Study design and setting

This study employed an exploratory qualitative research design using open-ended interview question and a purposive sampling approach. This design uses participants’ narratives to best articulate and understand the phenomenon of interest, especially when the issue is poorly understood. This study was conducted in the capital of Ethiopia, Addis Ababa, which has a population of 8,938,683 million (30% of the country’s urban population). At the time of the research, the city had the largest number of reported COVID-19 cases in Ethiopia.

Ethiopia has a three-tier healthcare delivery system: level one is a Woreda/District health system comprised of a primary hospital (to cover 60,000–100,000 people), health centers (1/15,000–25,000 population) and their satellite Health Posts (1/3,000–5,000 population) connected to each other by a referral system. The primary hospital, health center, and health posts form a Primary Health Care Unit (PHCU). Level two is a General Hospital covering a population of 1–1.5 million people, and level three is a Specialized Hospital covering a population of 3.5–5 million people [24].

Study approach and period

We conducted a qualitative study to explore the health system response to sustain the provision of essential healthcare services during the pandemic. This study was conducted from February to March 2021.

Sample size and study participants

Participants were purposively selected based on the possibility of having a comprehensive information about the effect of COVID-19 and mitigation strategies and/ or the health system response to sustain the provision of essential healthcare service in their respective departments. Accordingly, study participants were based in Addis Ababa city where we purposively selected participants from the FMOH, Addis Ababa regional health bureau, and Sub-city health office who were assigned to the COVID-19 response team. In addition, we interviewed the medical director/focal person of the rapid response team and service providers at health centers from different contact points that provide essential health services. We determined the sample size based on the saturation of information. We included 60 key informants in this study and there were no non-participants in this study.

Data collection tool and procedure

Interview guides were developed with open-ended questions to ensure high degree of flexibility. Interview guides covered topics such as the effect of COVID-19 on essential healthcare service provision and utilization, activities undertaken to sustain the provision of essential healthcare service, infection prevention and control (IPC) measures taken to ensure safe delivery of essential healthcare service, and challenges faced by the health system while providing essential healthcare service during the pandemic. We pilot-tested interview guides and made the necessary refinements to make sure the guides could elicit the information we intended to capture. The pilot was conducted among healthcare providers that were not included in this study. Based on the pilot interviews, we paraphrased some of the questions and added probing questions to get detailed information.

Face-to-face interviews were arranged for a time and location that was deemed most favorable and comfortable for the participants. Interviews were conducted in a quite office and privacy was ensured to provide participants the freedom to express their opinions by conducting the interview in the presence of the interviewer and the respondent only. All interviews were audio taped and took about an hour on average. Five authors (4 female (EHF, BT, FBW, ML) and one male (AA) conducted the interviews. All interviewers are MPH holders, trained in qualitative methods, and have prior experience in qualitative data collection and analysis. Among the interviewers, three (EHF, BT, AA) were lecturers, one (FBW) was a director of knowledge translation directorate, and one (ML) was a project coordinator at the time of data collection.

During data collection, respondent validation was done by restating and paraphrasing information for respondents to determine the accuracy. Additionally, investigator triangulation was applied by using multiple investigators to study the same phenomena improving the credibility of findings of the study.

Data analysis

Thematic data analysis was used to describe and compare general statements as relationships and themes present on the data [25]. The interviews were first transcribed verbatim into Amharic then translated into English. Inductive and deductive processes were used to identify common themes. Data analysis started with reading and re-reading each transcript and listening to the audio recordings. Once this was done, a codebook was developed by a study team of five researchers/coders. The study team double-coded 4 randomly selected transcripts and reviewed them together. If coder disagreement revealed ambiguity in codes, code definitions were updated, and examples were added. Double coding proceeded until no new disagreements were identified and inter-coder agreement was reached. After this, the transcripts were single-coded using open code (V 4.02). Finally, decision was made on the themes that gave possible explanations and made meaningful contribution to answering the research question.

Rigor

The trustworthiness of the data was ensured using different approaches, namely, credibility, dependability, confirmability, and transferability. To ensure credibility, data was collected from different health system levels, starting from the Ministry of Health down to health centers, and included individuals from different contact points that provide essential health services. The last author, who has rich experience in qualitative research, oversaw the codebook development, data coding, and category development. Interviews were conducted by five primary and co-primary investigators. The interviewers are trained in qualitative methods and have prior experience conducting interviews and analyzing qualitative data. During data collection member checking was done by restating, paraphrasing, and summarizing information for respondents to determine the accuracy of data. To ensure the dependability of the findings, more than one individual was involved during both data collection and coding, and debriefing was conducted daily. Furthermore, the transcribed data were systematically reviewed against the audio files. Transferability of the data was maintained by a rich description of the method section of the study. The study finding confirmability was guaranteed with the research supported by the data collected. Data were collected from various contact points to ensure all views were represented. Interviews were continued until data saturation was reached.

Ethical consideration

Ethical clearance was obtained from the Research Ethical Committee of the School of Public Health, Addis Ababa University (reference number 070/20SPH) and the Ethical committee of the University of Liverpool (reference number 8049). In addition, permission was obtained from relevant federal and regional health offices. Written informed consent was obtained from the participants after the necessary explanation about the purpose, procedures, benefits, and risks of the study had been made. The informed consent also included permission to audio record the interviews and use anonymized quotes. The respondent’s right to refuse few or all the questions was respected at all times. In addition, participants’ privacy and confidentiality of the information obtained were always maintained.

Theme 1: Effect of COVID-19

Theme 2: Adaptation strategies to sustain the provision of essential healthcare services

Prioritization

Participants explained that health facilities prioritized certain services with the aim of freeing up the space and resource for the provision of other essential healthcare services. There were facilities that suspended services such as, medical service for a driving license, cervical cancer screening, minor surgeries such as circumcision. However, this only lasted for the first two to three months of the pandemic, and they resumed providing the service by taking the necessary precautions. One participant explained this as:

“…..In addition, we suspended services like circumcision service and cervical screening service since they were not urgent. Service for general medical checkup was also suspended at first but then we started providing it because we believed we can sustain the provision by taking the necessary precaution. Now every service is being given.” (Male, 29, Public Health)

Participants also explained that prioritization of clients was in place. All facilities prioritized clients who were believed to have high risk of getting seriously ill from COVID-19. These included patients with underlying medical conditions such as ART and NCD patients, pregnant women, and children. These clients were given priority to minimize waiting time to access services, thereby reducing their duration of stay at the HC. This is captured in the following quote:

“The medical card of ART patients is kept in the case manager office, so they don’t have to wait in line to get their card. Thus, clients are able go straight to the case manager office and get the service in ART clinic which is located right next to the case manager office. This means they immediately get the service and leave.” (Female, 35, Public Health)

The other strategy implemented by the health facilities to prioritize clients coming for ART services was through changes in work-hour arrangements. Providers started working early in the morning, during lunchtime and during Saturdays to provide ART service and to minimize crowding.

Ensuring availability of resources

Health centers took different measures to ensure adequate workforce and supply to sustain essential healthcare services. In terms of workforce, no service provider was allowed to take annual leave except for those who were high risk. These included service providers who were pregnant, elderly or who had comorbidities. In addition, providers on study leave were asked to return to work.

In terms of supplies, health centers used to borrow supplies from one another when they faced shortages. This includes supplies needed for essential services as well as personal protective equipment (PPE). Health facilities also reshuffled budgets such as shifting budget allocated for training to purchase PPE.

Support

According to our findings, different governmental and non-governmental organizations provided support for health facilities. Non-governmental organizations supported facilities by providing vehicles for CAG services, and mobile airtime cards used to trace clients and remind upcoming appointments. In addition, they provided logistic support such as masks, sanitizers, soap, and tents.

Federal and governmental offices also provided support for health centers by providing o masks and sanitizer. However, participants explained that the support from sub-city in terms of supportive supervision to ensure sustained provision of essential healthcare services reduced significantly.

“Higher offices should have their own contribution to ensuring that services are being provided properly. There should have been supervision to ensure that services are being provided properly and that the community is not inconvenienced but no one came from higher offices to supervise because of fear of contracting an infection when they come to health center. People might be scared, and it is understandable, however, they should have monitored whether essential services were being provided properly or not.” (Female, 40, MPH)

“We (personnel at sub-city) did not do that much monitoring and supervision. The professionals here at the sub-city did not want to go to health centers to monitor or supervise since there was frustration that resulted from fear of being infected from going to health facilities. So, one of the drawbacks of COVID-19 was the visible collapse of service deliveries, because of the poor support and supervision of health services.” (Male, 36, MPH)

Infection Prevention Control (IPC) measures

Initiative of renovating facilities for infection prevention including securing water supplies was done by sub cities by following strict application of guidelines.

Regarding providers’ IPC implementation practice, most healthcare providers explained that there was a stricter adherence to infection prevention activities like wearing masks and washing or sanitizing hands. However, the use of masks and sanitizers has loosened over time.

“We use sanitizer and mask, but we hug and kiss. We also used to eat lunch separately by keeping distance, but now we are eating in one dish, which was the custom before COVID-19; everything is getting back to normal.” (Female, 35, Public Health)

As for IPC measures taken for and by patients, initially, temperature screening at the health facility entrance, isolation of clients with symptoms of COVID-19, social distancing in waiting areas, mandatory mask use and hand washing was enforced. There was also repurposing of rooms for better ventilation if the previously used rooms were too small. Although the IPC measures were strictly implemented during the first few months, practices have become lapsed as indicated by our participants.

Participants explained the different challenges they have faced related with IPC measures. The challenges cited include, lack of well-equipped isolation room, shortage of PPE including mask, sanitizer, glove, and water especially during the first 2–3 months resulting in frustration among service providers.

“We had shortage of PPE and the budget that was allocated for health centers was very low this year. It was even lower than last year’s budget, so we were challenged to equip the health center with the necessary logistics. In addition, the price of materials has gone up this year. A single surgical mask that used to cost around 3 birr now costs around 9 or 10 birr. So, we struggled to purchase a mask. We used to give providers one mask for a weak and half liter sanitizer for a month. This has made the providers feel uncomfortable.” (Male, 37, MPH)

Other challenges related to IPC measures were a shortage of isolation rooms, poor infrastructure of health facilities to ensure the safe delivery of different services, and negligence from the community/clients towards prevention strategies. There were also challenges related to sustaining the implementation of different IPC measures, including screening at the gate, ensuring clients wash their hands, and ensuring social distancing at the health centers.

Poor preparedness was one of the reasons identified by participants that has resulted in the shortage of supply.

"We could have been more prepared. There was negligence starting from the higher health office. Otherwise, we would not have a shortage of logistics. We would not have such a shortage if we had thought about this earlier. If you go to the pharmaceutical supply agency, they will tell you that they don’t have mask even though they have, and this shows you that they were not ready and they didn’t have adequate stock." (Male, 29, Public Health)