Background information

We performed this study in Costa Rica, a middle-income country in Central America. In 2019, it had a population of five million people and the third-highest life expectancy in the Western Hemisphere, despite not spending as much on health and having a smaller GDP per capita than other countries in the region.11 Costa Rica’s health system is mostly funded by compulsory health insurance schemes (70% by 2019) and household out-of-pocket payments (22%), with government schemes and voluntary healthcare payment schemes representing only 4% each.12

The Ministry of Health is responsible for public health regulations, epidemiological surveillance, and research and development in health; it does not provide healthcare services. The most important healthcare provider in the country is the CCSS, an autonomous public institution funded by mandatory contributions from all formal workers, employers, and the state; the latter also covers vulnerable populations. By 2019, the CCSS’s health insurance covered 85% of the population, and its revenues represented approximately 70% of all health expenditures in Costa Rica.12 13 The CCSS provides preventive, curative, and rehabilitative healthcare services through a network of public hospitals and primary care clinics that cover more than 97% of the population.14 The network is present across three different levels of care. The first level is the system’s gatekeeper and provides basic healthcare services through 1 057 community healthcare teams (known as EBAIS) managed by 105 primary-level clinics (known as health areas). The second level supports the first level by providing healthcare services with varying degrees of complexity and specialty. It includes a network of 20 regional hospitals and 17 clinics. Finally, the third level provides healthcare with the highest level of specialisation and complexity. It includes a network of three national hospitals, six specialised hospitals and six specialised centres.15 16

To provide healthcare, the CCSS leverages a unified digital health record known as the Unique Digital Health Record (its acronym in Spanish is EDUS). The EDUS is a unified cloud-based medical record system, an ambitious warehouse of personal health data that records hospital emergency attendances, admissions and discharges, and surgical operations digitally. It also integrates personal health records (including diagnoses and treatments) with the national appointment system for planned elective care. The EDUS is accessed from the cloud by health workers and patients have access to their own clinical history. The CCSS rolled out the EDUS from May 2015 to December 2018. Since then, the EDUS has become a control tower that has made it possible to concentrate information, monitor the occupancy rate of beds by a health facility and manage transfers. By the time the pandemic began, the EDUS was widely used across all levels of attention and by CCSS health workers to record patient visits, making Costa Rica one of the first countries in the world to have a single, national electronic health record that is unified across all levels of care.17–19

Study design

In a pandemic context, essential services are those health interventions that should be unconditionally protected and delivered, even if substantial resources need to be diverted to respond to the outbreak. For this paper, we reviewed a list of 120 EHSs proposed for low-income and middle-income countries to prioritise during an outbreak based on the probable magnitude of the harm that would occur from interruptions or disinvestment in these services.20 As each country will need to define what is essential for them, according to its health system capacity, available resources, and epidemiological profiles, for Costa Rica, we focused on three categories from this list: maternal and newborn health, cancer, and cardiovascular and related disorders.

To understand how the CCSS sought to maintain EHSs during COVID-19, we followed an explanatory sequential design in which qualitative and quantitative data were collected and analysed to generate our findings.21 In the first phase, we collected and analysed qualitative data through a desk review and interviews with key informants (KIs). We used the data from the desk review to inform the design of the semistructured interviews. In the second phase, we validated our insights from the first phase with quantitative data from EHS performance.

We started by reviewing research papers (both peer-reviewed or unpublished), books, policy documents, and newspaper articles looking for practices for EHS maintenance adopted by the CCSS for COVID-19 in 2020 or 2021. For documents other than newspaper articles, we searched Google Scholar, PubMed, BINASSS (the local library on health and social security), and organisation websites, including the WHO and the United Nations Economic Commission for Latin America and the Caribbean (ECLAC), for scientific publications, reports, and policies from September 2021 to July 2022. The inclusion criteria comprised institutional practices, actions or programmes implemented by the CCSS at any of its different managerial levels in response to the COVID-19 pandemic in 2020 or 2021 and that affected EHS maintenance. We excluded reports that described actions to control COVID-19 implemented by local public institutions or ministries other than the CCSS. We excluded reports documenting the country’s economic and financial response to COVID-19, as well as reports documenting the country’s public health and healthcare responses, if they did not influence EHS maintenance. For news articles, we used Google News to find printed and online newspapers, magazines, and blogs published by local and international private and public organisations and institutions between January 2020 and December 2021, focused on those related to EHS maintenance and the CCSS. We gathered 13 newspaper articles, 13 journal articles, one book, 12 white papers or official institutional reports and six websites. We extracted data into a Microsoft Excel template to summarise the responses, their enabling factors, how they helped sustain EHSs and how they induced changes in the structure of the CCSS (such as changes in staffing mix or profile) and in the processes (such as how patient consultations are organised or the modalities for dispensing medications).22

Qualitative data were gathered through semistructured interviews with KIs from the health sector in Costa Rica. Interviewee selection was intentional, aiming for a representative sample of various actors. We included representatives from different levels of care (such as first-level care, clinics and specialised hospitals), sectors (including public, private, and civil society or patient groups) and work roles in the response (such as strategic, operational, and tactical roles, eg, directors of health areas, clinics, and hospitals, doctors, and coordinators). The research team contacted KIs via email and social media (eg, WhatsApp) to schedule interviews, including a statement about the research goals in the initial communication. At the start of each interview, the interviewer explained the project’s objectives, research background, and confidentiality terms, and asked for the interviewee’s verbal consent to participate in the research. Using an interview guide (online data supplement), we explored themes and practices related to COVID-19 response and maintaining EHSs through open-ended questions. Interviewers also welcomed emerging themes highlighted by interviewees as significant to the Costa Rican response. After each interview, we asked participants to suggest other key actors who could contribute valuable insights, facilitating a snowball effect in our recruitment process.

Between October 2021 and February 2022, we conducted 21 interviews of an average length of one hour, with 26 KIs using Google Meet due to COVID-19 restrictions. Online supplemental table A1 shows the profile of the interviewees. The interviews were carried out in Spanish by two experienced members of our research team, AAP (internationalist, male) and AMP (economist, female), skilled in qualitative research methods from various projects. The interviews were audiorecorded and transcribed. AAP also took field notes during the interviews and reviewed the transcripts to ensure alignment with the audio recordings. After completing the data collection, we analysed the data through content analysis by identifying and coding themes drawn from the interview transcripts.23 AAP used qualitative analysis software (Atlas.ti) to code and analyse the transcripts and audio, focusing on healthcare providers’ strategies to maintain services during the COVID-19 pandemic. In developing this analysis, we held a reflexive stance to understand how our experiences, biases, and background—as Costa Rican economists and internationalist professionals working in an academic institution—might influence the interpretation of the data, ensuring that it is as faithful as possible to the participants’ perspectives. We applied deductive coding to the interventions using the Donabedian model, categorising them into modifications related to structure, processes and outputs/outcomes.22 KIs did not participate in designing the study, interpreting the results, writing the findings or reporting the research. The results are presented with supporting quotations from KI interviews and guided by the Consolidated Criteria for Reporting Qualitative Research guidelines.24

In the second phase, we validated the qualitative data from the desk review and Kis with quantitative data on three dimensions. First, we address the outputs and outcomes of the CCSS between 2014 and 2022 (CCSS, 2019, 2020). Second, deaths according to the ICD-10 were recorded from 2014 to 2022 from the Costa Rican National Institute of Statistics (INEC). The third is monthly outpatient consultations during 2020.25 We used the quantitative data to confirm what the KIs mentioned and assess the potential impact of the responses.

There was no patient or public directly involved in the study.

Changes in structure

Changes in processes

The CCSS adopted administrative procedures, internal decision-making processes and mechanisms to provide patients access to health, seeking to maintain EHSs during the COVID-19 pandemic.

Changes in outputs and outcomes

This section describes changes in outputs and outcomes, relating them to structural and process changes. The outputs are immediate results, and the outcomes are second-level results. Table 1 compares the 2019 and 2020 output and outcome levels, and in online supplemental file 1, we present figures with annual time series data from 2014 to 2022.

COVID-19 disrupted routine work at the first level of care, impacting the follow-up of diseases and health promotion tasks due to personnel shortages. Health workers reduced patient care during the pandemic’s early stages, affecting referrals, counter-referral and follow-up for conditions such as high blood pressure and diabetes. Mobility restrictions and school closures strongly affect vaccination programmes in schools (such as vaccination against the human papillomavirus). KIs reported that emergency visits increased with patients previously treated at the primary level, such as decompensation for blood sugar levels or high blood pressure, and patients seeking medical attention in critical conditions, including increased cases of heart attacks.

Process changes, including the integrated network strategy, led to an increase in the rate of patient transfers between CCSS hospitals and clinics, which increased by two-thirds between 2019 and 2020 (see figure 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Percentage of hospital discharges with a transfer to another public hospital (2014–2022). Source: own elaboration from 45.

Changes in structure and processes also influenced auxiliary services. While the home delivery of medicines programme prevented a substantial decrease in medication dispatches, there were no partnerships for home delivery of other auxiliary services such as laboratory tests (see online supplemental figure A5). A reduction in laboratory tests was attributed to patient concerns about attending appointments during the pandemic. A coordinator of the cancer patients’ forum described the fears that the population had, including ‘… fear of dying from COVID-19 in the case of the population at risk. Fear of dying alone due to the isolation of patients with COVID-19… Fear because they did not know the next steps they had to take’ (personal communication, 8 February 2022). A communication strategy emphasising staying at home and mobility restrictions resulted in a generalised fear of visiting healthcare facilities and reduced consultations in some services.

Supporting alternative modalities of care led to a rapid shift from in-person to remote medical consultations after Costa Rica’s first confirmed COVID-19 case (see figure 3). Supporting alternative modalities of care contributed to sustaining some but not all EHSs, given that some procedures (such as surgeries) can only be provided in person (see online supplemental figure A6). The mandate to reduce in-person medical attention differentially affects medical consultations by specialty. Between 2019 and 2020, medical visits to cardiology and oncology specialists decreased. In contrast, orthopaedics, odontology and ophthalmology—which can only be provided in person and may be less life-threatening in nature—were heavily affected (also see online supplemental figure A7).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Monthly outpatient visits, in-person versus alternative modalities per 1000 people, during 2020. Source: own elaboration from 25.

Figure 4 compares the average number of deaths nationwide from 2014–2019 to 2020–2022 according to the ICD-10. Infectious and parasitic diseases showed the most significant increase (+830.4%), followed by endocrine (+93.1%), mental disorders (+72.5%) and blood diseases (52.0%). Alternative care modalities helped some patients, but not all.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Number of deaths by International Classification of Diseases version 10 (ICD-10) chapter, 2014–2019 average vs 2020–2022 average. The data label shows the absolute change between the 2014–2019 average and the 2020–2021 average. Source: own elaboration from online query to 46.

Our desk review and KIs interviews suggest that neither the CCSS nor the government had an official list of what constituted EHSs. Nevertheless, there was an implicit prioritisation process when the institution mandated reducing in-person services, and only certain services, such as ER, respiratory, cardiology and oncology services, remained active. Other services not implicitly considered a priority were heavily affected (eg, orthopaedics, odontology, ophthalmology and dentistry) or delivered virtually.

Enablers

Individually analysing changes made by healthcare systems to respond to COVID-19 is necessary but does not provide the whole picture for pandemic preparedness. Hence, we also go beyond describing specific structure and process changes and try to understand the enablers in the local healthcare system that allowed those changes to be possible or effective. We identified three key enablers.

First, the CCSS is a single institution that oversees a network of hospitals and clinics nationwide, covering more than 97% of the population. This integration enables a coordinated response to both COVID-19 and regional outbreaks. Notably, during the pandemic, the flexibility given to the CCSS, because it is a single and integrated healthcare system, distinguished it from fragmented ones (ie, many different health ‘subsystems’ coexist, providing care for different parts of a country’s population; each of these subsystems has its own way of financing and delivering healthcare, and each provides healthcare to different types of people), such as in New York, where challenges arose due to the lack of coordination across different health providers.44 The CCSS Medical Manager mentioned the importance of network management of the system; for instance, ‘I saw what was happening in other places, for example, what the governor of New York said, that he would have liked all the hospitals in New York to operate at the same time and in the same way. At a certain moment, not only were some hospitals filled to the brim and others not as much, but they also belonged to different systems. They didn’t have the flexibility we have’ (personal communication, 10 February 2022).

Second, the EDUS plays a vital role in patient follow-up, service continuity, and efficient patient transfers within the healthcare system. The EDUS facilitated data sharing among health facilities, supporting a coordinated response. These data encompass epidemiological statistics and resource availability at the regional level, such as the number of beds, number of physicians, number of ambulances, UCI capacity and testing capabilities. The successful implementation of the EDUS relies on appropriate administrative processes and governance structures to support decision-making.

Finally, the availability of financial resources, including a contingency fund, bolstered the pandemic response. However, recognising that not all healthcare systems can establish such funds, it is crucial to explore alternative funding sources and potential intersectoral collaborations to prepare for the future crises.

Our paper has strengths and limitations. One of the strengths is the broader understanding of the decisions made and the processes followed to maintain EHSs. We went beyond analysing statistics and performance indicators through the dominant qualitative nature of the data and analysis. One limitation of our paper is that the CCSS did not have an explicit definition of what the institution considered EHSs, which limits our research approach.