Abstract
Background. The role of primary healthcare (PHC) during a pandemic varies across European countries. The coronavirus disease 2019 (COVID-19) pandemic has altered the working practices of family medicine doctors and impacted the resilience of healthcare systems.
Objectives. This study aimed to examine European healthcare system responses to the pandemic, focusing on rural and urban differences.
Materials and methods. This cross-sectional, mixed-methods study used a semi-structured online questionnaire with 68 questions, including 21 free-text comments. Data were collected from May 2020 to January 2021. Key informants from 16 European Rural and Isolated Practitioners Association (EURIPA) member countries distributed questionnaires to 406 PHC doctors. Data were analyzed using descriptive statistics and nonparametric tests (χ2, Kruskal–Wallis, Mann–Whitney U) with a significance threshold of 0.05.
Results. A statistically significant difference was found between rural (36.4%, 55/151), semirural (19.4%, 24/124) and urban populations (29.8%, 39/131) regarding medicine shortages (χ2 = 9.91, degrees of freedom (df) = 4, p = 0.042). The semirural setting showed a statistically significant difference from the other settings (p = 0.004 in post hoc χ2 test). Significant differences were found between countries in resilience features including, effectiveness of triage, adapting to the rapidly changing requirements, government help, existence of a community resilience group, improved interprofessional collaboration, medicine shortage, and general practitioners (GPs) involvement in palliative care.
Conclusions. Medicine shortage was more prevalent in rural and urban areas compared to semirural areas. Differences were observed between countries in their responses to the pandemic, particularly in adapting to the rapidly changing requirements, effectiveness of triage, government help, and the existence of a community resilience group. These differences were confirmed with qualitative analysis. The results emphasize the need for tailored approaches considering diverse contexts in shaping effective healthcare system resilience.
Key words: rural health, COVID-19, resilience, primary care
Background
Caring for patients with novel coronavirus infections has placed an immense burden on healthcare systems and healthcare workers worldwide. The COVID-19 pandemic, unlike any previous health crisis, has demanded a prolonged and sustained response, akin to a marathon rather than a sprint. This has necessitated that leaders and managers within healthcare organizations carefully pace their strategies and responses to ensure long-term efficacy and sustainability of the healthcare systems.1
Rural areas tend to have a higher proportion of elderly patients and patients with multiple medical comorbidities.2 The shortage of medical specialists, hospital beds and intensive care unit (ICU) beds, as well as the financial problems of many rural hospitals, is a widespread problem worldwide, made more apparent by the COVID-19 pandemic.3, 4 The lack of personal protective equipment and medications further exacerbates such difficulties. Rural areas are particularly vulnerable during such pandemics due to a higher proportion of elderly patients and individuals with multiple comorbidities.5 The chronic shortage of medical specialists, hospital beds and ICU beds, compounded by the financial struggles of many rural hospitals, has been starkly highlighted by the COVID-19 crisis.4 Furthermore, the lack of essential resources such as personal protective equipment (PPE) and medications has exacerbated the situation.3 Since February 2020, when COVID-19 began to spread rapidly in Italy, general practitioners (GPs) worldwide have faced the daunting task of managing an unexpected and severe public health challenge.6 In the initial stages, health services struggled to recognize and treat suspected cases promptly, leading to increased infection rates and overwhelming healthcare systems. Over time, primary care physicians adapted to the new reality, incorporating teleconsultations, reducing face-to-face interactions and meeting the new demands imposed by the pandemic.7, 8 Training programs also underwent significant changes, with distance learning being rapidly and successfully implemented across Europe, even in rural areas.9
The pandemic has also exposed significant gaps in healthcare infrastructure and support systems. In countries like the UK, the inadequate provision of PPE to healthcare and social care workers during the pandemic was perceived as a betrayal by the government and National Health Service (NHS), undermining the culture of integrity, transparency and support essential for healthcare workers.10 It is important to recognize that epidemics have a significant impact on daily life, even for those who are not infected. Some recent studies have shown that community volunteers can help provide social distance and minimize disease transmission.11 The role of primary healthcare (PHC) in a pandemic varies across Europe.12 For example, in the UK, PHC has been effectively involved in managing post-acute COVID-19 conditions, often providing care remotely via telephone or video consultations.13, 14
Palliative care presents another critical opportunity for PHC during pandemics. According to Prof. Scott Murray and in line with the World Health Organization’s (WHO) statement, the COVID-19 pandemic has underscored the need to integrate palliative care into national healthcare services to enhance quality of life and alleviate suffering in the final stages of life.15
The COVID-19 pandemic has necessitated rapid innovation across healthcare systems. Family medicine practices, in particular, have faced unique challenges in maintaining care for medically complex older populations. They have had to adapt their traditionally hands-on model of care to accommodate the limitations on face-to-face contact while ensuring the safety of older, medically complex patients in their homes. This required significant flexibility, transparency, teamwork, and partnerships with external providers.16, 17
Objectives
This study aimed to examine the attitude of European healthcare systems in responding to the pandemic, with particular attention to PHC and the difference between rural and urban areas. Searching and understanding the specificity of differences in PHC between European countries, considering urban and rural settings, is essential for creating a more resilient, inclusive and effective healthcare system that can better withstand future challenges and emergencies.
Methods
Study design and setting
This study is based on a survey of key informants from 16 member countries of the European Rural and Isolated Practitioners Association (EURIPA) that operate under the umbrella of the World Organization of National Colleges, Academies and Academic Associations of General Practitioners/Family Physicians (WONCA Europe). More details on the study design and setting are described elsewhere.18
Procedure
The steering committee of this project, called the EURIPA COVID-19 study, developed a semi-structured questionnaire with 68 questions, 21 of these including free-text comments. The open-ended (free-text) responses to some questions were included to explore the issues pragmatically and at a deeper level, considering the complexity of COVID-19 pandemic management.19 The 1st draft of the questionnaire was based on the research objectives through an extensive literature review. Subsequently, a panel of PHC experts and 1 methodology expert used a Delphi method to evaluate the validity of the items and the length of the questionnaire, formulated suggested changes and identified missing items. The research team then discussed all feedback until consensus was reached, and a 2nd version of the questionnaire was developed. Data were collected between May 2020 and January 2021 using Google’s online survey tool – Google Forms.
Validity
The psychometric properties of the questionnaire were assessed, focusing on validity as a theoretical and an empirical construct. During the development of the questionnaire, face validity (whether the questionnaire measures at first glance what it purports to measure) and content validity (whether the items adequately represent the entire domain that the questionnaire attempts to measure) were tested. In each case, this was done by EURIPA PHC experts, who are all international authorities in the field of rural healthcare. The informants were contacted directly via email by the national coordinators, and the response rates were all above 50%.
Participants
The EURIPA, which currently includes 31 member countries, is a representative network organization founded by family doctors to address the health and wellbeing needs of rural communities and the professional needs of those serving them across Europe (https://www.euripa.org/page/home). Although there are several official definitions of rurality for example those developed by WHO and EU, all of them based on solid objective data but with severe limitations in capturing the difference in the mentality between urban and rural setting, for the purpose of our study we have preferred a self-reported statement of the respondents, who characterized the place of their practice as rural, semirural and urban, and this kind of procedure has been repeatedly used in other international studies in PHC.20
The informants were contacted directly by the national coordinators; they were required to be primary care practitioners (PCPs, i.e., any professional working in PHC such as a doctor, nurse, physiotherapist, or assistant) with a good command of English, as the survey was written in English and was not translated into other languages. The baseline characteristics of the informants, broken down by country, and some descriptive statistics are shown in Table 1.
Because a convenience sample of informants was used, the PCPs for each country may not be representative, although there was an attempt to achieve geographical variation.
Study size
When the number of respondents reached or exceeded 30 for countries with a population of 35 million or more, and 20 for countries with a population of less than 35 million, data collection for that country was terminated.
Main outcome measures
The questionnaire consisted of 68 questions including sociodemographic variables, length of clinical experience and experience in dealing with the COVID-19 pandemic, and geographic location (see Supplementary data). The questionnaire was divided into several sections. In this paper, we analyzed the questions related to healthcare system resilience. The other sections of the questionnaire will be analyzed in separate papers.
Participants with complete data could not be distinguished from the less than 10% of participants with incomplete or missing data, which were then considered “Missing Completely at Random” (MCAR) data. We used the “Complete Case Analysis” method in IBM SPSS v. 29.0 (IBM Corp., Armonk, USA) and removed all participants with incomplete data from the analysis.
Analyses
To describe baseline characteristics, proportions were calculated for dichotomized or categorized data and means for numeric variables. We used the mean also for non-normally distributed data to show some slight differences between countries.
Pearson’s χ2 test of independence was used to measure the association between categorical variables, followed by the post hoc χ2 test based on comparing of response frequencies between a given country and all the other countries together. To control the false discovery rate, the Benjamini–Hochberg correction was used with the aid of the R package ‘base’ (the command ‘p.adjust’). As the numeric data were not normally distributed, when comparing the scores between the groups, we used nonparametric tests such as the Kruskal–Wallis test with the post hoc Dunn’s test (with the Bonferroni correction) and the Mann–Whitney U test. The statistical significance threshold was set at 0.05.
Statistical package: statistics were obtained using IBM SPSS Statistics for Windows v. 29 (IBM Corp.), Microsoft Excel 2013 (Microsoft Corp., Redmond, USA) and the R environment v. 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria).
Because responses were limited to short sentences for the open-ended questions, a brief conceptual qualitative content analysis was conducted. Responses (direct quotes) from PCPs were independently reviewed by 2 members of the research team (D.K. and O.G.).
Ethical issues
Respondents’ answers were collected anonymously; no formal approval from an ethics committee was required in the countries participating in the survey. Informants were aware that they could stop the survey at any time. Informed consent was obtained. Confidentiality and anonymity were assured. The study was conducted in accordance with the principles of the Declaration of Helsinki.
The questions analyzed in Table 2 are as follows:
• Have you been able to receive help from community volunteers in this pandemic?
Yes, Maybe, No
• How is the triage of patients with suspected COVID-19 working in your rural area?
(5-point Likert scale)
• How easy have you found it to keep up to date with rapidly changing requirements?
(5-point Likert scale)
• How is your government addressing the needs of clinicians and patients in rural and remote areas?
(5-point Likert scale)
• Is there a community resilience group meeting to coordinate access to essential services including transport?
Yes, Maybe, No
The question analyzed in Table 3 is as follows: Did the cooperation and collaboration with other disciplines/services change during this pandemic?
Yes, Maybe, No
The question analyzed in Table 4 is as follows: Have you and your patients had any issues with access to supply of certain medicines during this pandemic?
Yes, Maybe, No
The question analyzed in Table 5 is as follows: Do you administer supportive and palliative care on your own in COVID-19 patients?
Yes, No, Other
Results
Participants
A total of 406 PHC informants from 16 European countries completed the questionnaires. Of these, 245 (60.5%) were female and 160 (39.5%) were male. Regarding practice location, 152 PHC informants were rural (37.5%), 124 were semirural (30.5%) and 130 were urban (32.0%).
Descriptive data
The mean age of respondents was 45.9 years (standard deviation (SD) 11.30) and the mean length of service was 18.2 years (SD 11.6). Three hundred and eighty-one (93.8%) of the respondents were GPs/family doctors. The remaining 24 respondents (6%) represented a range of other medical specialties, including nurses, doctors, managers, social workers, midwives, dentists, and physical therapists.
Main results
The baseline characteristics of the informants, broken down by country, and some descriptive statistics are shown in Table 1. The countries involved in the study are illustrated in Figure 1.
Community volunteers’ help
There were differences between countries on the questions related to community volunteer assistance. According to our informants, the percentage of those who stated that they have received community volunteers’ help varied from 10% (2/20) in Romania to 52.2% (12/22) in the Czech Republic (Table 2). The relationship between the variables “country” and “community volunteers’ help” was statistically insignificant (χ2 = 34.627, degrees of freedom (df) = 30, p < 0.256).
Qualitative analysis
Community help was differentiated within a country and organized locally. Negative comments dominated: “Nobody offers any kind of help. They run away from it” [Croatia_1]; “There are no volunteers in our region” [Greece_3; Romania_2; Slovakia_3,4]. The assistance was mainly related to volunteer support for patients staying at home: “We had sewing masks for patients in the first wave of COVID” [Czech Republic_7]; “PSAs from volunteers, companies or friends” [France_1; Slovakia_2,5; Ukraine_1,3]; “They delivered food and medicines to patients’ homes” [Georgia_1,2,3,4; Israel_2]; “Civil Protection and other volunteers delivered free drug prescriptions and medicines to patients with chronic diseases. They also distributed face masks to patients” [Italy_1].
Effectiveness of triage
In our study, most respondents stated that triage worked well, with an average score above 3 on the Likert scale, while only 3 countries, Slovakia, France and the Czech Republic, scored below 3. Respondents indicated that guidelines on clinical management of COVID-19 existed in all countries and that triage was organized similarly.
The lowest score was obtained in Slovakia (2.50, SD 1.40) and the highest score in Georgia (3.81, SD 1.00), but we must point out that this is the “subjective” perception of the success of the triage by the respondents and not an objective evaluation (Table 2, Figure 2). There was a significant difference across the 16 European countries (Kruskal–Wallis test: H (15, 400) = 33.683, p = 0.004). Post hoc Dunn’s test of the differences in the effectiveness of triage between the countries is illustrated in Table 6.
Qualitative analysis
Guidelines existed in all countries, and triage was organized similarly: “We have algorithms for this” [Croatia_5]; “We follow the described symptoms. If it is necessary, patients have to call the ambulance” [Czech Republic_8]; “Special places for those with suspicious symptoms, organization of remote consultations, for those with high suspicion of COVID-testing and isolation measures, contact tracing in coordination with health authorities via a special online application” [France_1; Greece_4; Hungary_2,3; Moldova_1; Poland_1; Romania_1,3; Turkey_1,3]; “Telephone interview before ordering an appointment, triage at the entrance of the clinic” [Israel_2]. However, “inadequate communication between GPs and hospitals was noted” [Czech Republic_9].
Adapting to the rapidly changing requirements
In our survey, responses on a 5-point Likert scale showed that the majority of our informants found it relatively easy to adapt to the rapidly changing requirements (score above 3). Only in France, Hungary, Italy, Romania, and Slovakia the average score was below 3. The free-text responses showed that GPs were rather confused by the “too many sources, too many changes.” The lowest average score was obtained in Hungary (2.35, SD 0.88 and the highest in Georgia (3.48, SD 0.85, Table 2, Figure 3. There was a significant difference in the responses to this question across the 16 European countries (Kruskal–Wallis test: H (15, 402) = 42.772, p < 0.001).
Post hoc Dunn’s test results of the differences in “adapting to the rapidly changing requirements” between the countries are provided in Table 7.
Qualitative analysis
The most common response from participants was “too many sources, too many changes” [Latvia_9,17; Slovakia_1,4,13,15,20; Poland_5; Romania_1,14; Czech Republic _1,7,18,21; Spain_16,17; France_1,3,10,19,25; Greece_10; Italy_14,28,29; Israel_9]), although in some countries, the problem was “not to be informed” or “late information” [Hungary _1,4,6,16,18,20; Latvia_19,21; Ukraine _15]. Through evidence-based practices, guidelines and patient education, the information storm has both negative and positive effects. While governments published “conflicting statements”, [Czech Republic _18,21], physicians sought to ease the burden for themselves and colleagues through webinars [Greece _12; Israel _10; Italy_ 2; Spain_ 3; Poland _12,13,23], professional associations [Poland_11; Romania _11], family medicine department websites [France_1], local hospital email groups [France_9], Facebook groups [Czech Republic _11], and WhatsApp groups [Italy_19,]. Only [Slovakia _12,19] reported that “the ministry has established a crisis team that issues guidelines and distributes them via app”, while [Latvia _11; Spain_3,13; Turkey_30] reported that “... daily information was provided via email.”
Government help
Most respondents (except in Greece) rated their answers less than 3 on the 5-point Likert scale, indicating a high level of dissatisfaction among GPs with their respective governments. General practitioners felt neglected and undervalued compared with hospital staff.
Our respondents seemed disappointed with the official help they received, and the average score above 3 was achieved only in Greece (3.15, SD 1.1) and Georgia (3.48, SD 0.8). All other countries were below 3, with the minimum in Romania (1.70, SD 0.9; Table 2, Figure 4). The differences between the countries were statistically significant in Kruskal–Wallis test (H (15, 400) = 61.981, p < 0.001). The results of the post hoc test for pairwise comparisons are shown in Table 8.
Qualitative analysis
Many GPs seemed dissatisfied with the help they received from their government: “They did not even answer my questions”; “Often we hear some information in the news rather than getting it officially”; “they did not take us important” [Croatia_2; Hungary_3; Spain_4]; “We often have to help ourselves”, “They do not think at all”; “The doctors are left to their own devices”; “It was chaotic”; “We need more staff”; “The health centers were closed and all appointments were centralized in one village. Three months later, the remaining health centers are still closed and patients have to travel several kilometers to reach a nurse/doctor when needed” [Czech Republic_2,3; Italy_4; Slovakia_4, Spain_5,6]; “It all depends on the financial possibilities of the office itself”; “Poor (you have to buy almost everything out of your own pocket)” [Croatia_2; Ukraine_2]; “Pandemic or not, the problems are the same”[France_9]; “Where is the government?”, “They have their heads in the clouds, they have no idea what is real!” [France_10; Romania_1]. “Not fast sometimes”; “slow”; “information, medical equipment and security for employees provides a long time”; “sometimes a small delay, also depends on the personal efforts of a particular worker, a pretty good situation” [Greece_1; Latvia_2; Poland_1; Slovakia_2]; “The government only cares about fines and restricts people” [Greece_2]; “I think they just issue some guidelines on a website and do not really know if we have read or understood them” [Greece_3].
Some respondents indicated that the situation is much worse in rural areas than in urban areas: “I feel that hospital staff are more visible and therefore more appreciated than general practitioners.” “The big cities are prioritized, but our needs in rural and remote areas are not adequately addressed.” “Paradoxically, the needs of physicians and patients in rural areas are less important than those in large cities and hospital centers, since infections are sporadic in rural areas COVID-19”; “Rural areas neglected”; “They think of urban patients, not rural, our needs are different”; “No facilities have been provided, deficits in rural health and primary care”; They think of hospitals or GP in cities” [Croatia_4,5; Greece_4; Italy_1,2,5,6; Spain_3,7,9]; “We have 10 liters of disinfectant, 2 boxes of disposable gloves, and 2 boxes of surgical masks, that’s all”; “We do not have enough mobile sampling units and it’s difficult to perform tests, especially in the elderly, because only PCR tests are accepted” [France_5,7; Poland_4; Slovakia_6]; “Financial support for new clinicians in rural areas” [France_6]; “no material excluding mask”; “free availability of protective material” [France_8; Poland_5].
Other respondents indicated that the situation was not so different between urban and rural areas: “Similar to urban areas, chaotic, not enough”; “Needs are not met worse than in urban areas”; “As in cities”; “I have not noticed any difference between rural and urban areas”; “I think there are no differences between rural and urban areas [Czech Republic_1; Italy_3, Poland_3; Slovakia_5; Spain_1,8]. “For national communication, I did not see any difference between rural areas. But for overseas areas, there were specific elements due to different aspects of the epidemic there”; “No difference between rural and urban areas” [France_1; Latvia_1].
Some respondents indicated that local organizations seemed to be more helpful than the central government: “Mainly focused on hospitals, more local and syndicate initiatives that are more helpful”; “local organization seems more efficient”; “patients and community really help doctors with PPE supplies from local companies” [France_2,3,4; Spain_2]’ “Through different associations and organizations”; “Our country doctors association keeps us informed”; “the local SANEPID is responsible for this”; “The COVID guide is regularly updated”; “they submit it to local governing bodies” [Croatia_1; Latvia_3; Poland_2; Turkey_1; Ukraine_1].
Some respondents acknowledge that governments have done their best: “They provide enough instructions and equipment”; “They try to do everything”; “They try to do the best, but sometimes it is not enough” [Greece_5; Hungary_1,2]; “Supply well delivered help provided by the army for the elderly housing”; “Psychological, financial support”; “The very precarious help came after the acute phase of the pandemic”; “They quarantined whole villages or neighborhoods, with policemen and soldiers at the borders, with a mobile hospital of the army on site”; “Good” [Israel_1; Moldova_1; Romania_2; Slovakia_1,3; Spain_10].
Existence of a community resilience group
Regarding the presence of a resilience group, our study found that only Moldova, at 64.80% (13/19), Croatia, at 60% (12/20), and Slovakia, at 50% (10/20), met or exceeded the 50% threshold. The majority of respondents indicated that there was no resilience group in the community while in Hungary, only 10% declared the existence of a resilience group (2/20); Table 2.
The relationship between these variables “country” and “existence of a community resilience group” was statistically significant (χ2 = 69.80, df = 30, p < 0.001; see Table 9 for post hoc test results).
The analysis of the free text answers showed that GPs were not sure about the existence of these resilience groups, or they denied their existence and often stated that there were some local initiatives organized by the community.
Qualitative analysis
Sometimes GPs were unsure about the existence of these community groups or denied their existence: “I do not know. Maybe local initiatives.” [Croatia_1]; “There are none in my practice” [Croatia_2]; “Everything works as it used to” [Czech Republic_1]. They often said that there were some local initiatives organized by the community: “Family, friends and community are enough”; ‘’through the local city council”; ‘’Yes, the responsibility lies with the city hall”; “In the city, a crisis team is set up to deal with local problems”; “In the municipal administration” by RÚVZ, self-government, welfare, army, police” [France_1,2,3; Moldova_1; Poland_2,5; Slovakia_1,2]; “Well-organized municipal and emergency services” [Israel_1]; “In a municipality, the mayor delegates social services and emergency services to help everyone” [Romania_1]. Some respondents pointed out the existence of home help: “The program helps at home”; “There is online medical training” [Greece_1; Poland_1].
Some respondents mentioned well-known civil organizations such as the Red Cross and Civil Protection: “Civil Protection provides access to important services in certain cases”; “Civil organizations such as the Red Cross, etc.” [Italy_1].
Regarding assistance for transportation, there were conflicting answers. Sometimes there was help, sometimes this was a burden placed on family members. Respondents reported implemented solutions for the “transport of COVID-positive or suspected patients” [Poland_3,4].
Other respondents mentioned that “transport services, we had them also before COVID” [Slovakia_4,5,6,7,8]. However, in Spain “transportation belongs to families” [Spain_1,2].
Improved interprofessional collaboration
While only 15.8% of informants in Greece (2/20) indicated this improvement, in many countries, informants indicated an improved interprofessional collaboration of 50% or more: Croatia, Hungary, Latvia, Moldova, Poland, Romania, Slovakia, and Spain, with a maximum in Moldova (73.7%, 14/19; Table 3). The relationship between “country” and “improved interprofessional collaboration was statistically significant (χ2 = 57.085, df = 30, p =0.002; see Table 10 for post hoc test results).
Qualitative analysis
Most respondents pointed to difficulties in communication between different levels of healthcare provision: “Minimized” [Croatia_1]; “System collapsed.” [Croatia_2]; “We were able to solve most problems on our own” [Croatia_3]; “Postponement of surgeries, prolonged waiting time for appointments with specialists” [Czech Republic_1,3,4,6,8; Hungary_1; Italy_2,3,4; Latvia_3; Poland_2,3; Slovakia_2]. These difficulties existed before the pandemic: “poor cooperation before and during the pandemic” [Italy_5]. However, there have been improvements in some areas: “More correspondence and understanding” [Croatia_4]; “More mails and phone calls” [Czech Republic_2]; “Very good relations with pharmacies and medical analysis laboratories” [France_3]; “Better cooperation with nurses” [France_4]; “More active communication with public health institutions; phone consultations from doctor to doctor were introduced” [Latvia_1,2]; “Very good cooperation with the Department of Epidemiology” [Slovakia_1].
Medicine shortage
The majority of informants denied the existence of medicine shortage, except in Romania (Table 4), but there were statistically significant differences between countries (χ2 = 72.819, df = 30, p < 0.001; see Supplementary Table 1 for post hoc test results).
Georgia, Israel, Italy, Romania, and Spain reached the statistical significance of the existence of medicine shortage, but the pattern of the differences was inconsistent among these countries. In Romania, 60% of respondents noticed medicine shortage (vs 25.9% in other countries), while in Israel and Georgia, over 90% of respondents (vs about 50% in other countries) did not.
Qualitative analysis
Some respondents confirmed the existence of problems with certain medications or PPE: “Another type of medication. For hypertension, anti-tetanus shots” [Croatia_1]; “Antiseptics” [Croatia_2]; “Paracetamol was not available for a short time” [Czech Republic_1]; “Depending on what was said on TV about what works against COVID, these drugs were sold out immediately, e.g., vitamin D or isoprinosine” [Czech Republic_3]; “Gloves, hydroalcoholic solution, disinfectant, PSA” [France_1]; “No hydroxychloroquine, hydroxychloroquine for patients with rheumatoid arthritis for a short time, plaquenil was only available in hospital and off label, shortages of oxygen and hydroxychloroquine”; [Georgia_1; Italy_1,2,3,6,7,8,9; Romania_5; Slovakia_5; Spain_4; France_3]; “No paracetamol, paracetamol is limited” [France_4,5; Poland_8]; “We have a general shortage of almost all types of medicines due to the COVID-related lock” [Hungary_1].
The drug shortages may have been the result of stockpiling and panic buying of drugs due to fears of drug shortages during the pandemic COVID-19: “In March, people started hoarding their medicines. But this was only a temporary problem” [Hungary_2]; “some antihypertensive drugs” [Hungary_3] “sometimes some drugs (for example, some antihypertensive drugs) were missing for a few days” [Hungary_4,5]; “pneumococcal vaccine is sometimes unavailable due to high demand” [Latvia_1]; “anticoagulants” [Latvia_2,3]; “Many medicines were not available, e.g., SABA – inhalation medicines or medicines for treatment of RA” [Poland_3]; “Problems with availability of metformin, metformin (no idea why?)” [Poland_4,5,6,7,8,9,10; Romania_1,2,6]; “Levothyroxine sodium” [Poland_6; Romania_1,6]; “Insulin, zinc” [Poland _7]; “C vit” [Romania_3,4; Turkey_1]; “Supportive therapy like high dose vitamin C infusion” [Slovakia_2; Turkey_1]; “paracetamol” [Romania_3,4; Slovakia_6,7,8]; “ceftriaxone, ventilation” [Spain_1]; “no access to certain drugs like antivirals” [Spain_2,4]; “there was a shortage of midazolam” [Spain_3]; “One of my colleagues (who is a resident) got a positive result, had shortness of breath and had to wait 3–4 days to get medication. Lack of medication? Health workers?” [Turkey_2].
These problems occurred mainly at the beginning of the pandemic: “At the beginning of the pandemic, there was a longer waiting time for medicines” [Poland_1,2; Slovakia_1,4].
Some respondents denied any problems with lack of medication: “No problems with medicines, on the contrary, we can prescribe more medicines without restrictions” [Italy, 4].
Problems with examinations were also cited by respondents: “Special care was limited only for acute cases, so there were problems with thyroid sonography, ECHO or hospital care for addicted patients” [Czech Republic_2]; “Many examinations/surgeries/blood tests were delayed... and for chronic patients, it was sometimes difficult to get them to the hospital (e.g., for cancers)”; “Difficulties in surgery” [France_ 2,6].
Difficulties in hospital admission were also reported: “If the patient had a saturation of less than 90% and was elderly or demented, he was not admitted to the hospital or transported to the psychiatric hospital, even for a chest X-ray or a smear test for confirmation” [Italy_5].
General practitioners involvement in palliative care
There was a considerable variation between countries in the palliative care of COVID-19 patients in PHC (χ2 = 55.818, df = 30, p < 0.001; Table 5). In most countries, primary care physicians were not directly involved in palliative care, with the exception of Moldova, Spain, France, Greece, and Hungary (Table 5). However, the differences between the given country and the others were statistically insignificant in the post hoc test (Supplementary Table 2), except for Israel, where many respondents chose the response “probably” (80% vs 28.6% in other countries).
Qualitative analysis
This type of care was mainly provided by hospital departments: “it was in the hospital” [Czech Republic_5; Greece_1; Moldova_2; Romania_4]; “patients with acute COVID-19 are treated in the hospital. We care for them after recovery for 2 weeks” [Georgia_5]; “There are special services that deal with this” [Israel_1]; “We do not manage this type of care” [Italy_4].
In France and Ukraine, guidelines were used: “We have received guidelines to help us when needed” [France_2]; “Supportive care according to protocol” [Ukraine_2].
In some countries, some assistance was provided in PHC and continued in hospital wards: “oxygen support when transferring the patient to the nearest hospital” [Greece_2]; “psychosomatic support, consultations, medications” [Hungary_4]; “I provide telephone consultation when needed, the family calls the emergency medical service” [Poland_4; Turkey_2,4]; “We see these patients and give them all the treatments they need at home when possible” [Spain_1]; “but I have help from people in the hospital” [Spain_2].
We also analyzed the influence of the rural, semirural and urban setting and gender on the responses.
Settings
Medicine shortage
A statistically significant difference was found between rural, semirural and urban populations regarding medication shortage, with 36.4% (55/151) of rural, 19.4% (24/124) semirural and 29.8% (39/131) urban respondents reporting medication shortage (χ2 = 9.91, df = 4, p = 0.042) (Figure 5).
Existence of a community resilience group
No statistically significant difference was found between rural, semirural and urban populations regarding the existence of community resilience group, with 29.8% (45/151) of rural, 42.7% (53/124) semirural and 35.1% (51.9/131) urban respondents reporting the existence of a community resilience group (χ2 = 3.416, df = 4, p = 0.491).
Improved interprofessional collaboration
No statistically significant difference was found between rural, semirural and urban populations regarding improved interprofessional collaboration, with 47.7% (72/151) of rural, 19.4% (24/124) semirural and 29.8% (39/131) urban respondents reporting improved interprofessional collaboration (χ2 = 5.961, df = 4, p = 0.202).
Effectiveness of triage
No statistically significant difference was found between rural, semirural and urban populations regarding the evaluation of COVID-19 triage, with a mean (SD) of 3.13 (1.12) for rural, 3.11 (1.2) for semirural and 3.33 (1.07) for urban respondents (Kruskal–Wallis test: H (2, 400) = 2.668, p = 0.263).
Adapting to the rapidly changing requirements
No statistically significant difference was found between rural, semirural and urban populations regarding the evaluation of adapting to the rapidly changing requirements, with a mean (SD) of 3.01 (0.99) for rural, 3.11 (1.2) for semirural and 3.15 (1.00) for urban respondents (Kruskal–Wallis test: H (2, 402) = 1.488, p = 0.475).
Government help
No statistically significant difference was found between rural, semirural and urban populations regarding the evaluation of government help, with a mean (SD) of 2.58 (1.10) for rural, 2.66 (1.16) for semirural and 2.73 (1.14) for urban respondents (Kruskal–Wallis test: H (2, 400) = 1.254, p = 0.534).
Gender
Government help
No statistically significant difference was found between genders regarding the evaluation of addressing needs by the government, with a mean (SD) of 2.59 (1.14) for the male gender and 2.68 (1.12) for the female gender (Mann–Whitney U test: z = –0.824, p = 0.410).
Adapting to the rapidly changing requirements
No statistically significant difference was found between genders regarding the evaluation of adapting to the rapidly changing requirements, with a mean (SD) of 3.07 (0.96) for the male gender and 3.11 (0.99) for the female gender (Mann–Whitney U test: z = –0.712, p = 0.477).
Effectiveness of triage
No statistically significant difference was found between genders regarding the evaluation of COVID-19 triage, with a mean (SD) of 3.15 (1.11) for the male gender and 3.22 (1.15) for the female gender (Mann–Whitney U test: z = –0.793, p = 0.428).
Discussion
Summary of the main findings
We found statistically significant differences between countries (confirmed with post hoc test) in: effectiveness of triage (Table 6), adapting to the rapidly changing requirements (Table 7), government help (Table 8), existence of a community resilience group (Table 9), improved interprofessional collaboration (Table 10), medicine shortage (Supplementary Table 1), and GPs involvement in palliative care (Supplementary Table 2).
In terms of setting, we found statistically significant differences between rural, semirural and urban settings in terms of drug shortages, with semirural settings performing better than the other 2 settings, whereas there were no differences between settings in terms of: existence of a community resilience group, improved interprofessional collaboration, effectiveness of triage, adapting to the rapidly changing requirements, and government help.
Regarding the gender, no statistically significant difference was found between genders in government help, adapting to the rapidly changing requirements and effectiveness of triage.
Community volunteers’ help
The COVID-19 pandemic has tested the resilience of healthcare organizations (and those who depend on them), especially when services were limited. In a qualitative study which involved 10 European countries, the authors concluded that the growing demand for services, the overload of staff with new tasks (such as digitalization), and the growing threat of infections rarely found adequate support from healthcare authorities and therefore, the vulnerability of the limited (voluntary) workforce and organizational management capacity became evident.21 Our outcomes align with this cited study and the free-text responses show the desperate complaints about the lack of any form of assistance in some countries. There is an urgent need to consider the future of volunteering, including opportunities for virtual volunteering, micro-volunteering and appealing to a younger demographic.22
Effectiveness of triage
In our study, most respondents stated that triage worked well. They indicated that guidelines existed in all countries and that triage was organized similarly. Sometimes there was inadequate communication between GPs and the hospitals. Regarding participation in triage, we must keep in mind that only a small percentage of coronavirus patients are admitted to the hospital. Primary care, as the first contact with the healthcare system, has handled most of the COVID-19-related care and will continue to be the “first in and last out”.23, 24, 25
According to the literature, the combination of multiprofessional staff within PHC practices improves preparedness to manage social challenges such as a pandemic at a structural level, and therefore maintains quality and safety for patients and staff. Future policy developments should prioritize promoting such collaboration through multiprofessional teamwork.26
Adapting to the rapidly changing requirements
The free-text responses showed that GPs were rather confused by the “too many sources, too many changes.” According to the literature, it was not easy to adapt to the rapidly changing demands: more teleconsultation and less face-to-face contact with patients. However, it seems that even those GPs who were pushed to work in teams and had to get used to the departmental/organizational structures within the self-organized physician teams came to appreciate them over time.8
In Europe, general practices have responded to the COVID-19 pandemic with an adapted organization of their practice, safety regulations and the use of video consultations.26
Despite the absence of staff due to illness or quarantine, the GPs have continued to provide care.27 However, according to GPs interviewed, triage, remote care, and a lack of clarity about access to care may have led to patient safety incidents.27, 28
Government help
Most respondents (except in Greece) rated their answers less than 3 on the 5-point Likert scale, indicating a high level of dissatisfaction among GPs with their respective governments. In our free-text responses, GPs felt that they were neglected and undervalued compared with hospital staff, and this is consistent with much of the literature.20, 27
In the UK, GPs were so disappointed with the governement that the Doctors Association UK and the Good Law Project have taken legal action to force the UK government to launch a public inquiry into the failure to procure and distribute sufficient PPE for healthcare and social care workers during the COVID-19 pandemic.29 In Italy, a qualitative study of PHC showed difficulties in communicating with other local services, and a lack of coordination between services and PHC, with the latter perceived as undervalued and neglected.6
The aftermath of the COVID-19 pandemic offers a valuable opportunity to enhance the utilization of digital health tools, with a particular emphasis on integrating PHC data.30 This effort should also prioritize making knowledge accessible, including within PHC. It is important to explore the accessibility of PHC data in the European response to COVID-19 to develop key indicators for managing future pandemics with less stress.31
Existence of a community resilience group
The lack of a community resilience group is a problem that is particularly severe in communities that are already chronically underserved, such as rural areas.11, 32
Non-professional staff and those who work in less visible areas of the hospital, such as laundry and facilities, usually receive less information, making them feel isolated and disempowered.1 If staff are not supported, COVID-19 trauma can lead to symptoms of stress and burnout and affect their ability to function effectively.33
Around the world, there are excellent examples of resilience groups being formed quickly. At Johns Hopkins Medicine, e.g., a unified command center was activated shortly after the WHO declared COVID-19 “a public health emergency of international concern”.34 When staff feel they have support in the event of a disaster, they are more resilient. Therefore, staff support was included alongside other important services such as infection control and supply chain management.1, 35
Improved interprofessional collaboration
According to our survey, 73.7% (14/19) of Moldovan respondents saw improvement in interprofessional collaboration, but only 15% (3/20) of Greek respondents did the same. Only a few countries reached the 50% threshold indicating an improved interprofessional collaboration (Table 3). The qualitative analysis of our data confirms the difficulties in interprofessional collaboration, especially in communication between different levels of healthcare system, but these problems existed before the pandemic, and the comments also suggested a need for future improvement.
According to some authors, assigning roles and responsibilities to individual team members was particularly challenging during the most critical phase of the pandemic because of the rapidly changing work environment, the diversity of expertise between professions and the inconsistency of staffing, but team members’ organisational skills from previous experience working for civilian relief organizations proved helpful.7
Medicine shortage
In our survey, 60% (12/20) of respondents in Romania reported a shortage of medicines, but the majority of respondents in the other countries denied the existence of a shortage of medicines during the pandemic, although this problem was somewhat more evident in rural areas. In our study, the post hoc test showed that the semirural setting scored better than the other 2 settings with fewer problems of drug shortages.
In the free-text responses, some respondents complained about the lack of medicines and/or delays in obtaining medications. Other problems cited were postponed surgeries and delayed cancer treatment and screening.
It has been postulated that some pharmacists may have begun to procure the medications they required to treat a surge in critically ill patients, particularly sedatives, opioids and paralytics.36, 37
One of the factors contributing to “healthcare system resilience” is health sustainability. In the free text responses, some respondents mentioned a shortage of paracetamol, which is an important antipyretic for patients with fever. In a crisis, being able to prescribe a needed medication and not being able to find it is a paramount issue.38, 39
General practitioners involvement in palliative care
In our study, only in France, Greece, Hungary, Moldova, and Spain did the majority of informants indicate that PHC was included in COVID-19 palliative care. In other countries, according to the free-text responses of our informants, this type of care was mainly provided by hospital departments or started in PHC and continued in hospital departments. In France and Ukraine, there are guidelines in this regard.
According to the literature on the management of supportive and palliative care for COVID-19 patients in PHC in most countries, the management of post-acute COVID-19 patients takes place in PHC and this includes palliative care.40, 41
Limitations
We conducted a mixed-methods study on a small scale, and the transferability of our findings may be limited. Nonetheless, the objective of examining healthcare professionals’ perceptions of resilience in their daily clinical practice was achieved. Because we used a random sample of informants, the representativeness of PHC staff for each country may be questionable, although we attempted to achieve geographic variation. National coordinators attempted to avoid bias and recruit practicing PHC staff with diverse interests. Our questionnaire was refined after an initial pilot study. However, apart from face validation, it was not validated against other measures. We cannot rule out the possibility of confounding or alternative explanations for our results because survey responses reflect attitudes rather than actual performance. It is also important to note that discrepancies in the number of responses to each question, the online questionnaire and the selection process may contribute to the potential for independent bias in the generalizability of the results.
We found quite solid differences between countries in the responses to our questionnaire, although not so many between settings and gender of respondents. Large sample size would have been needed but due to the dramatic conditions (the study was conducted during the 2nd wave of the pandemic, when healthcare workers were struggling against the pandemic), we only managed to get a few responses from each country, and some EURIPA member countries did not manage to participate at all.
Conclusions
This decade began with one of the most significant pandemics in human history: the COVID-19 pandemic. Hence, the problems of resilient health and healthcare systems have become urgent. Public health emergencies have a high impact in countries with weak healthcare systems and inadequate preparedness and surveillance mechanisms. Therefore, better healthcare system preparedness is required to absorb the impact, respond to the consequences and adapt for future crises.
Our study found disparities between pandemic responses in different countries, particularly in triage effectiveness, adapting to the rapidly changing requirements, government help, existence of a community resilience group, improved interprofessional collaboration, medicine shortage, and GPs involvement in palliative care. The results emphasize the need for tailored approaches considering diverse contexts in shaping effective healthcare system resilience.
Various organizational and work-related measures can mitigate the impact of a COVID-19 pandemic in the workplace, such as improving workplace infrastructure, implementing appropriate and universal infection control measures, including the regular provision of PPE, and implementing resilience training programs. Another post-pandemic priority is to improve effective teamwork based on mutual respect.
Supplementary data
The Supplementary materials are available at https://doi.org/10.5281/zenodo.13902230. The package includes the following files:
Excel file. Raw dataset.
Supplementary Table 1. Results of the post hoc Pearson’s χ² independence test or relationships between the country and shortage of medicine. Adjusted p-values were corrected using the Benjamini–Hochberg correction to control for false discovery rate. Frequency of the responses are given for adjusted p < 0.05.
Supplementary Table 2. The results of the post hoc Pearson’s χ² independence test or relationships between the country and GP administering palliative care in COVID-19 patients. Adjusted p-values were corrected using the Benjamini–Hochberg correction to control for false discovery rate. Frequency of the responses are given for adjusted p < 0.05.
Data availability
The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.
Consent for publication
Not applicable.
