Streptococcus pneumoniae causes invasive infections such as meningitis and bacteremia as well as pneumonia, otitis media (OM) and sinusitis. Infants and young children under the age of two years have an increased risk of pneumococcal disease. The 23-valent polysaccharide vaccine, available since the 1980s, is not immunogenic in children of this age group. In the year 2001, the European Agency for the Evaluation of Medicinal Products (EMEA) granted market authorisation to the heptavalent pneumococcal vaccine Prevenar (Wyeth). The approved indication of Prevenar is the active immunisation of infants and young children under the age of two against invasive disease caused by Streptococcus pneumonia serotypes 4, 6B, 9V, 14, 18C, 19F and 23F. At the time of this study the German vaccination schedule advises the immunisation with Prevenar only for children at high risk. Reasons behind this decision are the limited experience with Prevenar in general and the compatibility with other combined vaccines when administered at the same time, the limited coverage of serotypes prevalent in Germany and the limited evidence on replacement with other serotypes at the time of decision. In its position paper on the heptavalent conjugated pneumococcal vaccine the German Paediatrics Academy states that about 200 cases of invasive disease and 5 to 10 deaths caused by this disease could be prevented each year in children aged between six months and two years if all children in this age group were vaccinated.
ÖBIG (Austrian Health Institute) was commissioned by DIMDI (German Institute for Medical Documentation and Information) to undertake an evaluation (Health Technology Assessment, HTA) of the heptavalent conjugated pneumococcal vaccine.
Streptococci pneumoniae (pneumococcus) are bacteria with a polysaccharide capsule. The polysaccharide capsule is considered to be the main reason for pathogenicity and virulence. Depending on their chemical structures, about 90 different serotypes can be defined. Each serotype has its own incidence and prevalence that differs from one to the other country. In a human being one can normally find pneumococcus in the upper (nasopharygial) respiratory part. There are not conclusive data about the dispersion of pneumococcus in a healthy population, but it is estimated that 40% to 70% of all children under six years carry pneumoccus.
Streptococci pneumoniae are a major cause of bacterial infections. German data on epidemiology of invasive pneumococcal disease is not comparable to international epidemiologic data in this field. This is due to the lack of obligatory reporting and the lack of standardised blood culture supported diagnoses. German data collected by the ESPED (Survey Unit for Rare Paediatric Diseases in Germany) show low incidence rates of pneumococcal infections in international comparison. But as the incidence rate of meningitis, caused by streptococcus pneumoniae (a condition that needs an indoor care), is within the average range of European incidence rates, it is concluded that the German incidence rate of all invasive pneumococcal diseases also has to be within the range of the European average - thus the incidence rate of invasive pneumococcal disease (IPD) in infants and young children under the age of two years is estimated at 19 cases per 100,000 children per year. 55% of these IPD can be attributed to meningitis; sequelea occurs in approximately 13% of these cases, lethality is an estimated 5%. There are no specific German incidence rates available for otits media. An estimated 600,000 cases per year can be attributed to streptococcus pneumoniae, approximately 90,000 cases of these occurring in infants and children under the age of two years. Furthermore there are no detailed incidence data for pneumonia in Germany, there are an estimated 50,000 cases per year occurring in the age group of children under five, with the total number of all pneumonia cases in all age groups being an estimated 63,000 to 105,000 cases per year. Antimicrobal-resistance of streptococcus pneumoniae strains is an issue in the treatment of pneumococcal infections. Whereas the antimicrobial-resistance is a major issue in Spain and France, the antimicrobial resistance rate in Germany is one of the lowest in international comparison. In the long run, prophylactic measures are indicated to keep the antimicrobial-resistance low, with the most effective measure being the prophylactic vaccination. As the 23-valent polysaccharide vaccine is not immunogenic in infants and children under the age of two years, there has been extensive research on conjugated vaccines. The active substances of the heptavalent conjugated pneumococcal vaccine are pneumococcal saccharides of the serotypes 4, 6B, 9V, 14, 18C, 19F and 23F, each conjugated to a non-toxic mutant of diphtheria toxin (CRM 197 protein) as a carrier protein. Thus, Prevenar acts by stimulating a T-cell immune response against these seven serotypes of streptococcus pneumoniae.
Whilst the immunogenicity and reactogenicity of Prevenar was investigated in double-blinded, randomised, placebo-controlled clinical trials, the overall benefit of a general immunisation of all children - healthy as well as at risk - is still under discussion. The inclusion in the general vaccination schedule - implicating in most cases public funding of the vaccine - would benefit a birth cohort of approximately 730,000 children per year. The vaccine costs around € 220 per child (four-dose immunisation schedule, year 2004) if funded by the compulsory health insurance. These costs do not cover the administration of vaccine. Thus, overall publicly financed vaccination cost for children would increase substantially.
The objective of the study is first to determine the efficacy and effectiveness of the immunisation of all children with the heptavalent conjugated pneumococcal vaccine in Germany and secondly to determine whether a general recommendation for vaccination of all children would be cost-effective from the healthcare-payers' and society's perspective. Ethical and juridical questions are considered.
A systematic literature search was performed in 29 relevant databases for the period of January 1999 to June 2004. Thus, 1,884 articles were identified which were then assessed according to predefined selection criteria. Some articles were added through references in full texts and internet research (e.g. ESPED Data). 341 articles were ordered as full text, 90 of which were either used for assessment (21) or as background literature (69), the others were excluded either because they were not available in full text (approx. 50) or according to predefined selection criteria. Relevant economic articles were assessed according to Drummonds' ten point checklist for critical assessment of economic evaluation. A systematic review of relevant economic studies is given. Relevant parameters of the economic studies are presented in tables. An overview table puts relevant parameters of all studies like country, viewpoint, immunisation strategy, cost per dose, cost-effectiveness (cost per life-year-gained (LYG)) and breakeven prices against each other.
5.1 Medical assessment
There is evidence for medical effectiveness of Preve-nar against invasive pneumococcal disease caused by the covered serotypes from a major double-blinded randomised controlled trial (RCT) undertaken in California. The vaccine shows lower values of effectiveness against OM and pneumonia, as demonstrated in the FinOm-Trial (FinOM = Finnish otitis media). There is a growing body of evidence that herdimmunity could occur by this vaccination. It remains unclear and still a hazard if there could be a replacement through not covered serotypes. Though values for effectiveness of the vaccine in Germany are below the data of California because of the different incidence of serotypes, Prevenar should remain a high protective vaccine.
5.2 Economic assessment
The cost-effectiveness ratios for an immunisation of all children with Prevenar vary across different countries. One reason - besides different Health Systems - can be seen in the uncertainty about the duration of protection, another in the assumptions on regional serotype coverage of the vaccine. The cost-effectiveness ratios range from € 46,212 per LYG in the UK to € 155,630 per LYG in the USA from the healthcare-payers' perspective. From the societal perspective the studies show the intervention to be cost-saving in Spain or Germany, up to a cost effec-tiveness ratio of € 132,380 in Australia. The figures correspond to the four-dose immunisation scheme. The economic studies discuss the possible positive effects of cross-reactivity, thus covering an even greater proportion of serotypes then included in the vaccine, the effect of a reduction in nasopharyngeal carriage thus rising the protection of non-vaccinated persons (indirect protection, herd immunity) as well as the effects on antibiotic-resistance, which might be reduced by the vaccination due to the reduction of the use of antibiotics. These effects are not included in the studies. On the other hand the negative effect of replacement, an increased nasopharyngeal colonisation with other serotypes, is not included in the economic studies as well.
The cost-effectiveness studies prove to be sensitive to a variation in the price of the vaccine and the cost of administration, the regional serotype incidence, the lethality and the discount rate. Depending on the included cost and perspective the incidence of certain diseases becomes more important.
There is one economic evaluation of the heptavalent conjugated pneumococcal vaccine for Germany. The study - which was published in 2003 - calculates cost-effectiveness ratios for a general vaccination of all children of a birth cohort from the healthcare-payers' perspective as well as from the society's point of view. The study does not calculate incremental cost as compared to the actual advice of the German vaccination scheme (vaccination for children at high risk), instead vaccination of all children (children at risk as well as healthy children) is compared to conventional medical treatment. Some other assumptions potentially biasing the results in favour of the vaccination strategy are taken, e.g. a compliance of 100% was assumed and the duration of protection was assumed to last for ten years. Furthermore, there are uncertainties regarding the total number of IPD cases as well as of non invasive pneumococcal diseases. With respect to OM and pneumonia the study points out that there are no specific German incidence data, also the higher uncertainty of efficiency data of the vaccine against these diseases is stressed. Still the major proportion of cost savings both from the healthcare-payers' and the societal perspective is attributed to the huge number of cases of these diseases. With respect to the assumptions on efficiency of the vaccine against IPD in Germany, incidence data from a single year were taken thus overestimating the number of potentially preventable IPD.
The price of the vaccine in the German study was assumed to be € 69.02 per dose - in the year 2004 the price per dose was lower. Taking into account that discounts are granted to the healthcare-payers (compulsory health insurance), the price could be assumed to amount to € 54.45 per dose (2004) if publicly financed.
From the healthcare-payers' perspective the cost of the general immunisation strategy would amount to € 72,866 per LYG, i.e. additional costs per child of € 129.30 under the assumptions of the study authors. From the societal perspective the savings slightly outweigh the cost of immunisation.
In the authors multivariate sensitivity analysis a best-case and a worst-case scenario is calculated. The variables compliance, duration of protection, discount rate, cost of medical treatment in the outpatient sector as well as cost for medicinal products for multiple sequelea and indirect cost for productivity losses are varied simultaneously. In the worst-case scenario cost per LYG rise to € 88,946 in the best-case scenario costs drop to € 17,711 per non-discounted LYG (both from the healthcare-payers' perspective).
5.3 Ethical and legal considerations
Ethical considerations include the question of the vaccination being recommended or obligatory (can parents be forced to let their children being vaccinated?), as well as the question of the vaccination being financed publicly or privately: some children may be excluded from vaccination if it is not covered by the Social Health Insurance. Especially the latter question is of particular importance considering the high price of Prevenar.
Important legal considerations mainly include implications (e.g. concerning financing) following a modification of the German vaccination scheme towards an inclusion of all children for pneumococcal vaccination as compared to the present recommendation of the vaccination only for children at high risk.
Prevenar is a high effective vaccine against invasive diseases that are caused by serotypes that are covered from the vaccine. It also reduces the colonisation through pneumococcus which can lead to herdimmunity. Prevenar shows a low effectiveness against OM, mainly because OM is only partly caused by pneumococcus. Therefore Prevenar is definitely not a vaccine against OM. It remains unclear how long-lasting the protection against pneumococcus infections is and if there will be a replacement of covered serotypes through uncovered ones.
Different to other European countries there are enough effective antibiotics in Germany available at the moment to treat pneumococcus infections. This situation can easily change within Europe without frontiers, and should therefore be carefully observed. Additionally, more epidemiological data should be collected to confirm the estimated epidemiology.
International studies show an inconsistent picture of cost-effectiveness of the conjugated vaccine. Reasons can be seen in a lack of international guidelines on economic evaluation, differing healthcare systems and the range of uncertainties on the vaccine. The low evidence of information on the herd-immunity-effect of pneumococcal immunisation, the occurrence of serotype-replacement and the effects on the prevalence of antibiotic-resistant strains shall be considered when deciding whether the pneumococcal vaccination for all infants and young children should be added to the German vaccination scheme. There is also little information on the duration of vaccine effectiveness and regional effectiveness because of different serotype incidence. The economic models thus incorporate some uncertainties.
The economic analysis for Germany calculates the incremental cost-effectiveness-ratio over the "do nothing" (conservative medical treatment) strategy. The study does not evaluate the incremental costs over the actual advice of the vaccination schedule which is vaccinating children at high risk. The authors reference the model of von Kries. The given effectiveness against IPD overestimates the proportion of preventable IPD due to the fact that it only takes into account data of a very limited time span. Regarding OM and pneumonia, effectiveness data from Finland and California were taken, thus incorporating further uncertainties in the model. On the other hand, the price of Prevenar decreased to € 54.45 in 2004, in comparison with the assumed price of € 69.02 in the model. Besides the costs and benefits incorporated in the model further effects which might influence the outcome of the economic assessment have to be considered: There is little information on the possible positive effect of herd immunity, as well as on negative effects of replacement. Furthermore, the effects of vaccination on the prevalence of antibiotic-resistant pneumococcal strains have to be observed.
From the healthcare-payers' perspective a general immunisation of all children turns out not to be cost-effective, from society's perspective the general immunisation of all children in Germany would be cost-saving according to the economic-model for Germany. Because of uncertainties summarized above and the decrease of the price of the vaccine against the price assumed in the study no clear recommendation in favour of a general vaccination strategy can be derived at the time of this study.
The ethical question whether some children are denied access to vaccination because of financial reasons gains importance because of comparatively high costs of this vaccination.
At present relatively few pneumococcal strains in Germany show antibiotic resistance. This situation shall further be observed while improving the data evidence for future decisions (epidemiologic data of incidence of pneumococcal diseases and serotyping of pneumococcal bacteria). From the economic perspective no distinct recommendation to add the conjugated vaccination for all children on the German vaccination scheme can be given. This situation may change if the price for the vaccine further decreases. Furthermore a future cost-effectiveness analysis for Germany should incorporate the effects of the replacement phenomenon, the herd immunisation effects and the effects of the vaccination on the antibiotic-resistant pneumococcal strains.