Vaccine wastage in The Gambia: a prospective observational study

Vaccination will save more than 20 million lives in low and middle income countries between 2001 2020 [1]. In recent years a number of new vaccines have been added to national vaccination programmes and consequently the cost per fully immunised child has increased considerably [2‐4]. These growing costs make the considerate use of vaccines pressing and interest to minimise vaccine wastage has risen.

Vaccine wastage, usually measured as rate, is the proportion of vaccine doses supplied but not administered. Wastage is categorized primarily into two types: a) wastage of the remaining doses in opened vials at the end of an immunisation session, and b) wastage from unopened vials generally due to problems related to the cold chain, breakage or expiry [5, 6]. A number of factors are known to influence vaccine wastage. These broadly include the vaccines themselves, syringes, logistics, immunisation practices, and national policies [6].

In 2005, the WHO estimated that approximately half of the vaccines produced globally are wasted and therefore recommended that countries strengthen local vaccine wastage monitoring [6]. Parmar et al. reported in 2010 that only 19 (26%) of 72 GAVI eligible countries had submitted to WHO wastage data that could be analysed [7]. In the absence of local data, countries use WHO projected wastage rates to estimate their vaccine needs [8].

Due to restricted cold storage capacity in many developing countries, multi-dose vials are commonly used. In 2014, WHO revised its multi-dose vial policy (MDVP) to advice countries on minimising vaccine wastage while ensuring vaccine safety [9]. Under the policy, multi-dose vials with preservatives may be kept for up to 28 days after opening in contrast to the lyophilised vaccines, which do not contain preservatives, and should be discarded 6 h after reconstitution [9].

Despite the recommendations and renewed policy, recent data on vaccine wastage are lacking and published studies have relied primarily on mathematical models [10]. In this study, we prospectively quantified wastage rates and type for the vaccines used in the Gambian Expanded Programme on Immunization (EPI) and assessed the knowledge, attitude and practice (KAP) of Gambian health workers on general knowledge on immunisation and vaccine waste management.

To our knowledge, this is the first prospective study to show vaccine wastage rates in a GAVI eligible country. Target wastage rates projected by the national EPI were met for the liquid vaccines in multi-dose/single vials. However, for the lyophilised vaccines, targets were only met at the major urban health facility with the largest catchment area of the four facilities (target population 150,000).

Wastage rates were highest for BCG. Overall, the rates were higher than the national target of 35% but within range of the WHO target of 50% [6, 13]. Vaccine wastage for lyophilised vaccines including BCG has been shown to correlate with the size of the vaccination session, as the number of children vaccinated decreases, the wastage increases because unused doses at the end of the session are discarded [14, 15]. Wastage in open vials has also been shown to be related to the vial size, with larger vials size leading to higher wastage [16]. The health workers in our study reported not to batch children for vaccination, but rather most would ‘open a vial for every child’ which is in keeping with the high national BCG coverage and the WHO guidelines [6, 11], but contrary to the Nigerian experience where vaccinators batch for a lyophilised vaccine [17]. Still, wastage rates in The Gambia were lower than in Bangladesh [14] with different size of the vaccination session and similar to India even though the facilities in India used 10-dose vials [15, 16, 18] compared to our 20-dose vial. A UNICEF review of the cost-effectiveness of changing from a 20-dose vial to a 10-dose vial BCG concluded that it may be more economical to waste than to reduce vial size because the given price for a 10-dose vial was only 2–8% lower than that of a 20-dose vial [19]. More recent models comparing 10 versus 5-dose vials for other vaccines have shown that a change to smaller vials could reduce wastage but this does not necessarily imply a reduction in costs as other technical reasons such as higher manufacturing and storage costs need to be taken into consideration [20, 21].

Our data supports existing evidence that wastage rates are low for vaccines that follow the MDVP. Wastage rates were lower in centres where the MDVP was practised in Cameroon [22] and Bangladesh [14] compared to those that do not. Similarly, in urban Indian health facilities, wastage of OPV reduced by 50% after introducing the MDVP [23]. The wastage rates reported for the liquid vaccines, Pentavalent, bOPV and Hepatitis B in our study were lower than that from Asian studies [14, 24].

Our study confirms mathematical models showing that wastage in single dose vials PCV are minimal [7]. PCV13 in 4-dose vial presentation is now available in addition to the single dose vial [25]. Earlier studies have shown that for immunisation programmes in developing countries, while per-dose price of antigens in multi-dose vials are lower than single-dose vaccines, even moderate wastage rates can quickly negate this price gain [7, 26]. As the MDVP is applicable to the 4-dose vial, wastage is expected to be low in countries where it is introduced.

Doses wasted in unopened vials are not acceptable and should be minimised with better planning [6]. The observed peaks in wastage rates for the Pentavalent, Tetanus toxoid and Rotavirus vaccines may be one off events during our study. Continuous monitoring is needed to quickly detect unwanted errors that could compromise the programme. It was reassuring that there was no wastage due to freezing and doses discarded due to changes in the VVM were few, perhaps reflecting the efficiency of the cold chain system with regular supervisory visits [2]. We noted that bOPV type of wastage were mainly ‘missing’ or ‘other’ for example spillage and spitting of the vaccine. This is not surprising as bOPV is oral and it may be difficult to quantify the actual number of doses that remain in a vial at the end of a vaccination session.

The KAP survey showed high general knowledge on vaccine waste management among PHO in The Gambia. We did not take into account the time from training which would have been useful particularly for those who did not know EEFO and the wastage rates projected targets. Regular refresher trainings will assist to improve their knowledge on ways to reduce wastage.

The main limitation of our study was that the wastage survey included only two regions of the country, the two with cold rooms. Though the results may not be generalizable to the other five regions, it is noteworthy that the EPI operates similarly in all regions with the RHT coordinating all vaccination activities. On the contrary, by obtaining data from both major and minor health facilities in rural and urban Gambia we showed a wide range of wastage rates across different settings. We conducted the wastage survey prior to the KAP to avoid changes in attitudes among health workers as this may have affected reporting and observations of vaccine wastage, the primary objective of the study.

This operational research shows that, in general, national targets for vaccine wastage in The Gambia were consistently met for liquid vaccines in multi-dose/single vials during the six months of study period. For lyophilised vaccines such as BCG, high wastage rates were detected from unused doses at the end of immunisation sessions. Given the generally high coverage of vaccines in the Gambia, wastage in unopened vials may be considered acceptable. Extra effort should be made to ensure wastage data, which were incomplete at the facilities are captured routinely. A longer period of surveillance will detect if the wastage rates from expiration and breakages are a recurrent issue.

The results from our study should prompt new estimates for the global wastage rates. In the era of new and more expensive vaccines, data on vaccine wastage are crucial for local vaccine forecasting and to validate models. We propose studies in different settings to compare data and contribute to the review of global vaccine wastage estimates. Moving forward, studies to assess if the concern of wastage hinders maximum vaccine coverage would be helpful.