Investigación original / Original research

Economic impact of a vaccination program in Mexico

Dagna Constenla,1 F. Raúl Velázquez,2 Richard D. Rheingans,1 Lynn Antil,1 and Yolanda Cervantes 3

Suggested citation Constenla D, Velázquez FR, Rheingans RD, Antil L, Cervantes Y. Economic impact of a rotavirus vac- cination program in Mexico. Rev Panam Salud Publica. 2009;25(6):481–90.

ABSTRACT Objectives. To evaluate the cost and benefits of a national rotavirus childhood vaccination program in Mexico. Methods. A decision-analysis model was designed to take the Mexican health care system’s perspective on a comparison of two alternatives: to vaccinate against rotavirus or not. Us- ing published, national data, estimations were calculated for the rotavirus illnesses, deaths, and disability-adjusted life years (DALYs) that would be averted and the incremental cost- effectiveness ratios (US$/DALY) of a hypothetical annual birth cohort of 2 285 000 children, with certain assumptions made for cost, coverage, and efficacy rates. Results. With 93% coverage and a vaccine price of US$ 16 per course (2 doses), a rotavirus vaccination program in Mexico would prevent an estimated 651 deaths (or 0.28 deaths per 1 000 children); 13 833 hospitalizations (6.05 hospitalizations per 1 000 children); and 414 927 outpatient visits (182 outpatient visits per 1 000 children) for rotavirus-related acute gastro- enteritis (AGE). Vaccination is likely to reduce the economic burden of rotavirus AGE in Mex- ico by averting US$ 14 million (71% of the overall health care burden). At a vaccine price of US$ 16 per course, the cost-effectiveness ratio would be US$ 1 139 per DALY averted. A re- duction in the price of the rotavirus vaccination program (US$ 8 per course) would yield a lower incremental cost-effectiveness ratio of US$ 303 per DALY averted. Conclusions. A national rotavirus vaccination program in Mexico is projected to reduce childhood incidence and mortality and to be highly cost-effective based on the World Health Organization’s thresholds for cost-effective interventions.

Key words Rotavirus, rotavirus vaccines, cost-benefit analysis, mass immunization, gastroenteri- tis, Mexico.

Rotavirus is a virus that infects the various factors, including limited access 1 Department of Global Health, Rollins School of Public Health, Emory University, Atlanta, Georgia, small intestine and can lead to life- to health care facilities, lower mean age United States of America. Send correspondence to: threatening acute gastroenteritis (AGE) at contraction, and malnutrition. Due to Dagna Constenla, Ph.D., Department of Global in infants and young children world- the widespread nature of rotavirus, its Health, Rollins School of Public Health, Emory University, Atlanta, GA, USA; telephone/fax: +01- wide. Unlike many other enteropatho- high concentration in the feces, and its 970-266-1153; e-mail: [email protected] gens, rotavirus affects children in both ability to survive in the environment, the 2 Unidad de Investigación Médica en Enfermedades developed and developing countries, prevention of rotavirus spread is virtu- Infecciosas y Parasitarias, Hospital de Pediatría, Centro Médico Nacional Siglo XXI, Instituto Mexi- regardless of socioeconomic status (1, ally impossible, even where the highest cano del Seguro Social, Ciudad de México, DF, 2). However, the risk of dying from hygiene standards are followed (2). México. 3 GlaxoSmithKline Biologicals S.A., Mexico City, rotavirus-related gastroenteritis (GE) is In Mexico, rotavirus is one of the prin- Mexico. greater in lower-income countries due to cipal causes of GE and occurs mainly

Rev Panam Salud Publica/Pan Am J Public Health 25(6), 2009 481 Original research Constenla et al. • Cost-effectiveness of rotavirus vaccination in Mexico during the fall and winter months. It is of FIGURE 1. Model framework depicting rotavirus (RV) vaccine introduction and subse- particular concern in pediatric hospitals, quent events that may be experienced by each child, including life-years gained (LYG) emergency rooms, and outpatients clin- and disability-adjusted life years (DALYs) ics. However, improved availability of Epidemiology potable water, access to oral rehydration Health Benefits therapy, and enhanced surveillance of • RV rates • RV proportion • Events all diarrheal episodes have reduced the • Healthcare utilization • Deaths overall burden of GE in young children • LYG Vaccine • DALYs (3). Morbidity rates associated with all • Coverage diarrheal diseases decreased from 4 588 • Efficacy per 100 000 children under 5 years of age • Cost in 1999 to 3 383 by 2002, with the great- est reduction observed among those under 1 year. There has also been a sig- Cost- nificant decrease in severe diarrheal Economic Burden Effectiveness episodes among children under 5 years • $ per cohort • $/DALY averted of age, from 24.8 per 100 hospitalizations • Medical cost • $/deaths averted • $/LYG in 1990 to 9.2 in 1995. This represents a • Patient cost 63% reduction (3) and is the most accu- Economics Net Cost rate information to date. Overall mortal- • Hospital per diem Savings • Visit costs • $ per cohort ity due to AGE in children under 5 years • Direct costs also decreased from 125.3 per 100 000 in • Parent expenses 1990 to 19.7 by 2002, an 84% reduction (3). From 1990–2002, a significant reduc- tion in mortality was observed in April– August, from 13.4 per 100 000 children effectiveness of rotavirus vaccination is DALY estimates were discounted at a under 5 years in 1990 to 1.6 in 2002 (3). shown in Figure 1. The boxes identify the 3% rate, as recommended by the U.S. While overall GE morbidity and mor- primary inputs needed for the model Panel of Cost-Effectiveness in Health tality have declined, the proportion at- (epidemiological, economic, and vaccine and Medicine (7) and the World Bank tributable to rotavirus appears to be data). The circles represent the intermedi- Global Burden of Disease Project (8). increasing (3, 4). The highest mortality ary outputs (health benefit, economic The annual birth cohort considered rates for children under 1 year coincided burden, and net cost savings). The pri- was 2 285 000 children (9). The age dis- precisely with the fall and winter season. mary outputs of the model are displayed tribution of rotavirus illness was esti- A retrospective, cross-sectional study of in hexagons as cost-effectiveness ratios. In mated for each one of the disease out- 15 large cities in Mexico reported that, this study, these ratios are expressed as comes (deaths, hospitalizations, and annually, 42% of the children seeking the cost per death averted, per disability- outpatient visits) using published stud- AGE treatment at pediatric hospitals adjusted life years (DALYs) averted, or ies from Mexico and Latin America (10– were rotavirus positive (3). per life year gained (LYG). This diagram 20). The estimated number of outcomes This paper presents detailed analyses provides only a broad representation of for an annual birth cohort was distrib- of the disease burden and costs of the inputs and outputs of the model. uted as follows: 0–2 months of age, 3–5, rotavirus disease among Mexican chil- A decision-analysis model was devel- 6–8, 9–11, 12–23, 24–35, 36–47, and 48–59. dren, and the expected cost-effectiveness oped in Microsoft Excel® to estimate the of introducing a national rotavirus vacci- health and economic outcomes of a na- Model inputs nation program. These results are part of tional rotavirus vaccination program in a a larger study being carried out in Latin hypothetical annual birth cohort of chil- Rotavirus disease burden. Disease bur- America and the Caribbean, specifically dren for a 5-year period. The net, med- den was estimated as the expected num- Argentina, Brazil, Chile, Dominican Re- ical, direct-cost of the vaccination pro- ber of rotavirus-associated outcomes public, Honduras, Mexico, Panama, and gram was determined to be the cost of for a single birth cohort. The risk of Venezuela (5, 6). In contrast to this Re- the vaccination program, minus the ex- rotavirus-related outpatient visit, hospi- gional study, in which estimates were pected averted medical costs (the sav- talization, and death were based on the derived from extrapolated data, this ings from reduced use of health care cumulative risk of each AGE event dur- paper focuses solely on Mexico and uses resources). ing the first 5 years of life and the pro- country-level epidemiological data and The primary perspective for this portion of these events attributed to ro- cost estimates. analysis was that of the health care sys- tavirus. The cumulative risk of each tem. Costs such as non-medical direct event in Mexico was estimated from na- MATERIALS AND METHODS costs or productivity losses to caregivers tional administrative data for the year were not included in these cost-effective- 2002 (21, 22), and refers to the proportion Analytic overview ness calculations due to paucity of data. of children in a defined population in- All cost estimates collected in local fected with rotavirus at least once before A diagram of the model used to esti- currency were converted to 2007 United reaching their 5th year of life. Risk refers mate the economic burden and cost- States dollars (US$). Future costs and to the proportion of a population for an

482 Rev Panam Salud Publica/Pan Am J Public Health 25(6), 2009 Constenla et al. • Cost-effectiveness of rotavirus vaccination in Mexico Original research event and cumulative is the risk for each The cumulative risk of a rotavirus- 2004 and made available to the private year of age multiplied together. related death was based on 2002 national sector in January 2005. In March 2007, In brief, the cumulative risk of a rota- data from the Instituto Nacional de Es- the vaccine was incorporated into the virus outpatient visit was calculated tadística Geografía e Informática (Na- routine immunization schedule, first by using data provided by the epidemiol- tional Institute of Statistics, Geography, the IMSS and later, in August 2007, by ogy unit of the Instituto Mexicano del and Informatics, INEGI) (4). Approxi- the Ministry of Health. Vaccine coverage Seguro Social (Mexican Social Security mately 2 157 deaths due to diarrheal dis- has not yet reached the expected 90%. Institute, IMSS). The IMSS reported a eases occurred annually in children less The efficacy of the human rotavirus total of 950 231 emergency room/outpa- than 5 years of age. By dividing these vaccine to prevent rotavirus-related acute tient visits for diarrhea among children (2 157) by the population less than 1 year GE (requiring hospitalization) was 85% less than 10 years of age, of which an es- of age (2 056 853), the cumulative risk of (32). Efficacy against outpatient visits timated 93% were in children less than a death due to AGE was calculated to be was assumed to be 78%. This represents 5 years (22). In addition, in 2002, the Sta- 1.05 deaths per 1 000 live births. Since the midpoint of 85% efficacy against tistics and Information Division of the diarrheal mortality data were not spe- rotavirus AGE (32) and 70% efficacy Secretaría de Salud/Servicio Social de cific to rotavirus, the proportion attribut- against any rotavirus GE (33). Efficacy Salud (Secretary of Health and Social able to rotavirus was assumed to be the against death was assumed equal to that Services, SSA), reported 974 129 outpa- same as the proportion of hospitaliza- of the most severe cases requiring hospi- tient visits for the same age group (21). tions attributed to rotavirus (3, 28). The talization (31). For the baseline analysis To calculate the 5-year cumulative risk cumulative risk of death due to rotavirus it was further assumed that one dose of of an outpatient visit for diarrhea, the AGE (0.42 per 1 000) was calculated by the vaccine would have the same effec- total number of diarrheal outpatient multiplying the risk of death due to GE tiveness as a full course during the in- visits for children less than 5 years by the proportion of GE deaths attrib- terdosing period, as demonstrated by a (883 715 + 974 129 = 1 857 844) was di- uted to rotavirus. clinical trial in Latin America (32). For vided by the proportion of Mexico’s total the baseline analysis, it was further as- under-1 year population (2 056 853). The Disability-adjusted life years (DALYs). sumed that effectiveness would not de- rotavirus-specific outpatient rate was The disease burden of morbidity and cline in the second and subsequent years then calculated assuming that 25% of di- mortality was also estimated in terms of following vaccination. arrheal outpatient visits were attributed DALYs (8). The DALY estimate included It was assumed that children would to rotavirus. This was an average esti- Years of Life Lost (YLL) due to prema- receive the at the same mate based on data from five published ture mortality and Years Lived with Dis- time as the , tetanus, acellular studies (23–26). The cumulative risk of ability (YLD). The DALY loss from mor- pertussis, , inacti- an outpatient visit for rotavirus AGE tality (YLL) was calculated based on the vated vaccine (DTPa + Hib + IPV), (265 per 1 000) was calculated by multi- average country-specific life expectancy and the vaccine (HBV), which plying the cumulative risk of an outpa- at birth (29). The average life expectancy in Mexico are given at 2 months, tient visit for AGE by the proportion at- for males and females is 73.6 and 74.4 4 months, and 6 months of age. In the tributed to rotavirus GE. years, respectively. Only morbidity from baseline analysis, national coverage of The cumulative risk of rotavirus AGE disease severe enough to require med- the third dose of DTP at 1 year of age for hospitalization was based on national ical care was considered for the calcula- the year 2003 was estimated to be 92.8% data for the year 2002 (21, 22), using tion of YLD. Default disability weights (34). Rotavirus vaccination would occur codes A00–A09 of the International Classi- from the global burden of disease study with DTPa + Hib + IPV doses one and fication of Diseases, 10th Edition (27). The (8) and the World Health Organization’s two; however, standardized data are SSA reported 15 513 hospitalizations for guidelines for cost-effectiveness studies only available for the coverage of the diarrhea among children less than 5 (30) and estimated rotavirus illness du- third dose (Table 1). As a result, this may years of age. In addition, the IMSS re- ration of six days (31) were used to cal- be an underestimation of the actual cov- ported a total of 16 347 diarrheal hospi- culate YLD. A discount rate of 3% and erage for doses one and two. talizations for children less than 10, of age weighting were included (30) to en- The effectiveness of vaccination was which 93% occurred in children less than sure comparability. assessed by combining information on 5. The number of children less than 5 the expected vaccine coverage and tim- years of age hospitalized for diarrhea Rotavirus vaccine efficacy and cover- ing of illness. An annual birth cohort was (30 716) was then divided by the popula- age. Estimates of the effectiveness of a followed through the same age periods tion less than 1 year (2 056 853) covered complete vaccination course for rota- previously described. During each pe- by both health care systems. The cumu- virus disease were based on the results riod, the number of outcomes in the ab- lative risk of diarrheal hospitalization of the phase III clinical trial of the live- sence of vaccination was estimated. The was 0.0176 per child by the 5th birth- attenuated human rotavirus vaccine reduction in the number of outcomes day. The rotavirus proportion (40%) was (GSK Biologicals S.A., Rixensart, Bel- during the same period was estimated based on published estimates (3, 28). The gium) that was given as two oral doses at based on the proportion of children with cumulative risk of hospitalization for ro- 2 and 4 months of age. This was a multi- one or two doses, and the efficacy of one tavirus AGE (7 per 1 000) was then cal- center trial conducted in 11 countries of or two doses. The expected number of culated by multiplying the cumulative Latin America, including Mexico. events averted prior to the 5th birthday risk of AGE hospitalization by the pro- The human attenuated rotavirus vac- was calculated as the sum of outcomes in portion attributed to rotavirus AGE. cine was first licensed in Mexico in July each period. The global efficacy of the

Rev Panam Salud Publica/Pan Am J Public Health 25(6), 2009 483 Original research Constenla et al. • Cost-effectiveness of rotavirus vaccination in Mexico vaccine against severe rotavirus GE was based on geographical region and mortal- was representative, the mean length-of- also considered. For this reason, infor- ity stratum. The per-diem and visit cost- stay was based on the average of 2.8 days mation on specific rotavirus serotypes estimates include items such as the cost (32) and 4 days (derived from physician was not considered in the analyses. of the facility, personnel, equipment, and interviews, 2003). food, and excludes items such as medica- The cost-per-outpatient visit (US$ 14.99) Vaccination costs. All costs are ex- tions and diagnostic tests (40). All prices was calculated as the mean cost of visit- pressed in 2007 U.S. dollars. Vaccination were adjusted for inflation (41). A hospi- ing a pediatrician, emergency room, or costs included the cost of administering tal per-diem rate was estimated as an av- rehydration room based on the propor- the vaccine, the price of each dose, the erage of estimates for public primary, sec- tion seen in each of three outpatient number of doses given (based on cover- ondary, and tertiary-level facilities, using settings (derived from physician inter- age level), and the expected waste losses the proportions from the multicountry views, 2003). Approximately 30% of pa- from the vaccination program (10%). The study (42). The cost of hospitalization tients with rotavirus AGE visit a pedia- costs for administering the vaccine in- was calculated as the product of the per- trician; 30%, an emergency room; 40%, a cluded the cost of health care personnel diem cost and a duration of 3.5 days (25, rehydration room. Costs for these three and training, cold chain, storage space, 43, 44). Country-specific outpatient visit types of outpatient visits were based on and public education. Mexico has the re- costs were developed by adjusting the the average given by the Hospital Gen- quired infrastructure for running a na- WHO-CHOICE regional estimate pro- eral de Tlalnepantla. tional rotavirus vaccination program, portionally to the country-specific hospi- Costs associated with diagnostic test- based on its extensive history of admin- tal per-diem cost. ing for rotavirus AGE were derived from istering oral polio vaccine. For this rea- Six physicians were interviewed to estimates provided by the finance de- son, the incremental administration costs characterize the typical management of partment of the Hospital General de are assumed to be low. A few studies rotavirus AGE. These physicians repre- Tlalnepantla. The total diagnostic cost- estimate the cost of immunization for sented the public and private sectors and per-patient was calculated by taking the the current WHO-recommended vaccines the IMSS. Interviews included questions number of times a particular test is per- (35–38); however, no data was found on regarding care for patients with moder- formed and multiplying it by its unit the incremental cost of adding another ate and acute rotavirus GE. cost, and then summing the total cost of vaccine. Based on the range of estimates For the present analysis, resource uti- all the tests performed on a single pa- found in the immunization cost studies lization data were based on the surveil- tient. A mean diagnostic cost was calcu- and the assumption that the incremental lance study that was conducted in three lated for both, inpatients (US$ 6.27) and cost of administering an additional vac- Mexican cities (32), since these were outpatients (US$ 4.95). The most common cine would be low, the model uses a cost considered to be the most reliable for diagnostic tests reported were blood of US$ 1 per course. The public sector Mexico. This surveillance study col- counts, electrolytes, blood gas, and stool price per course of the vaccine (2 doses) lected resource utilization and cost data specimens. was assumed to be US$ 16. in January–June 2003, which covers the Costs associated with medications The vaccine price is a key variable in months with the highest rotavirus inci- were based on estimates provided by the the cost-effectiveness estimates. For this dence (December–January) and with the pharmacy at the Hospital General de Tlal- reason, vaccine prices from US$ 8–22 lowest (March–August) (32). Seven hos- nepantla. The total medication cost per per course were assessed in sensitivity pitals participated: three in the state of patient was calculated by multiplying the analysis. The cost of adverse events re- Mexico (two tertiary level and one sec- cost per dose of each medication used sulting from the vaccine were not in- ondary); two in Morelos (both tertiary by the number of doses given and then, cluded because its safety profile is simi- level) and two in Durango (one tertiary summing the total cost for all the medica- lar to a placebo, therefore the cost, if any, level and one secondary level). tions given to a single patient. A mean would be negligible (32, 33). medication cost was calculated for both Health care costs. For this analysis, costs inpatients (US$ 9.63) and outpatients Cost-generating outcomes. The eco- associated with episodes requiring hos- (US$ 7.29). Medications included intrave- nomic burden of rotavirus disease in pitalization and outpatient visits were nous rehydration fluid, antiemetic, anti- Mexican children was determined by considered. Health care costs were cal- fever, and antibiotic therapies. The input combining the estimated number of each culated as the sum of the per-diem or variables used to perform the base case outcome with information on the costs per-visit cost, the cost of diagnostic tests, analysis are described in Table 1. associated with each. Cost generating and the cost of medication. The hospital outcomes were estimated based on three per-diem cost (US$ 57.45) was estimated Sensitivity and uncertainty analyses potential sources: the WHO CHOosing based on one public health care institu- Interventions that are Cost Effective tion, the Hospital General de Tlalne- For the present analysis, one-way sen- (WHO-CHOICE) project (39), physician pantla, which was considered to be rep- sitivity analysis of rotavirus disease-re- interviews, and a multicenter hospital- resentative of the public health sector lated and cost inputs was performed for based rotavirus surveillance study done and provided its actual per-diem cost. different scenarios that are likely to in- prior to a randomized vaccine trial (32). The cost-per-stay was calculated by mul- fluence the cost-effectiveness of vaccina- The WHO-CHOICE project provides tiplying the per-diem rate, which in- tion (45). These scenarios included high- estimates of the per-diem cost of hospi- cluded hotel costs, food, and non-allied and low-end estimates of vaccine price, tals, outpatient visits, and health center health personnel, by the mean length-of- disease incidence, and disease-related visits for 14 epidemiological categories stay (3.4 days). To ensure the estimate costs. Sensitivity analysis was performed

484 Rev Panam Salud Publica/Pan Am J Public Health 25(6), 2009 Constenla et al. • Cost-effectiveness of rotavirus vaccination in Mexico Original research

TABLE 1. Input variables for a cost-effectiveness analysis of a rotavirus (RV) vaccination program in Mexico

Variables Baseline estimate Data source

Demographics Birth cohort 2 285 000 PAHO, 2004 (9) Average life expectancy at birth (years) Male 73.6 WHOSIS, 2000 (28) Female 74.4 Cumulative risk of outcomes for RV gastroenteritis per 1 000 children < 5 years Outpatient visit rate 265 SSA, 2002 (21); IMSS, 2002 (22); extrapolation of five published studies (23–26) Hospitalization rate 7 SSA, 2002 (21); IMSS, 2002 (22); Velázquez et al., 2001/4 (3, 27) Mortality rate 0.42 SSA, 2002 (21); assumed RV proportion is the same as hospitalized Direct medical cost of RV gastroenteritis per patient a Ambulatory cases US$ 27.3 Financial department of Hospital General de Tlalnepantla; physician interviews; Hospitalized cases US$ 174.2 information not published Vaccine Efficacy (RV outpatient visits) 78% Ruiz-Palacios et al., 2006 (31); De Vos et al., 2004 (32) Efficacy (RV hospitalization) 85% Ruiz-Palacios et al., 2006 (31) Efficacy (RV death) 85% Assumes efficacy is same as for hospitalizations (31) Coverage 93% CONAVA, 2003–4 (33) Administration cost (per course) US$ 1.0 Author’s assumption Price per course (2 doses) US$ 16.0 Author’s assumption a These costs include the direct medical costs of hospitalization, outpatient visits, diagnostic tests, and medication, and are expressed in 2007 US dollars.

using the current price of US$ 16 per TABLE 2. Breakeven costs of rotavirus vaccine in Mexico using alternative criteria (best estimate course. and range) in 2007 US$ Additional results from the phase III clinical trial conducted in Latin America US$ / dose (32) suggest that the human rotavirus vac- Best estimate 95% CI cine may reduce AGE of all causes, not Breakeven cost limited to rotavirus, by 42% within the Direct, medical price 5.10 4.11–6.66 first year of life. Using age-distribution Retail price (direct, indirect costs) 6.24 5.49–8.47 data for diarrhea disease, the vaccine effi- Cost effectiveness cacy against all causes of AGE was con- gross national product 64.37 50.78–78.94 sidered as a separate scenario in the sensi- $1 000 (expanded oral rehydration therapy, water) 11.74 10.02–14.46 tivity analysis. $550–800/DALYa (World Bank) 9.15 7.86–11.49 An uncertainty analysis was con- $175/DALY (WHO low income) 5.16 4.44–7.07 b ducted to help evaluate the potential im- $75/DALY (HepB, Hib) 4.36 3.69–6.26 $30/DALY (Expanded Program on Immunization) 4.00 3.34–5.90 pact of this uncertainty on preliminary quantitative estimates and to assess the a DALY: disability-adjusted life year. b need for additional data collection. The HepB: Hepatitis B; Hib: Haemophilus type b. specific purposes included:

• assessing the range of likely outcomes lected from each of the distributions in tion attributable to rotavirus are addi- of immunization based on the prelim- a series of iterations, and results are cal- tional key contributors to uncertainty. inary quantitative analysis; culated and stored. A new set of input • identifying specific variables that are values are then randomly selected and RESULTS likely to be important and may merit a new set of results are calculated and additional data collection; and, stored. The process is repeated for 10 000 Disease burden and risk of rotavirus • providing a preliminary framework iterations. The final product is a distri- outcomes for developing confidence intervals bution of potential outcomes that de- for final national estimates. scribe the likely range of actual expected The estimated disease burden of ro- results. tavirus with and without a vaccination A Monte Carlo model was developed The output ranges developed through program in Mexico is presented in Table based on the preliminary model of rota- the uncertainty analysis are included in 3. Each year, due to rotavirus AGE, 1 in virus burden and immunization cost- Table 2. In addition, the uncertainty 4 children less than 5 years of age (265 effectiveness described elsewhere (5, 6). analysis allows us to identify the para- per 1 000) required a clinic visit; 1 in 140 In the Monte Carlo analysis, individual meters that contribute the most to uncer- (7 per 1 000) was hospitalized; and 1 in point estimates of parameters are re- tainty in the outcome estimates. For the 2 500 (0.4 per 1 000) died. An estimated placed with distributions of potential val- cost-effectiveness analysis, uncertainty 14 DALYs per 1 000 births were lost to ues. Individual values are randomly se- about diarrheal mortality and the por- these outcomes. For the base case, the

Rev Panam Salud Publica/Pan Am J Public Health 25(6), 2009 485 Original research Constenla et al. • Cost-effectiveness of rotavirus vaccination in Mexico

TABLE 3. Estimated disease burden of rotavirus in Mexico with and without a rotavirus mated that rotavirus AGE resulted in a vaccination program total health care cost of over US$ 19.6 million for each annual birth cohort, Events per Total annual events 1 000 children equivalent to US$ 8.58 per child. Of these costs, 83% were associated with outpa- Without vaccination tient visits. Vaccination significantly re- Outpatient visits 605 525 265 duced the economic burden of rotavirus Hospitalizations 16 0867 AGE in Mexico, averting US$ 14 million Deaths 923 0.40 DALYsa 31 768 14 (71% of the total health care burden). With vaccination Outpatient visits 190 598 83.41 Cost-effectiveness of the rotavirus vacci- Hospitalizations 2 2530.99 nation program. The estimates of the Deaths 273 0.12 DALYsa 9 436 4.13 cost-effectiveness of a rotavirus vaccina- Benefit of vaccination (averted events)b tion program in Mexico are presented in Outpatient visits 414 927 181.59 Table 6, and assume a basic vaccine price Hospitalizations 13 8336.05 of US$ 16 per course. The cost of the Deaths 651 0.28 vaccination program includes the price of DALYsa 22 331 9.77 the vaccine and its administration. More- a Disability-adjusted life years discounted at 3%. over, the vaccination program cost as- b Benefit of vaccination is based on the vaccine effectiveness, which incorporates information on the vaccine effi- cacy and coverage. sumes only the cost of vaccinating chil- dren who would receive the vaccine, based on vaccine coverage estimates and TABLE 4. Comparison of direct medical costs of treating gastroenteritis in Mexico using the 10% wastage. Benefits are expressed as hospital-based rotavirus surveillance study and alternative data, in 2007 US$ net medical costs from the health care sys- Hospital-based tem perspective and the cost-effectiveness surveillance Alternative Alternative ratio. All of these cost and benefit mea- (US$) data (US$) 1a (US$) 2b sures vary directly with vaccine price. At a price of US$ 16 per course (two Hospitalization Per-diem cost 57.45 66.83 57.45 doses), it would cost US$ 25.4 million Length of stay (days)3.4c 3.5 4.0 to vaccinate the Mexican birth cohort Total hotel cost 195.32 233.91 229.80 with the rotavirus vaccine (net medical Diagnostics/medications 15.90 77.19 27.04 costs). From the health care system per- Total direct cost/hospitalization 211.22 311.10 256.84 Outpatient visit spective, a rotavirus vaccination pro- Hotel cost 14.99 17.03 14.99 gram costs US$ 1 139 per DALY averted, Diagnostics/medications 12.24 11.07 5.87 US$ 39 102 per life saved, US$ 1 839 per Total direct cost/outpatient visit 27.23 28.10 20.86 hospitalization averted, and US$ 61 per a Alternative 1: WHO-CHOICE and CVI data. These values were originally based on 2002 prices and were adjusted for in- medical visit averted. Across all course- flation using the Consumer Price Inflation index (41). costs less than or equal to US$ 16, ro- b Alternative 2: Physician interview data. tavirus vaccination was projected to be c Average of 2.8 days (hospital-based surveillance study) and 4 days (physician interviews). cost-effective, using the WHO criteria and a per-capita GDP of US$ 6 121 for model predicts that rotavirus AGE is The costs of treating AGE using the Mexico (40). responsible for an annual 414 927 (182 hospital-based rotavirus surveillance per 1 000) outpatient visits; 13 833 (6 per study were similar to alternative cost- Sensitivity and uncertainty analyses 1 000) hospitalizations; and 651 (0.28 per estimates. In Table 4, the inpatient and 1 000) deaths, all of which could be pre- outpatient data estimates are compared The results of one-way sensitivity vented by rotavirus vaccination. This rep- to those using secondary data (WHO- analyses are presented in Table 7. The es- resents an overall reduction of 69%, 86%, CHOICE and Children for Vaccine Ini- timated medical cost per child is most and 70%, respectively. tiative cost studies) and physician in- sensitive to changes in assumptions re- terviews. The total direct medical cost garding the incidence of rotavirus outpa- Costs of rotavirus disease. Based on the for hospitalization ranged from US$ 211, tient visits and the cost of outpatient treat- estimates derived from the hospital- using the hospital-based surveillance ment. A 25% change in either of those based rotavirus surveillance study (32), data, to US$ 311, using secondary data— parameters results in 11–21% change in the total direct medical cost for inpa- a total difference of US$ 100. The differ- the medical cost per child. The incre- tients was US$ 211, with 92% of the cost ence between the high- and low-end es- mental cost-effectiveness rate estimates attributed to hospitalization and the re- timates of total direct medical costs for are affected by the rotavirus mortality mainder due to the cost of diagnostics outpatient visits was US$ 7.24. rate, vaccine efficacy against mortality, and medication. For outpatients, the The estimated economic burden of ro- and vaccine price. A 25% change in any of total direct medical cost was US$ 27, of tavirus GE in Mexico and benefits of a these variables results in 8–41% change which 55% was attributed to the cost of rotavirus vaccine are shown in Table 5. in the incremental cost-effectiveness ratio. the visit (Table 4). In the absence of vaccination, it was esti- In addition, assuming a vaccine efficacy

486 Rev Panam Salud Publica/Pan Am J Public Health 25(6), 2009 Constenla et al. • Cost-effectiveness of rotavirus vaccination in Mexico Original research

TABLE 5. Estimated economic burden of rotavirus-related gastroenteritis and benefits of a rota- TABLE 6. Estimates of costs, net benefits and virus vaccination program in Mexico, in 2007 US$ cost-effectiveness of a rotavirus vaccination program in Mexico, in 2007 US$ Direct medical costsa Mean cost per childa,b (US$) (US$) Vaccine program costa US$b

Without vaccination Vaccine administration 2 120 480 Outpatient visits 16 273 770 7.12 Vaccine ($16/course) 37 320 448 Hospitalizations 3 326 828 1.46 Total cost 39 440 928 Total 19 600 598 8.58 Averted medical costs 14 005 314 With vaccination Net medical direct cost 25 435 613 Outpatient visits 5 127 904 2.24 Incremental cost-effectiveness ratio Hospitalizations 467 3800.20 Cost per DALYc averted 1 139 Total 5 595 284 2.44 Cost per life saved 39 102 c Benefit of vaccination (averted costs) Cost per hospitalization averted 1 839 Outpatient visits 11 145 866 Cost per medical visit averted 61 Hospitalizations 2 859 448 Total 14 005 314 a Vaccination cost assumes only the cost of vaccinating chil- dren who would receive the vaccine, based on vaccine a The total discounted direct medical costs for a given birth cohort was calculated as the sum of costs for all facility types and coverage estimates and 10% wastage. age groups, adjusting future financial outcomes for the cohort back to a present value. The discounted costs were based on b The intervention cost, net medical cost and incremental hospital-based surveillance data (US$ 2007). cost-effectiveness ratio (US$/DALY) assume a vaccine b Calculation only includes direct medical costs to the health care system and pertains to the direct medical costs that each child price of US$ 16 per course. Values are based on hospital- under-5 years incurs as a result of a rotavirus infection. based surveillance data (US$ 2007). c Benefit of vaccination is based on the vaccine effectiveness, which incorporates information on the vaccine efficacy and c DALY: disability-adjusted life year. coverage. of 73% against AGE from any cause im- TABLE 7. The effect of rotavirus incidence, vaccine efficacy, and cost estimates on the medical proved cost-effectiveness, as reflected by cost attributed to rotavirus disease and the cost-effectiveness of a rotavirus vaccination program the lower cost-effectiveness ratio for Mex- in Mexico, in 2007 US$ ico. In general, an increase in rotavirus in- Incremental cost-effectiveness cidence and vaccine efficacy will result in Medical cost per child ratioa (US$) a lower incremental cost-effectiveness ra- Variable US$ Change % Change% tio making a vaccination program more cost-effective. Furthermore, a reduction 5-year cumulative risk of hospitalization in the price of the rotavirus vaccination due to RV GEb program (US$ 8 per dose) will yield a 0.005 (–25%) 7.22 –4 1 150 +1 0.007c 7.54 1 139 lower incremental cost-effectiveness ratio 0.009 (+25%) 7.87 +4 1 128 –1 of US$ 303 per DALY averted. 5-year cumulative risk of death due to For the uncertainty analysis, the cost- RV GE (× 100 000) effectiveness of vaccination using the 30.3 (–25%) No 1 515 +33 40.4c 7.54 1 139 baseline and alternative scenarios is 50.5 (+25%) No 911 –20 shown in Figure 2. The graph shows the 5-year cumulative risk of an outpatient visit cost effectiveness (US$/DALY) at differ- due to RV GE ent vaccine price levels for each of the 0.199 (–25%) 5.98 –21 1 207 +6 0.266c 7.54 1 139 scenarios. The point at which each line 0.331 (+25%) 9.11 +21 1 071 –6 crosses the horizontal access is the break- Vaccine efficacy (deaths averted) even price (medical cost). The break-even 0.80 (–14%) No 1 298 +14 c cost per dose, i.e., the cost per dose below 0.92 7.54 1 139 1.00 (+8%) No 1 048 –8 which the vaccine became cost-saving, Vaccine efficacy (hospitalization averted) was US$ 5.10 per course in an analysis 0.80 (–14%) No 1 150 +1 that considered health care system costs 0.92c 7.54 1 139 alone. The figure does not consider the 1.00 (+8%) No 1 139 0 Hospitalization per-diem cost combined impact of the scenarios. 38.24 (–25%) 7.25 –4 1 150 +1 50.98c 7.54 1 139 DISCUSSION 63.73 (+25%) 7.84 +4 1 128 –1 Outpatient visit cost 9.98 (–25%) 6.69 –11 1 173 +3 This is the first economic evaluation of 13.30c 7.54 1 139 the impact that a national rotavirus vac- 16.63 (+25%) 8.41 +11 1 105 –3 cination program would have on chil- Vaccine price ($/course) dren in Mexico. Morbidity and mortality $16 (–33%) No 1 606 –41 $24c 7.54 1 139 data were derived from national admin- $32 (+33%) No 672 +41 istrative data and efficacy data from the most recent results of the phase III effi- a The incremental cost-effectiveness ratio (US$/disability-adjusted life year) assumes a vaccine price of US$ 24.00 per course unless specified otherwise. cacy trial of the live-attenuated human b RV GE: rotavirus-related gastroenteritis. rotavirus vaccine performed in Latin c Base case scenario.

Rev Panam Salud Publica/Pan Am J Public Health 25(6), 2009 487 Original research Constenla et al. • Cost-effectiveness of rotavirus vaccination in Mexico

FIGURE 2. Cost-effectiveness of rotavirus vaccination in Mexico for alternative scenarios, in diarrheal deaths due to rotavirus is 2007 US$ unknown. For this reason, it assumed that the proportion would be similar to 3 000 Baseline hospitalized cases. Likewise, vaccine ef- Delayed coverage ficacy against death has been assumed 2 500 Reduced coverage of mortality risk (20% reduction) to be the same as that for hospitalized a 2 000 Reduced efficacy in subsequent years (10%) cases, because death has not been eval- uated as an outcome of a vaccine trial. Upper estimate of hospitalization rate

$/DALY 1 500 Efforts should be made to empirically Lower estimate of hospitalization rate ascertain the incidence of death due to 1 000 rotavirus AGE using active surveillance, especially in areas of the country where 500 there is limited access to emergency treatment facilities. 0 The sensitivity analysis also demon- 0 3 691215 18 21 strated that the evaluation of health care Vaccine price ($/dose) resource utilization and its ascribed costs a DALY: disability-adjusted life year. impact the health care costs associated with rotavirus AGE. Despite uncertainty America, including Mexico (32). Rota- Another criterion frequently used by in health care costs, the sensitivity analy- virus AGE results in a significant disease national decision-makers is to compare sis shows that varying costs by 25% has and economic burden from the perspec- the cost-effectiveness of a health inter- almost no effect on the cost-effectiveness tive of the Mexican health care system. vention to the cost-effectiveness of other of vaccination. Considering a societal per- Rotavirus vaccination would prevent similar health interventions; in this case, spective would therefore have made no 70% of deaths due to rotavirus, 86% of other strategies for diarrhea prevention, difference in the cost-effectiveness of associated hospitalizations, 69% of out- child nutrition, and immunization. Ide- vaccination. Since it is possible that a patient visits, and 70% of the associated ally, the most cost-effective intervention child with AGE had a co-infection that DALYs. This compares to rates observed would be chosen first. However, this required additional medications and di- in other Latin American countries where type of country-specific information is agnostic tests, there is some uncertainty rotavirus-associated mortality rates are not available in Mexico. Even if this in- about the actual cost of treating AGE. To lower (46, 47). Vaccination would also formation were available, comparing avoid overestimating the medication result in total, direct medical savings of data across studies must be done with and diagnostic costs, only the medi- US$ 14 million, 71% of the overall health caution since there may be differences in cations and diagnostic tests commonly care burden (burden of all diseases). the methods used. used to treat AGE were included in the Several criteria can be used for eco- Our analysis integrated empirically- analysis. However, there is no way to be nomic evaluations of a vaccination pro- collected country-specific cost data with certain whether a specific antibiotic was gram. The appropriateness of the different country-specific epidemiological esti- prescribed to treat AGE or some other approaches depends on the perspec- mates to develop the first national eco- infection. tive of the policymaker. As an approxi- nomic model of rotavirus vaccine intro- Another limitation is related to the mate benchmark and for comparison duction in Mexico. It complements other availability of data on rotavirus disease purposes only, we used the per-capita Regional efforts to develop further un- treatment costs. Estimates of resource Gross Domestic Product (GDP) as a derstanding of the value of a rotavirus utilization were developed using a sam- threshold and assumed that vaccination vaccination program. These include two ple population in select facilities and strategies that were less than this bench- Regional studies (5, 6) and six country- medical treatment costs were derived mark were potentially very cost-effective level studies (46–51) that describe the from only one hospital. However, costs (46). With a per capita GDP of US$ 6 121 disease and economic burden of rota- were varied by 25% in the sensitivity for Mexico (40) and a vaccine price of virus, and estimate the benefit and cost- analysis. The sensitivity analysis shows US$ 16 per course, a national rotavirus effectiveness of rotavirus vaccination at that varying costs by 25% has almost no vaccination program for Mexican chil- the Regional and country-levels. effect on the cost-effectiveness of vacci- dren would be considered a very cost- A few study limitations need to be nation. Alternative cost estimates were effective intervention since the cost per borne in mind. As shown by the sensi- described in the paper for comparison, DALY averted is US$ 1 139. It is impor- tivity analysis, the rotavirus mortality but were not used in the sensitivity tant to emphasize that even if vaccination rate and vaccine efficacy against death analysis because there was uncertainty can be cost-effective, i.e., that the incre- greatly affect the cost-effectiveness of about the various methods used to de- mental cost-effectiveness ratio is less than vaccination. Although these variables are rive these estimates. the per capita GDP of a specific country, important to the results of the analysis, In addition, outpatient visits were the it may not be affordable without assis- they have not been directly evaluated. main contributor to medical costs; how- tance. Future studies should focus their Diarrheal deaths are generally recorded; ever, outpatient and emergency room vis- efforts on providing estimates of the fi- however, vital statistics do not register its remain difficult to ascertain, espe- nancial costs to vaccinate a single birth co- the primary cause of diarrhea death as cially because rotavirus AGE is not hort at different coverage rates. rotavirus. Moreover, the proportion of easily diagnosed in these settings in

488 Rev Panam Salud Publica/Pan Am J Public Health 25(6), 2009 Constenla et al. • Cost-effectiveness of rotavirus vaccination in Mexico Original research

Mexico. These estimates could be im- recommended time. If high-risk popula- direct effects of vaccination, but it did not proved with additional data and larger tions were missed or vaccination were consider the indirect protective effect on samples. delayed, the effectiveness would be re- persons never vaccinated. The herd im- Additional study limitations are re- duced. Since rotavirus GE occurs in munity effect could be large and might lated to the coverage rate and timing of young children and protection is con- offset inefficiencies in delivery of a com- vaccination. Vaccine coverage is the ferred from the time of vaccine dose 1 plete course, and on-time vaccination to proportion of children less than 5 years until dose 2 is given (32), it is important all children. of age who would receive the rotavirus that the vaccine be given on time. Future In summary, rotavirus AGE results in vaccine. The coverage is based on cover- considerations should be made to ac- a significant disease and economic bur- age rates of other vaccines (e.g., DTPa, count for the fact that not all children den in Mexican children. The results of Hib+IPV, and HBV), which may not be will receive the vaccine at the recom- this evaluation provide to policymak- an accurate estimation of rotavirus vac- mended time. ers strong evidence that supports the cine coverage if a vaccine were to be in- A final limitation of the study is the implementation of a national mass ro- troduced. In the analysis, we assumed lack of data on the magnitude of herd tavirus vaccination program as a very that all groups within the country have immunity, which may be conferred by cost-effective health care intervention in equal likelihood of vaccination and all partial coverage of vaccination into a Mexico. children would receive the vaccine at the population. The analysis considered the

REFERENCES

1. Breese J, Glass RI, Ivanoff B, Gentsch J. Cur- 10. Barraza P, Avendano LF, Spencer E, Calderon 20. Velázquez FR, Calva JJ, Guerrero ML, Mass rent status and future priorities for rotavirus A, Prenzel I, Duarte E. Hospital infection D, Glass RI, Pickering LK, et al. Cohort study vaccine development, evaluation, and imple- caused by in infants, Santiago, of rotavirus serotype patterns in symptomatic mentation in developing countries. Vaccine. Chile. Bol of San Panam. 1986;101:328–38. and asymptomatic infections in Mexican chil- 1999;17:2207–22. 11. Bok K, Castagnaro N, Borsa A, Nates S, Espul dren. Pediatr Infect Dis J. 1993:12:54–61. 2. Kapikian AZ, Hoshino Y, Chanok RM. Rota- C, Fay O, et al. Surveillance for rotavirus in 21. Secretaría de Salud/Servicio Social de Salud. viruses. In: Knipe DM, Howley PM, et al., eds. Argentina. J Med Virol. 2001;65:190–8. Anuario estadístico 2000–2001. In: Sistema Volume 1: Fields virology. 5th ed. Philadel- 12. Bok K., Castagnaro NC, Diaz NE, Borsa A, Nacional de Información en Salud, 2001–2002. phia: Lippincott Williams and Wilkins; 2007. Cagnoli MR, Nates S, et al. Rotavirus labora- Ciudad de México: Ministerio de Salud; 2002. Pp.1917–74. tory network: results after one year of obser- 22. Instituto Mexicano del Seguro Social. Estadís- 3. Velázquez FR, Garcia-Lozano H, Rodriguez vation. Rev Argent Microbiol. 1999;31:1–12. ticas: Daños y egresos hospitalarios 2002. In: E, Cervantes Y, Gómez A, Melo M, et al. Diar- 13. Cardoso DDP, Soares CMA, Souza MBLD, Dirección de prestaciones médicas: Informa- rhea morbidity and mortality in Mexican chil- Azevedo MSP, Martins RMB, Queiroz DAO, ción estadística en salud, 2002. Ciudad de Mé- dren: Impact on rotavirus disease. Pediatr In- et al. Epidemiological features of rotavirus in- xico: Ministerio de Salud; 2002. fect Dis J. 2004;23(10):S149–55. fection in Goiânia, Goiás, Brazil, from 1986 to 23. O’Ryan M, Pérez-Schael I, Mamani N, Peña 4. Instituto Nacional de Estadística Geografía e 2000. Mem Inst Oswaldo Cruz. 2003;98:25–9. A, Salinas B, Gonzales G, et al. Rotavirus- Informática, Secretaria de Salud. Dirección 14. González FS, Sordo ME, Rowensztein G, Sab- associated medical visits and hospitalizations General de Información en Salud. Proyeccio- bag L, Roussos A, De Petre E, et al. Rotavirus in South America: a prospective study at three nes de la población de Mexico, 2000–2050. diarrhea. Impact in a pediatric hospital of large sentinel hospitals. Pediatr Infect Dis J. CONAPO; 2002. Buenos Aires. Medicina (B Aires). 1999;59(4): 2001;20:685–93. 5. Rheingans RD, Constenla D, Antil L, Innis BL, 321–6. 24. Ehrenkranz P, Lanata CF, Penny ME, Salazar- Breuer T. Economic and health burden of ro- 15. Gomez JA, Sordo ME, Gentile A. Epidemio- Lindo E, Glass RI. Rotavirus diarrhea disease tavirus gastroenteritis for the 2003 birth co- logic patterns of diarrheal disease in Argen- burden in Peru: the need for a rotavirus vac- hort in eight Latin American and Caribbean tina: estimation of rotavirus disease burden. cine and its potential cost savings. Rev Panam countries. Rev Panam Salud Publica. 2007; Pediatr Infect Dis J. 2002;21:843–50. Salud Publica. 2001;10:240–8. 21(4):192–204. 16. Perez-Schael I, Gonzalez R, Fernandez R, 25. Guardado JAA, Clara AW, Turcios RM, 6. Rheingans RD, Constenla D, Antil L, Innis BL, Alfonzo E, Inaty D, Boher Y, et al. Epidemio- Fuentes RAC, Valencia D, Sandoval R, et al. Breuer T. Potential cost-effectiveness of vac- logical features of rotavirus infection in Cara- Rotavirus in El Salvador—an outbreak, sur- cination for rotavirus gastroenteritis in eight cas, Venezuela: Implications for rotavirus veillance and estimates of disease burden Latin American and Caribbean countries. Rev immunization programs. J Med Virol. 1999; 2000–2002. Pediatr Infect Dis J. 2004;23(10): Panam Salud Publica. 2007;21(4):205–16. 59:520–6. S156–60. 7. Gold MR, Siegel JE, Russell LB, Weinstein MC 17. Urrestarazu MI, Liprandi F, De Suarez EP, 26. Salinas B, Gonzalez G, Gonzalez R, Escalona (Eds.). Cost-effectiveness in health and medi- Gonzalez R, Perez-Schael I. Etiological, clini- M, Materan M, Perez-Schael I. Epidemiologic cine. Oxford: Oxford University Press; 1996. cal, and sociodemographic characteristics of and clinical characteristics of rotavirus dis- 8. Murray CJL, Lopez AD. The global burden of acute diarrhea in Venezuela. Pan Am J Public ease during five years of surveillance in Vene- disease: a comprehensive assessment of mor- Health 1999;6:149–56. zuela. Pediatr Infect Dis J. 2004;23:S161–7. tality and disability from diseases, injuries 18. Velázquez FR, Matson DO, Guerrero ML, 27. World Health Organization. International and risk factors in 1990 and projected to 2020. Shults J, Calva JJ, Morrow AL, et al. Serum statistical classification of diseases and health Cambridge: Harvard University Press; 1996. antibody as a marker of protection against related problems: ICD-10. Second edition. Ge- 9. Pan American Health Organization. Regional natural rotavirus infection and disease. J In- neva: WHO; 2005. Core Health Data System-table generator fect Dis. 2000;182:1602–9. 28. Velázquez FR, Castellanos A, Luna G, Busta- 2004. Washington, D.C.: PAHO; 2004. Avail- 19. Villa S, Guiscafre H, Martinez H, Munoz O, mante ME, Gómez A, Cedillo R, et al. Impor- able from: http//www.pago.org/Project. asp? Gutierrez G. Seasonal diarrheal mortality tancia de los agentes virales como causa de SEL=HD&LNG=ENG&ID=379. Accessed 6 among Mexican children. Bull WHO. 1999;77: diarrea grave en los niños menores de cinco August 2004. 375–80. años de edad que requieren hospitalización, y

Rev Panam Salud Publica/Pan Am J Public Health 25(6), 2009 489 Original research Constenla et al. • Cost-effectiveness of rotavirus vaccination in Mexico

factores de riesgo asociados. 1a. Ed.: Sestante, 36. Miller MA, McCann L. Policy analysis of the vaccines: review of epidemiologic studies of S.A. de C.V. In: Las múltiples facetas de la in- use of hepatitis b, haemophilus influenzae type b, rotavirus diarrhea in Argentina. Rev Panam vestigación en salud: Proyectos estratégicos streptococcus pneumoniae-conjugate and rota- Salud Publica. 1998;3(2):69–78. del Instituto Mexicano del Seguro Social. Ciu- virus vaccines in national immunization 45. Baltussen R, Hutubessy R, Evans D, Murray dad de México: Instituto Mexicano del Seguro schedules. Health Economics. 2000;9:19–35. C. Uncertainty in cost-effectiveness analysis: Social; 2001. Pp. 133–52. 37. Walker D, Mosqueira NR, Penny ME, Lanata probabilistic uncertainty analysis and sto- 29. World Health Organization, WHO Statistical CF, Clark AD, Sanderson CFB, et al. Variation chastic league tables. Int J Technol Assess Information System (WHOSIS). Life tables for in the costs of delivering routine immunization Health Care. 2002;18(1):112–9. 191 countries. Geneva: WHO; 2001. Available services in Peru. Bull WHO. 2004;82:676–82. 46. Constenla D, O’Ryan M, Navarrete M S, Antil from: http://www3.who.int/whosis/menu. 38. Waters HR, Dougherty L, Simon-Pierce T, L, Rheingans R D. Potential cost effectiveness cfm?path=whosis,bod,burden_statistics,life Tran N, Wysonge CS, Long K, et al. Coverage of a rotavirus vaccine in Chile. Rev Méd Chile &language=english. Accessed 21 October 2004. and costs of childhood immunizations in 2006;134:679–88. 30. Baltussen R, Adam T, Tan Torres T, Hu- Cameroon. Bull WHO. 2004;82:668–75. 47. Constenla D, Pérez-Schael I, Rheingans RD, tubessy R, Acharya A, Evans D. Generalized 39. World Health Organization. CHOosing In- Antil L, Salas H, Yarzábal JP. Evaluación del cost effectiveness analysis: a guide. Geneva: terventions that are Cost Effective (WHO- impacto económico de la vacuna antirrotaví- World Health Organization; 2002. CHOICE). Prices for hospitals and health cen- rica en Venezuela. Rev Panam Salud Pública. 31. Liddle JL, Burgess MA, Gilbert GL, Hanson tres. Available from: http://www3.who.int/ 2006;21(4):213–22. RM, McIntyre PB, Bishop RF, et al. Rotavi- whosis/cea/prices/unit.cfm?path=evidence, 48. World Health Organization. The world health rus gastroenteritis: impact on young children, cea,cea_prices,cea_prices_unit&language= report 2002—reducing risks, promoting their families and the health care system. Med english. Accessed 20 January 2004. healthy life. Geneva: WHO; 2002. Pp. 108. J Aust. 1997;167:304–7. 40. World Bank Group. World development indi- 49. Constenla D, Ortega E, Rheingans R, Antil L, 32. Ruiz-Palacios GM, Pérez-Schael I, Velázquez cators online 2004. Available from: http:// Saez-Llorens X. Impacto económico de la va- FR, Abate H, Breuer T, Clemens SC, et al. www.worldbank.org/data/wdi2004/. Ac- cuna antirrotavírica en Panamá. Anales de Pe- Safety and efficacy of an attenuated vaccine cessed 16 August 2004. diatria 2008;68(2):128–35. against severe rotavirus gastroenteritis. N 41. U.S. Department of Labor, Bureau of Labor 50. Constenla D, Rivera M, Rheingans RD, Antil L, Engl J Med. 2006;354(1):11–22. Statistics. Inflation and consumer spending, Vásquez ML. Evaluación económica de una 33. De Vos B, Vesikari T, Linhares AC, Salinas B, 2003. Available at: http://www.bls.gov/bls/ eventual incorporación de la vacuna antirota- Perez-Schael I, Ruiz-Palacios GM, et al. A inflation.htm. Accessed 19 August 2004. virus en el calendario de vacunación infantil en rotavirus vaccine for prophylaxis of infants 42. Adam T, Evans DB, Murray JL. Econome- Honduras. Rev Med Hond. 2006;74(1):19–29. against rotavirus gastroenteritis. Pediatr In- tric estimation of country-specific hospital 51. Constenla D, Linhares A, Rheingans RD, fect Dis J. 2004;23(10 suppl):S179–82. costs. Cost Eff Resour Alloc 2003;1(3). Avail- Antil L, Waldman EA, da Silva LJ. Economic 34. Mexico, Consejo Nacional de Vacunación able at http://www3.who.int/whosis/cea/ impact of a rotavirus vaccine in Brazil. J (CONAVA). Informes trimestrales de vacuna- background_documents/pdf/Unitcost.pdf Health Pop Nutrition 2008. [In press]. ción y vigilancia nutricional. Available from: 43. Bittencourt SA, Leal Mdo C, Santos MO. Hos- http://www.conava.gob.mx. Accessed 3 No- pitalization due of infectious diarrhea in Rio vember 2004. de Janeiro State. Cad Saude Publica. 2002; 35. Brenzel L, Claquin P. Immunization pro- 18(3):747–54. grams and their costs. Soc Sci Med. 1994;39: 44. Gomez JA, Nates S, De Castagnaro NR, Espul Manuscript received on 24 March 2008. Revised version 527–36. C, Borsa A, Glass RI. Anticipating rotavirus accepted for publication on 24 August 2008.

RESUMEN Objetivos. Evaluar el costo y los beneficios de un programa nacional de vacunación infantil contra el rotavirus en México. Métodos. Se diseñó un modelo de análisis de decisión, desde la perspectiva del sis- Impacto económico de un tema de salud mexicano, para comparar dos alternativas: vacunar contra el rotavirus programa de vacunación o no vacunar. A partir de datos nacionales publicados se estimó el número de casos y contra rotavirus en México muertes por rotavirus, los años de vida ajustados por la discapacidad (AVAD) que se evitarían y la relación costo-efectividad incremental (US$/AVAD) de una cohorte anual hipotética de 2 285 000 niños; se partió de algunos supuestos sobre el costo, la cobertura y las tasas de eficacia. Resultados. Con una cobertura de 93% y un precio de la vacuna de US$ 16,00 por esquema (dos dosis), se estima que un programa de vacunación contra rotavirus en México evitaría 651 muertes (0,28 muertes por 1 000 niños), 13 833 hospitalizacio- nes (6,05 hospitalizaciones por 1 000 niños) y 414 927 visitas de consulta externa (182 consultas por 1 000 niños) por gastroenteritis aguda asociada a rotavirus (GAR). La vacunación podría reducir la carga económica por GAR en México al evitar gastos por US$ 14 millones (71% de la carga total por atención sanitaria). A un precio de US$ 16,00 por esquema, la relación costo-efectividad sería de US$ 1 139,00 por AVAD evitado. Una reducción en el precio del programa de vacunación contra rotavirus (US $8,00 por esquema) generaría una menor relación costo-efectividad incremental de US$ 303,00 por AVAD evitado. Conclusiones. Un programa nacional de vacunación contra rotavirus en México re- duciría la incidencia y la mortalidad infantiles y sería altamente efectivo en función del costo, según los umbrales de las intervenciones de costo-efectividad de la Organi- zación Mundial de la Salud.

Palabras clave Rotavirus, vacunas contra rotavirus, análisis costo-beneficio, inmunización masiva, gastroenteritis, México.

490 Rev Panam Salud Publica/Pan Am J Public Health 25(6), 2009