The cost of clinical management of SARS-COV-2 (COVID-19) infection in Ghana: A cost of illness analysis

Background : As the global strategies to fight the SARS-COV-2 infection (COVID-19) evolved, global response strategies impacted the magnitude and distribution of health-related expenditures. Although the economic consequence of the COVID-19 pandemic has been dire, and its true scale yet to be ascertained, one key component of the response is the management of infected persons which its cost has not been adequately examined, especially in the context of African countries. Methodology : To partly fill gaps in context-specific cost of treating COVID-19 patients, we adopted a health system’s perspective and a bottom-up, point of care resource use data collection approach to estimate the cost of clinical management of COVID-19 infection in Ghana. Findings : We found that resource use and average cost of treatment per COVID-19 case varied significantly by level of disease severity and treatment setting. The cost of treating COVID-19 patients in Ghana from the perspective of the health system ranged from US$282 (GH ¢1,629) to US$23,382 (GH ¢135,149), with an average of US$11,925 (GH ¢68,929). The cost of treatment increased by at least 20 folds once a patient moved from home management to the treatment centre. PPEs and Transportation were the main cost drivers for institutionalised care, whilst investigations (COVID-19 testing) and staff time were for home-based care. Conclusion : Cost savings could be made by early detection and effective treatment of COVID-19 cases, preferably at home, before any chance of deterioration to the next worst form of the disease state, thereby freeing up more resources for other aspects of the fight against the pandemic. Policy makers in Ghana should thus make it a top priority to intensify the early detection and case management of COVID-19 infections.


INTRODUCTION
In December 2019, an unknown respiratory illness was reported in the Hubei Province of China, a disease later named COVID-19 caused by Severe Acute Respiratory Syndrome Coronavirus type-2 (SARS-COV-2) (1). The disease subsequently spread globally and was declared a pandemic by the World Health Organization (WHO) in March 2020 (2). As of 17 September 2020, there have been 29,679,284 confirmed cases globally and 936,521 deaths while the WHO African Region has recorded 1,127,164 confirmed cases and 24,294 deaths (3,4). In Ghana, the disease was first reported on 12 th March 2020 and by 17 September 2020 about 45,655 have been reported with a case fatality rate of less than 1% (5).
Given the alarming basic case reproduction rate of the virus globally, the WHO implored countries to "aggressively double down to contain it" (6). Learning from the West Africa Ebola crisis between 2013 and 2016, effective case management, public health containment measures, community empowerment and partnership became a central part of the pandemic response (7). As most were not yet fully prepared, varied risk-based containment strategies were employed (8); initially centred on travel restrictions, evolving into partial and complete lockdowns to drastically limit physical interactions with the intent of curbing the spread of the virus while aggressive treating infected persons (9). These strategies have evolved over time and tend to vary substantially from country to country.
Although the economic consequence of the COVID-19 pandemic had been dire, and its true scale yet to be ascertained, one key component of the response is the management of infected persons which its cost has not been adequately examined. Earlier studies in China estimated the mean cost of treating COVID-19 patient to be about US$6827 which roughly amounted to US$ 0.49 billion for the entire clinical management of COVID-19 in China (10), whilst those of the African Region focussed on generating regional averages of US$3.65 for a mild case and US$15.56 for a hospitalized patient (11) . However, these estimates were premised on effective prevention measures (wide scale and intensive social distancing) which were not met.
There are very limited country-specific analyses that estimate the cost of treating COVID-19 cases in Africa. One Kenyan analysis estimated the cost of treating COVID-19 patients to range from US$282 in home management to US$5,707 in critical care settings (12). In another attempt to estimate the probable cost of the entire response to COVID-19, Rueda et al extrapolated data from costing of other diseases in South Africa, Ethiopia and Pakistan to low-and-middle-income countries. The authors estimated the cost of treatment between US$147 in home-based management to US$1,082 per day for critically ill patients managed in resourceintensive settings (13). However, as the characteristics of the pandemic evolved, so did the global strategies, which ultimately impacted on the case management and overall response of countries. The wide variations in case management protocols between countries has made it even imperative for context-specific estimation of the cost of treatment as part of the needed evidence towards the adoption of sustainable policies and priorities on COVID-19 interventions. To partly fill gaps in context-specific cost of treating COVID-19 patients, we adopt a health system's perspective and a bottom-up, point of care resource use data collection approach to estimate the cost of clinical management of COVID-19 infection in Ghana.

METHODOLOGY
We employed a costing approach guided by the costing framework proposed by Drummond (14) and based on the approved national COVID-19 treatment protocol for Ghana (15). We adopted a bottom-up, point of care resource use data collection approach. Resource use for each type of patient (according to disease severity) was identified and quantified using patientlevel data guided by the advice of case management experts in the frontline of COVID-19 management in the country.
The perspective of cost: The Ministry of Health fully bore the cost of COVID-19 treatment; hence the Ministry's perspective of the cost was adopted for this analysis. The costs related to loss of income to patients were not taken into account, same for loss of revenue to health facilities due to reduced utilization of other routine services arising from the suspension of such services or patients not demanding for them for fear of the COVID-19 pandemic.

Resource use identification and quantification per patient by the level of severity:
The COVID-19 treatment protocol of Ghana as of June 2020 was reviewed together with clinicians to identify resource use elements for the different levels of disease severity and setting of treatment (home or institutionalised care) which culminated in the development of resource use identification template. The resource template was used to collect data, with the assistance of frontline clinicians, on the resources used to treat each type of patient (by the level of disease severity and treatment setting), and the quantity of the resources used. The identified resource use with quantities was validated by two of the country's leading experts in COVID-19 case management and further reviewed by the case management Team Lead of the West African Health Organisation (WAHO). The resource use was categorised into five (5): overheads (patient accommodation, utilities, feeding and toiletries, as well as set of personal protective equipment used by the health professionals), investigations, medications, in-patient care and human resources (staff time). In the case home isolated patients, we considered costs relating to provided items (thermometer, cost of visits and staff time). Ghana's treatment protocol for COVID-19 specified that all moderate, severe or critically ill patients be treated in hospitals while those with no or mild symptoms are supported to manage at home or at isolation centers where the home environment is deemed inconducive for isolation.
Assigning unit cost for each unit of resources used: The unit cost of the resources used in treating COVID-19 patients were triangulated from Ghana's National Health Insurance Scheme (NHIS) price list for medicines and services (16, 17), average prices from the Public Procurement Authority (18), the Government salary structure for public sector health workers (19) and invoices of goods that were procured that were hitherto not regularly procured (example Personal Protective Equipment, PPEs). Where the unit costs could not be obtained from these sources, local open market prices were used except for resources that were not on sale in the local market where the international prices were adopted.
Cost estimation per patient tenement: Using the resources identified and quantified for each patient type, and the unit cost, the cost of treatment per patient (by the level of severity) was computed using the following formula:

Total cost of treatment for patient i = åj (Resourceij x Unit costj)
Where • Resourceij is the amount of resource j used by patients with disease severity i • Unit costj is the unit cost for resource j We compared the cost of treatment by the different categories of resource use within the same level of disease severity and across the different levels of disease severity as well as treatments settings -home-based care or institutionalised care. The estimated costs were not discounted or adjusted for inflation since they were cross-sectional with no long-term extrapolation. The official exchange rate of 5.78 Ghana Cedis to US$1 (as of end of August) was used for converting the cost from one currency to the other.

Resource Use by the level of severity
Patients that received institutionalised care at isolation centres or hospitals were transported from their homes or point of referral to the treatment centre or hospital and sent back upon discharge. This was an essential part of the overhead cost alongside patient accommodation, feeding and toiletries, as well as set of personal protective equipment used by the health professionals. The resource need differentials in the overhead category were driven by the average length of stay, which was for up to 21 days for severely or critically ill patients and 19 days for the rest. Home managed patients received a thermometer for self-monitoring, and were paid visits by clinicians (staff time) which formed the overhead costs from the health system perspective.
Resources needed for investigations and monitoring of patient prognosis included materials for sample taking and reagents like test kits for SARS-COV-2 test, full blood count, blood gases, chemistries and coagulation profile. Other investigative procedures included x-rays, computer tomography (CT) scans, electrocardiogram (ECG) and ultrasound scans for pregnant women.
The resources needed for, and frequency of, these investigations varied markedly depending on the severity of the disease. For instance, while all these were necessary for critically ill patients, analysing blood gases were not indicated for those classified as severe unless they were put on a ventilator (where oxygen saturation was less than 90% with a continuous downward trend in spite of optimal oxygenation or when significant lung changes were detected on x-ray or CT scan). Similarly, blood gases, CT-Scan, chemistries, ECG and coagulation profile were not included in the monitoring protocol for those with mild illness or those that were asymptomatic. For patients managed at home, only routine temperature checking, and the SARS-COV-2 test were needed.
Another category of resource use was medications, which the treatment protocol outlined the use of vitamin C with zinc, hydroxychloroquine (or chloroquine) and azithromycin for all patients regardless of the treatment settings. Patients with severe illness or those who required high dependency or intensive care received additional antibiotics such as ceftriaxone and thrombolytics (commonly enoxaparin).
Resources needed to facilitate the in-patient care of all institutionally managed patients included oxygen for patients who experienced difficulty in breathing, and mechanical ventilation for critically ill patients, patients with severe symptoms and oxygen saturation of 90% or less, as well as for in-patients with significant lung changes on x-ray or CT scan. The use of oxygen and mechanical ventilation was concomitantly associated with the use of syringes, needles, oxygen masks, endotracheal tubes, among others.
The health workforce needed for the management of each case depended on the level of severity and availability of other resources. For example, severely ill patients required up to 4 hours of medical specialists' time and 6 hours of nurses' time per day -of various skill-mix.
Supplementary appendix 1 provides details of the resource needs identified for each level of the disease severity and the associated unit cost. The rest of the analysis was based on these resource use and unit costs.

Average cost of treatment per COVID-19 case -by level of disease severity.
The cost of treating COVID-19 patients in Ghana irrespective of the setting and level of severity from the perspective of the health system ranges from US$282 (GH ¢1,629) to US$23,382 (GH ¢135,149), with an average of US$11,925 (GH ¢68,929). For patients that were treated in hospitals or treatment centres, the overhead cost notably for PPEs and Transportation were the main cost drivers, followed by the cost of in-patient care (see Table 1 for cost summary in USD and supplementary tables S1 -S5 for resource use and unit cost in local currency unit, the Ghana Cedis). For home management, the main cost drivers were investigations (COVID-19 testing) and staff time. As shown in table 2, overhead cost on average accounted for 55% (6-71%) of the cost of clinical management COVID-19 cases in Ghana, followed by in-patient care which accounted for 19% (17-22%). Across all cases, the third cost driver was staff time, accounting for 18% (4-42%) of the treatment cost, followed by investigations which accounted for 11% (1-47%) of the total cost. Interestingly, medications constituted only 2% (0.02-5%).
It is worth noting that there was a wide range in the proportion of cost attributable to overheads, staff time and investigations. For example, there was a 65% cost differential between the overhead cost of patients managed at home and those managed in institutional settings; and across the different levels of disease severity of those that were institutionally managed. Eighty  Meanwhile, the present study found that once a patient's condition deteriorated from mild to moderate, the cost escalated by 155% but the cost-mix shifted from 71% overheads in the case of patients with mild symptoms to 50% for cases with moderate symptoms, while the cost of in-patient care increased from 22% to 35%. Similarly, deteriorating from moderate to severe resulted in more than doubling the costs of treatment but only a marginal difference of 15.2% was found between the cost of patients with severe symptoms versus those that were critically ill.
The current study also estimates that in institutionally managed patients, overhead cost accounted for 63% to 71% of the overall cost of treatment of which 81% were attributable to PPEs, 28% for accommodation and utilities, and 3% for transportation. The cost of drugs accounted for just up to 1% in institutionally managed patients and 5% for patients managed at home. These findings, however, contrasted sharply with those of the Chinese study in which the cost of drugs was observed to be the major cost driver, accounting for 45.1%, of the overall mean (10).
From a cost containment perspective, these findings underline the need for early detection and effective treatment of COVID-19 cases, preferably at home, before any chance of deterioration to the next worst form of the disease state. The few studies that have reported on the cost of treating COVID-19 also collaborate these findings in the context of Kenya, and broadly lowand middle-income countries. Nevertheless, the substantial cost jumps also raise concerns if there was still room for efficiency gains in the resource use in the management of severe and critically ill patients that may accrue as better evidence on the management of COVID-19 evolves.

CONCLUSION
Cost savings could be made by early detection and effective treatment of COVID-19 cases, preferably at home, before any chance of deterioration to the next worst form of the disease state, thereby freeing up more resources for other aspects of the fight against the pandemic.
Policy makers in Ghana should thus make it a top priority to intensify the early detection and case management of COVID-19 infections.

LIMITATIONS
This study has some inherent limitations that must be taken into account when using the same for policy or decision making. First, it is worth noting that the estimates reported in this paper Finally, although every effort was made to use the prevailing market prices as unit costs for resources needed for COVID-19 treatment, volatile pricing resulting from the COVID-19 itself is one factor that could make these estimates quickly outdated. Also, hydroxychloroquine which was part of the treatment protocol was not available on the local market; hence its price was taken from international sources.
Nevertheless, this study, to the best of our knowledge, represents one of the first attempts to Hospital for their contribution to the data collection.
Funding: This paper did not require any funding to be realised.

Availability of data and materials:
The datasets supporting our conclusions are publicly available and will be provided upon request. Ethics approval and consent to participate: The work is based on publicly available data and did not involve the use of human subjects or animals. No ethical approval was required.

Consent for publication: All authors have approved the manuscript for submission.
Competing interests: The authors declare that they have no competing interests.
Supplementary Tables S1 -S5: Resource Use and Unit Costs for various levels of severity of COVID-19