Evidence Gaps and Challenges in the Fight Against COVID-19 in Africa: Scoping Review of the Ethiopian Experience

Plain Language Summary

Africa in general and Ethiopia, in particular, contributed very little to COVID-19 research. As a result, most of the actions taken by the continent to curtail the pandemic were trial-and-error approaches. In this review, we conducted the first complete overview of the evidence gap and body of literature from Ethiopia on COVID-19. We tried to assess how the available evidence and the pattern of the outbreak affected the countermeasures taken by the country. With our review, we found that Ethiopia generated limited local scientific evidence. Moreover, the actions taken by the country were not tailored to the outbreak pattern and local contexts. This created a vicious circle of early actions that led to serious collateral damage which in turn forced the government to relax the restrictions later on. As a result, Ethiopia, like many African countries suffered from high spread of the disease after relaxing the interventions. The findings in our review highlight the values of locally generated evidence and context-specific actions in the pandemic control. Thus, Africa in general and Ethiopia in particular should invest in generating and making proper use of local evidence in the control of existing and future outbreaks and pandemic diseases.

Introduction

An outbreak of a new infectious disease is a global concern no matter where it occurs because of the risk of its spread to other countries.¹ Data and information sharing between countries are essential to contain the spread of an outbreak; and generating local data is essential to tailor targeted actions and sustain the control measures.^2,3

China tried to cut off Wuhan, the epicenter of the outbreak, from the rest of the country and implemented a complete lockdown to curb the spread of the infection during the early months. That helped the country control the outbreak within a few months.^4,5 Almost all other countries tried to copy these approaches but only a few managed China-like success.⁶

Every country has its peculiarities in terms of population characteristics, environment, and resources to manage its healthcare system. Thus, any effort to contain an outbreak of any disease should put into consideration the pattern of local disease transmission, the capabilities of its healthcare system, the feasibility of actions taken, and its population characteristics.⁷ Hence, any mitigation strategies destined to succeed should be customized to domestic needs based on scientific evidence generated locally.⁸

Like many African countries, Ethiopia took immediate actions to contain the coronavirus disease 2019 (COVID-19) outbreak during the early months of the pandemic. However, the pandemic control initiatives by the country were challenged by multiple factors.^9,10 The country lifted most of its COVID-19-related restrictions in September 2020 despite the steady increase in the number of cases and COVID-19-related deaths by then.¹¹ On top of that, there has been widespread public fatigue, reluctance to implement preventive measures, and conspiracy beliefs distracting the public.¹²

Actions taken to curtail the pandemic in Africa were mostly copied from elsewhere and never tested locally.¹³ Although most of them were known to be effective to control the pandemic, they might severely affect lives and livelihood in economically constrained settings. These in turn may affect the communities’ compliance to implement the recommended actions.¹⁴

It is thus essential to strengthen the quality of national data collection, handling and management procedures regarding COVID-19 to produce evidence that will help design contextualized actions.¹⁵ The overall aim of this review was to identify knowledge gaps and scope a body of literature about COVID-19 in the Ethiopian context during the early phase of the pandemic.

Methods

In this scoping review, we tried to (i) assess the existing scientific evidence on COVID-19 from Ethiopia; (ii) compare peer-reviewed scientific evidence on COVID-19 from Ethiopia against countries from Africa that were highly affected by COVID-19 and/or comparable population size; and (iii) evaluate if the action taken by the Government of Ethiopia was guided by scientific evidence and disease pattern.

Review of Existing Scientific Knowledge

The review was performed based on the methods described by Arksey and O’Malley¹⁶ by implementing the five stages of the scoping review process: (i) identifying the research question; (ii) identifying relevant studies; (iii) study selection; (iv) charting the data; and (v) collating, summarizing and analyzing the included literature. We have reported this review following the guidance in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR).¹⁷

Identifying the Research Question

This review was conducted to answer the following research questions: (i) What were the scopes, nature, and quality of local scientific evidence on COVID-19 from Ethiopia? (ii) Were the national actions against COVID-19 in Ethiopia guided by local scientific evidence and disease patterns? (iii) How did Ethiopia fare in terms of locally generated scientific evidence compared with other African countries with a comparable population and/or COVID-19 outbreak patterns? We defined “local scientific evidence” as any research conducted within Ethiopian territory and focusing mainly on COVID-19 in Ethiopia. The setting included community, healthcare facilities, and healthcare systems.

Identifying Relevant Studies (Search Strategy)

We performed a comprehensive literature search aimed to find both published and unpublished studies conducted from January 1, 2020 to October 10, 2020. All relevant articles were identified through a systematic search of MEDLINE via PubMed and LitCovid, Web of Sciences, and Embase. Pre-print servers such as MedRx, ChemRxiv and BioRx were also searched to include non-peer-reviewed articles. The following search terms were used to retrieve articles: “COVID-19” OR “SARS-CoV-2” OR “2019 novel coronavirus” OR “2019-nCoV” OR “novel coronavirus” AND “Ethiopia”. The search for unpublished studies/gray literature was performed in Google Scholar and through the review of reference lists and input of content experts. All literatures that met the keywords were included in this review. The search strategies used for PubMed and gray literatures are detailed in Supplementary Material as Annex – I.

Study Selection (Inclusion and Exclusion Criteria)

All studies focusing on COVID-19 in Ethiopia during the specified period were included. There was no exclusion based on the type of participants (individual patients, community, healthcare workers, and healthcare system). Only original research published as peer-reviewed articles or pre-print versions during the period of January 1, 2020 and October 10, 2020 were included in this review. However, we did not exclude articles based on the study design. Only studies published in the English language were included.

Identified articles were screened for duplications. Two independent reviewers screened the articles using titles and abstracts based on the inclusion criteria for the review. Full texts of potentially eligible studies were retrieved and assessed in detail against the inclusion criteria. Full-text studies that did not meet the inclusion criteria were excluded. Any disagreements that arose between the reviewers were resolved through discussion and consensus. The search of the literature was conducted from October 15, 2020 to November 30, 2020. The results of the search were reported in full in the final review and presented in a Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) flow diagram.

Assessment of Methodological Quality

Studies meeting the inclusion criteria were assessed for methodological validity before inclusion in the review using standardized critical appraisal instruments from Joanna Briggs Institute for all types of studies. All studies regardless of their methodological qualities underwent data extraction and synthesis.

Data Extraction, Synthesis, Analysis, and Presentation of Key Findings

We conducted a qualitative synthesis of the characteristics of the included literatures. We described the source where we found the article, publication year, type of article/study, and topic of article/study or guidance/guideline on COVID-19 to examine the existing gaps in research. Data from included articles were extracted using a standardized data extraction format from the Joanna Briggs Institute.

We categorized the literatures based on their focuses and key findings into (A) Knowledge, attitude, and practice about COVID-19 and its prevention; (B) Disease epidemiology and clinical profile of the cases; (C) Mental health issues during COVID-19 pandemic; (D) Preparedness for COVID-19 control; (E) Impacts on essential healthcare; and (F) Violence against women during COVID-19 lockdown. Findings were synthesized only descriptively and presented narratively and summarized in proportions and frequencies in tables. The extracted data were summarized into a table using the following contents: author, year, and title; source/journal; aims; study design; participants; and key findings (Annex – II in Supplementary Material).

Results

Study Inclusion

The initial comprehensive literature search yielded 573 articles. Of these, 105 were excluded due to duplications. From the remaining 468 articles, 280 were excluded after reviewing their titles and abstracts confirmed non-relevance to this review. With further scrutiny, 100 of the remaining 188 articles were excluded after reading the abstracts. The remaining 88 articles were retrieved for full-text review (Figure 1). Among these, 64 articles were included in the current narrative synthesis (Annex – II in Supplementary Material) and 24 were excluded for various reasons (Annex – III in Supplementary Material).

Figure 1 PRISMA flow diagram of study selection and inclusion process. Note: Adapted from Tricco AC, Lillie E, Zarin W, et al.PRISMA extension for scoping reviews (PRISMA-ScR): checklist and explanation. Ann Intern Med. 2018;169(7):467–473. doi:10.7326/M18-0850.17

Characteristics of Included Studies

Only 25 of the included full-text articles were published in peer-reviewed journals;^18–42 the rest 39 were preprints at the time of this review^43–81 (Annex – II in Supplementary material). Most, 50 (78%), of the included studies were cross-sectional descriptive studies. The 64 articles were thematized into six categories based on their major focus areas.

Knowledge, Attitude and Practice (KAP) About COVID-19 and Its Prevention

Nearly half (n = 30) of the articles fell into this category. A total of about 20,000 participants, comprised of community and healthcare workers, participated in these studies. More than 95% of the participants reported to have repeatedly heard about COVID-19 and its prevention methods.^21,22,69 As a result, the perceived knowledge level was found to be high even during the first few weeks of the outbreak among HCWs^{21,28,43,48,49} and the community.^{19,20,22,23,45,73} HCWs in general^27,43,51 and those working at university hospitals in particular⁴⁹ showed a favorable attitude towards COVID-19 and its prevention. The trend was similar among community participants.^45,56

However, the proper infection prevention and control (IPC) practices were low for both HCWs and the community. While self-reported practices of prevention methods were acceptable (62–100%) among HCWs,^{26,28,48,49,51} it was found to be low (<50%) among the general public.^{22,44,52,54–56,58,59}

Concerning specific prevention practices, most of the recommended methods were poorly implemented by HCWs and the community. For instance, mask use at workplaces and public gatherings was reported to be low.^{28,29,47,56,58} Only a few reported avoiding social gatherings and crowded settings⁵⁸ and practicing social distancing.^45,47,56 Most importantly, 21.2% of the participants opposed the wearing of face masks²³ and 28.3% had low intention to carry out those prevention measures.⁵⁷ Even though HCWs and community participants reported frequent hand washing practice,^24,25,52 only 16.3% of HCWs were observed to follow the correct sequence in hand washing.²⁵ Moreover, only <5% reported availability of enough soap, water, and hand sanitizers even at healthcare facilities.²⁵

KAP of HCWs and communities about COVID-19 and its prevention was associated with several factors. Among HCWs, younger age,²⁶ having college level education,^18,27 being a physician,¹⁸ owning TV/radio at home,¹⁸ access for IPC training and guideline,²⁶ prior knowledge about pandemic diseases,¹⁸ being urban resident,²⁶ and social media use²⁶ were associated with better knowledge and favorable attitudes. Exposure to IPC-related training, having a better knowledge and a positive attitude about COVID-19, having a chronic medical illness, and being married were associated with better IPC practices.^26,53 On the other hand, younger age,⁴⁷ rural residency, working at low level healthcare facility, shortage of personal protective equipment (PPE), and high workload were associated with poor practice of infection prevention measures among HCWs.²⁶

For the general public, younger age,^50,55,56 urban residency,⁵⁰ college level education,^50,55,56 access to TV/Radio and internet,^20,56 having large family size,^50,55 and having chronic medical illness⁵⁶ were associated with good level of knowledge and favorable attitude for COVID-19 prevention. Similarly, high level of education, urban residency, female sex, older age, having a good knowledge and a positive attitude, intention to seek care, having a chronic medical condition, and perceived mortality were positively associated with good IPC practices.^55,56,58,59 However, male sex, young age, lack of formal education, and rural residency were associated with poor intention to implement IPC practices.^19,57

Disease Epidemiology and Clinical Profile of the Cases

Studies on disease epidemiology were mainly descriptive. Slow but sustained increase in the reported number of cases³¹ and PCR test positivity rate^30,61 were observed in the first six months of the outbreak. Serological studies done in April⁶⁸ and May 2020³⁴ showed that there was already a high community prevalence in cities (up to 9%). Almost all of the studies showed that young men were affected more than women with median age of 34–36 years.^31–33,61 Most of the COVID-19 patients in the country during the first two months of the outbreak were imported cases.^32,33,61 In terms of geographic distribution, most cases were concentrated in the capital Addis Ababa and surrounding towns.^31,61 Data regarding the overall fatality rate of the outbreak were difficult to obtain, however, the case fatality among admitted cases was reported to be 1.9%.³³ Severity of the disease and COVID-19-related deaths were reported to be higher in the elderly,^32,63,65,66 those with diabetes,^63,64,67 and those presenting with cough and dyspnea.^62,65 Mortality among patients admitted with severe COVID-19 was very high (33.3%)⁶⁷ and outcome of intubated patients was reported to be very low (survival of <20%).³³

Mental Health Issues During COVID-19 Pandemic

The perceived psychological toll of the pandemic was reported to be very high among HCWs,^36,51,70 college students,⁷³ and the general public^{35,37,69,71,72} alike. Depression symptoms (12.4–78.4%),^35,70 anxiety (36–81.9%),^35,70,73 and stress (18–51.6%)^{35,36,69,70,73} were widely reported. Whether these were related to fear of COVID-19, economic difficulties during COVID or because of the underlying mental health problems remain to be known. Female sex,^35,70,71,73 low-income status,^35,37,71 and having a large family^37,71 were associated with psychological problems.

In addition to moderate to severe psychological problems, HCWs reported a high rate of perceived vulnerability for COVID-19 infection²⁹ and were worried about the risk of becoming infected with the virus and the risk of transmitting it to their families.⁶⁰ As a result, most (97%) of the HCWs reported that they were not ready to give care for COVID-19 patients.²¹

Preparedness for COVID-19 Control

Despite having a good level of awareness, HCWs,^74,76 the healthcare system,⁷⁶ and the general public⁷⁷ were not prepared for the COVID-19 pandemic. Major reasons mentioned by HCWs were lack of training, critical shortage of PPE, lack of scientific evidence at hand, inability to access guidelines and recommendations, limited access for literature, low local support, poor communication and collaboration between stakeholders, and fear of contracting the infection.^39,74–76

Impacts on Essential Healthcare

Few studies tried to assess the impact of COVID-19 control actions on access and delivery of essential healthcare services. Maternal and child health were major areas of focus for most of the studies. Compared with prior years, there were reductions during the first three months of COVID-19 in Ethiopian antenatal care, institutional deliveries, safe abortion care, family planning utilization, and neonatal admissions.^{40,41,79–81} A modeling study based on these findings indicated an excess of at least 10,252 under-five deaths, 379–4038 maternal deaths, and 1673–13,294 stillbirths during the first few months of the outbreak. Such humanitarian crisis attributable to disruption of the healthcare system was more than the mortality caused by the pandemic itself.⁸¹ There were also disruptions in the care of patients with chronic non-communicable diseases such as hypertension and diabetes.⁷⁸ Emergency medical services including emergency surgeries have also shown significant decline during the first three months of the outbreak.⁴¹

Fear of acquiring the infection,^78,79 service interruptions, and restrictions of movement⁷⁸ were mentioned as major reasons for disruption of essential healthcare services.

Violence Against Women During COVID-19 Lockdown in Ethiopia

One study⁴² tried to assess if there was an increase in violence against women during COVID-19 pandemic. The prevalence of intimate partner violence was reported to be 24.6%. Psychological and physical violence were commonly reported among the affected women. Housewives, younger women (< 30 years), women with arranged marriage, and women with a younger husband (31–40 years) were more likely to suffer from violence.

Comparison of Peer-Reviewed Articles from Ethiopia Against Highly Affected Countries in Africa

In order to assess the level of scientific evidence generated from Ethiopia, we also tried to compare peer-reviewed articles on PubMed as of October 10, 2020. Accordingly, as compared with five other highly affected African countries, Ethiopia had the fewest peer-reviewed articles on PubMed (115) compared with other countries; South Africa (774) had the highest number of articles (Table 1).

Table 1 Comparison of COVID-19-Related Articles from Ethiopia in the First Nine Months of the Pandemic with Other African Countries as of October 10, 2020

Timeline of COVID-19 Outbreak and Counteractions Taken in Ethiopia

We were also interested in comparing actions to curb the pandemic against the patterns of the outbreak. From this evaluation, we found that most of the actions were taken during the first three months of the outbreak when the rate of the disease transmission was low. These actions gradually declined and most of the restrictions were relaxed in September when the test positivity rates and daily reported number of cases were high (Figure 2). Moreover, the actions taken did not seem to be guided by the pattern of the outbreak and available local scientific evidence. For instance, during the period of March to June 2020, reported cases were predominantly imported^32,33,61 and concentrated in and around Addis Ababa.^31,61 However, no action was tailored to address the geographic distribution of the outbreak and pattern of spread. Instead, blanket restrictions were put in place for the whole nation even when essential healthcare services were reported to be extremely compromised at primary care facilities and regional hospitals.^{40,41,78–81}

Figure 2 COVID-19 outbreak timeline in Ethiopia and actions taken to curb the pandemic. The figure was constructed based on information released on different occasions by COVID-19 pandemic preparedness and response weekly bulletin (Ethiopia Public Health Institute) and Federal Ministry of Health of Ethiopia.

Discussion

Ethiopia, like many African countries, implemented several measures to contain the spread of COVID-19 during the early months of the outbreak.⁸² The outbreak appeared to be well contained during this period.¹⁰ However, the impacts of the actions taken as well as the threat imposed by the virus were not properly evaluated.^84,85 As a result, the country was not able to take a measured and tailored action to contain the pandemic. The actions put in place during the early months were gradually relaxed and almost completely lifted by the time the outbreak started to unfold.

Studies conducted soon after the first reported cases have shown good level of awareness^{19–23,28,43,45,48,49,73} and favorable attitudes^{27,43,45,49,51,56} by community and HCWs towards COVID-19 and its control. However, the IPC practices by HCWs and the community remained low.^{22,25,44,52,54–56,58,59} Adherence to social distancing and mask use by both HCWs and the general public was reported to be poor.^{23,28,29,45,47,56–58} Moreover, HCWs, the healthcare system in general, and the community were not prepared for COVID-19 pandemic.^74,76,77 Lack of proper training and guidelines, critical shortage of PPE, lack of scientific evidence at hand, low local support, lack of proper communication and collaboration between stakeholders, and fear of contracting the infection were reported by the HCWs to be the major limiting factors.^39,74–76

Though limited, the studies have clearly depicted that the collateral damage was felt more than the direct impacts of the outbreak during the first few months. Maternal health service utilization declined significantly compared with the previous months.^{40,41,79–81} Overt mental health issues such as anxiety, depression, and stress were also reported during the time.^{35–37,51,69–73}

Evidence about the real pattern of the outbreak in Ethiopia is scarce due to limited testing capacity⁸⁶ and lack of local scientific evidence. The outbreak unfolded at a slower rate than initially feared. Nevertheless, the number of cases and PCR test positivity rate showed steady increase after the first three months of the initial report.^30,31,61 Because most of the studies were limited to Addis Ababa,^{31–34,61,63–68} much cannot be said about the pattern of the outbreak in Ethiopia. These studies have shown a very high fatality rate among patients admitted with severe COVID-19 (as high as 33%)^32,63–67 and those treated with mechanical ventilator (80% fatality).³³

Research outputs on COVID-19 from Ethiopia were limited in both quantity and quality. For instance, the number of published articles containing COVID-19 and country name as of October 10, 2020 on PubMed yielded fewer publications than from other high COVID-19 burden countries in Africa – Egypt, Kenya, Nigeria, and South Africa. Furthermore, the existing research outputs were more of descriptive studies and focused mainly on human behavior. Important issues such as diagnostics, care of patients with the disease, and assessment of effectiveness of interventions were overlooked.

Looking closely at the actions taken by the country against the outbreak pattern, it is hard to explain the mitigation strategies scientifically. Most of the restrictions were put in place between March and April 2020, when the reported cases were very low and limited to imported cases.^32,33,61 The restrictions were then relaxed after August 2020 when the outbreak was peaking (Figure 2). It is possible to speculate that the country tried to correct the panic actions taken before the outbreak kicked off and took a pragmatic action to minimize the collateral damages emanating from the restrictions. This might be intended to minimize the impact of the restrictions on Small and Medium Enterprises, which are critical to the country’s economy. A similar pattern of initial prompt actions leading to economic difficulties forced many African countries to relax restrictions by the time the outbreak pattern was changing.⁸⁷ This later led to major community outbreaks in many countries in the continent.⁸⁸

However, such decisions should trade-off between the economy and the negative social impacts of downplaying the pandemic. This may ultimately lead to a sustained disease transmission and large community outbreaks. This has practically happened in Ethiopia in the first quarter of 2021 when the country registered the highest peak, test positivity rate, critical care admissions, and deaths due to COVID-19.^89–91

The findings in our review indicate that understanding the social, economic, and environmental context of the country is important to design and implement public health interventions. Particularly in countries such as Ethiopia, where multiple cultural factors, financial constraints, and health system challenges were documented as major barriers,⁹² public health interventions should be tailored to cultural and social contexts of the society.

Strengths and Limitation

We did a comprehensive assessment of studies conducted in Ethiopia on COVID-19. However, systematic analysis and presentation of the findings suffered from extreme heterogeneity of the findings. Besides, most of the included articles were not peer-reviewed and suffered from methodological problems and doubtful scientific validities. Furthermore, some of the manuscripts in pre-print versions during our review might have been peer-reviewed and published in much changed form after our submission.

Conclusion

Ethiopia, likely other African countries, might have avoided the feared COVID-19-related crisis. However, the true burden of the pandemic is likely to be underestimated due to limited testing and surveillance systems. The country might have been faced with the dilemma of balancing between minimizing the humanitarian crisis due to the pandemic and limiting the economic impact of the public health measures. The strategies taken to control the outbreak might have also suffered from lack of local scientific evidence, cultural conflicts, and political interference. Moreover, adopting interventions designed for settings with different cultural context, economic situations, and pandemic risk might have also resulted in low public trust.

Hence, countries such as Ethiopia should prepare themselves for future similar pandemics even more than countries with robust healthcare systems. They should put a system in place to anticipate, detect, and early control outbreaks. Actions that are based on local scientific evidence, culturally sensitive, and locally acceptable should be prioritized.