We adopted a case study approach to present a descriptive analysis of the COVID-19 epidemic in Zimbabwe, including policy responses to the pandemic and its impact on health and healthcare. In this country case study, we performed a more detailed examination of the urban provinces of Harare and Bulawayo, using both qualitative and quantitative data collected concurrently. The two cities are major urban centers in Zimbabwe, with the former as the capital and the latter as the second-largest city.

To produce a chronological narrative account of policies related to COVID-19 from February 2020 to August 2021, using a policy review checklist we collected data from government policy documents and announcements, newspaper articles, social media, and the Oxford COVID-19 response tracker (OxCGRT), a project that collated data on COVID-19 control measures from countries across the world (19). The policy review checklist was basically an excel sheet which captured COVID-19 policy data. The tool captured the name of the policy document, the type of policy document, its date of publication, the content of the document, and its implications. Five research assistants completed the checklist, and it was reviewed by the study coordinator and the two Principal Investigators (PIs) to enhance rigor.

To present an account of the COVID-19 pandemic, we obtained official data on the number of COVID-19 tests performed, confirmed cases, and deaths from daily situational reports and the official Twitter (now X) handle of the Ministry of Health and Child Care (MoHCC). Given the limited disease surveillance and low testing rates, we also drew on data from other studies to assess COVID-19 transmission in Zimbabwe.

To describe the effects and impacts of COVID-19 and its control measures on maternal and child healthcare, we extracted routinely collected health service data on antenatal care (ANC) and growth monitoring of children and childhood vaccinations from District Health Information Software version 2 (DHIS2) for the period January 2016 to August 2021 in the Bulawayo and Harare metropolitan provinces. Harare, the capital city, and Bulawayo, the second largest city in the country, have estimated populations of 1,896,134 and 676,650, respectively, and together account for approximately 30% of the country’s recorded COVID-19 cases (20). Primary health services in the two cities are primarily administered by the city council, whereas tertiary and quaternary hospitals are administered by the central government. We selected 20 public primary health clinics in Harare and 18 clinics in Bulawayo to study the trends in healthcare provision and utilization.

We selected two child and two maternal services: growth monitoring (GM), child vaccination, antenatal care (ANC), and HIV testing of pregnant women. Several data items are routinely collected to monitor child vaccination services indicators, and we chose to analyze trends in the number of children aged 9–12 months who have received their final measles, mumps, rubella (MMR) vaccine, which is captured in the DHIS2 as ‘primary course complete’ (PCC). For maternal health services, we analyzed trends in the number of pregnant women attending a 4^th ANC visit (a proxy indicator of adequate ANC visits according to the WHO), as well as the number of women who received an HIV test at their first ANC visit. However, data for the latter were only available from 2019 onwards.

We also collected primary qualitative data between April 2021 and December 2021 in Harare and Bulawayo. We conducted face-to-face 28 in-depth interviews (IDIs) of key informants drawn from community-based organizations (CBOs; n = 5), city health managers (n = 5 Harare, n = 3 Bulawayo), national program managers and policymakers (n = 5), and healthcare workers (n = 3 Harare, n = 7 Bulawayo). We further conducted two focus group discussions (FGDs) with healthcare workers from selected primary health facilities in Harare (n = 12) and Bulawayo (n = 11) and four FGDs with members of the public (two in Harare with nine and 10 participants, and two in Bulawayo with 11 and 13 participants). In addition, we conducted 40 household interviews evenly distributed across the three economic strata (n = 20 Harare and n = 20 Bulawayo), four households all from the higher income bracket declined to be interviewed. We purposively identified economically stratified residential areas and then randomly selected households within these strata. Topic guides were used for all qualitative interviews, with interviews lasting an hour and FGDs lasting an average of 2 h. Primary qualitative data were collected by five experienced and trained (three research assistants, one study coordinator, and one principal investigator) fluent in English, Shona, and Ndebele. Separate topic guides were used for different informant groups, including questions about their experiences and views on COVID-19 and its control measures as well as about health services and treatment access during the pandemic. The focus was on understanding the impact of the COVID-19 control measures on maternal and child health.

Finally, for policy documents, we drew data from the selected published literature to describe the historical and socioeconomic context of the pandemic and added further information about the impact of COVID-19 on maternal and child health in Zimbabwe. We identified this literature from the PubMed and Google Scholar databases for papers published up to January 2023 using the search terms ‘maternal and child health’, ‘COVID-19,’ and ‘Africa’ or ‘Zimbabwe’.

To guide our analysis, we developed a conceptual framework, (see Figure 1) that broadly captures the hypothesized impacts of pandemic control measures, including lockdowns, travel restrictions, and the reallocation of healthcare resources. We hypothesized that, in addition to disrupting transmission as intended, these would also disrupt health service access and delivery, in turn affecting the utilization and outcomes of maternal and child health (MCH) services. Mediating and modifying factors interacting to produce observed impacts include socioeconomic vulnerability (including poverty, dependence on informal employment, and urban overcrowding), governance and policy processes (such as centralized decision-making, enforcement strategies, and community engagement), and health system resilience (encompassing human resources, financing, supply chains, and information systems). The solid arrows in the model depict the direct hypothesized causal pathway; the dashed arrows represent non-linear mediating and modifying dynamics unpredictably shaping the magnitude and directionality of effects.

Changes over time in the four indicators of healthcare utilization are shown graphically for each clinic, stratified by city, with trend lines using loess smoothing and 95% confidence intervals. We used interrupted time series models January 2016 to the end of 2019. The ratios were color-coded using shades of orange to signify counts lower than the pre-COVID-19 average and shades of green for higher counts. Some clinics had missing values for some months, especially in 2021, and these are colored gray. These missing values included a combination of instances where there was no service uptake (e.g., the clinic was open but there was no service utilization, or the clinic was shut) and when there was service uptake, data were not recorded or uploaded onto the DHIS2. In the pre–COVID-19 with generalized least squares to assess monthly trends, accounting for seasonality and autocorrelation, and estimated immediate and slope changes at the onset of the COVID-19 lockdown. Clinic-level changes in healthcare utilization are shown as ratios computed by dividing monthly counts by the clinic-specific monthly average for the period from COVID-19 period, data were missing for 6 months in 6 clinics. For each missing month, values were imputed by averaging the observations from the preceding and following months. Data for HIV bookings were not reported to DHIS2 before January 2019 in Bulawayo and February 2019 in Harare and were not analyzed.

All interviews and FGDs were recorded digitally, transcribed verbatim, and transferred to NVivo14 software for analysis. In the field, notes were taken and interview summaries were prepared daily. A codebook was developed through a review of all transcripts by three researchers, who contributed to the development of a thematic framework of codes through consensus. Thematic analysis was used to identify patterns and themes in the data, with some codes determined beforehand and others emerging during coding.

All data sources (qualitative, quantitative and policy data) were triangulated at the interpretation stage to trace policy events and assess the impact of the COVID-19 pandemic and its control measures on the provision and use of healthcare services for women and children.

Zimbabwe is a low-income landlocked country with approximately 15 million people, 42% of the population under the age of 15 (20). Approximately 71% of the population lives below the international poverty line of $1.90 per day (21) leading to poor health indicators. Over 80% of income earners operate in the informal sector (22) which contributes to over 60% of the gross domestic product (GDP) (22, 23). One in four of the urban population (about 1.25 million people) lives in settlements with poor water and sanitation infrastructure, as defined by UNICEF (24).

The country experienced severe political and economic challenges for the three decades before the COVID-19 pandemic. An unsustainable external debt burden and structural adjustment program resulted in a severe contraction in public expenditure in the 1990s, which weakened most public services (25). The turn of the millennium saw that the country entered a period of protracted political turmoil and the imposition of targeted sanctions by several European countries (25, 26) further undermining the economy. Just before the pandemic, Zimbabwe’s GDP declined by 11.3% in 2019 compared to 2018, and continued to contract in 2020 (27).

Although Zimbabwe’s health system was once considered among the best in Africa (28) it deteriorated due to reductions in public spending, out-migration of health professionals, mismanagement, and industrial action (29, 30). By 2021, the doctor and nurse-person ratios had deteriorated to 1:12,000 and 0.39:1,000, respectively, far below the World Health Organization’s (WHO) recommended minimum threshold density of 2.28 doctors, nurses, and midwives per 1,000 people in developing countries (31). Today, there are high levels of out-of-pocket payments and donor dependency (17). The maternal mortality rate was 458 per 100,000 live births (32) and the child mortality rate was 55 per 1,000 live births in 2020 (33).

Zimbabwe faced social, economic, and political challenges that impacted the country’s health system long before the COVID-19 pandemic. In addition to this chronic deterioration of the health system, there were a series of acute crises during the run-up to the pandemic (see Figure 2). This included doctors and nurses engaging in intermittent industrial action (30), typhoid and cholera outbreaks in 2018, a major natural disaster (cyclone Idai) (22) in 2019, and an acute worsening of inflation since April 2020, causing shortages of fuel and water.

The precarious state of the health system during the pandemic was noted in the qualitative data (Figure 2).

According to our key informants, a notable feature of the response to COVID-19 was that policies were implemented in a centralized and top-down manner with little bottom-up and contextualized input.

Informants also noted that public health communication (mainly conducted through the Zimbabwean broadcasting cooperation’s television and radio channels) was not fully effective, and that misinformation and disinformation on various social media platforms contributed to poor compliance with COVID-19 control measures, despite rapid response teams being established to respond swiftly to rumors and false information, albeit with limited capacities to deal with all rumors due to a shortage of human and material resources. Others have indicated that many people feel overloaded with conflicting information (40).

Another theme that emerged from our primary data was the impracticality and impossibility of implementing social distancing measures, especially in high-density and overcrowded informal settlements with inadequate water and sanitation infrastructure (56). Other measures such as the use of hand sanitizers were unaffordable. According to one study in Harare and Mashonaland East, between 18 and 53% of healthcare workers reported a lack of soap, water, and masks between June 2020 to the middle of 2021 (57). Other studies have similarly described how communities cannot comply with social distancing measures because of the need to secure food, water, and income (40).

The inability to monitor and enforce travel restrictions was another cause of poor compliance with the lockdown measures. Despite the deployment of the police and the army, resources were insufficient to enforce quarantine and ensure adherence to travel restrictions. For example, despite major points of entry being manned by armed security, many people were able to enter the country illegally without being quarantined (58).

The relative ineffectiveness of lockdown measures may also be deduced from seroprevalence studies, which found that, by the end of March 2021, a high proportion of the population was infected (59). A systematic review of seroprevalence studies in Africa between January 2020 and December 2021 also estimated that the seroprevalence in Southern Africa was 56.1% (95% CI 44.6–66.9%) by the third quarter of 2021, with lower rates in rural areas and high heterogeneity between countries (59). In a study of randomly selected households from three high-density communities in Harare, seroprevalence increased from 19% in November–December 2020 to 53% in February–April 2021 (58), suggesting high levels of community transmission during the second wave. A survey of healthcare workers as early as June 2020 reported that 39.1% had a history of COVID-19 symptoms and a SARS-CoV-2 seroprevalence of 8.9% (60). The reported data also indicated a less severe disease profile in Africa, with more asymptomatic cases than in other parts of the world (61).

Another feature of the policy response to COVID-19 was the government’s efforts to mitigate the social and economic harm of the lockdown. In May 2020, the government announced a COVID-19 Economic Recovery and Stimulus Package worth approximately 9% of the GDP to mitigate the lockdown’s social and economic impacts. This package included US$5 monthly income support for vulnerable households and fiscal support for sectors like manufacturing, agriculture, mining, and tourism (62). Temporary measures were also introduced to defer rent and mortgage payments. However, the income support was not fully implemented, with only 202,000 of the 1 million eligible households receiving it 4 months into the lockdown (23). A compensation scheme for civil servants who lost their lives on duty was similarly incomplete, but all civil servants were granted a US$75 monthly COVID-19 allowance for over 36 months, in addition to their regular salaries.

Our analysis of routinely collected data from July 2016 to July 2021 in Harare and Bulawayo found evidence of a reduction in healthcare utilization. Figure 3 presents data on the number of fourth antenatal care visits by pregnant women and the number of HIV tests conducted among pregnant women at their first antenatal visit.

The data show a reduction in the number of fourth ANC visits in both provinces, with a decrease of 29% in Harare and 34% Bulawayo following the COVID-19 lockdown. However, there was no observable reduction in the number of HIV tests among the women attending antenatal care (Supplementary Table 1).

Key informants explained that the HIV program, which is largely donor-funded and relatively well resourced, was better able to maintain their pre-COVID-19 level of service. In a study conducted in Harare, the number of pregnant women accessing HIV services remained the same, except for a slight decline during the first lockdown (5).

Figure 4 presents the data on the number of GM visits by children under 5 years of age and the number of child PCCs. The data show that the number of GM visits decreased by more than 50% in both Harare and Bulawayo after the first COVID-19 lockdown (p < 0.001; Supplementary Table 1). The number of PCCs decreased substantially in Harare from March 2020 onwards, but not in Bulawayo.

Other studies also report that healthcare utilization declined in 2020 and 2021 compared to earlier years (63, 64). A study at Mpilo Hospital in Bulawayo, although the mean number of monthly deliveries reduced from 747 in the first quarter of 2020 to 681 in the second quarter, the overall reduction was not substantial (65).

Our qualitative data pointed to several reasons for the reduction in healthcare utilization. In some places, clinics were closed because the staff went sick because of COVID-19 (66) or because of industrial action. Some informants explained that healthcare utilization also dropped because health facilities were perceived as “hotspots” for COVID-19 transmission, especially those that had been repurposed for COVID-19 services.

Police roadblocks and checkpoints were highlighted as barriers to health care because of fear and unwillingness to disclose one’s health status. Healthcare workers reported that it was difficult to access public transport because they were perceived to be at risk to others, especially when wearing a uniform.

Health services were especially disrupted by operating times being reduced from 07:30–17:30 to 08:00–14:00 and by capping the number of patients seen per day. In some instances, operating hours were reduced even further because healthcare workers struggled to get to work or had to leave early to comply with curfews.

The pandemic diverted resources and attention away from MCH services. Informants reported that in some clinics, GM was completely suspended to allow for the prioritization of other health services. Stockouts particularly affected immunization and family planning services (67).

Access to healthcare was also restricted by some facilities implementing a policy of patients, including pregnant women, showing a negative SARS-CoV-2 test result before admission. This led to patients being turned away from the hospitals, and ambulances refused to transport patients.

Some healthcare workers believe that an increase in the number of home deliveries without a skilled birth attendant resulted in an increased number of largely unreported maternal and neonatal deaths.

Other studies have reported similar findings in women and children struggling to access MCH services during the pandemic (15, 65). In Nyanga, Manicaland province, Nyashanu et al. describe transport problems, roadblocks, shortage of medication and personal protective equipment, and lack of routine healthcare services affected routine health access such as antiretroviral therapy (40). Chimhunya et al. viewed strikes by healthcare workers at Sally Mugabe Hospital in Harare as an indirect consequence of COVID-19, which caused a decline in neonatal admissions (64).