Health care workers hospitalized for COVID-19 in Liberia: who were they, and what were their outcomes?

Introduction

An umbrella review of 13 systematic reviews from 1,230 studies including 1,040,336 COVID-19 confirmed cases revealed a global infection rate of 14.5% among health care workers, and an overall case fatality of nine deaths per 1,000 infections.¹ These rates could have been accentuated by health-related challenges among one-third of the health care workers such as poor sleep quality, mental health issues, work-related stress and skin injuries associated with use of personal protective equipment.² There was also limited access to COVID-19 vaccines.³

In the African region, the World Health Organisation (WHO) estimated that over 10,000 health care workers were infected with COVID-19. Liberia reported a total of 8,090 confirmed COVID-19 cases and 404 healthcare worker infections to WHO, ranking it among the top three with the highest country-specific health care worker infection rates in the region.⁴

Prior to the COVID-19 pandemic, the 2014-2016 Ebola outbreak in Liberia took a devastating toll on health care workers with 372 infections and 184 Ebola-related deaths which reduced the health workforce by 8.1%.⁵ The country has one of the lowest core health care worker density ratios in the world with 12.8 core health care workers per 10,000 population (WHO minimum = 44.5 per 10,000 habitants). For example, in the USA, this figure stands at 82 per 10,000 (WHO - health-workforce). Thus, preventing health care worker deaths is a top priority for sustaining the Liberian health system and achieving Universal Health Coverage.⁶

The dire health care worker situation in Liberia underpins the importance of health care worker safety and the need to effectively monitor health care worker infections using standardized data surveillance and clinical management tools. During outbreaks, such data collection tools are often not available, and countries may be “caught unprepared”. Thus, there is the need to establish timely standardized and regular surveillance, case management and reporting systems. Data validation during outbreaks is also a challenge as the health system gets rapidly overwhelmed and busy health care workers end up often unable to devote time for data validation. The situation was no different in Liberia which, in 2020, introduced a paper-based data collection form for health care workers hospitalized with COVID-19. To date, there has been no formal audit of the completeness and usefulness of this data in informing outbreak management.

An operational research priority in Liberia is to better understand the health care worker cadres that were affected by COVID-19, their clinical course, and outcomes. The existing data collection forms provide an opportunity to analyze such information using an operational research lens. Such information could contribute towards sustaining a ‘fit-for-purpose’ health workforce and improving the monitoring system in preparation for future outbreaks.⁷ A PUBMED search revealed no studies from Liberia (or neighbouring countries) that assessed the quality of COVID-19 data for health care workers hospitalized to health facilities and their outcomes.

We thus aimed to describe the characteristics of hospitalized health care workers with COVID-19 in Liberia and their outcomes. In all hospitalized health care workers with confirmed COVID-19 in Liberia between March 2020 and May 2023, our specific objectives were to determine a) the COVID-19 hospitalization and case fatality rates by year, b) the type of health care worker cadres affected along with their demographic and clinical characteristics, c) the medical interventions administered and d) the health facility exit outcomes. We also assessed the proportion of missing and unknown data.

Methods

Statistical analysis

Data was analyzed using (Stata software version 16) (College Station, TX, USA: Stata Corp LLC. Free alternatives to Stata are RStudio or JASP). Countrywide geographic distribution of health care workers infected with COVID-19 was presented graphically per county using QGIS software 3.36.2 (Figure 1).

Figure 1. Countrywide geographic distribution of health care workers hospitalized for COVID-19 in Liberia (March 2020-May 2023).

The demographic characteristics, clinical characteristics, vaccination status and hospital exit outcomes were summarized using frequencies and percentages. Hospitalization rates for COVID-19 among health care workers were calculated by standardizing the numbers of health care workers hospitalized by year with the total number of health care workers registered in that year and expressed by 1,000 health care workers.

Results

Discussion

This first country-wide study from Liberia on health care worker hospitalizations for COVID-19 showed that clinical and laboratory personnel were the most affected, the main source of infection was the health facility and only 20% of health care workers were vaccinated. The majority (91%) of all hospitalizations were for mild symptomatic infections and this applied as well to the 35% of hospitalizations in specialized centres. Among those who died with moderate to severe infection, none had benefited from hospitalization in specialized centres.

The study findings are important as they are a clarion call for improving health care worker safety, strengthening infection prevention and control, and offering health care workers priority access to vaccination. In the early part of the pandemic, when vaccines were in short supply and/or equity issues were in play,¹⁰ vaccinating health care workers should have been a top-level priority to sustain a ‘fit for purpose’ health workforce. This strategy is akin to “treating the pilots in a plane flying at 30,000 feet before attending to passengers”. However, Liberia like most other African countries did not have sufficient access to COVID-19 vaccines.¹¹ This study also highlights the importance of early institution of triage and referral systems to ensure that individuals with moderate to critical infections and most in need of hospitalized care, are adequately catered for during an outbreak.

This study has a number of policy and practice implications. First, clinical and laboratory staff constituted 70% of all hospitalizations and their predominant source of infection was the health facility. This highlights that health facilities were ‘hot spots’ for the acquisition and transmission of infection. The way forward is to ensure regular IPC assessments in all health facilities¹² and ensure the availability of on-site ‘safety stocks’ of Personal Protective Equipment (PPE) material as part of the preparedness strategy for future outbreaks. Furthermore, the fact that there were nine health care worker deaths from COVID-19 compounds the already low health care worker density in Liberia and this may have contributed to disruptions in routine health service delivery.

Second, the fact that the majority of hospitalizations (91%), including a considerable proportion in specialized centres, were for mild symptomatic infections is concerning. It implies that COVID-19 treatment sites largely served as ‘isolation facilities’ rather than treatment centres. From a health system perspective, this is disadvantageous for two reasons: a) available beds get unnecessarily filled and those genuinely meriting hospitalized care may risk being turned away, and b) large numbers of individuals with mild infections hospitalized in health facilities increases the risk of disease transmission to both health care workers and the community at large.

Clogging up the health system will also have a crippling effect resulting in those with moderate to severe disease not being promptly identified and referred to higher level health facilities. This seems to have been the case since nine health care workers who died should have been referred to specialized centres but this did not happen. Five individuals identified with complicated sepsis were also not referred to specialized treatment centres. Such shortcomings in the triage and referral system should be corrected for future outbreaks and public health emergencies.

Third, during outbreaks, there is need for data and information in ‘real-time’ to inform and enable decision makers to institute prompt actions that can uphold the capability and accountability of health systems.¹³ A survey conducted by the WHO in 2022 showed that 92% of all countries were still facing COVID-19 related disruptions which affected all health service delivery platforms.¹⁴ In Liberia, as of July 2021, utilization of essential maternal and reproductive health services was lower than pre-pandemic trends. Similarly, there were substantial disruptions in outpatient services (down by 17%), family planning consultations (down by 36%), post-natal care first visit (down by 21%), ante natal care first visit (down by 19%).¹⁵ ‘Real-time intelligence’ would help steer the health system out of such trouble as has been demonstrated in some African countries.^16,17 The use of mobile phone applications (such as EpiCollect5) and cloud-based data storage can provide a useful adjunct to emergency responses and should be explored in Liberia and neighbouring countries.^16,17

The study strengths are that the research subject was a national operational research priority and thus relevant, and all health facilities in the country were included making it likely that the study was representative of the situation on the ground. We also ensured that data were cross-validated with individual patient cards at all study sites, and we adhered to the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines.¹⁸

One of the study limitations is that we had missing and/or unknown variables (one to ten percent) in the patient cards, highlighting the need for enhanced oversight for improving the quality of data by the national and county incident management teams. The fact that these teams have already taken corrective steps with health facilities shows the useful synergy developed between operational research and the strengthening of the current monitoring and evaluation system. Another limitation is that we did not know the exact reasons for the observed decline in numbers of hospitalized cases by year and whether this was related to increasing vaccination coverage over time and/or changes in hospitalization practices or the virulence of the circulating SARS-CoV-2 virus strains.

Conclusions

In conclusion, this country-wide study from Liberia showed that hospitalized health care workers for COVID-19 were predominantly clinical and laboratory personnel who were mostly unvaccinated, and health facilities were hot-spots for contracting infections. The triage and referral system was weak with unnecessary hospitalization of mild infections. This study provides useful insights for outbreak preparedness including priority vaccination and improving health care worker safety in Liberia.