Self-reported delays in care-seeking in West Africa during the first wave of the COVID-19 pandemic

The COVID-19 pandemic resulted in delays in care-seeking behavior in countries globally [1, 2]. Particularly during the beginning of the pandemic in March 2020, there were significant decreases in primary and specialty care visits due to fears of COVID-19 infection [2‐4], government restrictions on movement [1, 3, 5], and intentions of reducing the use of medical care to reserve capacity for the most severely ill [5]. Delays in care-seeking resulted in challenges to individual and public health, including reductions in primary care visits [6], immunizations [7‐9], prenatal care [10‐12], pediatric care [13], cancer screenings [5], mental health visits [5], and emergency care [14], all of which have long-reaching economic and social consequences.

Studies thus far have yielded mixed results regarding differences in care-seeking across countries and settings during the COVID-19 pandemic. Several studies have suggested that low-income countries were disproportionally affected by delays of care-seeking due to COVID-19, particularly early in the pandemic [15, 16]. Forty-five percent of low-income and 30% of low-middle income countries (compared to 23% globally) reported partial disruptions in at least three quarters of their health services [15]. While another large, multinational study found no clear relationship between healthcare disruptions and national wealth, it suggested that urban regions may have been more impacted [17].

Studies that have emerged during this pandemic have also shown mixed findings regarding what populations and risk factors have been linked to decreased care-seeking, particularly in communities that already experienced significant barriers to care-seeking prior to COVID-19. One study comparing pre- and post-COVID-19 care-seeking in slum communities in four countries saw major decreases in care-seeking related to healthcare cost increases, income decreases, limited transportation to healthcare, and high fears of COVID-19 infection [18]. Another community study in Portugal suggested that those with decreased income since COVID-19, with less confidence in the pandemic government response, and who identify as women were more likely to delay care-seeking [19]. In Ethiopia, there is evidence that those who experienced transportation difficulties, had strong fears of COVID-19, and were over the age of 60 years were much less likely to present to care [20]. Those who experienced food insecurity were significantly more likely to delay care-seeking in the United States during the first wave of COVID-19 [21].

However, many patient-level factors that may influence care-seeking remain unknown, specifically in West Africa, where there were several known risk factors for the underutilization of healthcare prior to COVID-19. Previous research in this region suggested that access to healthcare is generally hindered by lack of transportation [22], long distances from clinics [23‐26], and location in rural areas [27]. Lack of health insurance [23, 28] and low income [29] are significant financial barriers, and other patient factors, such as older age [27] and being female [24, 30], have also been linked to lower use of healthcare. Like other areas, West Africa reported major decreases in healthcare utilization during COVID-19 [31], with demonstrated reasons for delayed care-seeking being distrust of medical facilities [32], fear of contracting COVID-19, and lack of face masks to enter buildings [33]. Recent research in West Africa has begun to elucidate specific risk factors for delays of care-seeking during COVID-19. In particular, evidence exists that women were more severely affected by lost employment and lockdown restrictions during the pandemic than men [34], likely affecting their ability to seek care. Pregnant women in Ghana also showed decreased likelihood of seeking hospital antepartum care, reporting COVID-19 fears about illness as well as concerns about food and income loss [35]. However, very little is known on how other preexisting risk factors or socioeconomic concerns (e.g. financial stability, food insecurity) affected care-seeking during the early COVID-19 pandemic in this region.

Evidence regarding access and delivery of care during this pandemic is still emerging. Globally, many countries developed increased resources for routine, non-COVID-19 care in response to the first wave of the pandemic, but there have been mixed results on the relationship between pandemic stage, patient access to care [36], and likelihood of seeking care [37, 38]. Some global studies showed utilization and perceived access to care in some countries improving or returning to near pre-pandemic levels after lifting nationwide lockdowns [39‐42], while others showed lasting effects, particularly in acute care [43]. A multinational study including Ghana suggested regional variation in healthcare disruption patterns during the first year of the COVID-19 pandemic, with Ghanaian healthcare services largely returning to baseline at the end of 2020 except for emergency room and total outpatient visits, which remained lower than in the years prior [17]. Considering the ever-evolving impact of COVID-19, it remains important to understand the risk factors that most impacted early care-seeking, before healthcare systems and government programs were able to adapt to pandemic circumstances and aid those most at risk.

The Research for Effective COVID-19 Responses (RECOVR) survey was developed by Innovations for Poverty Action (IPA), a nonprofit organization, and administered via telephone in ten countries between May and June 2020 [44]. The purpose of the survey was to assess the early economic impact of COVID-19 in real time and identify potential policy implications across the world. The survey included data from Burkina Faso, Ghana, and Sierra Leone in West Africa, all of which reported their first COVID-19 cases during the second week of March 2020.

We analyzed the survey results to understand the rates of self-reported delays in care-seeking and explore factors linked to delays in Burkina Faso, Ghana, and Sierra Leone during the first wave of the COVID-19 pandemic. We used these results to describe variability in factors associated with delays in care-seeking in these West African countries that, while geographically close, differ in economic and human development, with Ghana having a substantially higher human development index. Our aim was to use early data from the COVID-19 pandemic to explore community-reported themes related to the challenges in care-seeking that emerged during this pivotal time. Understanding the impact of COVID-19 on care-seeking behavior in West Africa, particularly at the beginning of this pandemic, is ultimately important to inform early interventions to address barriers to care at the individual, household, and health system levels during current and future pandemics.

We found that there were significant population-level rates of self-reported delayed care-seeking during the first wave of the COVID-19 pandemic in these three countries in West Africa. In bivariate analyses, factors associated with increased reported delays in care-seeking in at least one country included being female, concern about someone in the household contracting COVID-19, and food insecurity. Multivariate analyses ultimately showed higher rates of delayed care-seeking by those with self-reported food insecurity in Burkina Faso as well as those with self-reported fears of household COVID-19 infection in both Ghana and Sierra Leone. Among the households who reported delays in care-seeking, both cost and fear of infection were broadly noted as reasons for delays in care-seeking by a large number of participants in all three countries.

Food insecurity was common among survey respondents and was frequently linked with delaying care-seeking and, among those delaying in Ghana and Sierra Leone, with financial reasons for the delays. There is newly emerging research investigating the relationship between food insecurity during COVID-19 and delayed care-seeking in other parts of the world. In the United States, those with food insecurity had significantly higher odds of delaying or forgoing medical care, including specifically due to cost concerns [21]. Additional studies have shown links between food insecurity and decreased antenatal care in Kenya [51] as well as delayed healthcare by parents of low-income families in California. Considering that a large majority of respondents in Ghana and Sierra Leone (72.2% and 81.7%, respectively) and a substantial minority in Burkina Faso (36.3%) reported income loss, it is possible that food insecurity served as an additional marker for economic vulnerability during the early lockdowns. Such a correlation between food insecurity and delays in care-seeking is particularly alarming considering the worsening of poverty and food insecurity since the beginning of the pandemic [52]; an estimated 25 million people in West Africa could not meet their food needs in 2021, a 34% increase from 2020 [53]. Early research during the COVID-19 pandemic suggested that poor and rural households were most at risk of significant worsening of food insecurity during the pandemic and that these effects have the potential to impact countries for up to 10 years [54]. Our data suggest that food insecurity may be linked to delays in care-seeking; thus, a major, lasting destabilization of food security related to COVID-19 presents a major threat to ensuring continued healthcare. Regarding the preservation of healthcare utilization during times of stress, our results support the need to address not just the direct medical impact of COVID-19 but also the economic and social impact of lockdowns and income loss, which ultimately may indirectly affect patients’ access to healthcare due to competing economic demands. Thus, coordinated interventions on a national and international scale that target the social determinants of health have the potential to reduce economic vulnerability to delays in care-seeking in future global health crises.

Participants’ perceived risk of themselves or their family members contracting COVID-19 was also significantly associated with decreased care-seeking behavior in Ghana and Sierra Leone. Our findings support existing evidence that fear of COVID-19 infection contributed to delays in care-seeking during the early COVID-19 pandemic in Ghana [55] and Sierra Leone [31]. Worries about COVID-19 infection and resulting decreases in medical care have been seen internationally [4, 56‐60], with some studies suggesting fear of infection to be the prevailing reason for decreased care-seeking by patients [11]. Patients may not have felt safe visiting medical clinics or leaving the home; while decreased movement is beneficial in slowing spread of COVID-19, it has serious unintended medical and social consequences, such as reduced treatment of infectious diseases (e.g. tuberculosis) and chronic diseases like diabetes [61, 62]. Effective communication from healthcare workers [63] and public health authorities [38, 64] to the general public is essential to balancing the spread of COVID-19 with the necessary utilization of regular medical care, and information dissemination has and will continue to influence continued COVID-19 vaccine uptake. Ensuring adequate patient education and access to personal protective equipment is an important area of future study regarding patients’ perceived safety in medical settings.

Sierra Leone had a lower overall rate of delays in care than Burkina Faso and Ghana. Of note, Sierra Leone was one of the primary locations for the Ebola outbreak in 2014, resulting in the enhancement of disease surveillance and mitigation systems. It is possible that such developments helped mitigate the delay in healthcare-seeking due to effective communication and trust in healthcare authorities in Sierra Leone [65]. Sierra Leone had a higher reported percentage of urban participants (83.7% compared to 60.5% in Burkina Faso and 22.9% in Ghana), although this contrast may reflect the manual coding of non-urban versus urban in Burkina Faso and Ghana, which may have underestimated urban residence. Ultimately, we did not see a difference in reported delays in care by urban versus non-urban settings, contrasting some pre-COVID [24, 66] and post-COVID [16, 17] health services research that found that variance in healthcare usage based on distance from clinics or metropolitan area. Other factors may also have been important, including strength of the primary care system. For example, the Ghanaian healthcare system has increased care delivery through Community-based Health Planning and Services (CHPS) zones, which have allowed for mobilization and increased proximity of care [67], potentially decreasing need to delay care in the context of COVID-19. Further research should investigate variation in the impact of COVID-19 by rural versus urban status by specifically assessing location status.

In contrast to a number of telephone surveys that have included primarily men (e.g. 33% female [68]), our weighted population was majority female (54–58%), suggesting adequate sampling. Previous research suggests that females have higher rates of delay globally during pandemics [69], and more recent research in Switzerland, Kenya, and Ghana has also suggested a link between female gender and delays in healthcare during the current COVID-19 crisis [34, 35, 70, 71]. However, in our sample, only Burkina Faso showed an association between female gender and increased delays in care-seeking. While we did not explore further, a previous study in Burkina Faso suggested that women have lower bargaining power in their households and work environments, leading to less control over their healthcare usage [72], perhaps contributing to the association seen in our study. Studies directly comparing delays in care-seeking across countries will be able to better characterize these observed associations.

Bivariate analyses showed that in Ghana, older age groups showed increased rates of delays in care-seeking. We do not have the knowledge to understand the cause of these delays, but possible reasons include that during COVID-19, older people may have needed additional financial and social support, as they may have been disproportionately impacted by social distancing and the decreased wages of those in their social network [32, 73]. In contrast to studies pre-pandemic and some studies during the COVID-19 pandemic [16], we did not see any differences based on education. Previous studies have found that patients with lower education completion were more likely to present later to healthcare for a variety of health conditions [24, 66]. Recent studies of delays in patient presentation for medical conditions caused by COVID-19 have also found that those with lower education were more likely to delay care-seeking [70]. It is possible that in our study, food insecurity was a strong marker of economic vulnerability with a more robust link with delayed care as compared to education; however, there was still no link between delays in care-seeking and education level, even when controlling for other variables in multivariate analysis. Our sample had high proportions of each education level, but it is still possible that it was underpowered or that there is an unmeasured confounder masking a link between education and delayed care-seeking.

There were a number of limitations to our study. The survey data were limited to participants with a mobile phone, affecting generalizability of results for individuals with less access to technology and mobile phone coverage. However, 75% of Ghanaians aged 15 and older reported owning a cell phone for personal use [74]. Cell phone coverage is also high in Burkina Faso and Sierra Leone (106 and 86 mobile phones per 100 people, respectively) [75], although the survey sample of Sierra Leone was significantly more urban compared to the two other countries, making it less representative of the general population. A prior study of phone surveys in sub-Saharan Africa during COVID-19 found that while survey participants are typically older and more educated than the general population, weighting based on likelihood of being interviewed (using individual and household characteristics) improved gender and educational representativeness of the samples [76]. Our use of sampling weights thus is likely to have improved generalizability of our results.

Additionally, the surveys in Burkina Faso and Ghana did not explicitly assess urbanicity of participants. Our approach of inferring urban versus non-urban location based on reporting living in the largest city of certain regions may have led to misclassification by underestimating the number of people in urban environments but smaller cities. This notable limitation may have affected our power to detect a relationship between location and delays in care-seeking.

Importantly, the surveys were collected from participants at a single time point. The available data cannot determine causality and does not assess any relationship between delays in care-seeking and harmful outcomes such as missed preventative care, disease diagnoses, and treatment of illness. Additionally, all analyses of factors were done within countries without statistical comparison between countries. Further studies that assess socioecological factors and policy in greater detail are needed to statistically assess potential differences between countries.

Many countries recognized challenges related to food insecurity early during the COVID-19 pandemic and provided food and other support. Ongoing and future research should assess if and how these factors or rates in delayed care-seeking changed throughout the subsequent waves of the pandemic and as a result of changes in government supports, health system changes, and infection prevention responses. Our power to identify associated factors was low due to the low percentage of participants reporting a delay in care-seeking. This may have limited our ability to detect links between care-seeking and other factors, such as education. Finally, our ability to better understand the causes of delays in care-seeking was limited by the data available from survey questions, which did not delineate whether healthcare was needed in the studied time frame, and, if so, what were the types, length, and outcomes of delayed care-seeking. Further exploration of these causes as well as the purpose and consequences of the delayed care will be vital to recognize those most at risk for poor outcomes and inform to support uninterrupted care-seeking in future health system shocks.

In conclusion, we found that there were notable rates of self-reported delays in care-seeking and several factors that related to these delays during the early COVID-19 pandemic in West Africa. Food insecurity and perceived risk of developing COVID-19 infection had significant relationships with delays in care-seeking, with those reporting increased food insecurity and/or perceived risk of developing COVID-19 having higher rates of delays in care-seeking. Increasing trust and safety of the health systems and patient education while addressing social determinants of health, such as ability to access and afford care, is necessary to improve overall care-seeking before, during, and after future pandemics. Further investigation into these factors and the sustained public health impacts of COVID-19 is ongoing and will likely also show variations in healthcare access and delivery over time in the ever-evolving landscape of COVID-19.