What do we know about the cost-effectiveness of pneumococcal conjugate vaccination in older adults?

ABSTRACT

The cost-effectiveness of 13-type pneumococcal conjugate vaccine (PCV13) use in older adults, and the relative merits when compared to the 23-type polysaccharide pneumococcal vaccine (PPV23), has been a topic of much debate. Although a number of economics evaluations have been conducted many of these were completed before the availability of critical data on PCV13 efficacy in older adults. Recent studies using this data have found conflicting results. This may in part reflect differences in the level of herd protection from infant pneumococcal vaccination programs in different countries. The costs and benefits of pneumococcal vaccination in adults are likely to rest on several critical parameters: the magnitude pneumococcal disease in older adults and the serotypes responsible for it, the efficacy of each vaccine against invasive and non-invasive pneumonia, the duration of vaccine protection, and differences in vaccine price. The ongoing changes in pneumococcal disease patterns highlight the need for economic evaluations to use recent serotype-specific disease estimates from the setting under consideration. In countries that do recommend PCV13 use in adults, post-implementation economic evaluation (using data from after a program is implemented) may be useful to help inform potential future changes to vaccine recommendations as well as the maximum price that should be paid for the vaccines in future negotiations.

KEYWORDS:

• adults
• cost-effectiveness analysis
• economic evaluation
• elderly
• immunization; pneumococcal
• Streptococcus pneumoniae

In 2015, a large randomized clinical trial (RCT) of 13-type pneumococcal conjugate vaccine (PCV13) use in older adults was published (CAPiTA).¹ This trial was conducted in the Netherlands and demonstrated the protective efficacy of PCV13 against both vaccine-type non-bacteremic pneumococcal pneumonia and vaccine-type invasive pneumococcal disease.¹ The cost-effectiveness of PCV13 use in older adults and the relative merits of PCV13 when compared to the 23-type pneumococcal polysaccharide vaccine (PPV23) has been a topic of much debate.^2-11 The costs and benefits of pneumococcal vaccination in adults are likely to rest on several critical parameters: the magnitude pneumococcal disease in older adults and the serotypes responsible for it, the efficacy of each vaccine against invasive and non-invasive pneumonia, the duration of vaccine protection, and differences in vaccine price.

For both vaccines there is relatively strong evidence to support protection against vaccine-type invasive disease in older adults.^1,12,13 The latest CAPiTA trial data¹ suggests that PCV13 may have a higher efficacy (75%; CI:41.4%-90.8%) against vaccine-type invasive disease in older adults than PPV23, although this potential advantage may be outweighed by the broader serotype coverage of PPV23. It should be noted that questions have been raised about the population impact of existing PPV23 programs⁴ due to the persistence of vaccine-type invasive disease in targeted groups and difficulties in detecting declines coinciding with changes in vaccine recommendations.^14,15 However, difficulties in establishing declines in observational data may be partly explained by the longstanding use of polysaccharide vaccine in some settings, which has meant that recent changes to PPV23 uptake were sometimes modest (Australia¹⁵) or focused on lower risk older adults (UK¹⁴).

For the prevention of vaccine-type non-bacteremic pneumococcal pneumonia, the CAPiTA trial provides solid evidence of the efficacy of PCV13 (45%; CI:21.8–62.5) in older adults.¹ Whereas, for PPV23 the evidence for protection against this outcome in older adults is inconsistent.¹⁶ Efficacy against this outcome was reported in the latest 2013 Cochrane review for adults.¹² However, this result was heavily informed by studies from the 1940s¹⁷ and 1970s¹⁸ which used lower valent polysaccharide vaccines and, in the case of Smit et al.,¹⁸ focused on younger adults. There is more recent data for PPV23 from a RCT in Japanese elderly nursing home residents which found a significant result¹⁹ and from cohort data based on older adults in Spain.²⁰ Together these data suggest that PPV23 may provide a level of protection against non-bacteremic pneumococcal pneumonia in older adults, but the level of any protection remains uncertain. The lack of reliable data on this outcome for PPV23 is unfortunate, as the relative efficacy assumed for each vaccine against non-invasive pneumonia (particularly cases resulting in death) is very influential when estimating their cost-effectiveness.²¹

The duration of vaccine-induced protection is likely to be equally important in estimating the cost and benefits of alternative pneumococcal vaccines. Most economic evaluations have evaluated the use of the vaccines in older adults of various ages under 75 years,²¹ as the very old are more likely to have reduced immunogenicity to vaccination. However, the risk of pneumococcal disease increases with age so that the assumptions around the ongoing vaccine protection provided as vaccinated individuals age will substantially change the benefits of vaccination. The CAPiTA trial found that PCV13 protection was maintained over the 4 year period of follow-up¹ but the level of protection after this period is uncertain. For PPV23, the exact duration of protection against disease in older adults is also uncertain and has been found to be influential in cost-effectiveness models.²² It is likely that for both vaccines the duration of protection and the initial efficacy provided may be age dependent in older adults^23,24 and also differ for those with some pre-existing conditions that place them at higher risk of severe disease. Model assumptions related to this heterogeneity in vaccine protection between individuals are likely to play an important role in determining the cost-effectiveness of different potential vaccination strategies.

Another major consideration when evaluating the cost-effectiveness of adult vaccination strategies is herd protection effects from infant programs targeted at the same pathogen.¹¹ In many countries after the introduction of an infant PCV7 programs there has been a reduction in invasive disease in adults caused by the serotypes included in the infant vaccine.²⁵ Similar herd protection effects for invasive disease have now been shown after the switch to PCV13 programs in US infants.²⁶ Therefore, in countries with high uptake infant PCV13 programs, there is likely to be less benefit from adult PCV13 programs due to the already reduced levels of invasive disease caused by these 13-types in adults. Infant PCV programs may also reduce the benefits of PPV23 programs in adults but to a lesser extent given the additional serotypes included in PPV23.

The evidence available to inform potential herd protection effects from infant PCV programs on adult non-bacteremic pneumococcal pneumonia is less clear-cut. A number of studies have found observed declines in pneumonia hospitalisations and/or mortality in older adults coincidental with introduction of infant PCV7 programs.^27-29 However, the pneumococcus bacteria is only one of many causes of pneumonia and the total burden of pneumonia is known to fluctuate independently of vaccination over time.^30,31 For example, seasonal variation in the strain's of influenza can change the magnitude of severe pneumonia disease even without changes in vaccination coverage.^30,31 Recently, several studies have found declines in the proportion of adult non-bacteremic pneumococcal pneumonia caused by PCV7 types after the introduction of infant programs.^32,33 These serotype-specific data give support to a herd protection effect on non-invasive pneumonia from PCV7. However, evidence of these effects from PCV13 use in infants has yet to be observed and may not be equivalent to those found from PCV7.³²

An alternative approach to choosing between the PCV13 and PPV23 is to administer both vaccines in sequence. This approach was recommended in 2015 by the US Advisory Committee on Immunization Practices (ACIP) for adults aged over 65 years, with PCV13 to be administered first followed (after a delay) by PPV23.³⁴ While providing greater protection, this dual vaccination strategy is likely to be less cost-effective than the individual use of each vaccine, as with the exception of serotype 6A, all 13 serotypes included in PCV13 are also included in PPV23. Therefore, those receiving subsequent immunisation with PPV23 after PCV13 are likely to be already partially protected against 12 of the serotypes in PPV23. This means that there is likely to be less incremental benefit from a subsequent vaccine dose when compared to the use of each vaccine in isolation.

Before the publication of the CAPiTA trial, there were a number of economic evaluations that evaluated PCV13 use in older adults.²¹ As a result, these evaluations had to make assumptions about the efficacy of PCV13 against both serotype-specific non-bacteremic pneumococcal pneumonia and invasive disease. The efficacy assumed for PCV13 against vaccine-type non-bacteremic pneumococcal pneumonia in these analyses tended to be higher than subsequently found in the CAPiTA trial.²¹ Likewise, the proportion of all pneumonia assumed to be caused by pneumococcus was often relatively high (30%-40%)²¹ when compared to some recent data.³⁵ To reliably inform vaccine policy, new economic evaluations are required that use the best data now available, particularly the CAPiTA trial data and up-to-date data on the evolving serotype-specific disease burden in each setting.

Although many evaluations are likely underway, to date there have been relatively few published economic evaluations that have made use of the recent CAPiTA trial data. One of the first analyses to use this data was an economic evaluation of adult PCV13 use in the Netherlands.³⁶ This analysis found that PCV13 was likely to be cost-effective in many of the adult groups considered for vaccination. However, the Netherlands use PCV10 rather than PCV13 in infants and estimates used suggested a relatively high proportion (35%-50%) of invasive disease was still due to PCV13 types in those aged over 50 years.³⁶ The results for countries that have seen greater herd protection effects from PCV13 may be less positive. Another important limitation of this analysis was that it did not compare PCV13 to the potential use of PPV23 and so did not allow an assessment of the relative costs and benefits of the 2 main alternative vaccines available.

In contrast, a recent economic evaluation conducted for Belgium found that PCV13 use in the older adults was unlikely to be cost-effective.²² In Belgium, PCV13 had been used in infants for several years resulting in herd protection effects in older adults against covered serotypes.²² The evaluation also found that the use of PPV23 was unlikely to be cost-effective but that it may provide better value for money due to the lower vaccine price. The authors highlighted 3 influential (but uncertain) parameters: negotiated vaccine price, duration of vaccine protection, and the serotypes causing adult pneumococcal disease.²² Where data was lacking, they made relatively conservative assumptions, such as assuming the immunity from PCV13 was limited to 5 years in the base case model.²²

The ongoing changes in pneumococcal disease patterns highlight the need for economic evaluations to use recent serotype-specific disease estimates from the setting under consideration. Alongside data on invasive disease, setting-specific data on the proportion of non-invasive pneumonia caused by PCV13 types would help to inform future economic evaluations. The development of urine-based antigen tests¹ may help provide these much needed data. However, due to the shifting nature of pneumococcal disease, adult pneumococcal vaccination programs would benefit from ongoing evaluation. This appears to be the approach that ACIP has adopted, with a 2018 review already planned for the 2015 recommendation for older adults.³⁴ While the cost-effectiveness of health interventions is often thought of as a static value, we know that cost-effectiveness is likely to vary over time.³⁷ Based on experience with PCV7, it is likely that the full extent of the herd protection effects from PCV13 use in infants may take several years to accrue.²² This means that even if economic evaluations use the most up-to-date serotype-specific disease data, there is still potential for continued changes in the cost-effectiveness of adult pneumococcal vaccination. Another important consideration is the potential for serotype replacement which may reduce the cost-effectiveness of both of infant and adult vaccination programs over time.

Finally, when considering the cost-effectiveness of alternative pneumococcal vaccination strategies it is worth remembering that the price negotiated for the vaccine's is a critical determining factor. For example, rather than asking if PCV13 is cost-effective in older adults, it may be more appropriate to ask, at what cost per dose is PCV13 use cost-effective in adults of a given age? The price negotiated for vaccines is not a direct function of the cost of production and cost-effectiveness analysis can help in the negotiation process by estimating the maximum vaccine price that should be paid by governments.³⁸ In countries that do recommend PCV13 use in adults, post-implementation economic evaluation (using data from after a program is implemented^29,39) may be a useful addition to pre-implementations analyses. These retrospective economic analyses could help to inform any future program changes and, if the value for money achieved is less than predicted, they may also help in future price negotiations with industry.^29,39

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Funding

This work was supported by an Australian National Health and Medical Research Council Project grant (1081344).