Abstract
This rapid scoping review has informed the development of the November 2020 United Nations Research Roadmap for the COVID-19 Recovery, by providing a synthesis of available evidence on the impact of pandemics and epidemics on (1) essential services and (2) health systems preparedness and strengthening. Emerging findings point to existing disparities in health systems and services being further exacerbated, with marginalized populations and low- and middle-income countries burdened disproportionately. More broadly, there is a need to further understand short- and long-term impacts of bypassed essential services, quality assurance of services, the role of primary health care in the frontline, and the need for additional mechanisms for effective vaccine messaging and uptake during epidemics. The review also highlights how trust—of institutions, of science, and between communities and health systems—remains central to a successful pandemic response. Finally, previous crises had repeatedly foreshadowed the inability of health systems to handle upcoming pandemics, yet the reactive nature of policies and practices compounded by lack of resources, infrastructure, and political will have resulted in the current failed response to COVID-19. There is therefore an urgent need for investments in implementation science and for strategies to bridge this persistent research–practice gap.
Keywords
Introduction
In January 2019, the world was introduced rather abruptly to COVID-19. Health systems around the world have had to rapidly adapt to increasing demand related to the novel coronavirus while concurrently attempting to maintain essential services. Fear, stigma, misinformation, and the many unknowns surrounding this pandemic have further disrupted the delivery of health services, leading to many health systems globally, even within high-income countries (HICs), being overwhelmed. While no region has been spared from COVID-19, countries and communities that were already vulnerable and struggling to reach the United Nations (UN) Sustainable Development Goals (SDGs) by 2030 are inevitably hit the hardest, reversing years of progress.
With the ongoing state of the pandemic, there are lessons learned and evidence acquired from previous crises that can be helpful. The Ebola Virus Disease (EVD), for example, first appeared in December 2013. Because of its unfamiliarity, EVD was not detected until March 2013 and did not reach the surveillance systems until May 2013. Distrust of governments and health workers, particularly rampant in areas where EVD hit the hardest, slowed down the response when it finally did occur. Routine services were severely impacted and deteriorated throughout the outbreak, in part due to shortages of health workers and dysfunctional surveillance systems. Failure of leadership emerged at various levels, with culturally insensitive public health messaging and operations in which patients were taken away and removed forcefully from communities, and burials were banned. EVD clearly demonstrated that in order to manage crises more effectively, certain areas needed to be strengthened, particularly with regard to (1) detection (better surveillance system), (2) governance (political will and trust), and (3) protection of health care workers (HCWs). Unsurprisingly, these three domains remain at the heart of the current COVID-19 pandemic.
The Ebola outbreak was in many ways a stark wake-up call to build and maintain resilient health systems. Kruk et al 1 define resilience as “the capacity of health actors, institutions, and populations to respond to crises; maintain core functions when a crisis hits; and, informed by lessons learnt during the crisis, reorganize if conditions require it.” A scoping review led by Nuzzo et al 2 went one step further and translated available definitions of resilience into 16 themes (#1068, Type VIII). Many of these themes are addressed by this scoping review and are highlighted throughout. Others, due to the rapid nature of the scoping review, did not emerge but merit further investigation. For example, evidence around crisis financing, leadership, and command structure (arguably the most complex function of a health system); collaboration with partners within and outside of the health system; and legal preparations, were unfortunately not captured by the chosen search strategy. As Nuzzo et al 2 note, much of the literature on health systems resilience tends to address high-level themes that do not necessarily translate into clear implementation strategies that governments and health facilities can adopt. Different actors—global, national, and local—also tend to have varying priorities when it comes to the elements of resilience in which they would like to invest, 3 further challenging implementation (#776, Type IV). While the World Health Organization (WHO) has devised a framework with specific questions and indicators for health systems to adopt, “Toolkit for Assessing Health System Capacity for Crisis Management,” 4 no evidence has emerged as to its use or implementation. As the research world plans for COVID-19 recovery, the hope is that investments are made to better understand the implementation of strategies aimed to improve resilience of health systems, the evaluation of such programs and strategies, and the protection of vulnerable populations during a crisis and its aftermath.
Methods
This rapid scoping review was conducted in response to the April 2020 UN framework for the immediate socioeconomic response to COVID-19, with a focus on Pillar 1, Health First: Protecting Health Services and Systems During the Crisis. The structure of the report and the search strategy loosely followed the two-pronged strategy instilled by the UN development system (UNDS): (1) evidence regarding which essential services are impacted and how they may be maintained and (2) evidence regarding health systems preparedness. Given UNDS’ emphasis on primary health care (PHC) and universal health coverage (UHC), evidence (or lack thereof) of these issues was further emphasized.
Given the broad scope of the review, and to ground the research in practice, expertise was first sought from health systems experts at Johns Hopkins Bloomberg School of Public Health, the Center for Humanitarian Health (n = 3), the Center for Health Security (n = 1), the Center for Global Health (n = 1), the Center for Livable Future (n = 1), and experts in health systems strengthening from the Bill and Melinda Gates Foundation (n = 2). Three of these experts had directly worked to manage the EVD response in West Africa; all of them are involved with the ongoing COVID-19 research response. Snowballing technique was used to identify experts through the author's connection at the school and their recommended parties. Experts were asked 2 open-ended questions about (1) where research investments should be prioritized in health systems for COVID-19 recovery and (2) what separates the COVID-19 crisis from previous outbreaks. These conversations and the UNDS’ focus on essential services and health systems strengthening and preparedness helped shape the search strategy (Supplemental Appendix 1, created with the help of librarians at the Johns Hopkins Welch Medical Library) and the eventual structure of the report. Given the emphasis on UHC and PHC in the UNDS report, and the emphasis on surveillance systems and HCWs for systems strengthening by the experts, these four themes had additional secondary search strategies embedded into the larger strategy to ensure that relevant studies would be captured.
PubMed was used to identify the relevant literature. In addition, websites of major relevant public health organizations (WHO and Centers for Disease Control and Prevention) were used to search for gray literature. All search results were filtered to include only those articles published during or after 2009, to capture literature emerging in the wake of the H1N1 pandemic, as well as the EVD outbreak and the COVID-19 pandemic. Although the severe acute respiratory syndrome (SARS) outbreak occurred years earlier, any studies referring to SARS were included. Studies referring to Middle East Respiratory Syndrome (MERS), even if not explicitly referred to by the search strategy, were also included. The end of the study period was the day of the search: July 22, 2020. Only English- or French-language studies were considered.
Commentaries, reports, perspectives, and opinion pieces lacking empirical evidence were excluded. Other exclusion criteria included: clinical guidelines or medical interventions for treatment; clinical features of the illness; mathematical models and projections; studies that did not refer to SARS, EVD, H1N1, MERS, or COVID-19 (eg, natural hazards, conflicts, or other smaller outbreaks); studies that did not directly contribute to the areas of priorities referred by the UNDS framework or conducted nonhuman research; and/or studies with <150 participants for a quantitative study or <35 participants for a qualitative study. Studies assessing the effect of COVID-19 on non-essential services (eg, elective surgery) were also excluded. These restrictions were necessary given the rapid 1-month timeline requested for the review.
Covidence was used for the management of articles. Each article's title and abstract were reviewed by the author for relevancy and tagged thematically (eg, essential services; infection, prevention, and control [IPC]). Studies were also marked if they were deemed to be a high priority. All articles deemed relevant after the title and abstract screening were read by the author, with the most relevant information extracted and assigned into the appropriate section of the report. This was an iterative process, with new sections emerging as additional literature was processed. The results and discussion section were merged in one for this review.
Included studies in the report (as opposed to background references) are followed with the study number (see Supplemental Appendix 4 for the full list of included studies) and the level of evidence (Supplemental Appendix 3). It should be noted that health systems research usually does not use randomized controlled trials. Qualitative studies are often very informative, particularly when investigating attitudes, perceptions, and behavior regarding uptake of services and access to care. The framework used to classify studies should therefore not be seen as a hierarchical level of evidence but rather as a depiction of the heterogeneity of study types included in this scoping review.
Results
See Prisma Chart (Supplemental Appendix 2). Of 1,619 studies imported, titles and abstracts of 1,446 were assessed once duplicates were removed. A total of 188 full-texts were reviewed, for a final inclusion of 121 studies.
Vulnerable Populations and Inequity
Crises seldom affect everyone equally; COVID-19 is no different. Despite the calls of COVID-19 as the great equalizer, evidence indicates that already vulnerable and marginalized populations are feeling the effects of the pandemic more forcefully, exacerbating existing disparities. Underlying systemic inequalities in health systems globally have been exposed due to the COVID-19 pandemic, clearly delineating inequities. While health systems and health services are triaging and reorganizing their resources and infrastructure to prioritize COVID-19 patients, their response should take into account these inequities and protect the most vulnerable members of our society, the leaving no one behind principle set by the SDGs: women, people of color, LGBTQIA populations, ethnic minorities, migrants, indigenous communities, incarcerated populations, the elderly, people with underlying comorbidities, and people with disabilities. It is important to recognize the intersecting power differentials and conduct intersectional analyses, where vulnerable groups are not seen as being homogenous, but rather with intersecting oppressions that shape their experiences and their health outcomes. This scoping review has demonstrated a large gap in intersectional research.
In addition, as the pandemic reaches certain lower- and middle-income countries (LMICs) and refugee settings, inequities may be more rampant and longer-lasting given their difficulties in responding to the pandemic aggressively. 5 Shortages of HCWs, increased risk of existing HCWs due to lack of personal protection equipment, difficulties in maintaining social distancing, and short supply of essential resources for patients are but some factors whose solutions will need to be further investigated. A global effort in research and programming is necessary to support faltering health care systems and economies.
Socioeconomic Status
Pandemic planning and responses should consider socioeconomic disparities and the impact of inequalities. Vulnerable groups tend to be disproportionately affected by crises and are more likely to experience mortality and morbidity. For example, examining H1N1 pandemic variations in mortality between different socioeconomic groups in England demonstrated that individuals in the “most deprived” quintile had an age- and sex-standardized risk that was 3 times higher than those in the “least deprived” quintile 6 (#902, Type II). A similar study in New York City, United States, demonstrated comparable results, suggesting that individuals in a lower socioeconomic status were more vulnerable to severe H1N1 illness 7 (#1542, Type IV). A more specific examination of disparities associated with the H1N1 pandemic indicated that disparities in exposure, susceptibility, and access to health care were clearly evident among racial and ethnic minorities in the United States 8 (#1553, Type V) and in Mexico 9 (#1578, Type V). This disproportionate burden is confirmed with COVID-19 10 (#189, Type IV). Some of this may be due to accessibility barriers, as was observed with H1N1: for example, associations between socioeconomic status and barriers to access of information led to varying levels of knowledge related to the pandemic and misconceptions, thus influencing the adoption (or lack thereof) of preventative behavior among Americans 11 (#527, Level V).
Evidence intentionally investigating this space seems to mostly be from non-LMIC settings. Understanding variation in socioeconomic status and their health impacts in LMICs, therefore, need to be further investigated. Targeted communication efforts across different socioeconomic groups as part of the response and recovery are vital.
Gender
Public health measures that are launched in response to a public health emergency include siphoning limited resources toward the emergency, school and business closures, quarantines, which often results in a lack of access and resources placed toward women-centered services; these put the lives of women and girls at increased risk, exacerbating gender-based and domestic violence. 12 This was previously seen with EVD. 13 Existing gender norms, as well as roles establishing women as caregivers, often put them at the forefront of the frontline during emergencies. The resulting reduction in provision and use of critical sexual and reproductive health services will be disastrous for women. Progress made with the SDGs in improving maternal health and reinforcing family planning will likely be reversed with the pandemic, as women are among those bearing the heaviest burden of COVID-19’s effects. 14 While no specific study captured by the search strategy exclusively presented evidence on the direct and indirect burden on women and girls as a result of the ongoing COVID-19 pandemic or previous outbreaks, many of the studies captured pinpointed gender as a risk factor for increased morbidity and mortality. There is a need, however, to conduct gender analyses of the outbreak, showcasing differential vulnerabilities between and within genders and include this in countries’ pandemic preparedness plans. In addition, research consideration for gender implications of outbreak measures—such as quarantining and lockdown and how individuals’ different cultural, security, sanitary, and physical needs are recognized—is necessary. This review has highlighted a major research gap in integrating gender norms, roles, and relations as they pertain to the current crisis.
Race and Ethnicity
Health disparities are immensely influenced by race 15 ; this is not a novel statement. While there is no evidence that there is a genetic or biological factor increasing the risk of COVID-19 among racial and ethnic minorities, social determinants of health—conditions in which people are born, live, work, and age—provide important insights into these observed inequities. Just as in previous crises, associations between race, socioeconomic status, and health outcomes are evident with the current one. 16 Compounded with this are existing and significant prepandemic, socially embedded discriminations caused by racism and segregation (eg, overrepresentation in essential work, lower health insurance access, and unconscious bias from health providers). For example, a retrospective cohort study in California showed non-Hispanic African American patients having 2.7 times the odds of hospitalization as a result of COVID-19 17 (#356, Type II). Yet, individuals belonging to at least one racial/ethnic group were shown to be more receptive to preventive behaviors (eg, hygiene, social distancing, and hand washing) during the H1N1 pandemic 18 (#1564, Type V). Inequitable access to health care (vaccines or antivirals), however, remains a key issue18,19 (#1560, Type VII). Furthermore, consistent gaps in gathering data on race and ethnicity further exacerbate our ability to better understand the effects of the pandemic in certain communities. For example, preliminary data from the Connecticut Department of Public Health observed that 55% of COVID-19 laboratory case reports (n = 3141) had missing data on race and ethnicity. 20
When conducting research, being more gender- and race-intentional and incorporating intersectionality is vital. Acquisition of gender- and race-disaggregated data is an initial step that researchers and funders can demand, in order to best identify gender- and race-based gaps and outcomes and to tailor policy and program responses accordingly. There is also a need to further examine different behaviors during times of crisis that may be economically driven. Further investigations on mistrust and disbelief of the gravity of the pandemic, particularly in communities that are marginalized and may not have access to the latest resources or data, need to be conducted. Compliance with responses in settings with limited resources (eg, access to water) needs to be better understood. Better methods of communication to reach geographically distant health facilities and communities are needed.
Environmental Sustainability
SDG13 pinpoints the need to take urgent action to combat climate change and its impacts. Changes in climate will likely lead to wide-scale movement of vulnerable populations, migrants, refugees, and internally displaced individuals across the globe. Health systems will have to be better equipped to handle the changes and increases in population needing their services. In addition, health systems are directly affected by external shocks (eg, hurricanes, earthquakes, and flooding), the frequency of which is likely to increase due to changes in climate. None of the papers reviewed led to any notable research considerations regarding environmental sustainability. Further investigation on how health systems can be designed and how health services can be delivered to be more sustainable is needed.
Interruption of Essential Services
The third SDG, to ensure healthy lives and promote well-being for all at all ages, had seen progress in many health areas, although acceleration was needed to reach the 2030 goals. With COVID-19, health care disruptions are severe: For example, childhood immunization programs are interrupted in 70 countries alone, 14 exacerbating preventable morbidity and mortality. Health systems tend to be paralyzed when faced with crises, particularly in countries where resources are limited. These indirect effects of disease outbreaks may be overlooked in the context of needing to provide immediate COVID-19 health services. LMICs in particular, whose coverage of essential health services was lacking even without crises, tend to be further overwhelmed and overstretched.
The UNDS’s first strategy of Pillar 1 focuses on protecting essentials due to COVID-19. This section of the scoping review lays out existing evidence on the essential services that have been observed to be impacted by COVID-19, EVD, and other previous crises. For example, an assessment of changes in provisions of surgery and non-EVD admissions in the first year of the outbreak in hospitals in Sierra Leone indicated a reduction of both indicators, although government facilities had a lower reduction compared to private facilities 21 (#623, Type V). A similar trend in 2 non-LMIC hospitals is observed, where neurological emergency admissions during the lockdown in Germany declined by 44.7%. The consequences of this morbidity and mortality remain to be seen 22 (#206, Level II). Emergency department (ED) visits declined by 23% for myocardial infarction, 20% for stroke, and 10% for hyperglycemic crisis between March and May 2020 compared to 10 weeks prior to the emergency declaration for COVID-19 in the United States 23 (#150, Level VII). Long-term consequences of bypassing essential services remain to be seen and are a source of concern. For example, interviews with general practitioners in Belgium pointed to their worry about postponing chronic essential care as a result of COVID-19 measures 24 (#1338, Level IV).
Health Care Workers
Scarcity of HCWs remains a large problem in some countries, particularly in rural settings, and directly prevents the provision of essential services during and after outbreaks. For example, staffing levels before, during, and 16 months after the EVD outbreak in rural Sierra Leone remained in deficit throughout (67%, 65%, and 60% among medical staff members, 92% for all 3 time periods among non-medical staff members). High levels of EVD-related HCW deaths also question current approaches (or lack thereof) of HCW safety in LMICs 25 (#1224, Type V), suggesting that urgent measures are needed to improve the state of human resources for health systems in LMICs. Even then, while scarcity of staff remains an ongoing challenge, interviews in Sierra Leone pointed to nurses, midwives, medical staff, and managers understanding their increased risk but choosing to continue to provide care during the epidemic, with professional duty, responsibility to the community, and religious beliefs cited as the main rationales 26 (#754, Type IV).
Maternal, Health, Sexual, and Reproductive Health
The evidence indicates a clear delineation in EVD disproportionately affecting women, with heavy impacts felt in maternal, sexual, and reproductive health. Indirect mortality and morbidity effects of a crisis due to the lack of resilience of a health system and its inability to provide essential care may be just as important as the direct mortality effects of the crisis. Women and children seem to be the most affected, with maternal and reproductive health depicting a significant drop in utilization and extremely slow recovery, even in the post-outbreak period. The use of essential maternal and child health services was significantly reduced during the EVD outbreak and, more importantly, have not recovered to their pre-crisis levels, suggesting a lingering negative effect on access and utilization of care in the long term. For example, a retrospective cohort study examining maternal and child health indicators (antenatal care, institutional delivery, receipt of five infant vaccines, yellow fever, measles, and tuberculosis [TB]) used interrupted time series models to estimate trends pre-Ebola, during the epidemic, and post-epidemic. Increasing trends of pre-EVD were reversed during the crisis and have not recovered to the pre-EVD thresholds in the post-epidemic period 27 (#668, Level II). Similarly, positive trends in utilization of reproductive and maternal health services in primary care facilities pre-Ebola in Sierra Leone were erased and experienced a negative and significant decline due to EVD, with antenatal care coverage being most affected 28 (#695, Type VII). In-hospital deliveries and C-sections also significantly declined in all open facilities offering emergency obstetrics in Sierra Leone, a 20% national decrease early in the EVD outbreak, mainly attributed to the closing of private facilities 29 (#589, Level II). A decline in utilization was also observed in Guinea, where an ecological study of all health facilities before, during, and after the EVD epidemic pointed to a 51% decline in family planning utilization during the epidemic. Antenatal care visits and institutional delivery declined during the outbreak and only partially recovered in the aftermath (2 years post-crisis), reaching 63% and 66% of pre-EVD levels, respectively 30 (#694, Level VII). Another study in Guinea showed a substantial and significant reduction of 41% in antenatal care visits in EVD-affected centers. In addition, an increase in the proportion of women who tested human immunodeficiency virus-positive (HIV+) but who did not receive antiretroviral was observed (12% in 2013 vs 44% in 2013). Services regarding prevention of mother-to-child transmission in HIV were severely neglected due to the crisis 31 (#844, Type V).
Sexual and reproductive health services are also affected by societal responses to crises because they may often not be considered essential. If family planning services are not provided, the rate of unintended or dangerous pregnancies could ascend, significantly, yet an additional burden on women and girls. For example, a nationwide, community-based survey in Liberia investigating health access during the Ebola epidemic indicated that 67% of urban respondents and 46% of rural respondents felt it was impossible or very difficult to access health care. Prenatal, obstetric care, and emergency services were the most impacted 32 (#827, Level V). The question of trust and fear is also relevant here: Interviews with HCWs in Sierra Leone pointed to fear of Ebola and public mistrust keeping women from seeking and accessing care at a health facility 26 (#754, Type IV). While evidence on COVID-19 specifically is still limited, a survey of health facilities in India observed that maternal and child health outpatient services are the most disrupted due to COVID-19 33 (#747, Type V).
Malaria, Human Immunodeficiency Virus/Acquired Immunodeficiency Syndrome, and Tuberculosis
Malaria, HIV/acquired immunodeficiency syndrome, and TB were also significantly impacted. Presenting a unique challenge to EVD control due to symptom overlap, individuals may deter from seeking care for malaria for fear of being stigmatized by EVD, of being transmitted with EVD in-facility, or due to a lack of availability of HCWs. For example, a retrospective cross-sectional study of 19 health centers in two rural, malaria-endemic health districts in Guinea showed significant declines in the number of visits and antimalarial treatment of children under 5 years during the EVD epidemic. Stock-outs of rapid diagnostic tests were evident and prolonged 34 (#838, Type V). Similarly, a cross-sectional survey of public health facilities in Guinea pointed to a significant reduction in provision, treatment, and access to malaria care in EVD-affected prefectures compared to non-EVD-affected prefectures, threatening malaria control in Guinea 35 (#587, Type V).
HIV incidence had declined globally from 2010 to 2018 and efforts to reach pregnant women had reduced incidence among children. TB mortality among HIV-infected individuals had also declined in the last several years. 14 Due to COVID-19-related disruptions in prevention, detection, and treatment, all of these trends can easily be reversed. A rapid online survey of 1,051 men who have sex with men in the United States indicated that COVID-19 had adverse impacts on their general well-being, with some reporting challenges in accessing HIV services (testing, prevention, and treatment) 36 (#1569, Type V). Similarly, in Liberia, the number of visits, number of new patients, and proportion of patients with follow-up delay among HIV patients decreased in both hospitals, with an expected long-term impact on HIV care 37 (#900, Type VII). With regard to TB, an encouraging retrospective comparative cohort study of a TB program pre-EVD and during EVD in 10 health facilities in Guinea indicated that TB diagnosis remained consistent in both periods, treatment success and antiretroviral therapy (ART) improved during the EVD outbreak, and uptake of HIV testing was maintained. Facilities had resources for better health system planning and support (Ebola training, quality control of laboratory services, screening prior to entry at the health facility, support and EVD counseling for workers, etc), allowing them to remain open during the outbreak and thus, resulting in sustained TB program performance 38 (#1056, Type II). While this is encouraging, not all facilities may have the level of resources necessary to continue offering services during a crisis. The scoping review was not able to capture the effect of COVID-19 on other communicable diseases and neglected tropical diseases. In addition, the limited number of studies captured cannot be extrapolated to all settings; further investigations on the impact of crises on TB and HIV care need to be conducted.
Immunization
Routine immunization programs are disrupted when crises are introduced into the health care system, increasing preventable morbidity and mortality. With COVID-19, vaccination campaigns for measles and polio vaccines, in particular, are said to have been suspended, with vaccine shortages due to border closures further exacerbating the burden in LMICs. 14 The evidence suggests a clear delineation in the impact between LMICs and non-LMICs, with recovery from pandemic-associated vaccination decline much faster among the latter. For example, a vaccination coverage survey showed that coverage rates of measles vaccine and pentavalent vaccine, already low before the EVD outbreak, decreased further during the outbreak in Sierra Leone. Coverage rates had still not recovered to the preepidemic coverage levels 1-year post-epidemic 39 (#598, Level V). However, a review of 3 years postpandemic indicated that there was some rebound in immunization coverage 40 (#1086, Level VI). In comparison, in the United States, facilities participating in the Vaccines for Children program found that 90% of practices remain open during the COVID-19 outbreak and that 82% are still offering immunization services to their pediatric patients. Despite the pandemic, there seems to be an existing capacity to maintain this essential service in the country 41 (#1488, Type V). In England, regional variations in vaccination were observed following COVID-19 physical distancing measures. Measles, mumps, rubella vaccination counts were 20% lower in 2020 when distancing measures were first introduced, compared to the same period in 2019. Counts were seen to have improved thereafter, suggesting a change of vaccine messaging during the pandemic 42 (#1536, Level IV).
With immunizations, mechanisms for effective vaccine messaging and uptake are needed. For example, barriers and drivers to immunization uptake were addressed in a rapid qualitative assessment in 4 counties in Liberia post-EVD. The strongest barriers included fear of EVD and suspicion and lack of trust in health services and vaccines. The strongest driver was knowledge and experience in measles/polio immunization pre-EVD and a trusted source for receiving vaccines. The need to involve community leaders in immunization campaigns was heavily emphasized 43 (#1601, Type IV).
Cancer and Other Noncommunicable Diseases
The limited literature on cancer-related essential services focused on non-LMIC settings, pointing to a clear gap in research needs in these areas in LMICs. Patients with cancer are at higher risk of developing complications after COVID-19 infection. At the same time, they face increased barriers to access care and treatments, resulting in cancer-related health outcomes that may be suboptimal. 44 Surprisingly, the search strategy did not capture the effects of previous crises on cancer patients or their outcomes. A model using the current pandemic context predicts that cancer-related cases and deaths attributed to COVID-19 will be a significant burden on health systems in Latin America, 45 but studies using empirical data are lacking in this space. One study tackled cancer care in the era of COVID-19: A survey of oncology health professionals in Italy indicated that 93% of their departments had reorganized routine clinical activity due to COVID-19, but only 41% felt adequately trained about the new COVID-19 required procedures 46 (#621, Type V). The heterogeneity of responses and the lack of evidence regarding responses in LMICs need to be further investigated.
Even without a crisis, noncommunicable diseases (NCDs) are a leading cause of morbidity and mortality globally. The introduction of an infectious disease stressor into the health system often leads to mismanagement of NCDs, resulting in compromised care. One of SDG3’s targets is to reduce premature mortality by 33% between 2015 and 2030, a lofty goal that might be even further removed due to COVID-19. Previous crises certainly seem to suggest this: For example, a cross-sectional survey in Sierra Leone pre-, during, and post-EVD reported a decrease in NCDs during the outbreak. Post-EVD trends had not recovered to pre-outbreak levels. Interestingly, hypertension cases did not significantly change between periods; these were predominately managed in peripheral health units, suggesting that decentralization in care might be an effective strategy in NCD management 47 (#1077, Type V).
Mental Health
Understanding the burden of the outbreak on mental health is fundamental; psychopathology might reduce resilience and endurance during lockdowns and adherence to treatments. Economic uncertainty and heightened risk of infection add additional tolls on mental health during a crisis and in recovery, particularly in LMICs, for vulnerable groups at risk or those working at the frontline. The limited evidence of mental health implications due to COVID-19 is still emerging but implies a severe hidden pandemic within the larger pandemic. For example, a longitudinal study on mental health during COVID-19 in China pointed to a prevalence of 8%, 29%, and 17%, respectively, of moderate-to-severe stress, anxiety, and depression among the population in the initial assessment. Women faced a higher psychological burden 48 (#1679, Type II). Another assessment of the mental health of the general public in China during the initial phase of the outbreak indicated that more than half of study respondents (54%) reported psychological effects of the outbreak being moderate or severe. Female gender, being a student, physical symptoms, and poor self-rated health status were all significantly associated with greater psychological impact 49 (#1076, Type V). In India, an online survey in 64 cities reported more than 33% of the sample population suffering from significant psychological impact related to COVID-19 50 (#716, Type V, 453 responses with some missing data were excluded; findings may be biased). In the United States, non-binary respondents, female respondents, and those with mental and physical health conditions reported the highest levels of depression and anxiety among young adults 51 (#1489, Type V). Previous crises had also indicated similar trends, with risk perception of the crisis (SARS in this case) being correlated with psychological distress 52 (#70, Type V).
Finally, children, particularly in LMICs, may be directly and indirectly hit by pandemics and epidemics, due to the clinical manifestation and associated comorbidities but also due to the crumbling of social structures (overcrowded housing, inadequate sanitation, school closures, and income insecurity). Additional research is needed to better understand how children have been impacted and are coping with lockdowns and bypassed essential services.
Suggestions for Continuity of Care
Ideally, effective policies and mechanisms of containing community transmission, as depicted by Taiwan, through a strong public health network, cross-sectoral collaborations, and a strong surveillance and information technology capacity are key in reducing the pandemic's impact, thus ensuring continuity of care 53 (#1252, Type IV). This might include needing to establish a new infrastructure for better triaging: for example, opening of Ebola management centers in Tonkolili, Sierra Leone, was associated with earlier admission of cases to the appropriate health facilities 54 (#1584, Type II). Similarly, when the capacity of Ebola treatment units filled up, establishment of community care centers was helpful to reduce community spread and facilitate access to care closer to communities in Sierra Leone 55 (#1619, Type IV). Medecins sans Frontieres's (MSF) implementation of a 6 monthly appointment spacing approach (Rendez-Vous de Six Mois = RCM) has also proved to be efficient (60% reduction in the rate of attrition) for HIV care in EVD-ravaged Guinea. This approach may reduce staff workload and attrition of patients even in the midst of restricted access to care due to EVD 56 (#723, Type II).
For COVID-19, revised triage protocols by emergency physicians and surveillance methods with external facilities for imaging were shown to be effective in isolating COVID-19 patients early, thus decreasing the necessity for ED closures in Korea 57 (#1481, Type VII). Digital solutions also show promise: A patient portal-based COVID-19 self-triage and self-scheduling tool (linked to electronic health record systems) made available to primary care patients in California has proven to improve triage efficiency and avoid unnecessary in-person visits during the pandemic 58 (#238, Level IV). This type of digital solution is only likely to be readily accessible or available in wealthier health systems. There is a need to investigate better-triaging approaches in LMICs. In short, additional research is needed to find innovative ways (rapid diagnostics for better testing, better-triaging systems and flow of patients, hospital and clinic management, resource allocation, etc) to control transmission in order to safely continue care.
The role of community HCWs is significant in LMICs for provision of essential services but needs to be further investigated, particularly with regard to training, infection prevention guidelines, and alternative methods to supervision due to lockdowns and restricted movements. Existing programs include a Partners in Health program aimed at improving clinical outcomes, strengthening health systems, acting as the bridge between the community and PHC systems, and supporting community health workers (CHWs) trained to provide treatment support for TB, HIV, and leprosy post-EVD in Maryland County, Liberia. Evaluation of the program indicated community-based treatment support is associated with improvement in lost-to-follow-up rates, treatment success for TB, and improved retention of HIV clients 59 (#39, Type V). The role of CHWs and the trust they hold within communities (thus allowing for continued services during outbreaks) was further confirmed in another study in Liberia 60 (#724, Level IV). Informal health providers were also seen to have played a large role in managing morbidity during the EVD outbreak in Liberia 61 (#811, Type V). Additional investigation is needed to better understand informal networks of health care providers in curbing outbreaks and providing essential care. The quality of services and satisfaction of patients during a crisis would need to be further investigated in this domain.
Shift in Service Delivery: Telehealth and Contact-Free Options
Enabling telemedicine and other digital health innovations to bridge the gap in care, particularly among the most vulnerable populations, will allow ongoing access to essential health services and drugs while minimizing the potential for virus transmission. Numerous studies have shown promise in this transition to digital health. For example, clinicians in Boston were able to rapidly set up (<1 week) an inpatient telehealth program during the COVID-19 pandemic, an evaluation of which indicated that clinicians were able to build rapport and conduct a reasonable physical exam of patients through this modality 62 (#447, Level VI). The transition of diabetes management into a virtual care model did not change glycemic outcomes for patients in North Carolina, United States, suggesting that virtual care was a feasible and efficient approach for diabetes care 63 (#1349, Level V). In the United Kingdom, an assessment of telemedicine in a sarcoma unit during the pandemic pointed to high patient satisfaction, with clinicians finding the new modality efficient despite the lack of physical examination 64 (#125, Level V). Finally, in Spain, a survey of 185 hospitals in the National Health System pointed to 88% of hospitals implementing pharmaceutical teleconsultations before delivery of medications during the state of COVID-19 emergency. Expansion of telepharmacy, including remote dispensing and delivery of medication, suggests an important shift in service delivery to maintain continuity of care 65 (#108, Level V). However, feasibility in low-resource settings and accessibility due to digital inequity need to be further investigated.
Health Systems in Crises
There is an opportunity today to rethink existing health system models globally and to build a stronger global health infrastructure. Ensuring continuity of essential services and provision of care during the crisis and in recovery requires that health systems be resilient. Factors that constitute community resilience, key in addressing health shocks such as EVD and COVID-19, were identified through a qualitative study as being the following: strong leadership, community kinship, trusted communication, and trust among different health stakeholders 66 (#1073, Level IV). Resilience also incorporates preparedness ahead of time. This includes acquiring enough essential medicines and treatments, equitable vaccine delivery, protection of HCWs, and a strong surveillance system to prevent the emergence of a novel disease and its spread. Surveillance and monitoring systems in LMICs are absent or severely neglected; evidence is therefore lacking in this space. In addition, a commitment to PHC as a framework for organizing health systems and for delivering health services may be the best approach forward. Strong health systems also require significant investments in human resources of health, particularly in LMICs. Financing models and universal health care also need to be further investigated; evidence on the latter seems to be lacking.
Primary Health Care
Primary care is the first line of defense, particularly in a pandemic, with its ability to triage, test, treat, and educate patients and communities. PHC is often the gateway to health systems as a result. Linking public health and primary care approaches together, as was done in South Africa’s early and unique response with community screening and testing linked back to primary care, 67 has the potential to be quite effective in a pandemic response. Yet, while reflections on the importance of PHC in times of crises are available,68,69 empirical evidence is needed to enhance our understanding of the availability and response of PHC during public health crises, particularly in LMICs. Current evidence suggests that outbreaks have lingering effects on PHC systems; the length, severity, and both direct and indirect burden of these effects need to be further investigated.
For example, a 7-year analysis of key PHC indicators in Liberia before, during, and post-EVD outbreak across 379 facilities showed that EVD was devastating to PHC services during the outbreak (loss of 35%-67% of essential PHC outputs in 4 months). As of 2016, all health indicators seem to have recovered to pre-EVD thresholds. Nevertheless, the loss during the EVD outbreak will continue to affect population health long-term 70 (#833, Type V). In India, a cross-sectional survey among PHC facilities during COVID-19 reported that PHC facilities are constrained due to their weak infrastructure (eg, limited physical space, inadequate ventilation, suboptimal infection prevention, and control measures) 33 (#747, Type V). Similarly, provision of pediatric primary care could not be sustained throughout the EVD outbreak in Sierra Leone due to hospital closure and HCW infections 71 (#425, Type II). Effective primary care mental health services in low-resource settings can be challenging, as demonstrated in Liberia post-EVD 72 (#95, Type IV). Mental health training of PHC HCWs, availability of infrastructure, support for patients to access these services, and approaches to reduce mental health stigma should all be prioritized as part of the PHC strengthening.
With H1N1, a survey of more than 3,000 members of the American College of Obstetrics and Gynecologists indicated that the majority (86%) reported offering the 2009 H1N1 influenza vaccine to their pregnant patients. Obstetrician-gynecologists who considered primary care and preventive medicine to be a core aspect of their practice were more likely to offer this vaccine. Barriers to offering vaccination included insufficient reimbursement, storage limitations, or deferring to another provider 73 (#1288, Type V). A cross-sectional study on knowledge regarding the H1N1 pandemic in a PHC clinic in Malaysia pointed to a significant association between knowledge and adoption of preventive practices and attitudes during a pandemic 74 (#1305, Type V). Thus, as expected, not involving primary care physicians at the beginning of the H1N1 outbreak in France likely led to a dissonance between the public health messaging and the low perception of risk among individuals 75 (#1306, Level V). Similarly, despite the commitment of CHWs to maintain essential and primary care services during EVD, because they were engaged late in the response and did not receive sufficient support, care continuity was disrupted 76 (#1094, Level IV). The lack of support felt by primary care practitioners is further confirmed in Japan during the H1N1 pandemic 77 (#1318, Type V).
A better understanding of underlying social determinants is also necessary to assess the vulnerabilities of certain countries and their capacity to respond to outbreaks. Finally, more research needs to be done to understand integrated primary care models in the context of pandemics, where “whole-person” care, as opposed to specific specialties, might need to be prioritized.
University Health Coverage
Reducing barriers of access to care is key, particularly for vulnerable populations during a crisis and in the recovery phase. The SDG Report 2020 emphasizes the need for UHC such that “people have access to the services they need, when and where they need them, without financially hardship.” 14 UHC is pinpointed as the ideal for strengthening health systems, yet existing evidence on what universal health care looks like, which financing mechanisms are shown to be most effective, how to establish and maintain these mechanisms, and how the system as a whole can operate in the presence of multiple and sustained stressors seems to be missing. For example, while Sierra Leone is committed to providing free care to its women and children, field research has indicated that corruption and a lack of accountability within the health sector have resulted in patients having to pay for the alleged free care 78 (#295, Type IV). A declaration by the government to free access to care is therefore not sufficient: quality of care and power differentials within the facilities and the country need to be further investigated, particularly in the context of crises. In addition, financing approaches to maintain universal health care and the state of out-of-pocket expenditures as a result of COVID-19, particularly important in LMICs, were not captured by this scoping review and need to be further investigated.
Trust
Difficult to measure, observe, and capture, trust is key when it comes to health systems strengthening and preparedness, particularly during crises. It encompasses trust from the population that their governments and their health facilities have patients’ priorities at heart, trust from the HCWs that their wellness is valued and that they are supported to keep facilities operating during times of crisis, and trust in health systems to be resilient enough to care for populations when they most need health care. The success of one of the most important strategies in a pandemic response, contact tracing, also relies in large part on trust between the community and the health system.
Trust Through Better Communication
Disregard for public health recommendations has been said to be partly as a result of information pollution on social media; the uncertainty associated with the virus, the economy, and people’s livelihoods; and mistrust with politicians and government representatives 79 (#1091, Level VI). Part of building trust, therefore, requires appropriate culture- and context-specific, targeted communications before, during, and after the crisis. This is sorely lacking with the COVID-19 response and seems to have also been missing during previous crises. During the Ebola crisis, for example, distrust from communities in Guinea led to their reluctance to participate in efforts to fight EVD. Perceptions a year post-EVD about level of preparedness of their facilities remain low 80 (#720, Level IV). Similarly, fear related to reporting EVD as well as misconceptions about the disease likely resulted in delays in care-seeking in Sierra Leone 81 (#881, Level IV). In Uganda, an assessment of the community’s perspective of the EVD outbreak in central Uganda demonstrated that the credibility of biomedical explanations of the crisis was decayed due to insensitivity to local culture and understanding 82 (#444, Level IV). Technological innovations will not be sufficient if cultural sensitivity is not included in pandemic preparedness plans and response.
Fear, myth, and misconceptions during crises are common, not just within the larger population 83 (#1377, Level V) but also within HCWs, as indicated by a survey in Nigeria 84 (#577, Level IV). No difference in knowledge and attitude was observed between private and public, as depicted by another study in Lagos, Nigeria 85 (#1322, Level V). Support, strategies, and clear communications can play a role in alleviating some of these. Risk communication frameworks 86 need to be further investigated. For example, using trained volunteer health advisors in rural settings may be effective to disseminate information about the crisis, as demonstrated in Nigeria 87 (#1227, Type V). Additional research on how best to enhance trust in response systems, therefore, needs to be investigated.
Finally, targeted, context-specific health communications are key. An analysis of primary sources of information of the H1N1 pandemic among vulnerable groups along the Mississippi Gulf Coast in the United States demonstrated that television was most commonly used by participants, followed by the newspaper, and lastly, the internet 88 (#1526, Type IV). A survey on perceptions of COVID-19 among poor households in the Philippines indicated that most (86%) received information about the pandemic through television 89 (#733, Type V). Non-targeted communication methods and assumptions about wide internet usage when disseminating information can therefore be limited, result in misinformation, and harm pandemic response.
Trust: Conducting Research Appropriately, Ethically, Fairly, and Sensitively
Additional research on conducting research is required; power differentials have long-lasting effects not only on the current crisis but also on how research is viewed and accepted in the future. Inclusivity (or lack thereof) in research studies of all groups should be further investigated. For example, given the high maternal and fetal mortality and morbidity related to EVD, women should be given equal opportunities to participate in trials so that drugs and vaccines are well-understood to inform future use for this group.
Using a clinical trial of an Ebola vaccine candidate in Sierra Leone during the EVD epidemic, the importance of integrating social scientists and community liaison officers into the study was seen as being necessary in order to take into consideration notions of power, fairness, and trust in scientific studies 90 (#534, Level IV). Community engagement and appropriate communication are therefore key to build trust and to confront political and social dynamics in times of crisis. There is a need to place these formally within the global response mechanism in order to effectively support all public health emergencies, which are a health and a social phenomenon 91 (#549, Level IV). Health systems research in particular should not be siloed and should not have a clear line of separation between researchers and participants. Using efforts such as the Participatory Action Research Framework, which acknowledges that power is directly related to knowledge, to conduct research among vulnerable populations should be further explored. One such study investigating the experiences of Australian indigenous communities during the H1N1 pandemic demonstrated its ability to incorporate the social aspect of communities in the research process 92 (#547, Level IV). Always, but particularly during crises, biomedical research should be carefully carried out, in consultation with communities and civil society to prevent mistrust and misunderstanding.
Preparedness
A sustained level of preparation is needed for countries to be prepared for emerging pandemics. This includes mechanisms in place for drugs and equipment scarcity, management of patient flow, infrastructural preparedness within health facilities, training of HCWs, a strong surveillance system, and community engagement at the root of it. A qualitative study of 7 countries within the WHO European Region indicated that preparedness was indeed helpful in responding to the H1N1 pandemic, with communication, coordination, capacity building, flexibility, leadership, and mutual support cited as essential elements for preparedness 93 (#1052, Level IV). The cost-effectiveness and sustainability of preparedness initiatives need to be further investigated, particularly in LMICs and resource-limited settings. Implementation of incident management systems considered key in the management of public health emergencies, needs to be further researched.
Preparing for the Next Outbreak: Preparedness Plans for Disaster Risk Management
Preparedness plans are the first step in preparing for outbreaks, but even in the United States, this seems to be lacking. A survey sent out to hospitals in the United States indicated that prior to the H1N1 pandemic, more than 40% of hospitals were lacking preexisting pandemic plans 94 (#169, Type V). However, having a plan alone is not sufficient. A study evaluating the quality of existing preparedness plans in the WHO African region prior to the 2009 H1N1 pandemic showed a large heterogeneity and inadequateness in the preparedness of member states 95 (#837, Level IV). Similarly, an evaluation of countries’ preparedness using indicators from the International Health Regulations State Party Annual Reporting tool indicated that there was wide heterogeneity in the ability of countries to prevent, detect, and respond to outbreaks 96 (#1059, Level IV).
Preparedness plans also need to take into account access to health facilities: Geospatial analysis can be used to identify facilities near vulnerable and at risk populations in order to prioritize detection, treatment, and response 97 (#291, Level IV). Data-driven approaches for resource allocation should be used by countries when developing plans. In addition, preparedness plans need to take into account the vulnerabilities of the populations they serve and adapt accordingly. For example, differences in susceptibility in rural counties of the United States as compared to metropolitan areas would require 2 different approaches to limit transmission 98 (#1643, Level IV).
An evaluation of the implementation of preparedness plans and response seems to be an existing research gap.
Preparing for the Next Outbreak: Strong Surveillance and Response Systems
A strong surveillance system is the best first line of defense, with an ability to detect outbreaks and determine their significance; to quickly describe the epidemiological characteristics of the pathogen, ideally in real-time; and to track the spread and measure the success of control strategies. Delays in detection and reporting can significantly threaten the success of pandemic responses, limiting the ability of health facilities to respond rapidly in an effort to prevent widespread transmission. The availability of a surveillance system is therefore not enough. Due to lack of infrastructure and resources, LMICs, in particular, are often left out of this effective surveillance equation. For example, post-EVD, Sierra Leone implemented the Integrated Disease Surveillance and Response, 99 but evidence regarding its operations and effectiveness are unknown. An assessment of EVD surveillance in northern Ghana pointed to gaps in delayed reporting, low-quality protective equipment, inadequate staffing, and laboratory capacity 100 (#773, Level IV). A discrepancy was observed between the completeness of Malawi's Integrated Disease Surveillance and Response strategy and its actual implementation 101 (#760, Type III). Another study indicated that while the integrated disease surveillance and response strategy was implemented in Ghana years ago, major challenges remain such as ill-equipped laboratories, absent supervision, unreliable diagnostic testing, and surveillance being low-priority among HCWs 102 (#187, Level IV).
There is a need to investigate how LMICs can be better prepared for pandemics with surveillance systems, perhaps with novel and digital approaches to screening and reporting. For example, a clinical decision support tool that can be widely and rapidly adopted was implemented by the military health system for EVD in primary care settings 103 (#901, Level IV). A pilot assessment of feasibility in using short message services to report daily Ebola cases in Guinea-Bissau shows high compliance and may be a rapid way to strengthen surveillance in a low-resource setting 104 (#334, Level II). Recruitment and training of staff, as well as the establishment of an EVD laboratory facility in Liberia with specific diagnostic platforms, led to an 8-fold increase in testing and a quick turnaround time, resulting in a better surveillance system that detected and contained two EVD clusters subsequently 105 (#1072, Level IV). Enhanced surveillance and better laboratory capacity are integral in pandemic response to ensure fast, affordable sequencing of pathogen genomes and rapid diagnostic capacity, for example. Research on emergency laboratory medicine to strengthen surveillance and health systems seems to be largely missing.
Finally, if the surveillance system is not linked to public health decision-making, the value-added may be negligible. An assessment of Ontario’s public health system pointed to syndromic surveillance having limited impact on decisions regarding the operation of immunization clinics and school closures during the H1N1 pandemic and was not used for its main purpose—that is, early outbreak detection 106 (#389, Level IV).
Preparing for the Next Outbreak: Community Engagement and Community Health Workers
In the absence of a strong infrastructural or resilient health system, communities and local people are crucial in a pandemic response, as evident with EVD 107 (#997, Type IV). Community-based management needs to be appropriately incorporated for effective responses.
Evidence on CHWs has mostly been in the context of EVD and LMICs, where their roles cannot be stressed enough: They were crucial in the EVD response, but need additional and sufficient support to be as effective as possible 76 (#1094, Level IV). Focus-group discussions with community members in Sierra Leone described their heavy role in fighting the EVD outbreak that included both labor-intensive tasks (eg, building isolation units) and community outreach (eg, contact tracing and building trust within the community). Yet, respondents indicated a lack of support and feeling neglected by the health system 108 (#669, Level IV).
CHW programs initiated during pandemics in both LMICS and HICs need to be further investigated to better understand their role and potential in pandemic preparedness and response. CHWs have also often been stigmatized and unsupported during pandemics despite working at the frontline and managing the shift in service delivery. This dynamic was not captured with this search strategy and is a clear gap.
Preparing for the Next Outbreak: Infection, Prevention, and Control
Hospital-acquired infections are often the result of poor strategies and human resources. 109 IPC measures are key to protect patients, visitors, and health care personnel alike and, if lacking, may contribute to propagating spread. For example, health care-associated transmission of MERS outbreaks has been linked to health facilities due to overcrowding, delayed diagnosis, and an ineffective IPC system. 110 Similarly, during the Ebola outbreak, patients with EVD who sought care in general health facilities without proper IPC (gaps in screening, isolation, and notification) may have initiated chains of transmission unknowingly, increasing the size of the epidemic. 111 This was likely due to weaknesses in the health system. An assessment of IPC resources in Sierra Leone pointed to severe gaps, with shortages in training, personal protective equipment (PPE), and IPC protocols 112 (#943, Level V). The availability of IPC protocols alone is also not sufficient; compliance by HCWs is necessary for IPC success and is often inadequate in outpatient clinical settings, as demonstrated in a study of 220 facilities in Tanzania 113 (#411, Level V).
Containment of EVD transmission (and other outbreaks) requires that a large investment in health systems strengthening be made, which includes not only availability of supplies but also integrating training to improve HCW behavior. For example, a self-reported survey of HCWs in China on IPC compliance pointed to better compliance in high-risk areas, but that compliance may not be as high when HCWs interact with suspected patients 114 (#340, Level V).
Facility readiness between rural and metropolitan health systems also needs to be investigated further. In the United States, 69% of hospitals indicated having airborne infectious isolation rooms in a 2006 survey; these tended to be large hospitals in metropolitan areas 115 (#377, Level V). Vulnerabilities of health systems to COVID-19 in rural areas are observed globally, including in Vietnam 116 (#985, Level V). Finally, facility readiness in IPC will require advanced planning and new protocols: For example, the implementation of rapid screening processes in health facilities, as was demonstrated during the H1N1 pandemic in the United States, helped improve patient flow considerably 117 (#1284, Level V).
Preparing for the Next Outbreak: Training and Protection of Health Care Workers
Clear guidance, training for changed roles in a pandemic, clear definitions of what constitutes essential activities, community engagement and awareness protocols, and contact tracing mechanisms need to be part of the larger training of HCWs. Protecting HCWs includes ensuring sufficient PPE and protection and support for their mental and physical health. Using frameworks such as the anticipate, plan, and deter responder risk and resilience model can be an approach to better understand and manage psychological impacts of crises among HCWs. 118 Evidence regarding the effectiveness of such frameworks is lacking.
During and in the aftermath of the Ebola outbreak, HCW training programs were observed to be successful in improving knowledge of EVD and reducing HCW infection 119 (#1075, Type IV). Similarly, IPC training of frontline HCWs during EVD in Guinea led to clear increases in knowledge 120 (#834, Level V). Implementation of an mHealth tablet application has led to some improved knowledge and attitude among HCWs in Ondo State, Nigeria 121 (#652, Level V). The perception among HCWs, however, was that the majority had not received training on EVD prevention and management. Fear and unwillingness to work in Ebola treatment centers were also expressed 122 (#708, Level IV). The lack of training among HCWs was further confirmed in Nigeria 123 (#799, Level IV).
Protection of HCWs also requires the availability of PPE, in short supply during pandemics, as demonstrated by a study among Jordanian frontline doctors during COVID-19 124 (#804, Level IV). This search strategy did not capture any evidence with regard to methods or approaches to ensure that PPE would be made available. A study conducted among nurses during the second wave of the H1N1 pandemic indicated that the majority would be willing to work during such crises, but that their willingness would decline if PPE availability declined and antiviral medication and vaccines were not offered 125 (#1276, Level V).
Evidence of the psychological impact of COVID-19 on HCWs is starting to emerge and requires some thinking as to how to best protect HCWs during and in the recovery phase of a crisis. A survey among HCWs in France pointed to 90% of young HCWs in training being more stressed due to the pandemic, affecting their work and training quality 126 (#250, Level V). Similarly, a cross-sectional assessment on the mental health status of health care providers in Wuhan right after the wake of the pandemic indicated that 37% had subthreshold mental health disturbances, 34% had mild disturbances, 22% had moderate disturbances, and 6% had severe disturbances. Young women were disproportionately affected 127 (#1671, Level V). Self-reported reasons for absenteeism of community nurses during the H1N1 pandemic were stress and fear of being infected 128 (#986, Type V).
Finally, recruitment and retention of HCWs are key, particularly in LMICs. Limited recruitment post-EVD in Guinea has led to unemployment of available HCWs. Decentralization of the health system was suggested by respondents as a potential avenue for better health workforce management and financing 129 (#1093, Level IV).
Preparing for the Next Outbreak: Vaccines
The full benefits of vaccines are widely known: They directly prevent diseases, limit community transmission, stabilize health systems, and promote better health outcomes overall. The availability and uptake of vaccines are key to contain a pandemic. Yet, mistrust in vaccines is rampant globally. Part of health systems strengthening today and in preparation for future outbreaks must incorporate better context-specific, culturally-sensitive vaccine messaging and vaccine delivery. Additional research on how this can best be done needs to be prioritized, not just for eventual COVID-19 vaccines, but also to ensure that the uptake of the many existing and life-saving vaccines remains high and accessible. This requires a combination of global consensus, political will and policies, in-country infrastructure, and financing mechanisms in place. Decisions by individuals to get vaccinated in the midst of an infectious outbreak require trust in the government, in the medical organizations, and ultimately, in science.
Vaccine Hesitancy and Trust
Focus-group discussions in the Netherlands, Poland, and Sweden pointed to important differences in responses regarding vaccine uptake: Low perception of risk, concerns about vaccine safety, and mistrust in health authorities were all cited as factors in reduced uptake. Importantly, negative experiences during the 2009 H1N1 pandemic have made (Swedish) respondents less likely to accept future vaccines 130 (#131, Level IV). Similarly, a survey among Americans during the H1N1 pandemic pointed to a low willingness to be vaccinated 131 (#30, Level V).
An investigation of low uptake of the H1N1 pandemic vaccine in the United Kingdom through 5 national telephone surveys reported low perceived risk of H1N1 influenza and doubt about the vaccine safety among participants. Being healthy, female, not being vaccinated with the seasonal flu vaccine previously, and belief in vaccine ineffectiveness were also cited as risk factors for not being vaccinated 132 (#1316, Level V). Similar results (low perceived risk and severity, fear about a new vaccine, and costs) were found in a qualitative study among low-income women in Georgia, Atlanta. Ineffective dissemination of information from HCWs was also an emerging theme 133 (#52, Level IV).
Focus groups investigating attitudes regarding vaccine among aboriginal communities in Manitoba, Canada, following the 2009 H1N1 influenza pandemic pointed to considerable variation. Knowledge deficits about the vaccine and the pandemic, concerns about vaccine safety, targeted testing within their communities (rooted by colonial legacy), and feeling like a “guinea pig” were stated as having deterred vaccine uptake. Doctor–patient trust, usually helpful to promote uptake, is more complicated in this setting with historically negative interactions 134 (#23, Level IV). This likely translates to LMIC contexts given the colonial nature of some global health vaccination campaigns in the past.
Improving Access to Vaccines
Rates of vaccination in Montreal for the H1N1 pandemic were shown to be lower in vaccination clinics located in areas with high violent crime rates and high residential density. Vulnerable communities with new immigrants and families living under the poverty line were observed to have lower vaccine uptake. Clinic location setup during mass vaccination campaigns as well as community messaging should therefore be further investigated to improve vaccine uptake 135 (#1547, Level VII). Barriers to access** of vaccination due to unavailability or to logistical or financial barriers in LMICs were not captured by this scoping review.
Vaccination Among Health Care Workers
HCWs are at higher risk of contracting and transmitting the virus, yet uptake and acceptance of a new vaccine seem to be alarmingly low. The main concerns among HCWs referred to concerns regarding a vaccine’s effectiveness and safety profile. For example, 6 months into the H1N1 vaccination campaign, a survey of hospital workers in Lyon, France, reported that their vaccination rate of the H1N1 influenza vaccine was only 54%. Reported arguments against the vaccine included a lack of studies on the vaccine and a low-risk perception of H1N1 136 (#177, Type V). Similar results were found among HCWs in London, United Kingdom 137 (#1319, Type V), as well as Saudi Arabia 138 (#1326, Type V), Singapore 139 (#1329, Stage V), and Australia 140 (#97, Type V). This trend translates into the flu as well: In a study of 742 interviews conducted with essential community workers in Alicante, Spain, only 21.5% reported having been vaccinated with the seasonal flu vaccine. Only 15.4% reported having been vaccinated with the H1N1 vaccine. Individuals who did not doubt vaccine safety considered pandemic influenza to be more severe than the seasonal flu, and those regularly vaccinated against the seasonal flu were more accepting of the H1N1 vaccine 141 (#1313, Type IV).
Additional Gaps and Recommendations for Research
Beyond the emerging themes and evidence presented, this rapid scoping review has illustrated additional gaps. Assessments for longer-term impacts on health outcomes and health systems using telemedicine are needed. While several models are proposed (eg, telerehabilitation, 142 telemedicine for vision centers, 143 telemedicine for pediatric weight management clinic, 144 and telepsychiatry 145 ) additional research is needed to understand best practices for patients and clinicians alike. Rapid diagnostic testing combined with a digital approach for contact notification is necessary for efficient contact tracing globally, which can allow for safer continuity of care of essential services. Evidence regarding the efficiency of such digital approaches was not discovered within the search criteria of this review. Evidence on contact tracing methodologies in LMICs was also missing.
It is also important to acknowledge that a shift in digital health may also introduce unintended consequences in health equity due to lack of availability and access to digital health, low digital health literacy, and poverty. Additional research in the equity and digital health is necessary before committing to shifts in service delivery. More broadly, data over a longer time period will need to be investigated. Most of the studies presented are of a cross-sectional nature, introducing major limitations in our ability to understand the full effects of crises on health systems over time and their ability to provide essential services. For example, longitudinal assessments of psychological needs are lacking.
In addition, the effects of crises on essential drug shortages and stockpiles in pharmacies—and their indirect effects on health systems and their people—do not seem to have been investigated globally. Supply chain logistics (eg, PPE and medications) are also key to ensure continuity of essential services, but evidence covering these topics was not captured.
Finally, understanding acceptance and hesitancy of vaccines among health care practitioners and the general population is key. Investigating how social and institutional trust can be built within communities is central to better prepare for future outbreaks and future vaccination campaigns, including that of COVID-19.
Limitations
This scoping review was of a rapid nature and thus has several limitations. The author was involved in the entirety of the research process, potentially introducing human error or biases. Inclusion and exclusion criteria had to be tightened for feasibility purposes; it is therefore possible that important and relevant articles may have been missed. A redundancy and overrepresentation of some themes were evident in the review; findings and gaps may therefore not include considerations of themes not captured by the search strategy. Given the non-systematic nature of the review, areas of evidence may also have been inadvertently bypassed. Nevertheless, findings and gaps identified provide an initial step in understanding health systems in times of crisis. There remains a need to further define the large scope that is health systems to investigate specific issues (either not covered at all or covered superficially in this review) in more detail.
Conclusion
Conversations with experts led to three main lessons from crises: (1) the necessity of community buy-in with regard to responses, (2) a centralized strategy for risk communication, and (3) distributive leadership and accountability. Many of the studies included in this scoping review further confirmed these as necessary for pandemic preparation and response, although evidence on leadership and accountability was not captured by this search strategy. One of the biggest pitfalls brought to light by this scoping review is the clear separation between research and practice. Previous crises had foreshadowed the inability of health systems to handle pandemics, but the reactive nature of existing policies and practices, as well as social and power dynamics, prevented the implementation of preventative measures in preparation for COVID-19. In LMICs in particular, this was compounded by the lack of resources, infrastructure, and political will. There is therefore an urgent need for implementation science and for strategies to bridge this research–practice gap.
As the SDG Report 2020 states, the need for greater public health preparedness has never been clearer. While reaching the SDGs by 2030 seems difficult given COVID-19’s disruption, we hope that more strategic investments in research will allow nations, particularly LMICs, to at least progress toward them.
Supplemental Material
sj-docx-1-joh-10.1177_0020731421997088 - Supplemental material for Health Systems and Services During COVID-19: Lessons and Evidence From Previous Crises: A Rapid Scoping Review to Inform the United Nations Research Roadmap for the COVID-19 Recovery
Supplemental material, sj-docx-1-joh-10.1177_0020731421997088 for Health Systems and Services During COVID-19: Lessons and Evidence From Previous Crises: A Rapid Scoping Review to Inform the United Nations Research Roadmap for the COVID-19 Recovery by Prativa Baral in International Journal of Health Services
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Institute of Population and Public Health (grant number DFD-170768).
Supplemental material
Supplemental material for this article is available online.
Author Biography
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
