key: cord-0876685-b6f9wtw1
authors: Ogunbameru, A.; Barrett, K.; Joda, A.; Khan, Y. A.; Pechlivanoglou, P.; Mac, S.; Naimark, D.; Ximenes, R.; Sander, B.
title: Estimating healthcare resource needs for COVID-19 patients in Nigeria
date: 2020-08-22
journal: nan
DOI: 10.1101/2020.08.19.20178434
sha: 26c0779e8a863c91c1d7131501be5c40c638dc7a
doc_id: 876685
cord_uid: b6f9wtw1

Background: Predicting potential healthcare resource use under different scenarios will help to prepare the healthcare system for a surge in COVID-19 patients. In this study, we aim to predict the effect of COVID-19 on hospital resources in Nigeria. Method: We adopted a previously published discrete-time, individual-level, health-state transition model of symptomatic COVID-19 patients to the Nigerian healthcare system and COVID-19 epidemiology. We simulated different combined scenarios of epidemic trajectories and acute care capacity. Primary outcomes included expected cumulative number of cases, days until depletion resources, and the number of deaths associated with resource constraints. Outcomes were predicted over a 60-day time horizon. Results: In our best-case epidemic trajectory, which implies successful implementation of public health measures to control COVID-19 spread, the current number of ventilator resources in Nigeria (conservative resources scenario), were expended within five days, and 901 patients may die while waiting for hospital resources in conservative resource scenario. In our expanded resource scenarios, ventilated ICU beds were depleted in all three epidemic trajectories within 60 days. Acute care resources were only sufficient in the best-case and intermediate epidemic scenarios, combined with a substantial increase in healthcare resources. Conclusion: Current hospital resources are inadequate to manage the COVID-19 pandemic in Nigeria. Given Nigeria's limited resources, it is imperative to increase healthcare resources and maintain aggressive public health measures to reduce COVID-19 transmission.

References: 26 (max: 40) . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 22, 2020. . https://doi.org/10.1101/2020.08. 19 .20178434 doi: medRxiv preprint

While western countries seem to be recovering from the COVID-19 pandemic, there is an increasing community spread of the virus in many African countries.

The limited healthcare resources available in the region may not be sufficient to cope with increasing numbers of COVID-19 cases.

Using the COVID- 19 Resource Estimator (CORE) model, we demonstrate that implementing and maintaining aggressive public health measures to keep the epidemic growth at a low rate, while simultaneously substantially increasing healthcare resources is critical to minimize the impact of COVID-19 on morbidity and mortality.

The impact of COVID-19 in low resource settings will likely overwhelm health system capacity if aggressive public health measures are not implemented. To mitigate the impact of in these settings, it is essential to develop strategies to substantially increase health system capacities, including hospital resources, personal protective equipment and trained healthcare personnel and to implement and maintain aggressive public health measures.

. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 22, 2020. . https://doi.org/10.1101/2020.08. 19.20178434 doi: medRxiv preprint

The World Health Organization declared the spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) a Public Health Emergency of International concern on January 30, 2020, and later a pandemic on March 11, 2020. 1 Africa recorded the first case of Coronavirus Disease 2019 (COVID- 19) on February 14, 2020, in Egypt 2 , and since then has reported over 945,165 cases and 18,476deaths as of August 17, 2020. 3 Nigeria reported the first case of COVID-19 on February 29, 2020. 4 By August 12, 47,743confirmed cumulative cases were reported. 4 The average daily growth rate of cumulative cases in Nigeria from when 100 cases were reported on March 29, 2020 (which correspond to the date lockdown was initiated in the three most affected states) to August 12, 2020 is 5%. 4 The lockdown ban was lifted on May 4, 2020. The mean growth rate from May 5, 2020 to August 16,2020 is 3%. 4 From June 1, 2020, to August 12, 2020, the mean growth rate furthered decrease to 2%. 4 The growth rate of COVID-19 in Nigeria is low compared to what was observed from March to May in Europe and North America. 3 Overall, the pandemic trajectory in Africa has been different compared to China, Europe and North America, with lower growth rates and lower case fatality rates. 5 However, the spread of COVID-19 in the region is rapidly increasing, with a shift from 9 to 42 countries affected from March to May 2020. 5 Factors that have been associated with the rapid spread of COVID-19 virus in the region include poor public health infrastructure, underreporting, and limited testing and contract tracing resources. 6 To mitigate the impact of COVID-19, aggressive public health measures must be pursued to prevent overwhelming the healthcare system which has limited resources. 7 To reduce the spread of COVID-19 in Nigeria, the Federal government initiated a Presidential Task Force to provide a high-level strategic national response to the disease; launched a campaign . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 22, 2020. . https://doi.org/10.1101/2020.08.19.20178434 doi: medRxiv preprint themed " Take Responsibility" -a call to encourage residents to take individual and collective   responsibility of implementing public health measures; quarantined all COVID-19 confirmed cases, test and track confirmed cases contacts; banned non-essential international and inter-state travel; initiated curfews to limit social interactions and restricted gathering to 20 people per workplace to encourage physical distancing. 4 Predicting COVID-19 population spread and healthcare resource needs for symptomatic patients using models is essential to prepare the health system and allow it to continue running efficiently during the pandemic, minimize morbidity, and reduce society disruption. 8 Simulation models can help to inform healthcare resource needs under different scenarios to improve planning and support procurement strategies, especially in resource-limited settings. 9 Nigeria, like many African countries, has limited healthcare resources and infrastructure. Before the start of the pandemic, Nigeria had an estimated 0.2 hospital beds per 1,000 population, 350 intensive care-unit (ICU) beds without ventilators (equivalent to 0.07 ICU beds per 100,000 population), and 450 ventilated beds. 10, 11, 12 With the country's cumulative COVID-19 cases exceeding 47,000 on August 11 2020, 4 and case numbers still increasing, forecasting how COVID-19 will affect hospital resources under different scenarios is critical for Nigeria's COVID-19 response.

Our objective is to predict the short-term effect of COVID-19 on hospital resources in Nigeria for a range of COVID-19 epidemic and hospital capacity scenarios using a health system simulation model.

. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 22, 2020. . https://doi.org/10.1101/2020.08.19.20178434 doi: medRxiv preprint We adopted the previously published COVID-19 Resource Estimator (CORE) model from Ontario, Canada, 13 to fit the healthcare system and COVID-19 epidemiology in Nigeria. Our primary outcomes included projected cumulative number of COVID-19 cases, number of days until depletion of ward bed and ventilated ICU bed resources and number of avoidable deaths assuming no resource constraint. Outcomes were predicted over a 60-day time horizon in daily time steps.

We used the CORE model that was available as an interactive software application online. 11 A detailed description of the CORE model is provided elsewhere. 9 Briefly, CORE is a discrete-time, individual-level, health-state transition model of symptomatic COVID-19 patients. CORE simulates a dynamic population of symptomatic COVID-19 adult patients (18 years and above) who present to the emergency department (ED), where they are either sent home to self-isolate or admitted to the hospital. If admitted, COVID-19 patients are assigned to either a general ward or ICU, depending on disease severity. Seventy-eight percent of ICU patients were assumed to require invasive mechanical ventilation. 9,10

Several assumptions were made in the CORE model. For simulated patients in the model, resource needs were based on the patient's health state (i.e. in a general ward or ICU). If there was a lack of ward or ICU beds, patients remained with their current available resources until the needed resource became available. Patients requiring mechanical ventilation but did not have access were assumed dead within 24 hours. Mortality risk was attributed to only ICU patients with or without . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 22, 2020. . https://doi.org/10.1101/2020.08.19.20178434 doi: medRxiv preprint a need for a ventilator. Patients awaiting an ICU bed (but not mechanical ventilation) were assumed to have the same risk of death as patients in ICU. Ward beds were assumed available upon patients' recovery. ICU beds with/without ventilators were assumed to be freed up upon recovery or death of patients. Available resources were prioritized based on the patient's location (e.g. ward beds will be prioritized to ICU patients before new admissions) and the length of waiting time of a patient since admission. The CORE model assumed that other essential resources, including personal protective equipment and essential medications, would be available in sufficient quantities. It also assumed all available beds are adequately staffed, and no staffing shortage would emerge with vast-increase in hospital resources.

We assumed that COVID-19 patients would only be treated in hospitals licensed by the Nigeria Centre for Disease Control. 14 Patients were assumed to be admitted to ICU only when in need of mechanical ventilation and that all ICU beds are vented.

All model parameters are listed in Table 1 .

probabilities of hospital and ICU admission obtained from reported Canadian COVID-19 data, reflecting clinical need without resource constraints. 9, 15 Mean length of hospitalization, and the probability of death for COVID-19 patients in the ward and ICU were based on reported data for patients admitted for moderate acute respiratory distress syndrome (ARDS), based on a similar in clinical manifestation. 9, 16, 17, 18 The proportion of ICU patients requiring mechanical ventilation was obtained from reported Canadian data. 9, 16 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 22, 2020. . https://doi.org/10.1101/2020.08.19.20178434 doi: medRxiv preprint Epidemic trajectories. We used reported data on daily and cumulative COVID-19 cases up to August 12, 2020 , when the number of cumulative cases exceeded 49,000 as published by the Nigeria Centre for Disease Control (NCDC). 4 We forecasted three scenariosbest case, intermediate case, and worst case -to predict the trajectory of COVID-19 cases in Nigeria for 60 days starting with data from August 12, 2020. In the three scenarios, we started our prediction with 47,743 confirmed COVID-19 cumulative cases and 453 new cases within 24 hours. For our bestcase scenario, we assumed an infection growth rate of 1%, to depict the successful implementation of aggressive public health measures (i.e., social distancing, school closure and travel restrictions).

To our knowledge, no African country has an average infection growth rate of 1%, from the day when 100 cases were reported. In our intermediate scenario analysis, we used a 2% infection growth rate, which is the average growth rate from June 1, 2020 to August 12, 2020 in Nigeria. 4 Similarly, Ghana recorded an average growth of 2% from June 1, 2020 to August 12, 2020. 19 For our worst-case scenario, we assumed an infection growth rate of 3%, which suggests a scenario where no further public health measures are introduced, economy re-opening is fast-tracked and greater community transmission is ongoing. Infection growth rates across countries were calculated using daily reported COVID-19 cases from the John Hopkins University repository. 19 We did not assume an epidemic peak in our study since our time horizon is short (60 days), and the daily number of new cases still fluctuates between low and high numbers. Case predictions for the three scenarios are shown in Figure 2 (daily number of cases are provided in Appendix 1).

We obtained data on the total number of ward beds, intensive care unit beds and ventilators before the pandemic from the published literature and press reports. 10, 11, 12 Our data estimate for total ICU beds and ventilators were obtained from a current press report by Nigeria health officials assessing the available health resources in Nigeria. We modelled three . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 22, 2020. . https://doi.org/10.1101/2020.08.19.20178434 doi: medRxiv preprint resource scenarios: conservative, expanded and surplus. For the conservative scenario, we assumed that 50% of ventilated ICU beds (175/350), and 25% of hospital beds (10,000/40,000) are available to treat symptomatic patients based on recent press reports and a national survey conducted by a pharmaceutical company. 10, 11, 12 In the expanded resource scenario, we added 200 ventilated ICU beds to the proportion available for treating COVID 19 ill-patients in conservative resource scenario and an additional 10,000 ward beds to the proportion of ward beds available in conservative resource scenario. The assumption was based on reduced clinical activity enabling re-allocation of privately-owned hospital resources, the recent increase in health resource supplies from donors, and the increase in government investment in healthcare services and resources since the start of the pandemic. 20 For the surplus scenario, we assumed an additional 1,000 ventilated ICU beds to the proportion available for treating COVID 19 ill-patients in expanded resource scenario, and an extra 15,000 ward beds to the proportion of ward beds in expanded scenario resources based on potential funding from foreign donors, high-income countries and international health organizations. 21 In the three hospital resource scenarios, we assumed adequate hospital staffing based on recent recall of retired health personnel to service by the Nigeria Federal government and the ongoing recruitment and training of volunteers across the country to boost staffing capacity. 22, 23 

For the best-case epidemic trajectory (1% growth rate), our study predicted 86,735 COVID-19 cases over 60 days (Figure 2) , of whom 4,782 patients would require hospital admission to ward and ventilated ICU bed. In all resource scenarios, ward beds would remain available; ventilated ICU bed would be depleted after 5 days and 27 days in both conservative and expanded resource scenarios ( Table 2) .

. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 22, 2020. Assuming the conservative resource scenario, the best-case epidemic trajectory projects 901 preventable deaths over 60 days. In the best case and intermediate epidemic trajectories, assuming the surplus scenario, no death due to the unavailability of resources was predicted to occur over 60 days.

We demonstrated that a substantial increase in healthcare resources is critical for Nigeria's health system to care for COVID-19 patients and improve health outcomes, even at low infection growth rates. Maintaining aggressive public health measures are therefore needed to effectively reduce transmission, reduce the number of new cases and deaths. As of August 12, the public health . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 22, 2020. . https://doi.org/10.1101/2020.08.19.20178434 doi: medRxiv preprint measures implemented in Nigeria are moderate, and the average infection growth rate is approximately 5% (averaging from March 29 when 100 cases were reported to August 12, 2020).

Since the implementation of public health measures, the average growth rate has further declined to 1-2 % within the last three months. While the Federal government acted promptly to initiate a lockdown in two of Nigeria's most affected states on March 31, 2020, this decision was reversed in May due to economic pressures, despite the fluctuating low to the high daily number of cases. 24 Our analysis shows that maintaining strong public health measures with vast-increase in resources is crucial to prevent hospital resource constraints and health system collapse in Nigeria.

Our study has limitations. Our estimated number of ward bed was based on a national survey by a pharmaceutical company published in 2007 and might not represent the country's current capacity.

While keeping with the current literature, we assumed that death would only occur in critically ill patients, which may underestimate mortality. Since the start of the pandemic, recruitment and training of new hospital staff have been ongoing, but actual data on the number of recruited staff was not available at the time of modelling, limiting our knowledge on the effect of vast expansion of hospital resources on staff capacity. Our study relies on reported cases to forecast future epidemic trajectories. The model does not account for underreporting of daily cases and long-time lag in case reports, which are common problems that occur in low resource settings due to limited testing capacity. 6 Due to the unavailability of detailed COVID-19 data, some parameters included in the model were estimates obtained from a study on ARDS, a disease with similar clinical manifestations with COVID-19. Our healthcare resource utilization probabilities parameters (i.e. length of hospital stay and probability for need of ward, ICU and ventilator admissions and probability of death) were estimates from Canadian setting. While these parameter values arguably deviate from observed data in Africa, they are likely estimates for COVID-19 disease severity.

. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 22, 2020. . https://doi.org/10.1101/2020.08. 19.20178434 doi: medRxiv preprint Our study has several strengths. We incorporated observed incidence data from Nigeria and other Africa countries with a range of epidemic trajectories to strengthen the validity of our epidemic predictions. The CORE model considers resource constraints within the health system and estimated deaths due to resource depletion. We stratified mortality between patients who receive adequate care in ICU and those who did not to estimate the number of patients who will likely die from an overwhelmed health system.

The COVID-19 epidemic trajectory is slower in Africa, 6 with many countries within 5-10% average infection growth rates since 100 cases reported. Our epidemic trajectory and resource scenario predictions could apply to other low resource setting countries, especially in the Africa region. Our findings are supported by modelling studies that predict an expanding increase in the number of COVID-19 cases in Africa region 25 and the need for strong government measures, abundancy of medical supplies and good personal protective behavior to mitigate the spread of COVID-19 in the region. 26 Our resource scenarios estimates could also be used as a guide to inform pandemic preparedness planning and policy development.

The epidemic trajectory of COVID-19 appears to have a low growth rate, and the implemented public health measures have helped to reduce community spread in Nigeria; however, the current hospital resources in Nigeria are still inadequate to manage the daily number of COVID-19 critically ill-patients. To mitigate the impact of COVID-19, implementing more aggressive public health measures is vital, and strategies to exponentially increase the number of health resources available in the country need to be put in place to prevent overwhelming the healthcare system.

. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 22, 2020 is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 22, 2020. . https://doi.org/10.1101/2020.08.19.20178434 doi: medRxiv preprint . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 22, 2020. . https://doi.org/10.1101/2020.08.19.20178434 doi: medRxiv preprint X-represents resource constraint for hospital beds; Y-represents resource constraint for ICU beds Z-represents resource constraint for ventilators; ICU-Intensive Care Unit When awaiting resources in the next health state, the patient remains in the previous state until the needed resource is available. A patient awaiting ventilator was assumed to die within the 24 hours. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 22, 2020. . https://doi.org/10.1101/2020.08.19.20178434 doi: medRxiv preprint Worst case -3% increase Intermediate case-2% increase Best case-1% increase . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted August 22, 2020. . https://doi.org/10.1101/2020.08.19.20178434 doi: medRxiv preprint

World Health Organization. WHO Timeline -COVID-19

Preparedness and vulnerability of African countries against importations of COVID-19: a modelling study

World Health Organization. WHO Coronavirus Disease (COVID-19) Dashboard

Nigeria Centre of Disease Control. An update on COVID-19 outbreak in Nigeria

The Coronavirus Disease 2019 strategic response plan in the WHO African Region

World Health Organization

The impact of COVID-19 on HIV financing in Nigeria: a call for proactive measures

Hospital capacity and operations in the coronavirus disease 2019 (COVID-19) pandemic-planning for the nth patient

Mathematical modelling of COVID-19 transmission and mitigation strategies in the population of Ontario, Canada

Estimation of COVID-19-induced depletion of hospital resources in Ontario

Winning in Nigeria: Pharma's next frontier. McKinsey and Company Pharmaceuticals and Products

Nigeria has 350 ICU beds for 200 million people. Premium Times 2020

African nations to get ventilators from Jack Ma Foundation, stress need for WHO help

COVID-19: Predicting Healthcare Resource Needs in Ontario. Developed by COVID-19 ModCollab

The Punch Newspaper. FG, private hospitals disagree over COVID-19 treatment

Public Health Agency of Canada. Coronavirus disease 2019 (COVID-19) daily epidemiology updated

Critical Care Services Ontario COVID-19 updates

LUNG SAFE Investigators; ESICM Trials Group. Epidemiology, patterns of care, and mortality for patients with acute respiratory distress syndrome in intensive care units in 50 countries

Characteristics of and important lessons from the coronavirus disease 2019 (COVID-19) outbreak in China: summary of a report of 72 314 cases from the Chinese Center for Disease Control and Prevention

COVID-19 Data Repository by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University-CSSEGIS and Data/COVID-19

COVID recovery offers opportunities to invest in Nigeria across the healthcare value chain

World Health Organization. COVID 19 Solidarity Response Fund. Help fight COVID 19

World Health Organization. As Nigeria battles COVID-19, volunteers remain critical to the outbreak response. Updated

COVID-19: Tomori blames FG, citizens for widespread community transmission. Punch Newspaper

COVID-19 pandemic in the African continent: Forecasts of cumulative cases, new infections, and mortality

Prediction of the COVID-19 spread in African countries and implications for prevention and control: A case study in South Africa