|Home | About | Journals | Submit | Contact Us | Français|
Vaccination of adults aged 60 years and older against Streptococcus pneumonia is not recommended in Brazil. The 23-valent polysaccharide pneumococcal vaccine (PPV23) is only available for institutionalized persons or with underlying diseases despite the substantial medical and economic burden related to pneumococcal infections in adults over than 59 years. The study aimed at evaluating the cost effectiveness of implementing a large PPV program in this population.
This analysis was performed using a static decision tree model. Demographic and epidemiological data were obtained from Brazilian official sources and international literature. Economic data were obtained from a study performed in 2007 in a public and a private hospital located in Sao Paulo. Vaccination was assumed to protect for 5 years with 60% effectiveness against bacteremic pneumococcal pneumonia (BPP) and 21% effectiveness against non bacteremic pneumococcal pneumonia (NBPP). Deterministic and sensitivity analyses were performed.
The pneumococcal polysaccharide vaccination saved 5,218 life year gained (LYG). The vaccination program was found to be cost effective in the social security and public health care perspectives with a mean incremental cost-effectiveness ratio of R$10,887 and R$8,281 per LYG respectively. Results were sensitive to the vaccine effectiveness against NBPP, the incidence and case-fatality rate of NBPP. From a societal perspective, PPV23 program for adults 60 and older was found to be cost-saving.
Pneumococcal polysaccharide vaccination is clinically and economically favored over the present vaccination strategy, in which persons aged over 59 years in Sao Paulo have not been vaccinated.
Streptococcus pneumoniae infections are a major cause of illness and death with about 1.6 million cases of fatal pneumococcal disease occurring worldwide annually, mostly in infants and the elderly.41 Streptococcus pneumoniae can cause both non-invasive and invasive infections.1 In adults and elderly, non-invasive disease can manifest as pneumonia, whereas bacteremia and meningitis are the most common invasive pneumococcal diseases.1 Although, pneumococcal disease can affect all age groups, the elderly and immunocompromised are at highest risk from infection.2 Moreover, since the introduction of pneumococcal conjugate vaccination in children in 2000, the epidemiology of invasive pneumococcal disease (IPD) has changed with the reported burden of IPD now highest in the elderly population. In the US, the annual incidence of IPD in those ≥65 years of age was recently reported to be 40/100,000 versus 21/100,000 in children <5 years of age.3 In Europe, there has been a similar shift of the burden of pneumococcal disease to the elderly population.4 Elderly patients are more vulnerable to infections due to physiological changes associated with aging process, therefore, the risk of an individual to develop pneumonia rises as they grow older. On a yearly basis, approximately one out of twenty persons over the age of 85 will have one episode of pneumonia.5 While in most cases, the etiological diagnosis of CAP will not have been obtained, the results of a study by Johnstone and co-workers6 showed that in 37% of the cases, Streptoccocus pneumoniae is the causative agent of bacterial pneumonia. In Latin America, S. pneumoniae was demonstrated to be also the most common pathogen implicated in adult CAP, accounting for 35% of cases.7 Consequently, mortality due to pneumonia also rises with age. In a study on the burden of community acquired pneumonia (CAP) reported by Jackson et al.5 in 2004, 12.5% of the patients over 80 years-old admitted to hospital for pneumonia die while in the hospital.
Resistance of pneumococci to commonly used antimicrobials is a serious and increasing problem worldwide, which complicates the treatment of infection.8 Antibiotic-resistance causes an increase in treatment failures and medical costs. Lynch and Zhanel reported in 2009 that worldwide, 15–30% of S. pneumoniae strains are multidrug resistant (i.e., resistant to ≥3 classes of antibiotics). These investigators also reported that six serotypes (6A, 6B, 9V, 14, 19F, 23F) account for >80% of penicillin- or macrolide-resistant S. pneumoniae, worldwide.9 Moreover in the US, an increase in the isolation of antibiotic resistant 19A strains has been reported, a strain associated with increased incidence of IPD.10 Effective immunization against S. pneumoniae is the most effective method to reduce the impact of pneumococcal infections, including those caused by antibiotic resistant strains. Therefore, physicians and policy-makers have high expectations concerning the value of pneumococcal vaccination of the adults and elderly.
Currently, the pneumococcal vaccine approved in adults is the 23-valent pneumococcal polysaccharide vaccine (PPV23). A growing number of national and international health bodies now recommend pneumococcal vaccination of elderly and at-risk groups. Indeed, PPV23 has been shown to be clinically effective in the reduction of the occurrence of both bacteremic pneumococcal pneumonias (BPP) and non-bacteremic pneumococcal pneumonias (NBPP).11
The world's population is progressively aging. This trend is also reflected in Brazilian population where both the proportion of the total population and the real number of elderly people (≥60 years old) have significantly risen for the past few years. Brazilian national statistics data show that the population over 60 years has grown from 8.8% in 1998 to 11.1% in 2008, and should became the sixth largest elderly population in the world by 2020.
In Brazil, presently the public national immunization program recommends PPV23 only for institutionalized elderly, but there is no funding for this vaccination program. As a result, the PPV23 coverage rate in Brazil is low. In fact, data from the Brazilian public health system indicate that only 160,508 doses were used in 2009, which represents a coverage rate of approximately 0.8% if we consider the total demography of adults aged more than 60 in Brazil in 2009. In US, PPV23 has been recommended for all people aged ≥65 years since 1983 and totally financed, and data shows that the coverage increased from 14.1% in 1989 to 60.1% in 2008.12 In the Latin America region, most countries recommend and finance PPV23 in the general elderly population considering the health and economic benefit of PPV23. These countries include Argentina, Chile, Ecuador, El Salvador, Mexico, Uruguay, Panama and Venezuela.
To date there have been no cost-effectiveness studies of pneumococcal vaccination of the elderly conducted in Brazil. Therefore, as a response to the need to decrease the burden of pneumococcal disease in Brazil's elderly population, we conducted a cost-effectiveness analysis of the PPV23 vaccination program in order to evaluate the medical and economic impact of a PPV23 public funding. This analysis was conducted in Sao Paulo State, which has the largest population and the most developed economy (approximately 30% of the Brazilian GDP) of the Brazilian States. Currently, Sao Paulo State offers PPV23 for institutionalized elderly and those with underlying diseases. The aim of this study was to evaluate the health and economic benefits of the PPV23 public funding to the general elderly population in Sao Paulo.
In the base case scenario where a routine PPV23 vaccination program is implemented in São Paulo State from age 60, vaccination markedly reduced the number of episodes of pneumococcal infections. With a cohort of 3,689,623 individuals, the number of NBPP and BPP cases avoided after 5 years due to pneumococcal vaccination was 12,469 and 2,967 respectively. The number of life years gained (LYG) was 5,218.
The incremental costs related to implementation of PPV23 program in elderly people in São Paulo State was: R$56.8, R$43.2 and R$-5.4 million respectively for scenario 1 using the costs paid by the Brazilian social security SUS, for scenario 2 using the total costs of a public hospitalization, and for scenario 3 taking into account the costs in a private hospital and the absenteeism costs (Table 1). The averted costs of pneumococcal infections outweighed the costs of a vaccination program in scenario 3. Consequently, the ICER related to a PPV23 vaccination program in São Paulo State in elderly varied according to the scenarios (Table 1): R$10,887 in scenario 1; R$8,281 in scenario 2 and cost-saving in scenario 3 suggesting a return on the investment.
In the univariate sensitivity analysis, we evaluated the individual effects of epidemiological and vaccination parameters on the cost-effectiveness of the vaccination strategy using the minimum and the maximum of the inputs range (Tables 2–4). As the various inputs should have relatively the same impact on results in each scenario, the univariate sensitivity analysis was performed only on scenario 1 (SUS hospitalization costs). Factors having the greatest impact on ICER were the vaccine effectiveness against NBPP, the case-fatality rate of NBPP, and the incidence rates of NBPP (Fig. 3).
A probabilistic sensitivity analysis of the cost-effectiveness results using the range or standard deviation of each input (Tables 2–4) is also shown as a cost-effectiveness plane in Figure 2. These results demonstrate that there is a 95% chance that the incremental cost per LYG is contained in the interval R$[6,273; 29,223] for scenario 1 (mean R$13,516); R$[-1,275; 24,653] for scenario 2 (mean at R$10,357), and R$[-76,509; 22,373], for scenario 3 (mean at -R$3,247).
Cost-effectiveness evaluations, such as the one described here, are necessary to define population strategies and help decision makers to choose the most relevant options in term of cost/benefit ratio of immunization programs.
In view of the variety of healthcare systems that exist throughout Brazil, the present analysis was performed from three different scenarios. From the cost study performed in Sao Paulo State, three different costs of hospitalization appeared: the lowest cost was that paid by SUS (R$462 for NBPP and R$830 for BPP), the intermediate cost was the amount actually paid by the public hospital (R$1,992 for NBPP and R$5,103 for BPP) and the highest was the cost of hospitalization in the private hospital (R$19,764 for NBPP and R$33,320 for BPP). Evaluation of cost-effectiveness ratios using the SUS and the public healthcare costs allowed coverage of the public perspective on the whole. In scenario 3, integrating larger costs (private hospitalization costs and absenteeism costs) allowed the inclusion of persons who used a private health care system, i.e., 46% of the total elderly population e.g., persons using a managed care organization.
In many developed countries, acceptable cost-effectiveness thresholds have been defined for planning healthcare policies. However, no such definition is available in Brazil, an emerging country facing limited healthcare resources. The World Health Organization (WHO) has suggested an acceptable cost-effectiveness threshold as one that is less than three times the yearly gross domestic product (GDP) per capita and an excellent cost-effectiveness threshold as one that is less than one times the yearly GDP per capita.13 In Brazil, where the yearly GDP per-capita was R$13,720 in 2007 14 at the time of the analysis, an intervention with a cost-effectiveness of up to R$41,160 may therefore be considered as cost-effective by WHO standards. Consequently, in the present analysis, PPV23 vaccination program from age 60 was either extremely cost-effective (inferior to R$13,720 for social security and public health care perspectives) or cost saving (for societal perspective), depending on the scenario considered. This evidence was in favor of a routine PPV23 vaccination program that would be offered by the government to all adults 60 years of age and older.
Using pessimistic values in deterministic sensitivity analyses, all ICER also were found to be below the WHO cost-effectiveness threshold of R$41,160 (3 times the GDP per capita). Specifically, the ratio with an effectiveness of PPV23 against NBPP at 0% was R$33,342 which is still under this limit. This means that even with no effectiveness against NBPP, a routine PPV23 vaccination program in elderly in Sao Paulo is a cost-effective option, which is in line with the other published studies in reference 15–17. In addition, in the probabilistic sensitivity analyses, taking into account all uncertainties around parameters and the thresholds of R$41,160, there were a probability of 99.8%, 100% and 100% that the funding of PPV23 vaccination is cost-effective compared to the current situation with no vaccination for scenarios 1, 2 and 3 respectively. If we considered the threshold of R$13,720, the probability would decrease to 60.9%, 74.8% and 87% respectively.
Our results are consistent with those of numerous cost-effectiveness analyses performed in US and Europe. As Postma et al.18 concluded in their literature review published in 2003 on the basis of the international literature in the elderly, the cost-effectiveness for the prevention of invasive pneumococcal disease, considering only direct medical costs, varies from cost-saving to more than €30,000 per LYG or per QALY.15,19–23 These results would justify local implementation of a pneumococcal polysaccharide vaccination program from a pharmacoeconomic point of view. In studies concerning both invasive and non invasive pneumococcal diseases, vaccination with PPV23 is again more cost-effective or cost-saving compared to no vaccination strategy among the elderly, considering only direct medical costs.24–27 This economic benefit increases with age of elderly targeted and vaccination of all individuals above the age 65 years is comparable in terms of cost-effectiveness to many accepted health care interventions.27
The study's limitations were related to the various inputs used. Firstly, in absence of specific local data, incidence and case-fatality rates were derived from international data. It is likely that the disease epidemiology and health service utilization might be different between US/Europe and Brazil because of differences in prevalent strains, antibiotic resistance within the populations, quality of health care and financial architecture of their respective health-care systems. However, the chosen values were discussed and accepted by local experts to ensure their consistency with the local epidemiology. Moreover, to be sure to not overestimate the impact of S. pneumoniae in Brazil, we used generally the most conservative value found. Epidemiological data related to pneumonia in Latin America was published just before the finalization of this manuscript.7 These local data were globally less conservative than those used in the base case and were all included in sensitivity analysis range. Using these data in the SUS base case scenario returned a cost-effectiveness ratio of R$9,709/LYG. The second limitation is related to the disease costs used in the present analysis. The sample size used to calculate the costs of hospitalization may be considered a limitation. It is enough large for NBPP cost calculation (N = 173 and N = 163 for public and private hospital respectively) but too small for BPP (N = 12 and N = 17 for public and private hospital respectively). Larger studies including larger numbers of patients with BPP will be necessary for confirmation. In addition, our analysis considered only the costs of hospitalization and not the outpatient costs and other costs related to pneumococcal infections such as transportation, diet, etc. Including these costs would increase the economic benefit of PPV23 vaccination. Thirdly, vaccination costs concerned only vaccine price and transportation. Promotion costs to ensure a high coverage rate also would have been estimated.
Over the last decade, the importance and benefit of pneumococcal polysaccharide vaccination in adults has been reviewed due to the decrease of pneumococcal incidence rate in countries with a high coverage rate of pneumococcal conjugate vaccines (PCV) in infants. Indeed, in the US, where the coverage rate of the 7-valent pneumococcal conjugate vaccine (PCV7) in children has achieved 80–90% over the last 6 years, herd immunity has globally led to a 38% decrease in the rate of IPD among elderly.28 However, an increase of the incidence of IPD in adults and elderly caused by non-PCV serotypes has been noted in many settings (“serotype replacement”); these increases range from minimal to substantial29 and could reduce the benefits of the PCV vaccination.30 Subsequently, PPV23 is the only protection available against these non-PCV7 serotypes at this time. Our analysis did not take into account the diminution of pneumococcal infections incidence in adults that could result thanks to the use of the 10-valent pneumococcal conjugate vaccine (PCV10), now routinely used in Brazil. The use of PCV10 in Brazil is recent and to date no herd immunity data are available. The future effect of a PCV10 program in Brazil will depend on coverage rates, serotype prevalence and serotype replacement inherent to PCV would not likely to be observed before the program is well established with a high vaccine coverage rate. However, even including an indirect effect of PCV, we can still expect cost-effectiveness results for PPV23 vaccination program regarding our univariate sensitivity analyses (with the most pessimistic incidence values for NBPP and BPP, ICER stayed cost-effective) and regarding a recent analysis performed in US that included this indirect effect.22
Within the Brazilian population aged more than 59 years, a large proportion is still economically active. Indeed 18.8% of people older than 60 years-old are still employed.31 For this reason, the analysis undertaken from the societal perspective was relevant in Brazil. Prevention strategies such as vaccination are extremely important to maintain the economic activity of adults aged 60 and older to maintain functional independence with advancing age. The aging of the population increases health costs and elderly people with acute diseases often require more complex health examinations and treatments. Elderly patients admitted for longer treatments are also much more susceptible to functionality loss. Therefore, a patient with pneumococcal disease can become economically inactive and incur high expenses with the initial disease and any subsequent sequelae.
A decision-analytic model was developed by Ezus Group (Lyon, France) to estimate the effectiveness and cost-effectiveness of a polysaccharide pneumococcal vaccination program in adult population (more than 18 years old) with a possibility to focus the analysis on an age-based or a risk-based program. This model also has been used in Turkey69 and Korea.33 Two identical cohorts, either vaccinated with the PPV23 or non-vaccinated (since the coverage rate in Brazil in this population is closed to zero), were designed to reflect the São Paulo State elderly population (≥60 years old). Each cohort was followed during a 5-year period divided into 5 periods of 1 year to provide a conservative estimate of the total duration of PPV23 effectiveness. Infections due to S. pneumonia were classified as either BPP or NBPP, which comprise around 90% of pneumococcal-related outcomes that are preventable by PPV23 vaccination.1 The structure of the model is illustrated in Figure 1.
Model assumptions are described below:
For each cohort, the number of life years experienced and the costs of pneumococcal infections were calculated and compared. To compare the costs and health consequences of PPV23 vaccination versus non-vaccination, an incremental cost-effectiveness ratio (ICER) was estimated as the incremental cost per life year gained (LYG). Since no health utility data are available in Brazil, the quality-adjusted life year (QALY) due to vaccination was not evaluated. The concept and methodology of cost-utility analyses is not well established in Brazil, the evaluation of LYG can be considered an initial approach in the analysis of cost-utility.
Representative demographic data on the São Paulo State elderly population was based on the national statistics data from 2007.34 The total size of the cohort was 3,689,623. The all-cause mortality rate from São Paulo State by age group was obtained from the Brazilian mortality information system.35
In the base case, 18.6% was used for the probability of hospitalization due to NBPP. This rate was calculated based on the assumption that the probability of hospitalization varies from 10% for 60–70 years old people to 30% for people more than 70 years old and based on the Sao Paulo population size of these subgroups. In the literature, the hospitalization rate related to CAP was reported to be 50–52.8%.7 Considering this high rate compared to the rest of the world where 80% of CAP patients are treated as outpatients, and the year of these published data, the value of 18.6% was selected for the analysis.36 For BPP, Vila-Corcoles et al. (2006) found that 4.5% of IPD cases were not hospitalized (1/22 cases) in Spain which gives a hospitalization rate of 95.5%. Previous reports noted similar rates: 100% from a report by Sisk et al. and 96% from a study by Robinson and co-workers. To be conservative, 95.5% was use in the base case and 90-100% for the range (Table 2).
International pneumococcal incidence rates and case-fatality rate (CFR) were used as there were no local data available (Table 2). Jackson et al. (2004) reported an incidence of hospitalized CAP in elderly (65+) in US of 1,150 cases per 100,000. Considering that 40% on average of CAP requiring hospitalization are caused by S. pneumoniae,2 we estimated an incidence rate of hospitalized pneumococcal pneumonia of 460 cases per 100,000. To determine the incidence of NBPP in general, we used also the fact that 80% of pneumococcal pneumonia are NBPP (BPp = ¼ NBPP),2 and we used the hospitalization rates defined above (18.6% of NBPP + 95.5% of ¼ NBPp = 460 cases). We obtained 1,083 of NBPP cases per 100,000. Weycker et al. also recently reported an incidence of 728 NBPP cases/100,000 in 65+ (using a mean of the moderate values weighted by population size). Another study conducted in five European countries reported a mean incidence of hospitalized pneumonia of 754/100,000 in elderly (65+).19 Using the same calculation as used by Jackson et al.5 we obtained a total NBPP incidence of 706/100,000. Vila-Corcoles et al.37 reported an incidence of hospitalized pneumonia in Spain of 1,048/100,000 in elderly (65+) that gave a total NBPP incidence rate of 987/100,000. The study from Ament et al.19 was used in the present analysis because it represents the data of five different countries. The mean European rate was used for the base case, and the minimum and maximum values (451 and 1,167 for incidence of hospitalized pneumonia in France and Sweden respectively) were used for the range. For BPP, range given by the WHO Weekly Epidemiological Record (WER) 2008 41 in people aged more than 65 years and leaving in developing countries were used. The mean value of the range was used in the base case analysis.
BPP CFR were retrieved from Vaccine 4th edition1 which reported a CFR range from 11% to 44% (mean = 26.6%, 16 studies included) in elderly in developed countries. This range is confirmed by many papers: Middleton et al. (2008) showed a CFR of 14.3% for 65-74 years old, 21% in 75–84 year old and 30.8% for 85+ years old in US; the WHO WER of Oct 2008 reported that the CFR attributed to BPP may reach 30–40% in elderly and in industrialized countries; Evers 2007 reported a mean CFR of 21% in 65+ among 10 European countries. For NBPP CFR, the CFR for CAP was used since we could fine no relevant data specific to S. pneumoniae and since assuming that the etiology of CAP to be S. pneumoniae should not impact the mortality rate. In Argentina, a country bordering Brazil, a 18% CFR in 346 patients with CAP (hospitalized or not) aged 18 to 102 years (median age 64) 43 is reported and a 13.3% in hospitalized patients.44 Other studies in more industrialized countries reported a CAP-related mortality averaged 10.7% in hospitalized elderly patients in US,45 3.6% in people aged more than 65+ (hospitalized or not) in US,5 and 12.7% in people aged more than 65+ in Spain.11 For the present analysis, a range between 3.6% and 18% was used for the sensitivity analysis and 12.7% (the more recent study-2009) in the base case.
For this analysis, a one-dose vaccination was simulated. A 60% coverage rate was estimated in the case that by the public system and also in consideration that an aged-based program is easier to implement than a risk-based program46–48 (Table 3).
Pneumococcal vaccine effectiveness against IPD is the most firmly established, however the level of effectiveness depends on risk status and age. Meta-analyses of five randomized clinical trials have shown that PPV23 is 68% (95% CI 53–78) effective against IPD among immunocompetent older adults.49 In our analyses, the values for vaccine effectiveness against were based on multiple observational studies which had the advantage of being conducted in large unselected natural populations11,37,50–55 (Table 3). One of these recent studies, conducted in Scotland in persons aged ≥65 years, demonstrated a protective effect of PPV23 against IPD of 61.7% (95% CI, 45.1–73.2).54 In Spain, a protective effect of 72% (95% CI, 46–95) and 70% (95% CI, 48–82) against IPD was reported by Vila-Corcoles et al.53 in persons over 60 years old, and by Dominguez et al.53 in persons over 65. In the US, IPD vaccine effectiveness of 44% (95% CI, 7–67) in persons 65 and older was reported by Jackson et al.55 in 2003, and 80% (95% CI, 51–92) in persons 65–74 years old. Shapiro et al.56 1991 values were in accordance with the 14-year nationwide surveillance study conducted by the US CDC in the elderly and in persons with underlying chronic disease (effectiveness of 75% in 65+).51 As a result of these studies, effectiveness against BPP in the base case was assumed to be 64% for all elderly, which is the mean effectiveness of the observational studies. In the sensitivity analyses, the minimum and maximum values given in observational studies were used. These values ranged from 44% from the study by Jackson et al.55 to 80% from reported by Shapiro et al.
The evidence for the effectiveness of PPV23 against NBPP is less firmly established mainly due to the difficulties in identifying S. pneumoniae as etiological agent of pneumonia.49,57 However, the results of a recent clinical trial conducted in nursing home residents considered at high risk for CAP (mean age: 85 years) showed that PPV23 reduced pneumococcal pneumonia by 64% (95% CI, 32–81) and all-cause pneumonia by 45% (95% CI, 22–61).58 A systematic review of observational studies,52 which may be considered to be better estimates of real-life vaccine effectiveness reported an effectiveness of 32% against all pneumonia. In addition, two recent cohort studies conducted in Spain, in which bacteria were identified by radiography, sputum culture, and Binex antigen test, demonstrated an effectiveness against NBPP of 42% (95% CI, 14–61) in persons >50 years old, and 39% (95% CI, −6–65) in elderly >65 irrespective of risk status.11,37 PPV23 has also been associated with reduced hospitalization of all-cause CAP by 24% (95% CI, 0.9–41) to 26% (95% CI, 8–41),59 lower risk of hospitalization and fewer deaths for pneumonia.26,60,61 Consequently in the current analysis, since our populations were aged ≥60 years, we considered a range of 0–42% effectiveness against NBPP for both populations in the sensitivity analyses (minimum and maximum value given in the literature). The mean value of this range was used for the base case: 21% (Table 2).
In addition, as reported in various publications,11,24,37,38,41,42,53,54,56,62,63 a 10% waning rate each year was applied and a total duration of effectiveness was fixed conservatively at 5 years (Table 2).
Vaccination costs including the vaccine public payer price and the door to door transportation cost, was fixed at R$27 Brazilian reais or US$ (US$1 = R$1.7955 − mean exchange rate in 2008) (Table 3). In this analysis, this cost was assumed to be paid entirely by SUS. Adverse events costs were not included as they can be considered negligible. Vaccine administration costs for people aged over 59 years were assumed to be null since it was expected that vaccination would be administered during a regular routine visit to the general practitioner.
The Brazilian health system is divided between public and private sector. The public system is called SUS and is a universal system. It means that every Brazilian citizen has the constitutional assured right to receive treatment funded by the government. SUS reimbursement coverage is based on procedure packages that are defined in term of value and included services by the Ministry of Health. The funding comes from the federal, state and municipal governments. The private health system (HMOs, Health Plans and Health Insurance Companies) is financed in two ways: fully paid by the patient or partially covered by the patient employer. Although both types of health care cover the treatment of pneumonia, there are differences in terms of access to treatment, medications available, and cost reimbursements routes.
Information about the cost of pneumococcal infections in Brazil was not available in the literature. A retrospective database study was therefore performed in two large hospitals in Sao Paulo State to data collection: one public hospital, Hospital São Paulo (HSP), and one private hospital, Hospital Israelita Albert Einstein (HIAE) were included. Both hospitals are among the largest in Sao Paulo, each with more than 500 beds (Fig. 4).
The inclusion criteria were patients aged over 60 years and patients hospitalized for diagnosis of NBPP or BPP due to S. pneumoniae with a laboratory confirmation. In these patients, resources consumed were retrieved from their files. Theses resources included: procedures, laboratory tests, medical examinations, treatments (medications and other therapies), and their length of stay in the intensive care unit and general ward. No outpatient resources were collected. As the cost data were not available with the medical records, official data sources were used to estimate the costs in both hospitals. For the public health care setting, procedures, hospitalization and medical test costs were taken from SUS (Sistema Único de Saúde-Unique Health System) reimbursement Table,64 and devices and drugs costs from Ministry of Health prices data banking.65 Since these two Tables present reference values of reimbursement and acquisition values in the public health care market displaying both individual and combined (packages) values, two methods for building costs were performed. The first was from the SUS payer perspective and the second from the hospital perspective. For the private health care setting, Brazilian Medical Association Table (LPM99) was used for procedures, hospitalization and tests costs66 and SIMPRO Brazil Magazine (wholesales list price 2007; 2009 number 51) for devices and drugs costs. Costs, expressed in Brazilian reais, were calculated and detailed in the Figure 4.
In the public hospital, two separate costs of hospitalization due to pneumococcal infections were retrieved. These included; (1) the cost paid by SUS corresponding to the reimbursement package as defined by the Brazilian Ministry of Health and; (2) the total cost of hospitalization corresponding to the exact cost for the hospital. As these two costs were very different (total hospitalization costs were approximately five times higher than SUS costs), we decided to test the impact of these costs in two different scenarios. The cost difference between the SUS reimbursement and the total hospitalization cost represents a deficit for the public hospital.
In addition to these two scenarios, another one was also considered, which included the total public health care costs (not only the SUS reimbursement package), the private costs of hospitalization, and the cost of absenteeism. Of persons over 60 years old, 18.8% were considered active31 with a daily salary estimated at R$37.59 (GDP per capita in 2007 divided by 365).14 Absenteeism duration was considered to be equivalent to length of stay in hospital.
Under base case assumptions, parameter values were varied individually in a one-way sensitivity analysis to identify which variables would have a major impact on the cost-effectiveness results. A variety of inputs were tested and their values are shown in Tables 2–4. Additional details pertaining to the values of the ranges are found in the epidemiological and vaccination inputs section.
In addition, parameters listed in Tables 2 and and33 (NBPP incidence, BPP incidence, NBPP case fatality rate, BPP case fatality rate, vaccine effectiveness against NBPP and vaccine effectiveness against BPP) were varied simultaneously in probabilistic sensitivity analyses, where random draws from each parameter's distribution were performed and the effectiveness and incremental cost calculated. This procedure was repeated 1,000 times. Parameter distributions were chosen based on the parameter type and level of certainty. Those parameters whose distributions were least certain (epidemiological and case-fatality rate data from international literature) were assigned uniform distributions, where all values in a range are equally likely to be chosen. Parameters whose distributions were most certain (cost data from local cost of illness study) were assigned log-normal distributions.
A 5% discount rate on costs and lives was used in base case as described in the Brazilian's health economics guidelines.68 In the deterministic sensitivity analysis, 0% and 10% were used respectively for the low and high value.
As recently described by Rijkers et al.32 pneumococcal vaccines may become the most important vaccines for adults and children worldwide in the future years. Healthcare providers and key decision makers should recognize the serious health impact of pneumococcal disease in adults, closely monitor the epidemiology of pneumococcal serotypes, particularly ‘serotype replacement’ and ensure increased vaccine coverage.
Allocation of public funds is one important recognized way to increase the coverage rate, facilitating people's access to vaccination. In Sao Paulo State, PPV23 vaccination program from age 60 would be highly cost-effective considering total public hospitalization costs as well as the public hospitalization costs reimbursed by SUS. It is even likely to be cost-saving and to generate a return on investment when private health-care and absenteeism costs are considered as suggested by the present analysis. This paper supports therefore a universal vaccination program funded by the Brazilian public system.
sanofi pasteur funded the cost of illness study.
Joao Tonolio Neto M.D., Ph.D.: Medical Doctor from Federal University of Sao Paulo, consultant for several other pharmaceutical industries;
Gabriela Tannus M.Sc.: independent researcher from Axia. Bio Consulting, consultant for several other pharmaceutical industries;
Anna Gagliardi M.D.: Medical Doctor from Federal University of Sao Paulo, consultant for several other pharmaceutical industries;
Amanda Pinho: employee from sanofi pasteur, Brazil;
Laure Durand, M.Sc.: employee from sanofi pasteur, France.
Marcelo Fonseca M.D., M.Sc.: Medical Doctor from Federal University of Sao Paulo and independent researcher from Axia. Bio Consulting, consultant for several other pharmaceutical industries.