants felt “forgotten” and at the same time overwhelmed by large amounts of contradictory information. This initial phase was followed by a realisation that a local response was needed, something that small, well-connected teams were able to rapidly deliver. Pragmatic innovation and flexibility were features of the local responses. Local leaders proved to be important facilitators, proactively managing external relationships and acting as a filter that adapted centrally generated policy and guidelines to the local context. A notable finding was the large variation in participants’ experiences of the rural hospital–base hospital interface, something that was largely determined by the quality of the pre-existing relationship with the relevant DHB. All participants raised concerns regarding ongoing uncertainty around the management and transfer of acutely unwell patients with COVID-19.

Although rural hospitals provide a spectrum of primary and secondary services, they are all small components of much larger urban-centric DHB structures and national professional bodies. Rural hospitals are not a homogenous group and there is no national strategy or policy that considers their role. It is not surprising that national and regional bodies failed to provide rural hospitals with clear direction during the pandemic. This left their senior RHM staff to synthesise copious volumes of only partially relevant primary and secondary care guidelines and policies. This experience adds weight to the statement in the Health and Disability System Review interim report that there is a “clear need for a better understanding of the form, structure and function of rural hospitals and their contribution to health service delivery, and have a strategy for their development.”[[26]]

Those rural hospitals with established clinical leadership roles entered the pandemic with a clear advantage over those where leadership was non-existent or distant. Clinical leadership and clinical governance have been slow to evolve in New Zealand’s rural hospital sector,[[27]] although there is evidence of progress.[[22]] The pandemic may have accelerated this process by highlighting the need for leadership at a time of crisis and by encouraging younger or informal leaders to step up into formal leadership roles. Prior to 2008 and the establishment by MCNZ of the RHM scope, medical staff in rural hospitals were poorly connected to their peers in other hospitals. The RHM professional structure has not yet matured to the point it can offer the level of clinical direction that came from other Colleges during the pandemic, but it created a network that facilitated information sharing and helped overcome some of the isolation experienced during the pandemic.

There was a universal sense among the participants of this study that the DHBs had a poor understanding of rural hospitals’ facilities and capabilities, but this was not what determined the quality of the relationship. More important was that the base hospital listened to their rural colleagues respected them as experts in the rural context and provided both the support and the autonomy to develop local solutions. The principle of subsidiarity is considered in the charter of at least one New Zealand rural health service and may usefully underpin all highly functional DHB–rural health service relationships.[[28]]

Equity concerns centred on access to advanced care. Transferring large numbers of highly infectious patients requiring ventilatory support would have presented a major challenge needing careful planning and agreed protocols. When considering rural hospital to base hospital transfers, consideration needs to be given to both the mode of transfer and the appropriate clinical thresholds for transfer. It is concerning that most participants felt that this issue (acute inter-hospital transfer) remained unresolved, and they were only able to express relief that this part of the system remained untested. Planning in the key area of inter-hospital transfer seems to have been complicated by the different perspectives of the three essential players: the ambulance service (centralised nationally), the base hospital referral service (DHB with a regional perspective) and the rural hospital (with a local perspective).

Limitations

The study’s perspective is that of individual medical staff and does not include the views of other rural hospital staff or the wider rural community. Nor does the study consider the perspectives of those working in base hospitals and DHB offices on the “other side” of the interface explored here. The early phase of the pandemic was an exceptional time characterised by uncertainty and anxiety. Care needs to be taken in extrapolating the findings in this study to more “normal” times.

Implications and future research

Rural hospitals matter to New Zealand rural communities, which is best demonstrated by the consistent community responses to threats of rural hospital closures or downgrades.[[29]] International studies have identified rural hospitals as important providers of healthcare that can benefit the health of rural populations by enhancing access to, and integration of, health services.[[17–19]] Although their role remains poorly understood, rural hospitals appear to be uniquely positioned to improve health equity for rural communities, particularly for Māori and Pacific peoples. Further research, strategy and policy at a national level is needed if they are to fully realise this potential.