The economic impact of premature mortality in Cabo Verde: 2016 - 2020

Mortality analysis studies in Cabo Verde are rare and those that are available are limited to short periods of analysis and to specific population groups. Traditionally used mortality data does not quantify the burden resulting from premature mortality. This study focused on the years of potential life lost (YPLL), years of potential productive life lost (YPPLL) and the costs associated with them in Cabo Verde from 2016 to 2020 and aimed to elucidate the patterns of early mortality in the period analyzed. Mortality data were obtained from the Ministry of Health, Cabo Verde. Deaths that occurred from 2016 to 2020, aged between one (1) and 73 years old, were analyzed by sex, age group, municipality and cause of death. YPLL, YPPLL and cost of productivity lost (CPL) were estimated using life expectancy and the human capital approach. A total of 6100 deaths were recorded in the sample population and males represented 68.1% (n = 4,154) of the deaths. The total number of deaths verified corresponded to 145,544 YPLL, of which 69.0% (100,389) were attributed to males. There were 4,634 deaths of individuals of working age, and these resulted in 80 965 YPPLL, with males contributing 72.1% (58,403) of the total YPPLL. The total estimated CPL due to premature death was 98,659,153.23 USD with trauma accounting for the highest estimated losses of 21,580,954.42 USD (21.9%), followed by diseases of the circulatory system 18,843,260.42 USD (19.1%), and certain infectious and parasitic diseases accounted for 16,633,842.70 USD (16.9%). The study demonstrated the social and economic burden of premature mortality. The YPLL, YPPLL and CPL measures can be used to complement measures traditionally used to demonstrate the burden and loss of productivity due to premature mortality and to support resource allocation and public health policies making in Cabo Verde.

Crude mortality data, however, have important limitations, such as the fact that 43 they do not reflect the age composition of the population, do not report on preventable 44 deaths in certain age groups and do not quantify the burden of premature mortality [2]. 45 To fill this gap, analysts have increasingly resorted to other measures, including the years 46 of potential life lost. 47 By definition, years of potential life lost (YPLL) represent an estimate of the 48 average time a person would have lived had they not died prematurely. As an impact 49 measure, YPLL seeks to quantify the socioeconomic burden of premature deaths [3]. 50 To quantify economic loss due to early mortality, one can also calculate the years 51 of potential productive life lost (YPPLL) and cost of productivity lost (CPL). The YPPLL 52 use the lower and upper limits of the productive life period as age cut-off and serve as a 53 basis for calculating CPL [4]. 54 Cost of illness studies are considered an essential evaluation technique in the 55 health sector. By measuring and comparing the economic burden of disease on society, 56 these studies can help health decision-makers establish and prioritize health policies and 57 interventions [5]. 58 Despite significant socioeconomic and health improvements, Cabo Verde still has 59 a high burden of mortality in the 15 to 49 age group [6]. Knowing that this age group falls 60 within the productive life span in Cabo Verde (15 to 65 years) it is extremely important 61 to quantify and understand premature mortality in the archipelago. 62 Mortality analysis studies in Cabo Verde are scarce and those found are limited to 63 short periods of analysis and to specific population groups [7,8]. Through a rapid 64 literature review, no studies were identified evaluating YPLL and/or costs associated with 65 premature mortality in the archipelago. This study sought to fill this gap, calculating and 66 analyzing the years of potential life lost (YPLL), by municipality, sex and cause of death 67 in Cabo Verde in the period between 2016 and 2020, as well as the costs associated with 68 early mortality in the country. 69 The results of the study intend to elucidate the patterns of early mortality in the 70 analyzed period, being able to subsidize public policies, mainly at the sanitary level. This 71 work focused on the calculation of years of potential life lost, years of potential productive 72 life lost and the costs associated with these in Cabo Verde from 2016 to 2020. 73 Methodology 74 General approach 75 This is an observational, cross-sectional, quantitative and analytical study using 76 secondary data. For data analysis, all deaths that occurred in Cabo Verde from 2016 to 77 2020, aged between one (1) and 74 years old, were considered. Deaths that occurred in 78 the first year of life were excluded due to the specificity of mortality in this period and 79 due to the social and economic value that this loss represents [9,10]. 80 In this study mortality data were presented by cause of death according to the 81

International Statistical Classification of Diseases and Related Health Problems 10th 82
Revision (ICD-10) [11]. 83 Data sources and data collection 84 An anonymized database, provided by the National Health Directorate of the 85 Ministry of Health, containing information collected from death certificates of all deaths 86 that had occurred in Cabo Verde from 2016 to 2020 was used. 87 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 22, 2022. ; https://doi.org/10.1101/2022.11.21.22282604 doi: medRxiv preprint Where: aiyears remaining to reach the age corresponding to the upper limit 118 considered, when death occurs between the ages of i and i + 1 years, using the adjustment 119 of 0.5 assuming that all deaths occurred in the middle of the year and dinumber of 120 deaths between age i and i + 1 years. 121 To determine the economic losses due to premature deaths during the study period 122 (2016-2020), the YPPLL and CPL method were used [4]: 123 (2) 125 126 Where: wjnumber of years left to complete the age corresponding to the 127 upper limit considered when death occurs between age j and j+1 year; djnumber of 128 deaths between the age of j and j + 1 year, using the adjustment of 0.5 when it is 129 assumed that all deaths occurred in the middle of the year; jmean age in 5-year age 130 groups for the productive population (15 -64 years); 65retirement age limit in Cabo 131 To determine the present value of the cost of productivity lost, the following formula was 141 used: 142 (4) 143 Where: VPpresent value, CPLfuture value, rdiscount rate, tmean 144 number of years remaining to reach the age corresponding to the upper limit 145 considered in each 5-year age group. 146 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 22, 2022. ; https://doi.org/10.1101/2022.11.21.22282604 doi: medRxiv preprint

Estimates of years of potential life lost in Cabo Verde, 2016 to
Of the total number of deaths reported from 2016 to 2020, 42.9% (6100) occurred 174 in the age group from 1 to 73 years, which corresponded to 145,544 YPLL. In the study 175 period, the highest number of YPLL (21,978) were in 2020. The results obtained showed 176 that the total mean YPLL per death in the five-year period varied from 22.6 to 25.0, Table  177 3. 178 Results showed that the leading contributors to YPLL between 2016 and 2020 181 were diseases of the circulatory system, certain infectious and parasitic diseases, external 182 causes of morbidity and mortality and injuries, poisonings and certain other consequences 183 of external causes, which accounted for 71% (103,336) of the total, of YPLL Table 4 In 184 the study period, the highest number of YPLL (31,509) was 2020. In this year, diseases 185 of the circulatory system accounted for 5,913 YPLL, followed by certain infectious and 186 parasitic diseases (5,312), external causes of morbidity and mortality (4,298), neoplasms 187 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 22, 2022. ; https://doi.org/10.1101/2022.11.21.22282604 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 22, 2022. ; Overall, males contributed a higher number of YPLL throughout the study period 205 (2016 -2020). The highest proportion of YPLL (62.2%; 62479) occurred in the age group 206 from 30 to 59 years for males. In the female population, the highest proportion of YPLL 207 corresponded to range 1 to 4 (11.3%; 5112) followed by 35 to 39 age group (10.5%; 4745) 208 (Fig 1). 209 for women, respectively, Table 6. 219 220 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 22, 2022. ; https://doi.org/10.1101/2022.11.21.22282604 doi: medRxiv preprint and certain other consequences of external causes 11.2% (9,038), neoplasms [tumors] 228 10.6% (8,583) and diseases of the respiratory system 8.1% (6,528), Table 7. 229 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 22, 2022. ; https://doi.org/10.1101/2022.11.21.22282604 doi: medRxiv preprint  123, was in the range 60 to 64 years. (Fig 2). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 22, 2022. ;

275
The sensitivity analysis was performed and showed total CPL of 134,724,671.62 USD (mean = 29 073.08 USD /death) at a discount rate 276 of 3% and 74,517,897.02 USD (mean = 16,080.69 USD /death) at 6%. Diseases of the circulatory system had the highest CPL observed at a rate 277 of 3%, and accounted for 24,297,098.13 USD or 18.0% of total CPL, followed by certain infectious and parasitic diseases with 22,809,540.03 USD 278 (16.9%). At a rate of 6%, the same trend was observed, with the two main causes of death having highest CPL, Table 11. 279

283
The present study estimated the loss of productivity due to premature mortality in 284 Cabo Verde, from 2016 to 2020. For this, the Human Capital approach method was used 285 in the analysis of the cost of the disease and the productivity losses due to premature 286 mortality, [4,18]. The current study is the first conducted in Cabo Verde, assessing the 287 indirect costs due to all-causes of premature mortality. 288 The results of the present study point to a significant difference between the sexes, in 289 terms of productivity losses. Males accounted for 71.8% of the total CPL. This result was 290 higher than the results obtained in Tanzania  In the present study, the mean CPL per death was 21,290.28 USD, a value higher 309 than the results of Rumisha et al.[4], and lower than the mean value per death reported by 310 Najafi et al. [18]. In the current analysis, the group "pregnancy, childbirth and 311 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint There is a limited number of studies that assess productivity loss due to all causes 360 of death, particularly on the African continent. In addition, most studies found analyzed 361 costs due to specific causes [17, [27][28][29][30][31][32][33]. Furthermore, no studies were found that 362 addressed the issue of productivity loss in small island developing countries, such as Cabo 363 Verde. Thus, the present study will contribute to the construction of scientific knowledge 364 on the subject at national and regional level. 365 The present study had limitations that should be taken into account when 366 interpreting the results. Firstly, it was assumed that each individual of working age would 367 contribute equally to society if they did not die by the defined age limit using the Human 368 Capital approach. Although this method is often criticized as it tends to overestimate the 369 losses resulting from premature mortality [12,17,34], however, the method is widely used 370 in economic studies, due to its relatively easy application. Moreover, there is still no 371 consensus on the best methodology to assess the indirect costs of mortality [5,12,35]. 372 Future economic studies evaluating the costs of premature death in Cabo Verde, could 373 apply other methods described in the literature [17,19,21]. 374 Other limitations include that fact that other indirect productivity losses related to 375 mortality were not considered in the analysis, such as intangible costs, among others, 376 which could lead to underestimation of the economic burden of the analyzed causes. Cost 377 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 22, 2022. ; https://doi.org/10.1101/2022.11.21.22282604 doi: medRxiv preprint analysis was performed by group of causes. Therefore, the loss of productivity due to 378 specific deaths was not demonstrated. This may constitute a topic for future 379 investigations. 380 Finally, it was found that 4% of the calculated costs were attributed to the group 381 of "symptoms, signs and abnormal findings of clinical and laboratory tests, not classified 382 elsewhere", which highlights the need to further improve the classification of causes of 383 death and registration in the country's health information system (SIS). 384

385
The study demonstrated the burden of premature mortality and associated costs in 386 Cabo Verde from 2016 to 2020 and elucidated patterns of early mortality in the period 387 analyzed. This study revealed that the main drivers of lost productivity in the period were 388 trauma, diseases of the circulatory system, certain infectious and parasitic diseases and 389 neoplasms. It also showed that a significant proportion of the costs of lost productivity 390 was attributable to males. Communicable diseases still represent an important burden 391 with regard to the costs of lost productivity in Cabo Verde, showing that the country is 392 still in an epidemiological transition. 393 The estimation of the burden of mortality and the cost of productivity lost for 394 society and the national economy can be another instrument for assessing the mortality 395 profile of the country, thus complementing the measures traditionally used to demonstrate 396 the profile of mortality and support decision-making and allocation of resources. The 397 implementation of policies aimed at reducing premature mortality, especially from 398 preventable causes, may result in reduced costs of lost productivity and improvements in 399 the well-being of the cabo-verdean population. 400 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 22, 2022. ; https://doi.org/10.1101/2022.11.21.22282604 doi: medRxiv preprint