Vaccine Wastage: Associated Risk Factors and Policy Implications in the Littoral Region of Cameroon: A Retrospective Data Review

Introduction: Immunization is an effective preventive health intervention. In Cameroon, the Expanded Program on Immunization aims to vaccinate children under ve years old for free, but vaccination coverage has consistently remained below the national target. Vaccines are distributed based on the target population size, factoring in wastage norms. However, vaccine wastage rates (VWR) may differ among various settings. Our study aimed to assess vaccine wastage for different site settings, seasonality, and vaccine types in comparison to vaccination coverage. As the government of Cameroon continuously seek to improve and vaccination coverage and vaccine wastage, an investigation into why vaccines are being wasted may provide solutions to minimize it occurrences. Methods: Retrospective data on immunization and vaccine wastage covering 2016 and 2017 in the Littoral Region, Cameroon, were collected. Health districts were classied as urban or rural, seasonality categorized as rainy or dry season, and vaccine types were grouped into liquid, lyophilized, oral, and injectable. VWR and vaccination coverage rates (VCR) were calculated, and the vaccine waste factor was investigated. Results: From 2,851,527 doses, children made up 2,640,077 in 2016 and 2017. VWR was the highest in the Bacillus Calmette-Guérin (32.19%) vaccine, followed by measles and rubella (19.05%) and yellow fever (18.34%) vaccines. Single-dose vaccine vials exhibited lower VWR than multi-dose vials. February and November experienced a decrease in the VCR. A negative correlation between VCR and VWR was presented. The VWR for most vaccines was higher during the dry season and remained persistently higher in rural than in urban districts. Conclusion: While the government of Cameroon has consistently been seeking to improve vaccination coverage in recent years by reducing vaccine wastage, understanding risk factors associated to vaccine wastage and coverage is essential for the immunization programs. Reasons for vaccine wastage were multifaceted: limited social infrastructure, cold chain, awareness of immunization, vaccine demand forecasting, and efforts to increase vaccination coverage. To achieve immunization targets, improved forecasting of vaccine demand is essential. Further studies on vaccine wastage and immunization coverage across Cameroon are needed for a comprehensive landscape analysis of immunization programs in Cameroon. in a campaign or special strategy setting and that they can tolerate ambient temperatures of at least + 40°C for a minimum of three days. The vaccines analyzed in this study are currently not available for CTC usage. Furthermore, capacity building of health workers involved in immunization programs is essential. This could include continued community engagement and sensitization, particularly for the rural population, regarding the importance of vaccination. Finally, increasing the size of immunization sessions by facilitating the transportation of vaccines and personnel close to the target population will reduce open vial vaccine wastage. Considering the diverse geographical and climatic characteristics of Cameroon in general and the Littoral Region in particular, better vaccine forecasting with more realistic wastage rates is recommended to prevent the inappropriate supply of vaccines. Further studies are warranted to generate a more comprehensive analysis of vaccine waste across Cameroon and in dynamic climatic changes. This will allow for more rened policy formulation and customized interventions in various settings.

during the dry season and remained persistently higher in rural than in urban districts.
Conclusion: While the government of Cameroon has consistently been seeking to improve vaccination coverage in recent years by reducing vaccine wastage, understanding risk factors associated to vaccine wastage and coverage is essential for the immunization programs. Reasons for vaccine wastage were multifaceted: limited social infrastructure, cold chain, awareness of immunization, vaccine demand forecasting, and efforts to increase vaccination coverage. To achieve immunization targets, improved forecasting of vaccine demand is essential. Further studies on vaccine wastage and immunization coverage across Cameroon are needed for a comprehensive landscape analysis of immunization programs in Cameroon.

Background
Immunization is strongly recommended by the global medical community as an effective preventive medicine to protect children and adults against infectious diseases (1,2). Although infectious diseases affect countries worldwide, the burden is higher in many low and middle-income countries, where vaccinepreventable diseases remain one of the major causes of child morbidity and mortality due to low vaccination coverage (3)(4)(5). Vaccination coverage also differs within countries, between rural and urban settings, whereby rural areas typically exhibit lower vaccination coverage rates (VCR) (2).
The availability of vaccines and vaccination-related commodities is an important factor that in uences vaccination coverage. However, in resourceconstrained countries, with often limited resources, (6, 7), effective vaccine management and utilization is critical in ensuring a regular supply of vaccines, safeguarding national vaccine security, and monitoring costs associated with immunization programs (8). This is further emphasized as the number of new vaccine introductions in developing countries has increased in recent years, coupled with the high pricing of new vaccines (6, 9,10). It has been argued that with newer vaccines for children, cost saving may not be possible at current prices unless the indirect costs to parents caring for a sick child are considered in addition to the direct costs of medical care (11). The importance of minimizing vaccine wastage without compromising vaccination coverage can be a challenge as recommendations to reduce missed opportunities to vaccinate may contradict efforts to reduce vaccine wastage (6), especially if the policy is not clearly explained to health personnel involved in immunization programs.
In 1997, the World Health Organization (WHO) o cially stated that 43% of vaccines delivered to developing countries were wasted due to poor infrastructure (12). Similarly, aggregated national statistics also show disparities in vaccine waste at a local level, such as rural-urban differences (13), inextricably associated with various challenges concerning infrastructure capacity. Other factors such as low quality of monitoring and tracking of vaccination programs (14), parents' objection or disagreement with vaccination, concerns about vaccine safety, accessibility to health facilities such as walking distances, waiting time in health facilities, low educational level of the local population including residents and health workers, population density, and logistical challenges to conducting vaccination programs contribute to vaccine wastage in both rural and urban settings (15)(16)(17).
In Cameroon, the Expanded Program on Immunization (EPI) began in 1976 as a coordinated pilot project of the Organization of Coordination for the Control of Endemic Diseases in Central Africa, which subsequently became operational nationwide in 1982 (18). The national EPI aims to prevent, control, and eliminate or eradicate vaccine-preventable diseases. Following the Declaration of the Reorientation of Primary Health Care in 1993, the EPI activities were integrated into the Minimum Package of Activities of health facilities nationwide and were given to children free of charge with an approach toward vaccination as a fundamental right of every child (18). Immunization coverage in Cameroon has been witnessing a gradual increase over the past decades, but it still falls short of the national target with documentation of missed or incomplete vaccination of eligible children (19). Several reasons explain this, including uptake of national EPI programs by the general population and challenges related to vaccine logistics and management of vaccination programs (20), which aims to not only increase the overall national vaccination coverage but also reduce vaccine wastage (21).
Immunization coverage is related to vaccine wastage as it translates to the availability of vaccines for use, especially in areas with poor access to vaccine storage facilities (6, 7). Even when accessibility to the vaccine storage facilities is guaranteed, high vaccine wastage increases the cost of immunization programs, as vaccine waste factors need to be considered when forecasting the number of vaccine doses. The government of Cameroon has consistently been seeking to improve vaccination coverage in recent years by reducing vaccine wastage (Table 1) (18). Vaccines are distributed based on the size of the target population and factoring in some wastage norms. Routine monitoring of vaccine wastage rates (VWR) and utilization of eld data in estimating vaccine needs is critical for the adequate management of vaccines for immunization programs. Here, we aimed to estimate VWR in various site settings in the Littoral Region of Cameroon, and further investigate vaccine wastage per season and vaccine types in comparison to VCR for respective vaccines that have been introduced in the region. As the government of Cameroon continuously seek to improve and vaccination coverage and vaccine wastage, an investigation into why vaccines are being wasted may provide solutions to minimize it occurrences especially a just a few of such publications exist in the country. The data analyzed and presented will support evidence-based policy decisions and public health measures to improve vaccine management.

Study setting
The Littoral Region is one of the most densely populated regions of Cameroon, with an estimated total population of 3.4 million and a surface area of 20,248 km² (22). There are a total of 189 health districts in Cameroon, of which 24 are in the Littoral Region. These 24 health districts comprised three urban, nine semi-urban, and 12 rural health districts (23). Health districts were classi ed as rural or urban based on their remoteness. Seasonal patterns were characterized as rainy and dry seasons, covering months from June to November and from December to May, respectively (24). The rainy season is typically associated with poor access to healthcare facilities because of deteriorating road networks and rampant power failure, especially in rural districts. This negatively impacts the vaccine supply chain and increases accidents that result in wastage of unopened vaccine vials during outreach sessions of immunization programs (25).

Data collection and analysis
Data were obtained from the government immunization records with authorization from the Ministry of Public Health of Cameroon (Reference N O : 814L/MINSANTE/SG/DSF/GTC-PEV). All data collected included dependent variables such as the number of children vaccinated and the number of doses received, in-stock, remaining, used, and wasted; and independent variables that included seasonality (rainy and dry season), setting (urban and rural), types of vaccines (liquid or lyophilized vaccines, and single-or multi-dose vaccines), and route of vaccine administration (oral or injectable vaccines) ( Table 2). Collected data were entered into an excel-based spreadsheet and analyzed using R version 3.6.0. The number of children vaccinated and the number of vaccine doses used were compared using the chi-square test of independence. The VCR and VWR were calculated using a set of formulas (Table 3) (26).   Comparative analysis of vaccine wastage and coverage rates showed a negative correlation in most vaccines (Fig. 1). VWR increased each time VCR decreased, except in 2016 between October and November, whereby both vaccination coverage and vaccine wastage rates decreased simultaneously. The vaccination coverage of three vaccines, BCG, IPV, and MR, started high in January but fell immediately in February before increasing again in the following months in both 2016 and 2017. Notably, vaccine coverage declined sharply in October and November for all three vaccines, but especially for BCG immunization in both years, although its coverage rate increased again in December.

Vaccine wastage in rural and urban settings
The VWR, categorizing vaccines according to their route of administration and form of preservation, was consistently higher in rural health districts than in urban health districts in both years (Fig. 2). Vaccine wastage was signi cantly different between rural and urban health districts in 2016 and 2017 for all vaccines, except for PCV and rotavirus (Table 5). Notably, the lyophilized vaccines exhibited higher wastage in both rural and urban health districts (Fig. 3).

Seasonality and vaccine wastage rates per types of vaccines
In 2016, more vaccines were wasted during the dry season in all vaccine categories except the lyophilized vaccines, while in 2017, more vaccines were wasted during the dry season in all vaccine categories (Fig. 4). In 2016, more lyophilized vaccines (BCG, MR, and YF) were wasted during the rainy season, whereas the liquid vaccines (pentavalent vaccine, OPV, and IPV) were wasted in the dry season (Table 6). Of all the vaccines, the biggest difference in vaccine wastage occurred with IPV in 2017, whereby the VWR in the dry season was 25.15% higher by over 50% than in the rainy season (12.16%) ( Table 6). Except for the single-dose vaccines (rotavirus vaccine and PCV) and only in 2016 with insigni cant differences in wastage rate between the dry season and the rainy season, the wastage rate of all other vaccines was statistically signi cantly different between the rainy and dry seasons (Table 6). In the Littoral Region of Cameroon, lyophilized vaccines showed a higher VWR, although within the WHO projected ranges. This nding is similar to that of an existing study in The Gambia (10), which showed higher wastage rates in lyophilized vaccines than in other types of vaccines. The VWR in our study was lower than in a study in Bangladesh (29), where the wastage rate for BCG was nearly 84.9%, followed by the MR vaccine at 69.7%, and PENTA at 44.4%.

Discussions
Notably, the liquid vaccine IPV also showed a high wastage rate (17.9%). This may be because it was introduced into the EPI in the Littoral Region in June 2015 (30), and wastage was high at the early stage of vaccine introduction as typically experienced in new immunization programs (31). Our study supports the existing literature on lower wastage rates for vaccines that follow the multi-dose vial policy (MDVP), as seen in other studies from the North West Region of Cameroon (32) and Bangladesh (29).
The MDVP recommendation on the use of opened vaccine vials for up to 28 days, provided the storage conditions are favorable (33), is expected to reduce vaccine wastages (34). However, for lyophilized vaccines (BCG, MR, and YF), their usage is limited to only six hours after reconstitution or at the end of the vaccination session whichever comes rst, after which they must be discarded irrespective of the doses that have been used in the vial (35). Thus, vaccine wastage is only avoidable in large enough sessions that last for less than six hours. Therefore, lyophilized vaccines have a higher wastage rate than liquid vaccines (OPV, IPV, PENTA, PCV, and rotavirus vaccines).
Understanding the relationship between vaccination coverage and vaccine waste is important to investigate the reason for vaccine wastage. Analyzing these two variables over time would allow a better understanding of the reasons for vaccine wastage. If vaccines are used to vaccinate the target population per the immunization plan and standard operating procedures for adequate vaccine management, wastage should remain at a minimum, and vaccination coverage should increase. Overall, our study showed a negative correlation between vaccination coverage and vaccine wastage, and causality may be multifaceted. A lower vaccination coverage may not necessarily be due to the unavailability of vaccines or high VWR. Conversely, a low vaccination coverage may cause an increase in vaccine wastage as vaccines can remain in health facilities and get damaged, resulting in an insu cient number of vaccines to immunize the target population. This is explicable as leftover vaccines taken to outreach sites may not return to the cold chain in their optimal conditions (36) and may be discarded. Notably, between October and November 2016, the wastage of all vaccines decreased as the coverage also decreased. This may be due to the lower number of available vaccines, or could also be related to adopting strategies that reduce vaccine wastage but compromise vaccination coverage (6). The former is the most likely cause in the Littoral Region, as no BCG was available even at the central vaccine storage facility in Yaoundé during this study period. The lack of a particular vaccine has a demotivating effect on healthcare workers in organizing vaccination sessions, as they will need to reorganize such sessions when the missing vaccine becomes available. Parents are demotivated to come for vaccination if they are aware that the vaccines are lacking.
Rural areas are characterized by a smaller population size that is sparsely distributed, resulting in conditions that favor a high VWR (34). This is the case with the Littoral Region, where over the two years, rural districts had higher VWR. Compared to the urban health districts that mostly employ a xed vaccination strategy (where children are brought to health facilities for vaccination), in rural districts, an outreach vaccination strategy is typically applied to reach people living in remote areas with limited access to health facilities. Usually, vaccine vials taken out for this strategy do not return to the vaccine storage facilities if the vaccine vial monitors (small stickers that adhere to vaccine vials and change color as the vaccine is exposed to heat, letting health workers know whether the vaccine can be safely used for immunization) are not in place. Furthermore, the possibility of accidents occurring in rural areas leading to unopened vial breakage is more likely than in urban areas, and less skilled personnel may be involved in the immunization activities (34). Not fully understanding the importance of vaccination due to low educational levels of rural populations often results in their negligent behavior toward meeting the vaccination appointment (17). This often leads to wasting open vials, especially in lyophilized vaccines. Notably, such differences in rural and urban vaccine waste were not signi cant in a study conducted in Gambia (10). This may be due to enhanced vaccine management and high vaccination coverage in Gambia. In the Littoral Region of Cameroon, attempts are being made to resolve the vaccine wastage problem in rural areas and nearby health facilities by planning immunization sessions more strategically, thus, increasing the vaccinated target population size.
The two major seasons, dry and rainy, in Cameroon have a distinctively different effect on immunization activities. Although the dry season is very dusty, it is favorable regarding weather, road conditions and energy supply. During the rainy season, parents are more likely to miss vaccination appointments, which result in increased vaccine wastage, especially for lyophilized vaccines. This is probably why vaccine waste for BCG, MR, and YF was higher during the rainy season in 2016. However, in 2017, the wastage rate for all vaccines was unexpectedly higher in the dry season. This may be due to the higher ambient temperature in 2017, which may have affected vaccines with inadequate cold chains.
In conclusion, investigating vaccine wastage concerning immunization coverage is important to better understand the reasons for VWR and to plan and design better immunization programs that address the identi ed challenges. To reduce vaccine wastage in the Littoral Region of Cameroon, emphasis should be placed on rural areas during the rainy season (especially for lyophilized vaccines) and the dry season. Better cold chain systems should be put in place by investing in basic social infrastructure, such as adequate energy sources for eld vaccine storage capabilities. To address some of the challenges associated with vaccine cold chain management, efforts are underway to develop vaccines that can tolerate extreme temperatures or be out-of-cold chains for a certain time under monitored and controlled conditions (37). This controlled temperature chain (CTC) is an innovative approach to facilitate vaccine management, aimed at reducing vaccine wastage and reaching at-risk vulnerable populations living in remote rural areas or hard to reach areas with limited cold chain conditions and infrastructure. However, this requires the vaccines to be used in a campaign or special strategy setting and that they can tolerate ambient temperatures of at least + 40°C for a minimum of three days. The vaccines analyzed in this study are currently not available for CTC usage. Furthermore, capacity building of health workers involved in immunization programs is essential. This could include continued community engagement and sensitization, particularly for the rural population, regarding the importance of vaccination. Finally, increasing the size of immunization sessions by facilitating the transportation of vaccines and personnel close to the target population will reduce open vial vaccine wastage. Considering the diverse geographical and climatic characteristics of Cameroon in general and the Littoral Region in particular, better vaccine forecasting with more realistic wastage rates is recommended to prevent the inappropriate supply of vaccines. Further studies are warranted to generate a more comprehensive analysis of vaccine waste across Cameroon and in dynamic climatic changes. This will allow for more re ned policy formulation and customized interventions in various settings.

Limitations Of The Study
The study reviewed secondary data. Inherent to secondary data, the accuracy of the result of the study depends on the accuracy with which the data was   shown as blue and red bars. Signi cant differences in VWR were observed between urban and rural areas for all vaccines in both 2016 and 2017, except for the single-dose PCV and rotavirus vaccine, with statistically insigni cant ndings (marked in red asterisk (*)).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Tables3.docx Tables5.docx Tables6.docx