Development of an information system to evaluate loss (wastage)

Samia Abdul Samad Ministry of Health of Brazil Brasília, Distrito Federal, Brazil

and

Antonia Maria Teixeira, Brendan Flannery, Ricardo Gonçalves, Consuelo Freiria

ABSTRACT public health system, referred to as the Unified Health System (or SUS). Vaccine wastage may occur due to physical damage In 2006, an information system for monitoring of vaccine including breakage or because fewer vaccine doses are utilization (referred to as AIU) was created in Delphi, with administered during an session than an Access database. In 2010, the system was upgraded to included in multi-dose vials. Methods: We analyzed Java language, using a PostgreSQL database with an reported data from 2,553 centers distributed in interactive site in PHP (Oracle database). The AIU 600 municipalities for 4 : measles-mumps-rubella information system is based on data provided by individual (MMR), diphtheria-tetanus-pertussis-Hib (DTP-Hib), vaccination posts (point-of-use), and provides information Bacille Camille Guerin (BCG) and oral . for routine evaluation of doses received at the vaccination Vaccine costs were provided by the immunization post, doses contained in opened vials (utilization of program. Results: Mean vaccine wastage was 65.7% for vaccine vials), doses applied and doses discarded MMR (range, 46.1% to 72.4%), 23.9% for DTP-Hib (wastage) for 44 immunobiologicals (vaccines and (range, 10.3% to 32.6%), 74% for BCG (range, 64.4% to antisera), as well as reasons for vaccine wastage. Data 79.9%), and 3.2% for rotavirus (range, 1.3% to 4.8%). from the system were used to calculate costs of The ratio of doses in opened vials to doses administered vaccination. was almost 3:1 for MMR, 1.3:1 for DTP-Hib, 3.8:1 for BCG and nearly 1:1 for rotavirus. Conclusion: Multi-dose As part of system implementation, data were analyzed vials are often preferred by immunization programs for from the first four states to begin using the AIU system requiring less space for transport and but (Amazonas [AM], Mato Grosso do Sul [MS], Rio Grande may result in higher cost per dose administered. do Norte [RN], Santa Catarina [SC]) to estimate the Calculation of vaccine wastage is an important component prevalence and causes of losses at the three levels of the of planning immunization program requirements; incorrect immunization program. The objective of this analysis was estimates can result in purchasing too few or too many to compare information on the movement of vaccines doses of vaccines, resulting in stock-outs or large obtained from four states in Brazil that have different quantities of expired vaccines. characteristics, and to evaluate the influence of these characteristics on vaccine utilization. The evaluation was Keywords: vaccines, wastage, utilization supported by Brazil’s National Immunization Program, which is responsible for purchase of recommended vaccines and delivery to state immunization programs. 1. INTRODUCTION More accurate data on vaccine utilization is important for planning, procurement and distribution. With the expansion and increasingly high profile of Brazil’s National Immunization Program (abbreviated PNI), information was needed to improve management of 2. OBJECTIVES vaccine supplies and control the movement of vaccines throughout the country. PNI prioritized the development of To evaluate and use data from the AIU system in the first an information system that would identify the percentage four Brazilian states to adopt the system, for calculation of of vaccine doses distributed that were acutally the prevalence and reasons for wastage of four administered, with the objective of estimating financial recommended vaccines provided by Brazil’s national costs involved in vaccine supply. The information system immunization program: MMR, DTP-Hib, BCG and oral needed to provide information on vaccine wastage at all rotavirus vaccine. three levels (national, state and municipal) in Brazil’s

3. METHOD

We abstracted AIU fields corresponding to the type of vaccine and presentation (number of doses per vial), vaccine stocks, opened vials (utilization), doses administered, physical losses (discarded, unopened vials) and technical loss (discarded doses in opened vials). Data were considered valid if the number of doses administered was equal to or less than the number of doses in opened vials, if the total number of vaccine doses utilized was equal to doses administered plus losses, and if doses registered in the AIU system were Mean vaccine wastage was 65.7% for 10-dose vials of equal to the number of doses administered at the MMR vaccine (range, 46.1% to 72.4%), 23.9% for five- vaccination post (according to the monitoring system for dose vials of DTP-Hib (range 10.3% to 32.6%), 74% for vaccination coverage, or API). We calculated the 10-dose vials of BCG (range 64.4% to 79.9%), and 3.2% percentage of technical loss according to the following for single dose vials of oral rotavirus vaccine (range 1.3% equation: to 4.8%).

100 x (n doses in opened vials) – (n doses administered) Table 1. Prevalence of losses by type of vaccine and UF in n doses in opened vials 2008

Physical loss is defined as the total number of discarded doses in unopened vials due to expiration, breakage and conservation outside recommended temperatures. Brazil does not use Vaccine Vial Monitors on vaccine vials to determine whether vaccines have been damaged by exposure to high or low temperatures.

Costs of wastage and costs of vaccine per dose administered were obtained by multiplying vaccine losses by the per-dose cost of vaccine vials, excluding storage and transportation costs. The per-dose cost of each vaccine in 2008 was US$3.50 for MMR in 10-dose vials, US$3.25 for DTP-Hib in 5-dose vials, US$0.31 for BCG in 10-dose vials and US$9.22 for oral rotavirus vaccine in single-dose vials.

4. RESULTS

We received data from 2553 registered vaccination posts in 600 municipalities. Non-valid data were excluded for Source: SI_AIU, MS 1254 vaccination posts for BCG vaccine, 668 posts for DTP/Hib, 716 for oral rotavirus vaccine and 648 for Legend: SRC (measles mumps and rubella); DTP+ Hib (diphtheria, MMR. For the four vaccines, and average of 1731 tetanus, pertussis and haemophilus influenzae tipo b); BCG (Bacilo vaccination posts had valid data for evaluation. Of these, Calmett Guérin); VORH (rotavírus). 502 (30%) vaccination posts submitted data for twelve months (January to December) in 2008 and 1229 (70%) submitted valid data for one or more months (range, 1 to The ratio of doses in opened vials to doses administered 11). was approximately 3:1 for MMR, 1.3:1 for DTP-Hib, 3.8:1 for BCG and 1:1 for oral rotavirus vaccine. For MMR, the ratio indicates that for opened, 10-dose vials, Figure 1: Flow rooms and selection of study vaccines two doses were discarded for each dose administered (or child vaccinated).

Table 2: Loss technique, due to the doses used doses technical losses and expenses, by vaccine, state, year 2008 physical loss. Physical loss due to problems with maintaining recommended temperatures in the cold chain had the lowest frequency. For DTP-Hib, physical losses due to interruptions in electricity and temperature changes were most common reasons in SC (50.7%) and MS (26.4%), expiration was most common in AM (36.2%) and “other reasons” were most common in RN (40.7%). For BCG, expiration of vaccine was the most common reason for physical losses in RN (65.2%) and AM (88%). In addition to low frequency of loss due to problems with transportation of vaccines within recommended temperatures (0% - 0.4%), inadequate procedure was an infrequent cause of wastage for BCG (0.3% - 3.5%). Physical losses of oral rotavirus vaccine were mainly due to lack of electricity (18% - 47.8%), expiration (38%) and refrigeration equipment failure (10.8% - 20.6%). In Source: SI_AIU, MS addition to vaccine wastage in opened multi-dose vials, principal reasons for vaccine loss were refrigeration Legend: SRC (measles mumps and rubella); DTP+ Hib (diphtheria, problems including interruptions in the supply of tetanus, pertussis and haemophilus influenzae tipo b); BCG (Bacilo electricity and past expiration dates. Calmett Guérin); VORH (rotavírus). Table 3: Proportional distribution of physical losses due to During the period of evaluation, a total of 150,000 vials of vaccines and state in the year 2008 MMR were opened (for a total of 1.5 million doses of MMR vaccine) in the four state immunization programs, while 550,000 MMR doses were administered. The actual vaccine cost per dose administered was US$10.06. The total expenditure for MMR vaccine in the four states was US$5.4 million versus costs in an ideal scenario without wastage of US$1.8 million (excluding costs of vaccine storage and transportation), a difference of US$3.6 million.

For 5-dose vials of DTP-Hib, cost per dose administered was US$4.40, which was US$1.15 more than the per-dose purchase price. Actual expenditures based on the number of DTP-Hib doses administered was US$1.0 million in excess of the ideal per-dose cost assuming no wastage (excluding storage and transportation costs). For BCG, the cost per dose administered was US$1.22. Of a total number of 770,000 doses in opened vials, only 200,000 doses were applied, resulting in a difference between ideal and actual spending of US$181,000. For oral rotavirus Source: SI_AIU, MS vaccine, actual per-dose costs were similar in the four states, with approximately US$0.50 difference between Legend: SRC (measles mumps and rubella); DTP+ Hib (diphtheria, ideal and actual costs, resulting in a total excess tetanus, pertussis and haemophilus influenzae tipo b); BCG (Bacilo expenditure of US$100,000. In the four states evaluated, Calmett Guérin); VORH (rotavírus). expenditure for the four vaccines totaled 20 million QF (loss of bottle breakage);EE (loss due to power outages); FE (loss Brazilian reais (~US$10 million), while expenditure due to equipment failure refrigeration); VV (loss lost validity); PI (loss whithout wastage would have been US$4.4 million. due to inadequate procedures); FT (loss for failure to transport) e OM (loss for other reasons than the above described). In relation to physical losses (discarded, unopened vials), the main reasons varied among the four states. For the states of Santa Catarina and Amazonas, physical losses for 10-dose vials of MMR resulted from lack of electricity 5. CRITICAL FACTORS (35.7% and 41.2% of physical losses, respectively). In Mato Grosso do Sul and Rio Grande do Norte, the most This analysis is subject to several limitations. Valid data common reason provided for physical loss of 10-dose were not provided from all vaccination posts, suggesting MMR vials was “other reasons” (38.6% and 40.4%, problems with the use of the information system or respectively), suggesting confusion about the concept of confusion regarding the concepts. Many vaccination posts did not send complete data throughout the entire vaccine wastage were not monitored to identify and correct period evaluated. The evaluation was only performed problems. The AIU represents a breakthrough in with data from four Brazilian states and is not controlling the movement of vaccines from central stores representative of all 27 state immunization programs. to the point-of-use at vaccination posts, enabling better Frequency of wastage and reasons for physical losses management and site assessment in the central cold chain. may vary between reporting and non-reportings sites within the states, as well as between state immunization Brazil’s National Immunization Program seeks an programs. appropriate use of vaccines and minimal losses, to reduce costs while expanding access to . Data from 6. CONCLUSION the AIU system in four states demonstrate the utility of the system for estimating the frequency of vaccine wastage and its causes. Ongoing evaluation of data from the AIU This was the first evaluation of vaccine utilization and system will provide input for planning vaccine losses to be conducted by Brazil’s National Immunization requirements for production and procurement, as well as Program. The creation of an information system for this for distribution of these products. purpose made it possible to estimate the frequency of vaccine wastage and its causes at the point-of-use.

Results of this evaluation showed that vaccine wastage due to technical losses (discarded doses in opened, multi-dose vials) were greated than expected. Wastage rates in the system are related to vial volume and technical specifications for use of multi-dose vials—i.e. how long a multi-dose vial may be used after opening. Prior to implementation of the AIU system, vaccine wastage was estimated by subtracting the number of doses administered from the number distributed. However, frequency of vaccine wastage in opened, multi-dose vials were almost two-times higher than estimates of wastage used by PNI for procurement of BCG and MMR. For DTP-Hib, estimated wastage from the AIU evaluation was similar to previous estimates used by PNI for vaccine procurement. Despite having the lowest wastage of the vaccines evaluated, the wastage of oral rotavirus vaccine was concerning due to the use of single-dose vials and high vaccine cost, indicating the need for better training and monitoring. Results of the AIU evaluation provided specific knowledge about vaccine utilization at public health care centers, which may inform decisions regarding vaccine supplies in the future.

Results show that for multidose vaccines with limited shelf lives after opening (BCG and MMR), technical losses due to discarded doses in opened vials were much greater than physical losses due to breakage, expiration of vaccine and temperature fluctuations (loss of cold chain). Two doses from opened vials are discarded for each one dose administered. On the other hand, multi-dose vials are preferred for use in the public network for facilitating distribution, reducing storage space requirements and assocciated costs. Losses due to transportation problems were minimal. Data from the AIU system will contribute to analyses that consider all the cost implications of vaccine formulations on distribution, cold chain capacity and wastage.

Previously, it had been assumed that not all vials distributed were opened, and that physical losses (discarded, unopened vials) predominated. The quantity of doses discarded for technical reasons and reasons for