The present value of human life losses associated with COVID-19 and likely cost savings from vaccination in Kenya

2.1 Study location and design

The valuation of human life cross-sectional study on Kenya was for the 5,670 deaths the government reported between 12 March 2020 and 25 July 2022.⁷ The 47 Kenya's administrative counties' share of COVID-19 deaths were as follows:

The indirect costs calculation was for the 5,586 reported deaths in the economically productive age bracket of 15 years and above.^7,8 Also, the direct cost estimation was for a cumulative total of 337,339 COVID-19 cases reported, as of 25 July 2022.^7,8

The direct cost savings estimations encompassed the projected 182,423 COVID-19 infections averted, assuming 30% coverage of the target population (15 years and above) of 31,786,253 with the Oxford-AstraZeneca vaccine.^52,53 The indirect cost savings calculations included the projected 29,872.27 deaths prevented, assuming 30% coverage of the target population with the COVID-19 vaccine.^52–54

2.2 Analytical framework

2.2.1 Model for estimating the present value of reported human lives lost

According to Culyer,⁴³ there are three main approaches for valuing human life: the human capital approach (HKA), the social decisions approach or implied values approach (IVA), and the contingent valuation approach (CVA) or willingness-to-pay approach. First, the HKA assesses the value of a human life lost from any cause (disease or injury) in terms of the discounted expected money worth of goods and services lost by society due to their premature death. Weisbrod⁴⁴ defines the present value of a human being as their discounted expected future income stream net of their consumption.

Second, the IVA (or revealed preference approach as observed in actual choices) infers values from actual past life-saving choices (or decisions) in the public sector.⁴³

Third, the CVA seeks to establish through a questionnaire survey the maximum amount of money individuals are willing to pay for small reductions in the risk of death they face concerning any cause, e.g., COVID-19.⁴³ Unfortunately, according to Robinson et al.,⁴⁵ there are few or no direct estimates of value per statistical life for most low- and middle-income countries which employ the willingness-to-pay (WTP) approach to assess the willingness of those affected by public health challenges (such as COVID-19) to trade their income for small reductions in risk of death.

Due to the availability of data on GDP per capita and current health expenditure per person, we applied HKA to estimate the total present value ($\mathit{TP}{V}_{\mathit{\text{KENYA}}}$) of human lives lost in Kenya due to COVID-19 as of 25 July 2022. A similar approach has been used in Brazil,²⁷ Canada,²⁸ China,²⁹ France,³⁰ Germany,³¹ India,³² Iran,³³ Italy,³⁴ Japan,³⁵ Mauritius,³⁶ South Africa,³⁷ Spain,³⁸ Turkey,³⁷ UK,⁴⁰ and USA.⁴¹

The $\mathit{TP}{V}_{\mathit{\text{KENYA}}}$ is a summation of the present value of human lives lost in age groups $\left({\mathit{PV}}_{i=1,..,7}\right)$ 0–9 years, 10–19 years, 20–29 years, 30–39 years, 40–49 years, 50–59 years, and 60 years and above. Formally^27–41:

The PV_i=1,…,7 for each of the seven age groups was a sum of the product of undiscounted years of life lost (UYLL), net GDP per capita, and COVID-19 deaths in a specific age group.^27–41 Formally:

Where: ${\sum }_{t=1}^T$ is the summation from the undiscounted year one of life lost [UYLL] (t = 1) to the final UYLL (T) in a specific age group, where the age group’s total number of UYLL equals average life expectancy at birth for Kenya $\left({\mathit{ALE}}_{\mathit{\text{KENYA}}}\right)$ minus average age at onset of death $\left({\mathit{AAD}}_{i=1,\dots ,7}\right)$; r is the discount rate of 3%; $\frac{1}{{\left(1+r\right)}^t}$ is the discount factor formula; ${\mathit{\text{COVIDD}}}_{\mathit{\text{KENYA}}}$ is the total number of COVID-19 deaths in Kenya between 12 March 2020 and 25 July 2022; ${\mathit{PD}}_i$ is the i^th age group share of COVID-19 deaths.

In Equation 2, all the years lost due to COVID-19, even those below the minimum working age of 15 years, are valued. We did this because Subsection 2.2.1 primarily concerns the monetary valuation of human lives lost at all ages, irrespective of productivity.

2.2.2 Model for estimating productivity losses (indirect costs) attributed to reported mortality from COVID-19

Kenya’s total indirect cost or productivity loss $\left({\mathit{TIC}}_{\mathit{\text{KENYA}}}\right)$ is a summation of indirect costs in economically productive age groups $\left({\mathit{IC}}_{i=1,..,6}\right)$, i.e. 1 = 15–19 years, 2 = 20–29 years, 3 = 30–39 years, 4 = 40–49 years, 5 = 50–59 years, and 6 = 60 years and above.

Formally^27–41:

The ${\mathit{IC}}_{i=1,..,6}$ for each age group, 1, 2, 3, 4, 5 and 6 equals the present value $\left({\mathit{PV}}_{i=1,..,6}\right)$ multiplied by the relevant employment-to-population ratio $\left({\mathit{EPR}}_{i=2,..,6}\right)$ and i^th age group share $\left({\mathit{\text{SHARE}}}_{i=1}\right)$. Formally:

The ${\mathit{PV}}_{i=1,..,6}$ was computed as explained in Subsection 2.2.1. The ${\mathit{EPR}}_{i=1,..,6}$, the proportion of a country’s working age population employed, was obtained from the Kenya National Bureau of Statistics (KNBS) quarterly labour force report.⁴⁶ ${\mathit{\text{SHARE}}}_{i=1,\dots ,6}$ is the proportion of the indirect cost to be apportioned to the age group, which varies from 0 to 1.

According to the International Labour Organization (ILO) Minimum Age Convention No. 138, the minimum working age “… shall not be less than the age of completion of compulsory schooling and, in any case, shall not be less than 15 years (Article 2)”.⁴⁷ Therefore, since the minimum working age in Kenya is 15 years, we assume that 50% of deaths from COVID-19 in age group 1 (10–19 years) were between 15 and 19 years. Therefore, the ${\mathit{IC}}_1$ for age group 1 (15–19 years) equals the present value $\left({\mathit{PV}}_1\right)$ multiplied by the group employment-to-population ratio $\left({\mathit{EPR}}_1\right)$ multiplied by 0.5, i.e., age group’s share $\left({\mathit{\text{SHARE}}}_{i=1}\right)$.

2.2.3 Model for estimating the total direct cost of reported COVID-19 cases care

The total direct costs of COVID-19 (${\mathit{TDC}}_{\mathit{\text{KENYA}}}$) encompass the value of quantities of inputs used by the NHS to provide appropriate health interventions to different disease severity categorises. For instance, ${\mathit{TDC}}_{\mathit{\text{KENYA}}}$ is the sum of direct costs across the four disease severity categories $\left({\mathit{DC}}_{s=1,..,4}\right)$, i.e. 1 = home-based isolation and care for asymptomatic cases, 2 = hospital/isolation centre care for mild/moderate cases, 3 = hospital high dependency unit care for severe cases, and 4 = hospital intensive unit care for critical cases.

Formally:

The ${\mathit{DC}}_{i=1,..,4}$ for each disease severity category 1, 2, 3, and 4 is the product of the number of COVID-19 cases in a severity category $\left({\mathit{\text{CASES}}}_{s=1,..,4}\right)$, average total direct cost per patient $\left({\mathit{ADC}}_{s=1,..,4}\right)$, and conversion rate from US$ to Int$ $\left(\mathit{CR}\right)$. Formally:

The ${\mathit{ADC}}_{s=1,..,4}$ estimates from Baraza et al.⁴² were used to estimate the cost of managing the four clinical categories of COVID-19 cases (see Table 2). The health system input costs by Baraza et al.⁴² included human resources for health, health worker transport, accommodation and overheads, pharmaceuticals (e.g. medicines), non-pharmaceuticals (fluids, oxygen, devices), COVID-19 tests, other laboratory tests, radiology, personal protective equipment, oxygen therapy, and capital items (e.g. buildings, medical equipment and vehicles).

2.2.4 Model for estimating potential direct and indirect cost savings due to COVID-19 vaccination

The potential savings associated with vaccination equals total direct cost savings $\left({\mathit{\text{TDCS}}}_{\mathit{\text{KENYA}}}\right)$ plus indirect cost savings $\left({\mathit{\text{TICS}}}_{\mathit{\text{KENYA}}}\right)$.

2.2.4.1 Direct cost savings model

The ${\mathit{\text{TDCS}}}_{\mathit{\text{KENYA}}}$ equals the number of COVID-19 cases averted with vaccination $\left({\mathit{\text{AVERTED}}}_{\mathit{\text{INFECTIONS}}}\right)$ multiplied by the average total direct cost per patient treated $\left(\mathit{\text{ATDC}}\right)$. In other words:

2.2.4.2 Indirect cost savings model

The ${\mathit{\text{TICS}}}_{\mathit{\text{KENYA}}}$ equals the number of COVID-19 deaths prevented with vaccination $\left(\mathit{\text{COVID}}19D_{\mathit{\text{PREVENTED}}}\right)$ multiplied by the average total indirect cost per death $\left(\mathit{\text{ATIC}}\right)$. Formally:

2.3 Data and sources

Table 3 shows the data and sources used in the Kenya analysis.

2.4 Data analysis

2.4.1 The total present value of reported human lives lost in Kenya due to COVID-19, as of 25 July 2022

Excel Software (Microsoft, New York) was employed to estimate Equations 1 and 2. The process involved seven steps.

Step 1: Computation of the undiscounted years of life lost

As depicted in Table 4, the UYLL for each of the seven age groups (1 = 0–9 years, 2 = 10–19 years, 3 = 20–29 years, 4 = 30–39 years, 5 = 40–49 years, 6 = 50–59 years, 7 = 60 years and above) were computed through subtraction of ${\mathit{AAD}}_k$ per age group from Kenya’s ${\mathit{ALE}}_{\mathit{\text{KENYA}}}$.

Step 2: Computation of the DYLL

Approximation of the DYLL at a 3% rate for each age group as a product of UYLL and the appropriate discount factor.^27–41 For instance:

The total number of DYLL in the age group 20–29 equals DYLL per human life lost (23.98190213) multiplied by the number of deaths (150) in the age group, i.e. 23.98190213 × 150 = 3,597.3. Table 5 depicts the DYLL per age group due to COVID-19 in Kenya at 3%, 5%, and 10% discount rates.

Step 3: Assessment of Kenya’s net GDP per person in 2022 International Dollars $\left({\mathit{F}}_{\mathbf{2}}\right)$

The net GDP per person $\left({\mathit{\text{NGDPPP}}}_{\mathit{\text{KENYA}}}\right)$ equals GDP per capita $\left({\mathit{\text{GDPPC}}}_{\mathit{\text{KENYA}}}\right)$minus current health expenditure per capita $\left({\mathit{\text{CHEPC}}}_{\mathit{\text{KENYA}}}\right).$^27–41 The 2022 ${\mathit{\text{GDPPC}}}_{\mathit{\text{KENYA}}}$ was Int$ 5762.003. Kenya’s most updated data on ${\mathit{\text{CHEPC}}}_{\mathit{\text{KENYA}}}$ were for 2019.¹⁹ The ${\mathit{\text{CHEPC}}}_{\mathit{\text{KENYA}}}$ for 2022 was forecasted utilising values of Int$185.41142273 in 2018 and Int$207.61849976 in 2019.¹⁹ Applying the annual growth rate of 11.9771892707702%, the forecast for the 2020 ${\mathit{\text{CHEPC}}}_{\mathit{\text{KENYA}}}$ equals Int$ 232.485360437389; forecast for the 2021 ${\mathit{\text{CHEPC}}}_{\mathit{\text{KENYA}}}$ equals Int$ 260.330572083807; and forecast for the 2022 ${\mathit{\text{CHEPC}}}_{\mathit{\text{KENYA}}}$ equals Int$ 291.510857431964. Thus, the ${\mathit{\text{NGDPPP}}}_{\mathit{\text{KENYA}}}$ = ${\mathit{\text{GDPPC}}}_{\mathit{\text{KENYA}}}$ – ${\mathit{\text{CHEPC}}}_{\mathit{\text{KENYA}}}$ = Int$5762.003 − Int$291.510857431964 = Int$5,470.49.

Step 4: Distributing the COVID-19 deaths across seven age groups

This was accomplished through multiplication of the 5670 reported cumulative COVID-19 deaths as of 25 July 2022 in Kenya $\left({\mathit{\text{COVIDD}}}_{\mathit{\text{KENYA}}}\right)$ by the respective age group’s proportion (PD_k).⁷ Therefore, the number of deaths accrued per age $\left({\mathit{\text{COVIDD}}}_{k=1,\dots ,7}\right)$ was:

(a). ${\mathit{\text{COVIDD}}}_{0-9}$ = ${\mathit{\text{COVIDD}}}_{\mathit{\text{KENYA}}}\times {\mathit{PD}}_{0-9}=5670\times 0.010934744=62;$

(b). 10–19 years = ${\mathit{\text{COVIDD}}}_{\mathit{\text{KENYA}}}\times {\mathit{PD}}_{10-19}=5670\times 0.007760141=44;$

(c). 20–29 years = ${\mathit{\text{COVIDD}}}_{\mathit{\text{KENYA}}}\times {\mathit{PD}}_{20-29}=5670\times 0.026455026=150;$

(c). 30–39 years = ${\mathit{\text{COVIDD}}}_{\mathit{\text{KENYA}}}\times {\mathit{PD}}_{30-39}=5670\times 0.072486772=411;$

(d). 40–49 years = ${\mathit{\text{COVIDD}}}_{\mathit{\text{KENYA}}}\times {\mathit{PD}}_{40-49}=5670\times 0.114991182=652;$

(d). 50–59 years = ${\mathit{\text{COVIDD}}}_{\mathit{\text{KENYA}}}\times {\mathit{PD}}_{50-59}=5670\times 0.181834215=1031;$

(e). 60 years and above = ${\mathit{\text{COVIDD}}}_{\mathit{\text{KENYA}}}\times {\mathit{PD}}_{60\mathit{\text{and above}}}=5670\times 0.585537919=3320.$

Step 5: Computation of total present value of human lives lost per age group (${\mathit{PV}}_{\mathit{k}})$

The ${\mathit{PV}}_{k=1,\dots ,7}$ was computed through the multiplication of DYLL per person in an age group, ${\mathit{\text{NGDPPP}}}_{\mathit{\text{KENYA}}}$, and the number of deaths in an age group $\left({\mathit{\text{COVIDD}}}_{k=1,\dots ,7}\right).$^27–41 For instance, the ${\mathit{PV}}_{20-29}$ for age group 20–29 years was obtained from the multiplication of DYLL per person in the age group of 23.982, ${\mathit{\text{NGDPPP}}}_{\mathit{\text{KENYA}}}$ of Int$5,470.49, and ${\mathit{\text{COVIDD}}}_{20-29}$ of 150. Therefore, ${\mathit{PV}}_{20-29}=23.982\times 5470.49\times 150=\mathit{Int}\mathit{\text{\$}}\mathrm{19,678,994}.$

Step 6: Distribution of Kenya's total present value by administrative counties

The ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ was shared across 47 administrative counties (${\mathit{TPV}}_{j=1,\dots ,47})$ through the multiplication of ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ by each county’s proportion of COVID-19 deaths.^27–41 For example, given ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ is Int$ 268,408,687 and PD_NAIROBI is 0.588382574 (Table 3), the share of ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ for Nairobi County equals Int$157,926,994.

Step 7: Sensitivity analysis

One-way sensitivity analysis was conducted through re-estimation of the economic model five times, assuming (i) a 5% discount rate, (ii) a 10% discount rate, (iii) Africa’s highest average life expectancy at birth of 78.76 years (Algeria females),⁴⁸ (iv) the world highest life expectancy of 88.17 years (Hong Kong females),⁴⁸ and (v) projected excess COVID-19 mortality of 180,217.4721 deaths as of 25 July 2022 in Kenya (COVIDD_KENYA).^7,8,50 How was the latter forecasted?

The COVID-19 Excess Mortality Collaborators⁴⁹ estimated that the actual number of COVID-19-associated deaths may have been far more significant than those reported due to Kenya's weak death registration system. According to COVID-19 Excess Mortality Collaborators,³⁶ by 31 December 2021, the reported COVID-19 deaths in Kenya were 5380 (5.7 per 100,000), and the estimated excess deaths were 171,000 (181.2 per 100,000). The ratio between excess mortality and the reported COVID-19 mortality rate was 31.78438662. The projected excess number of COVID-19 deaths as of 25 July 2022 was 180,217.4721, i.e. 5670 reported deaths as of 25 July 2022 multiplied by 31.784.

2.4.2 The indirect costs (productivity losses) attributed to reported mortality from COVID-19

Equations 3 and 4 were built into an Excel spreadsheet and used to estimate the indirect costs following the seven steps below.

Step 1: Search the ILO website for the minimum working age in Kenya.⁴⁷

Step 2: Delineate the economically productive age groups⁴⁶: 1 = 15–19 years, 2 = 20–29 years, 3 = 30–39 years, 4 = 40–49 years, 5 = 50–59 years, and 6 = 60 years and above.

Step 3: Extract the present values for each of the six economically productive age groups from the results obtained following procedures explained in Subsection 2.2.1.

Step 4: Extract the employment-to-population ratio for each productive age group $\left({\mathit{EPR}}_{i=1,..,6}\right)$ from KNBS quarterly labour force report of 2021.⁴⁶

Step 5: Ascertain the proportion (ranging from 0 to 1) of the indirect costs to be apportioned to the age group. Age groups coded 2 to 6 were allotted a share of 1 because all persons in those age groups were presumed to be potentially economically productive. Given that in the age group 10–19, only persons aged 15 to 19 years (i.e., five years) are potentially productive, we divided five years by 10 years (number of years in the age group) and obtained a value of 0.5, as the age group ${\mathit{\text{SHARE}}}_{i=1}$.

Step 6: Estimate the ${\mathit{IC}}_{i=1,..,6}$ for each age group 1, 2, 3, 4, 5 and 6 by multiplying the present value $\left({\mathit{PV}}_{i=1,..,6}\right)$, the relevant employment-to-population ratio $\left({\mathit{EPR}}_{i=1,..,6}\right)$, and the share of the indirect cost to be apportioned to the age group $\left({\mathit{\text{SHARE}}}_{i=1,..,6}\right)$.

Step 7: Sum up the six age groups' indirect costs to derive Kenya's total indirect cost or productivity loss $\left({\mathit{TIC}}_{\mathit{\text{KENYA}}}\right)$.

2.4.3 The direct cost attributed to reported COVID-19 cases

Equations 5 and 6 were built into an Excel spreadsheet and used to estimate the direct costs following the six steps below.

Step 1: Determine the share of the total COVID-19 cases reported in Kenya by four disease categories: 1 = asymptomatic cases on home-based care, 2 = mild/moderate cases on hospital/isolation centre care, 3 = severe cases on hospital high dependency unit care, and 4 = critical cases on hospital intensive unit care. According to the Kenya Ministry of Health COVID-19 daily report 940,⁵¹ out of the total number of COVID cases, 73.06% were from home-based care, and 26.94% were from various health facilities. Out of the total COVID-19 cases treated at health facilities, 72.9% were mild-to-moderate cases admitted in a general ward, 12.1% were severe cases treated in a high dependency unit, and 15.0% were treated in an intensive care unit.

Step 2: Estimate the number of COVID-19 cases per disease category $\left({\mathit{\text{CASES}}}_{s=1,..,4}\right)$ by multiplying the total number of reported cases (330,910) by the share for each clinical category (obtained from Step 1). For instance, category 1 = 330,910 cases × 0.7306 = 241,762.8; category 2 = 330,910 cases × 0.2694 × 0.729 = 64,988.3; category 3 = 330,910 cases × 0.2694 × 0.121 = 10,786.8; category 4 = 330,910 cases × 0.2694 × 0.15 =13,372.1.

Step 3: Search in ‘Pubmed.com’ for a published Kenyan study documenting the average total direct cost per patient per clinical category $\left({\mathit{ADC}}_{s=1,..,4}\right)$. The search revealed a study by Barasa et al.⁴² that reported ${\mathit{ADC}}_{s=1,..,4}$ (see Table 2).

Step 4: Derive a rate for converting $\left(\mathit{CR}\right)$ unit costs expressed in US Dollars (US$) into International Dollars (Int$) using GDP data from the IMF World Economic Outlook Database.³ As shown in Table 2 in 2022, Kenya’s GDP in 2022 was US$116.641 billion, equivalent to Int$293.423 billion.³ Thus, the CR from US$ to Int$ equals 2.51560771941256, i.e. Int$293.423 billion divided by US$116.641 billion.

Step 5: Estimate the direct cost per disease severity category $\left({\mathit{DC}}_{i=1,..,4}\right)$ by multiplying the number of COVID-19 cases in a severity category $\left({\mathit{\text{CASES}}}_{s=1,..,4}\right)$ from Step 2 by the respective average total direct cost per patient $\left({\mathit{ADC}}_{s=1,..,4}\right)$ from Step 3 and $\mathit{CR}$ from Step 4.

Step 6: Calculate Kenya’s total direct cost $\left({\mathit{TDC}}_{\mathit{\text{KENYA}}}\right)$ through summation of direct cost (obtained in Step 5) across the four disease severity categories $\left({\mathit{DC}}_{s=1,..,4}\right)$.

2.5 The potential projected direct and indirect cost savings due to COVID-19 vaccination

Our study estimates the potential savings from COVID-19 vaccination using actual population coverage of 30%, as of 25 July 2022, and the potential direct and indirect cost savings of the projected COVID-19 cases and deaths averted due to vaccination.

As explained by the COVID-19 Excess Mortality Collaborators⁴⁹ (Subsection 2.4.1), the actual number of COVID-19-associated deaths may have been underestimated by a ratio of 31.784. For this reason, we decided to base the estimation of potential direct and indirect cost savings from COVID-19 vaccination on the projected total number of cases and deaths as of 25 July 2022.

2.5.1 Expected savings in total direct costs due to COVID-19 vaccinations

Equations 7, 8, 9 and 10 were built into an Excel spreadsheet and used to estimate the expected total direct cost savings attributable to vaccination with the Oxford-AstraZeneca vaccine following seven steps.

Step 1: Obtain target population (15 years and above) of 31,786,253 for Kenya from the Kenya COVID-19 vaccination programme daily situation report dated 26 July 2022.⁵²

Step 2: Search PubMed for an epidemiological study on COVID-19 vaccine efficacy. The research revealed a study by Voysey et al.⁵³ that found "Overall vaccine efficacy more than 14 days after the second dose was 66·7% (95% CI 57·4–74·0), with 84 (1·0%) cases in the 8597 participants in the ChAdOx1 nCoV-19 group and 248 (2·9%) in the 8581 participants in the control group (p. 881)". Their study was a pooled analysis of four randomised trials (Brazil, South Africa, and the UK) with 8597 participants receiving the Oxford-AstraZeneca vaccine and 8581 receiving the control vaccine or saline.

Step 3: Use the evidence in Step 2 to estimate the COVID-19 infection risk without$\left({\mathit{IR}}_{\mathit{\text{Control}}}\right)$ and with Oxford-AstraZeneca$\left({\mathit{IR}}_{\mathit{AZ}}\right)$ in the Voysey et al.⁵³ pooled analysis of randomised trials in Brazil, South Africa, and the UK (Table 6). Infection risk in a group equals the number of infected persons divided by group size.

Step 4: Estimate the number of people in Kenya expected to contract COVID-19 without vaccination$\left({\mathit{\text{PoPI}}}_{\mathit{\text{without}}}\right)$. The ${\mathit{\text{PoPI}}}_{\mathit{\text{without}}}$ was obtained by multiplying the target population $\left(T{\mathit{PoP}}_{\mathit{\text{KENYA}}}\right)$ of 31,786,253 by the ${\mathit{IR}}_{\mathit{\text{Control}}}$ of 0.02890106.^52,53 Thus, ${\mathit{\text{PoPI}}}_{\mathit{\text{without}}}$ equals 918,656.405, i.e. ${\mathit{\text{TPoP}}}_{\mathit{\text{KENYA}}}\times {\mathit{IR}}_{\mathit{\text{Control}}}=\mathrm{31,786,253}\times 0.02890106.$

Step 5: Approximate the number of people in Kenya expected to contract COVID-19 even after being fully vaccinated with the Oxford-AstraZeneca vaccine $\left({\mathit{\text{PoPI}}}_{\mathit{\text{with}}}\right)$. The ${\mathit{\text{PoPI}}}_{\mathit{\text{with}}}$ was derived by multiplying respective ${\mathit{\text{TPoP}}}_{\mathit{\text{KENYA}}}$ by ${\mathit{IR}}_{\mathit{AZ}}$ of 0.00977085.^52,53 Therefore, ${\mathit{\text{PoPI}}}_{\mathit{\text{with}}}$ equals 310,578.71, i.e. ${\mathit{\text{TPoP}}}_{\mathit{\text{KENYA}}}\times {\mathit{IR}}_{\mathit{AS}}=\mathrm{31,786,253}\times 0.00977085.$

Step 6: The number of infections averted, assuming 30% population coverage, equals the difference between ${\mathit{\text{PoPI}}}_{\mathit{\text{without}}}$ and ${\mathit{\text{PoPI}}}_{\mathit{\text{with}}}$, multiplied by 0.30 (30% coverage).⁵² Since ${\mathit{\text{PoPI}}}_{\mathit{\text{without}}}$ = 918,656.41 (from Step 4) and ${\mathit{\text{PoPI}}}_{\mathit{\text{with}}}$= 310,578.71 (from Step 5), infections averted through 30% vaccination coverage with Oxford-AstraZeneca equals 182,423.31, i.e. (918,656.41 − 310,578.71) × 0.30.

Step 7: The total direct cost savings expected from vaccination equals the number of infections averted (from Step 6) multiplied by the average total direct cost per patient. For instance, the expected savings in Kenya equals Int$300,668,273, i.e. 182,423.31 infections averted (from Step 6) multiplied by the average total direct cost per patient of Int$1,648.19.

2.5.2 Expected savings in projected total indirect costs due to COVID-19 vaccination

Equations 11, 12, 13 and 14 were built into an Excel spreadsheet and used to estimate the expected savings in total indirect costs due to COVID-19 vaccination following the six steps below.

Step 1: A search in the PubMed.com database for COVID-19 vaccine effectiveness in reducing the risk of death revealed an article by Bernal et al.,⁵⁴ which attempted “to estimate the real-world effectiveness of the Pfizer-BioNTech BNT162b2 and Oxford-AstraZeneca ChAdOx1-S vaccines against confirmed covid-19 symptoms, admissions to hospital, and deaths” (p.1).

Step 2: Utilise the evidence in Step 1 to calculate the risk of COVID-19 resulting in death among the unvaccinated$\left({\mathit{DR}}_{\mathit{\text{unvaccinated}}}\right)$ and those vaccinated with the Pfizer-BioNTech BNT162b2$\left({\mathit{DR}}_{\mathit{PB}}\right)$ (Table 7). The risk of death in a group equals the number of deaths from COVID-19 in an age group divided by the total number of cases in the group.

Step 3: Estimate the number of people infected in Kenya expected to die from COVID-19 without vaccination $\left({\mathit{\text{PoPD}}}_{\mathit{\text{without}}}\right)$ as a product of ${\mathit{\text{PoPI}}}_{\mathit{\text{without}}}$ (918,656.41) and ${\mathit{DR}}_{\mathit{\text{unvaccinated}}}$ (0.131380546). Thus, ${\mathit{\text{PoPD}}}_{\mathit{\text{without}}}={\mathit{\text{PoPI}}}_{\mathit{\text{without}}}\times {\mathit{DR}}_{\mathit{\text{unvaccinated}}}=\mathrm{918,656.41}\times 0.131380546=\mathrm{120,693.58}.$

Step 4: Estimate the number of people in Kenya expected to die from COVID-19 even though vaccinated with Pfizer-BioNTech BNT162b2$\left({\mathit{\text{PoPD}}}_{\mathit{\text{with}}}\right)$ through the multiplication of the number of people expected to contract COVID-19 though vaccinated $\left({\mathit{\text{PoPI}}}_{\mathit{\text{with}}}\right)$ (from Step 5 of Subsection 2.5.1) by the probability of death in a vaccinated group $\left({\mathit{DR}}_{\mathit{PB}}\right)$ of 0.068. In Kenya, for instance, the ${\mathit{\text{PoPI}}}_{\mathit{\text{with}}}$ equals 310,578.71 persons multiplied by ${\mathit{DR}}_{\mathit{PB}}$of 0.068, i.e. ${\mathit{\text{PoPD}}}_{\mathit{\text{with}}}={\mathit{\text{PoPI}}}_{\mathit{\text{with}}}\times {\mathit{DR}}_{\mathit{PB}}=\mathrm{310,578.71}\times 0.068=\mathrm{21,119.35}.$

Step 5: The number of COVID-19-associated deaths prevented $\left(\mathit{\text{COVID}}-19D_{\mathit{\text{Prevented}}}\right)$, assuming 30% population vaccine coverage, equals the difference between the number of people expected to die of COVID-19 without $\left({\mathit{\text{PoPD}}}_{\mathit{\text{without}}}\right)$ and with $\left({\mathit{\text{PoPD}}}_{\mathit{\text{with}}}\right)$ vaccination, multiplied by 30%. In other words, $\mathit{\text{COVID}}-19D_{\mathit{\text{Prevented}}}=\left({\mathit{\text{PoPD}}}_{\mathit{\text{without}}}-{\mathit{\text{PoPD}}}_{\mathit{\text{with}}}\right)\times \left(30/100\right)$. The number of deaths averted through 30% vaccination coverage equals 29,872.27, which is 120,693.58 (from Step 3 in Subsection 2.5.2) minus 21,119.35 (from Step 5 in Subsection 2.5.2) multiplied by 0.30.

Step 6: The indirect cost savings expected from vaccination equals the number of deaths prevented $\left(\mathit{\text{COVID}}-19D_{\mathit{\text{Prevented}}}\right)$ (Step 5 in Subsection 2.5.2) multiplied by the average indirect cost per COVID-19 death in Kenya $\left({\mathit{\text{ATIC}}}_{\mathit{\text{KENYA}}}\right)$ (Subsection 2.4.1 Equation 9). For example, the expected savings from COVID-19-associated deaths prevented in Kenya equals Int$1,100,277,535.56, i.e. 29,872.27 deaths prevented (among 15 years and older) multiplied by the indirect cost per death from COVID-19 of Int$36,832.74.

3.1 The total present value of reported human lives lost in Kenya due to COVID-19

3.1.1 Findings from the present value of life analysis assuming Kenya’s average both sexes life expectancy of 67.47 years and a discount rate of 3%

As of 25 July 2022, Kenya had lost 5,670 human lives from COVID-19, translating to 66,980 UYLL, equivalent to 49,065 DYLL. As depicted in Table 8, the cumulative number of human life losses had a ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ of Int$268,408,687 and an average total present value of Int$47,338 per human life (i.e., about eight times the GDP per capita for Kenya).

Approximately 3.6% of the ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ was borne by 0–9-year-olds, 2.4% by 10–19-year-olds, 7.3% by 20–29-year-olds, 17.4% by 30–39-year-olds, 21.9% by 40–49-year-olds, 22.3% by 50–59-year-olds, and 25.2% by 60–year-olds and above. The persons between 20 and 59 years—the most economically productive bracket—incurred 68.9% (Int$185,000,542) of the ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$. The ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ decreases as the age of the person advances. For instance, the 0–9-year-olds average TPV of Int$154,025 was eight-fold higher than Int$47,338 among the 60-year-olds and above.

3.1.2 Share of the TPV by administrative counties in Kenya

Figure 1 depicts the share of the ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ across the 47 administrative counties in Kenya.

Figure 1. Distribution by county of discounted monetary value of human life losses associated with COVID-19 in Kenya as of 25 July 2022 in international Dollars (Int$).

The average ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ was Int$5,710,823 per county, with a standard deviation of Int$23,349,696. The size of ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ varied widely between counties, i.e., from a minimum of Int$14,934 (in Elgeyo Marakwet, Samburu, and West Pokot Counties) to a maximum of Int$157,926,994 in Nairobi County. Thirty-five (74.5%) counties had a ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ of less than Int$1,000,000; six counties (12.8%) had between Int1,000,000 and Int$10,000,000; four counties (8.5%) had between Int$10,000,001 and Int$20,000,000; two (4.3%) counties had over Int$20,000,000. Five counties (Kajiado, Kiambu, Machakos, Mombasa, and Nairobi City) bore 86% of the ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$. Nairobi city alone bore 58.8% (Int$157,926,994) of the ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$. The size of ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ borne by a county hinge on the number of COVID-19 life losses sustained.

3.1.3 Sensitivity analysis

3.1.3.1 Impact of changes in the discount rate

Table 9 shows that the re-run of the HKA model with a discount rate of 5% led to a decrease in the ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ from Int$ 268,408,687 to Int$226,791,171, which is a 16% (Int$41,617,516) decrease. The ${\mathit{\text{ATPV}}}_{\mathit{\text{KENYA}}}$ decreased from Int$47,338 to Int$39,998 per COVID-19-associated death.

Re-estimation of the HCA model with a 10% discount rate, all other factors held constant, reduced the ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ from Int$268,408,687 to Int$164,679,113, which was a 39% reduction (Int$103,729,574). The ${\mathit{\text{ATPV}}}_{\mathit{\text{KENYA}}}$ decreased from Int$47,338 to Int$29,044 per COVID-19-associated death.

3.1.3.2 Effect of changes in life expectancy at birth

As portrayed in Table 10, a re-estimation of the economic model with Africa's highest life expectancy at birth of 78.76 years (Algeria's females) grew the ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ from Int$268,408,687 to Int$480,899,177, which is 79% (Int$212,490,490) growth. Likewise, the mean ${\mathit{\text{ATPV}}}_{\mathit{\text{KENYA}}}$ grew from Int$47,338 per human life (obtained assuming a national life expectancy of 67.47 years) to Int$84,815.

Re-estimation of the economic model with the World's highest life expectancy at birth of 88.17 years (Hong Kong females) increased the ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ from Int$268,408,687 to Int$613,747,054, which is 129% (Int$345,338,367) growth. The ${\mathit{\text{ATPV}}}_{\mathit{\text{KENYA}}}$ grew from Int$47,338 per human life (obtained assuming a national life expectancy of 67.47 years) to Int$108,245.

3.1.3.3 Effect of changes in the number of deaths due to COVID-19

As portrayed in Table 11, a re-run of the economic model with the excess mortality of 180,215, instead of the reported 5670 COVID-19 deaths, increased the ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ by 3,078% (Int$8,262,796,784), i.e., from Int$268,408,687 to Int$8,531,205,470.

3.2 The indirect and direct costs of reported cases

3.2.1 The indirect costs (or productivity losses) of reported deaths

As shown in Table 12, the 5586 COVID-19-reported deaths among those within the economically productive age bracket of 15 years and above resulted in a total indirect cost of Int$ 205,747,692; and an average total indirect cost per death of Int$ 36,833.

All the 84 deaths that occurred below the age of 15 years, which were not within the working age bracket, were valued at zero. Out of the total productivity losses, 0.3% were borne by 15–19-year-olds; 5.6% by 20–29-year-olds; 19.2% by 30–39-year-olds; 24.3% by 40–49-year-olds; 24.5% by 50–59-year-olds; and 26.1% by 60-year-olds and above.

3.2.2 The direct cost of caring for reported COVID-19 cases

As depicted in Table 13, the estimated total direct cost of caring for the reported 330,910 cases was Int$545,401,259.29; and the average total direct cost was Int$1,648.2 per patient.

Of these, 25% was for home-based isolation and care for asymptomatic cases, 22.9% for hospital/isolation centre care for mild/moderate cases, 7.4% for hospital high dependency unit care for severe cases, and 44.4% for hospital intensive care unit care for critical cases. As expected, due to the resource-intensive nature of hospital intensive unit care, the care of critically sick COVID-19 patients accounted for almost half of the total direct cost.

3.4 The potential projected direct and indirect cost savings due to COVID-19 vaccination

We estimate that the 30% target population's COVID-19 vaccination coverage may have saved Kenya a total cost of Int$ 1,400,945,809. It consists of Int$300,668,273 direct cost savings associated with the prevention of 182,423 COVID-19 projected infections and indirect cost savings of Int$1,100,277,536 from 29,872 deaths averted among 15-year-olds and above.

4.1 Key findings

This study has seven key findings. First, the 5,670 human lives Kenya reported to have lost from COVID-19 had a ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ of Int$268,408,687, equivalent to 0.1% of Kenya's total GDP in 2022. Second, the ${\mathit{\text{ATPV}}}_{\mathit{\text{KENYA}}}$ of Int$47,338 per human life was eight times the per capita GDP of Kenya. Third, about 59% of ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ accrued only in Nairobi City County. Fourth, sensitivity analysis revealed that an increase in discount rate reduces ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$, increases in life expectancy at birth augment ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$, and increases in the number of deaths associated with COVID-19 grow the estimated ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$. Fifth, the 5586 COVID-19-reported deaths in the economically productive age bracket of 15 years and above resulted in a total indirect cost of Int$ 205,747,692 and an average total indirect cost per death of Int$ 36,833. Sixth, the estimated total direct cost of caring for the reported 330,910 cases was Int$545,401,259.29, and the average total direct cost was Int$1,648.2 per patient. Seventh, the 30% target population COVID-19 vaccination coverage may have saved Kenya a total cost of Int$ 1,400,945,809.

4.2 Comparison with COVID-19 related value-of-life studies

As depicted in Table 14, Kenya's ${\mathit{\text{ATPV}}}_{\mathit{\text{KENYA}}}$ was lower than all the 15 countries that also applied a similar human capital model.

For instance, the ${\mathit{\text{ATPV}}}_{\mathit{\text{KENYA}}}$ for Kenya of Int$47,338 is less than those of Spain by approximately 10-fold, Italy by 8-fold, China by 8-fold, France by 7-fold, Mauritius by 7-fold, USA by 6-fold, Japan by 6-fold, Canada by 5-fold, Turkey by 5-fold, UK by 5-fold, Germany by 4-fold, Iran by 3-fold, Brazil by 2-fold, India by 2-fold, and South Africa by 2-fold. Kenya’s lower ${\mathit{\text{ATPV}}}_{\mathit{\text{KENYA}}}$ might be related to the lower GDP per capita³ and the lower average life expectancy at birth.⁴⁸

4.3 Limitations of the study

First, the discounted monetary values of life reported in our paper hinge on the number of COVID-19-associated deaths reported by the Government of Kenya (GoK). COVID-19 Excess Mortality Collaborators estimated that the GoK may have underestimated excess mortality due to the pandemic by 31.784-fold.⁴⁹ Consequently, our ${\mathit{TPV}}_{\mathit{\text{KENYA}}}$ estimate of Int$268,408,687 might be underestimated by 31.784-fold.

Second, due to the unavailability of research resources, we could not compare our estimates using the HKA with those of alternative human life valuation methods (IVA and CVA) highlighted in the Methods section.³⁷

Third, our study uses the GDP per capita as a proxy indicator of the value the Kenyan society attaches to human statistical life. As discussed by Giannetti et al.,⁵⁵ Stiglitz et al.,⁵⁶ Fleurbaey⁵⁷ and Kahneman and Deaton,⁵⁸ the indicator is not an indicator of overall well-being (quality of life, happiness, wellness) of society as it ignores social-economic-political-ecological inequities, omits environmental costs (e.g. depletion of natural resources, global warming due to pollution), and excludes most non-monetary production (e.g. child and elderly care at home, household chores by full-time homemakers).

Fourth, the HKA omits a person's non-monetary value to the bereaved family,⁴⁴ the psychological pain of the loss of a loved one, takes account only of society's loss in national income and ignores the person's desire to live.⁵⁶

Fifth, our study captures only one of the adverse effects of the global COVID-19 pandemic, i.e. the associated mortality. It does not value non-fatal short-term and long-term effects on victims’ health, which could be significant.⁵⁹

Sixth, our study suffers the limitations explained by Baraza et al.⁴² because it used their direct unit cost estimates. Furthermore, in calculating direct cost, Baraza et al.⁴² did not consider the out-of-pocket expenses incurred by COVID-19 patients and their families and friends during diagnosis, isolation, management, and rehabilitation. Thus, in that respect, the total direct cost savings due to the COVID-19 vaccination reported in our paper might be underestimated.