Documentation of vaccine wastage in two different geographic contexts under the universal immunization program in India.

BACKGROUND
Government of India is introducing new and relatively costly vaccines under immunization program. Monitoring of vaccine wastage is needed to guide the program implementation and forecasting. Under pilot introduction of rotavirus vaccine in two districts both 5- and 10-doses vials were used, which was considered as an opportunity for documenting the wastage. The wastage rates for other routine vaccines were also documented.


METHODS
A survey conducted in two districts (Kangra, Himachal Pradesh and Pune, Maharashtra) covered 49 vaccine stores, 34 sub-centres and 34 outreach sessions collected vaccine receipt, distribution and usage data for two complete years 2016 and 2017.


RESULTS
The overall wastage rates for almost all vaccines were higher in Kangra district (BCG 37.1%, DPT 32.1%, Measles 32.2%, OPV 50.8%, TT 34.1% and pentavalent 18.4%) than Pune district (BCG 35.1%, DPT 25.4%, Measles 21.7%, OPV 14.3%, TT 23.1% and pentavalent 13.2%). Wastage for pneumococcal conjugate and measles-rubella vaccines in Kangra district were 27 and 40.5%, respectively. With transition from 5- to 10-doses vials for rotavirus vaccine, wastage at stores levels increased in both Kangra (29 to 33.2%) and Pune (17.8 to 25.7%) districts. With transition from intramuscular to intradermal fractional inactivated polio vaccine, the wastage increased from 36.1 to 54.8% in Kangra and 18.4 to 26.9% in Pune district.


CONCLUSIONS
The observed vaccine wastage rates for several vaccines were relatively higher than program assumption for forecasting. The observed variations in the vaccine wastage indicates need for state or region based documentation and monitoring in India for appropriate programmatic action.


Background
Under the Universal Immunization program (UIP) in India the vaccine portfolio has been expanded from 6 to 13 vaccines now with 10 vaccines [BCG; oral polio vaccine (OPV); inactivated polio vaccine (IPV); diptheria-pertusis-tetanus (DPwT); liquid pentavalent (LPV) combining DPwT, hepatitis B vaccine (HBV) and Haemophilus influenzae type b (Hib); tetanus toxoid (TT), measles and rubella (MR)] universally and three vaccines [rotavirus vaccine (RVV), pneumococcal conjugate vaccine (PCV) and Japanese encephalitis (JE)] in select areas. There is also progressive rise in the full immunization coverage, although slower than anticipated. Usually in the program the vaccine requirements are estimated based on the estimated number of beneficiaries, number of doses and the vaccine wastage. Vaccine wastage is an integral part of vaccine logistics. The Government of India have recommended the wastage rate for 50% (BCG), 25% (Measles, Measles and rubella, rotavirus, and Japanese Encephalitis vaccines), and 10% for all other multi-dose vial vaccines including DPT, LPV, IPV, OPV, TT, and HBV [1]. The vaccines are costly commodities and the newer vaccines are even costlier, putting a significant financial burden on the country. An UNICEF study across 5 states of India informed a high vaccine wastage rates for various vaccines ranging from 27 to 61% for various vaccines [2]. The wastage rates varied across the states for different vaccines. World Health organisation (WHO) estimated that about 50% of the vaccines produced globally are wasted [3]. According to the available reports for PCV pipeline planning in 2010, WHO had analysable vaccine wastage data for only 19 of the 72 Gavi, the Vaccine Alliance eligible countries [4]. Hence it was recommended to reinforce the vaccine wastage monitoring and documentation at national and sub-national levels for large countries like India, with new costlier vaccines being introduced in the UIP. The increase in cost of vaccines makes the solicitous use of vaccines and interest to minimize the vaccine wastage has risen. To reduce the wastage, India adopted open vial policy for liquid vaccines in 2011. Several factors influence the wastage including the vaccine vial, applicability of open vial policy, beneficiary base, session planning, syringes used, immunisation practices, storage and logistics, and national policies [3].
National Technical Advisory Group on Immunization (NTAGI), the apex advisory body for immunization in India recommended phased introduction of RVV in India. NTAGI also recommended a pilot rollout linked to surveillance of adverse events under monitoring of the relevant Government agencies. Pilot introduction of RVV was done in two districts (Kangra, Himachal Pradesh and Pune, Maharashtra) of India, prior to the national introduction linked to surveillance of intussusception, a rare adverse event. Under the pilot introduction, initially five doses RVV vials were supplied, which were changed to 10 doses vials synchronising with the national rollout. The PCV was introduced in Himachal Pradesh (May 2017) under first phase of national rollout. The state UIP program also requested to capture the wastage rates for the other routine vaccines simultaneously. This provided an opportunity to document the vaccine wastage for RVV (two different vial sizes) in two different geographic contexts and PCV under the UIP.

Background setting
Before national introduction of RVV in India in April 2016, a pilot introduction linked to surveillance of intussusception was planned. This pilot introduction was led by Department of Biotechnology in partnership with Ministry of Health and Family Welfare (MoHFW), monitored through an Interministerial-Interagency Coordination Group and funded by Bill and Melinda Gates Foundation. The RVV pilot introduction was launched in two districts Kangra (Himachal Pradesh) and Pune (Maharashtra) in December 2015, which preceded the national introduction (April 2016). The district Kangra continued to receive the RVV supply from the project for 2 years (till December 2017), although the RVV was launched in the state through national rollout in April 2016. The RVV was launched in Maharashtra state (and Pune district) through national rollout in July 2019. District Kangra with population 1.5 million had one district hospital (DH), one sub-district hospital (SDH), 13 community health centres (CHCs), 84 primary health centres (PHCs) and 440 sub-centers. In the Kangra district about 262 fixed sessions and 903 outreach sessions were organised every month. In Pune district, a selected part of the district covering 1 million population with a network of 29 PHCs and six rural hospitals (RH) was included. The immunization and vaccine logistics related health functionaries in these two districts and community mobilisers were trained for implementation through a cascade training system. The RVV was launched in December 2015 and routinely administered to the infants from January 2016. The birth cohorts targeted in these two districts were 41, 400 (Kangra 23,400 and Pune 18,000). Three doses of ROTAVAC® (Bharat Biotech) were given at 6, 10 and 14 weeks to the infants. Five doses vials of RVV were used till September 2016 and then 10 doses vials were used, similar to the national RVV supply guideline. The PCV was launched in Himachal Pradesh in May 2017 as part of the national phased introduction with two primary and one booster schedule. The PCV was administered to infants routinely under UIP from July 2017. IPV introduced under UIP underwent transition from 0.5 ml intramuscular to 0.1 ml intradermal (fractional IPV-fIPV) since January 2017. Availability of all these combinations in vaccine vial size, administration route across two different geographies (Kangra-hilly and Pune-plain regions) provided an opportunity to document the vaccine wastages. The state governments also requested to estimate the wastage of the other vaccines under UIP in the districts to inform the program.
The vaccine is supplied from district store to the subdistrict stores, cold chain points (CCPs) at CHCs and PHCs on monthly basis. From these CHCs/PHCs, for the sessions at sub-centres and outreach sites, vaccines are supplied using vaccine carriers on the day of session. The

Study design
This cross-sectional study was undertaken in two districts, Kangra (Himachal Pradesh) and Pune (Maharashtra). In each district, the facilities (all levels) providing vaccination and session sites (including outreach sessions) in both urban and rural areas were included to obtain a comprehensive picture of vaccine utilisation and wastage. In Kangra, the DH, 14 CHCS, 14 PHCs, 14 sub-centres were included. In Pune, the DH, five RHs, 4 CHCs, 10 PHCs and 20 sub-centres were included.

Data collection
These selected 49 facilities (vaccine stores and also service delivery point), 34 sub-centres (service delivery point) and 34 outreach session sites (one per sub-centre) were visited by the trained research staff for data collection. The vaccine supply, logistics and administration data for all UIP vaccines were collected for the period January 2016 to December 2017. Month and vaccine wise data were collected from the multiple registers (at vaccine stores: vaccine stock register-issue and return, vaccine distribution register for sessions and monthly report; at session sites: vaccinator's logistics diary and RCH register, session and monthly vaccination report) including sessions held, vaccines supplied and returned, vaccine utilisation, children vaccinated, and vaccines discarded or wasted using structured tools (Supplementary files 1 and 2). The interviews with cold chain handlers (CCHs) (n = 43), vaccinators (n = 31), store-in-charges (n = 23), medical officers (MOs) (n = 35) and district immunization officers (DIOs) (n = 2) were done to gain additional understanding the perceptions, practices and experience related to vaccine usage and wastage handling using semi-structured guide after obtaining informed consent (Supplementary file 3). Double data entry of the data collected was done using a MySQL and PHP based online data entry program followed by matching and corrections. The data was extracted into Microsoft excel format.

Data analysis
The descriptive analysis performed using Microsoft Excel 2013 and Stata 15.0. The wastage rate at store level and session site levels were estimated using the formula; Wastage rate = Doses wasted × 100 /Doses issued, expressed as percentage. For the stores the doses wasted for specific periods were estimated using the opening stock balance, new doses received, closing stock, and the children vaccinated. For the stores, all doses administrated at the sessions held at facility and outreach sites under the respective store were considered. For the outreach sessions, all doses administrated at the outreach session were considered. Wastage for each vaccine for every month for each facility, session site and store and for each different dose type (IPV and fIPV), vial size (5,

Frequency of doses administered during the vaccination sessions
According to the records, the number of doses administered per vaccination session varied widely by vaccine, which ranged between 0 to 45 doses per day. The distribution and dispersion of doses administered for different vaccines for two districts are shown in Fig. 1a and b).
Vaccine wastage at the outreach session level: The doses issued to the session sites and administered and vaccine usage and wastage rates for each vaccine not following open vial policy were calculated for the two districts ( Table 1).
The overall wastage rates for BCG and pooled RVV was higher in Kangra district than Pune. The vaccine wastage rates for the outreach sessions for various vaccines (RVV, Measles, MR and BCG) and pooled for the districts are summarised in Table 2. It was observed that while 45% of the outreach sessions in Kangra had < 25% wastage rates for vaccines, 80% of the Pune sessions had similar wastage level.

Vaccine wastage with transition of vial dosage sizes
At the cold chain point level, with transition from 5-to 10-doses of RVV vials, the wastage rate increased both in Kangra (29 to 33.2%) and Pune (17.8 to 25.7%) ( Table  3). The wastage rate for RVV at outreach session level increased with transition from 5-to 10-doses vial in Kangra (22.9 to 36.6%), while it declined in Pune (25.1 to 18%) ( Table 1).

Vaccine wastage with transition from IPV to fIPV practice
With transition of IPV dosage from 0.5 ml to fractional 0.1 ml (fIPV), the wastage rates at CCP level increased sizably in Kangra (36.1 to 54.8%) and marginally in Pune (18.4 to 26.9%) ( Table 3).

Vaccine wastage according to the vaccine types and vial sizes
The wastage rates for vaccines according to the type (liquid or lyophilised), reuse type (open vial policy), and vial sizes (number of doses) for oral and injectable routes for two districts for cold chain stores and outreach levels are summarised in Table 4. Generally the wastage rates for majority of the vaccines in Kangra district were higher than Pune, except for five dose (both oral and injectable) and 10 dose (only injectable) vials. The higher wastage rates in Pune district for five dose vials were due to measles and RVV. The lower wastage rate in Kangra district for 10 dose injectable vials was due to lesser wastage of MR vaccine at outreach sessions (Table 1). While the pooled wastage rates for liquid vaccines were higher than the lyophilized vaccines in Kangra, the reverse pattern was observed in Pune (Supplementary files 5 and 6). The wastage rated for single use vials and pooled oral vaccines were comparable for the two districts. For the five dose vials, the wastage was lesser in Kangra than Pune.

Perceptions and practices regarding vaccine wastage
Almost all health functionaries indicated that the vaccine wastage increased with transition of 5 to 10 doses vials of RVV and IM IPV to f-IPV in both districts. To minimise the wastage, the store-in-charges had been issuing the vials according to the anticipated beneficiary numbers. Majority of the vaccinators and CCHs from Kangra reported that the outreach sessions were rescheduled/ reorganised to minimise the wastages. The MOs and DIO were aware about the session rescheduling/reorganisations made by the vaccinators. Few vaccinators from sites with high beneficiary load, didn't find much difference in the wastage with the change in vial. The statements from various stakeholders are presented in the Supplementary document (Supplementary file 7).

Discussion
This report attempted to document the vaccine wastage status for the vaccines in UIP in Kangra and Pune districts representing two different geographies, the hilly and plain regions. The use of two different sizes of RVV vials provided an opportunity for documenting the change in wastage rates. Additionally the transition from regular to fraction dose of IPV in Kangra during the observation period was also captured. We estimated the wastage for UIP vaccines at CCPs and outreach session sites during 2 year period. This is the first report from India on wastage rate of RVV with both 5 and 10 dosage vial package versions. Additionally this is the first report from India on wastage for PCV and transition of IPV to fIPV dosage under UIP.
The present study documented higher wastage rates for several vaccines than the recommendation by MoHFW. In Kangra, the wastage rates for OPV and fIPV were above 50%, for most of the other vaccines except LPV were in the range of 26-50%. For most of the vaccines the wastage was >10% of the prescribed range, except Measles, LPV and RVV-5. In Pune, the wastage  The vaccine wastage levels at the cold chain points across the different months practices for storage, logistics and stock management at the CCPs. Rise in the wastage rate with transition from RVV 5 doses to 10 doses vial in both districts can be explained by the beneficiary case load. With rising institutional deliveries, fewer beneficiaries for BCG vaccine at outreach session level may explain little higher wastage. The wastage rates for vaccines under open vial policy like DPT, fIPV may be higher due to lower number of beneficiaries per session (outreach and facility based) within the usable window period of 4 weeks in the same session area. Despite higher wastage for fIPV (50 doses) dose practice, higher number of beneficiaries would have been vaccinated from the unit vial compared to IPV (10 doses) dose practice. Using the opened vaccine vial for sessions other than the session where it is opened may reduce the wastage of these multidose vaccine vials under open vial policy. Most of the stakeholders from both districts reported increased RVV wastage with change of vial size (5 to 10 doses vials). Recognising the increased wastage with 10 doses vial, the health functionaries had reorganised the sessions to increase the pooled beneficiaries number and reduce the wastage.
The available reports on vaccine wastage from outreach sessions, CCPs and immunization clinics of tertiary care facilities are summarised in Table 5. The wastage rates for Himachal Pradesh were higher compared to several other states. Compared to the earlier report, the wastage rates for most of the vaccines have declined during the current study period [2]. Similarly the wastage rates for several vaccines in Pune are lower than the earlier report from Maharashtra [2]. The wastage rates at outreach and cold chain store (including outreach sessions and facility sessions) were comparable as per the available reports.
The reported vaccine wastage rates from other developing countries varied widely. The vaccine wastage rates from Bangladesh (34.8 to 84.4%) are comparable to Indian figures [16].. The wastage rates in Nigeria were in the range of 21-35% at outreach sessions and 18-27% at the stores [17]. The wastage rates were lower in Gambia for most of the vaccines, except the BCG (54.9%) [18]. In Gambia the wastage of RVV (single dose vials) has been reported to be 5.2%, mostly due to expiry [18].
There have been several efforts to make immunization program cost effective through appropriate adjustment in vaccine logistics, operation including session planning, vial reuse policy, vial size package, and vaccinator practice [19]. It was estimated that with transition from 10to 5-dose vials of IPV the open vial wastage would be reduced by nearly half in all countries; Bangladesh by 56% (from 0.25 to 0.11), India by 53% (from 0.17 to 0.08), Mozambique by 53% (from 0.13 to 0.06), and Uganda by 44% (from 0.09 to 0.04) [20].
The two states (Himachal Pradesh and Maharashtra) are common with the Unicef vaccine wastage study in five states [2]. The vaccine wastage rates for most of the vaccines in both states have decreased overtime, which may be due to better organisation and logistics management, improved supervision and monitoring and open vial policy adoption. There has been wide variation in the vaccine wastage across the states and most of them have been from urban or tertiary care facilities. Limited information on vaccine wastage at outreach levels is available from different states in India to better understand the determinants. In our study, with transition from smaller to larger dose vial size, the wastage rate increases as documented for RVV and IPV in the current study. Reducing vial size has been an appealing strategy to reduce wastage, but cost-benefit analysis should be conducted considering the logistics and cold chain footprint, especially the costlier vaccines.
This study had several limitations. The study included only two districts and selected facilities in the districts. Some of the CCPs and outreach sessions had incomplete records, which may have introduced bias in the estimates. For the vaccines under open vial policy (DPT, LPV, HBV, TT, IPV, OPV and PCV), the doses returned from the session sites were not recorded. We used the doses issues and doses administered for estimation of the wastage, which

Conclusions
The study conducted in two different geography and population contexts in India documented that the vaccine wastage rates in both districts were relatively higher than the national program assumptions. The hilly district (Kangra) had relatively higher wastage for all vaccines compared to the plain district (Pune). The change in vial size for RVV from 5-to 10 doses increased the wastage in both districts. The transition from intramuscular IPV to intradermal fIPV has also increased the wastage rate, but the children vaccinated from one vial had also increased. The UIP in India face pressure to minimize the wastage and improve coverage with the available vaccine. With introduction of newer vaccines the coverage and vaccine wastage needs to be monitored for informing the program and operation. Close monitoring the wastage rates at different levels including the service delivery points; facilities with different case load and outreach levels in different geographies with various microplan types and stores at different levels. India provides opportunity for documenting the vaccine wastage across different contexts and lessons for appropriate programmatic action.