The broader impacts of otitis media and sequelae for informing economic evaluations of pneumococcal conjugate vaccines

ABSTRACT

Introduction

Otitis media (OM) is a common childhood infection. Pneumococcal conjugate vaccines (PCVs) prevent OM episodes, thereby reducing short- and long-term clinical, economic, humanistic, and societal consequences. Most economic evaluations of PCVs focus on direct health gains and cost savings from prevented acute episodes but do not fully account for the broader societal impacts of OM prevention.

Areas covered

This review explores the broader burden of OM on children, caregivers, and society to better inform future economic evaluations of PCVs.

Expert opinion

OM causes a substantial burden to society through long-term sequelae, productivity losses, reduced quality of life for children and caregivers, and contribution to antimicrobial resistance from inappropriate antibiotic use. The effect of PCVs on acute OM has been recognized globally, yet the broader impact has not been consistently quantified, studied, or communicated. Economic evaluations of PCVs must evolve to include broader effects for patients, caregivers, and society from OM prevention. Future PCVs with broader coverage may further reduce OM incidence and antimicrobial resistance, but optimal uptake will depend on increasing the recognition and use of novel frameworks that include broader benefits. Communicating the full value of PCVs to decision makers may result in wider access and positive societal returns.

KEYWORDS:

• Burden of disease
• clinical and social burden
• cost-effectiveness analysis
• economic evaluation
• healthcare costs
• hearing loss
• otitis media
• pneumococcal conjugate vaccines
• quality of life
• socioeconomic benefits

1. Introduction

1.1. Otitis media

Otitis media (OM), or inflammation of the middle ear, is one of the most common diseases in young children worldwide [1,2]. OM is caused by viral or bacterial infection, most notably by four major bacterial pathogens, Streptococcus pneumoniae, non-typeable Haemophilus influenzae (NTHi), Moraxella catarrhalis, and Streptococcus pyogenes, which are part of the normal flora of the upper respiratory tract. OM is classified into several clinical subtypes, including acute otitis media (AOM), an acute inflammation of the middle ear with effusion and signs and symptoms of acute infection; otitis media with effusion (OME), a chronic yet asymptomatic inflammation of the middle ear with effusion; and chronic suppurative otitis media (CSOM), a chronic infection of the middle ear and mastoid cells with chronic perforation of the eardrum and otorrhea [1]. OM is a leading cause of physician visits and antibiotic use in young children and places considerable strain on health systems globally [3,4]. Compared with other clinical subtypes, AOM is most frequently observed by healthcare professionals, with an estimated 709 million cases per year globally (global incidence rate, 10.9% or 10.9 new episodes per hundred people per year) [5], half of which occur among children under 5 years of age. Incidence varies by region, ranging from 3.6 to 43.3 episodes per 100 population per year in Central Europe to Sub-Saharan Africa, respectively [5].

Most cases of OM resolve spontaneously, but some children experience recurrent OM and CSOM, which can cause ongoing episodes of pain and fever as well as long-term distress to children and their parents [1,2]. These complicated forms are relatively prevalent; before the era of pneumococcal vaccination, approximately 20% of children experienced recurrent OM in their first year of life [6,7], and this may still be the case in countries with limited health resources. The incidence of CSOM varies from 1.7 to 9.4 cases per 1,000 people across the 21 World Health Organization (WHO) regional areas [5]. Recurrent OM, CSOM, and OM-related tympanic membrane rupture can lead to temporary or permanent hearing loss, which can affect social interactions and developmental outcomes in childhood [8–10]. In rare cases, the infection can lead to serious complications such as meningitis, mastoiditis, and intracranial and extracranial abscesses. In 2005, an estimated 21,000 individuals died from OM complications globally, with most of these deaths occurring in low-income countries [5].

Current clinical practice guidelines in the United States (US) recommend using antibiotics to treat AOM with severe symptoms or otorrhea and, for milder cases, analgesics and watchful waiting (i.e. delaying medical therapy or intervention in hope of spontaneous resolution) [11]. Antibiotics are not recommended for treating OME; persistent OME may be treated by inserting ear tubes to keep the middle ear aerated (i.e. tympanostomy) [12]. Eradication of CSOM often requires surgical procedures, namely mastoidectomy, tympanoplasty, or both.

1.2. Pneumococcal conjugate vaccines and their efficacy/effectiveness in preventing OM

Pneumococcal conjugate vaccines (PCV) have been shown to be clinically effective at reducing the burden of OM [13–18]. PCVs contain purified capsular polysaccharides of pneumococcal serotypes conjugated to a carrier protein. The selection of serotypes for the first PCVs was based on those which were the most common and invasive at the time (currently there are ~100 immunologically distinct S. pneumoniae serotypes). The first was a seven-valent PCV (PCV7) containing polysaccharides of seven serotypes (Figure 1) conjugated to diphtheria CRM₁₉₇ protein, which became available first in the US in 2000. Efficacy of PCV7 against invasive pneumococcal disease [19], pneumonia [20], and OM [21] in children was demonstrated in randomized controlled trials (RCTs). By 2008, PCV7 was licensed in approximately 90 of 193 WHO member states [22].

Figure 1. Current and investigational pneumococcal conjugate vaccines (not exhaustive). *For Synflorix® (10-valent PCV), serotype 18C is conjugated to tetanus toxoid and serotype 19F is conjugated to diphtheria toxoid; the remaining eight serotypes are conjugated to protein D from H. influenzae. Abbreviation: PCV, pneumococcal conjugate vaccine.

In 2010, to provide expanded S. pneumoniae serotype coverage, PCV7 was replaced by a 13-valent PCV (PCV13)¹ using the same carrier protein (Figure 1). PCV13 was licensed based on immunobridging to PCV7, which showed that the vaccine elicited a non-inferior immune response to PCV7 for the seven common serotypes, and based on a robust immune response to the six additional serotypes (1, 3, 5, 6A, 7 F, 19A) [23,24]. Because there is no correlate of protection for PCVs against OM, PCV13 is indicated in some countries (including the US) for preventing OM caused by the seven original serotypes in PCV7 (i.e. based on results from a PCV7 efficacy trial [25]). The clinical impact of PCV13 on OM has been established through post-licensure data and effectiveness in real-world settings [15,16,26].

A 10-valent PCV (PCV10), which uses the carrier protein H. influenzae protein D, has also been used in several countries since 2009 for infant vaccination. PCV10 was licensed based on non-inferior immunogenicity against PCV7, and its efficacy against AOM was confirmed in post-licensure RCTs [27].

Most recently, in late 2019, a CRM₁₉₇-conjugated 10-valent PCV, manufactured by Serum Institute of India, achieved prequalification by the WHO following clinical trials conducted in India and The Gambia [28–31]. An alternative 13-valent PCV, manufactured by Walvax Biotechnology, is also now licensed in China (Figure 1). While these two new PCVs showed non-inferior immunogenicity to existing products [28,32], their efficacy and effectiveness against OM and other pneumococcal disease is yet to be determined.

PCV7, PCV10, and PCV13 have reduced circulation of both invasive and noninvasive disease-causing serotypes since their introduction into pediatric national immunization programs (NIPs) [33–35], which has resulted in clinically relevant reductions in OM [13,14,16,36–39]. Both PCV10 and PCV13 have demonstrated an incrementally higher impact on OM than PCV7 due to their broader serotype coverage [13–18,40]. Since its introduction, PCV13 has considerably reduced the incidence of OM episodes [13,14] and has been linked with reduced carriage of vaccine serotypes [34,41–43], including two of the most serious disease-causing serotypes, 19A and 7 F [33,34,43]. Consistent with this, the transition from PCV7 to PCV13 is associated with decreased OM-related ambulatory and emergency department visits [44,45], decreased rates of tympanostomy among children [17,46], and decreased rates of antibiotic prescription for AOM [14,47,48] and antibiotic-resistant OM cases [18,49–53]. Similarly, PCV10 has been associated with delayed first OM episode in children [54] and lower outpatient antibiotic use [55] than in the PCV7 era.

PCV13 may also provide protection against OM caused by non-vaccine serotype pneumococci and other otopathogens. Protection from PCV13 may limit mucosal damage inflicted by early episodes of OM, which may in turn reduce biofilm formation and OM recurrence [36,37]. Specifically, some countries have found parallel decreases in pneumococcal and non-pneumococcal OM episodes, such as those caused by NTHi, Moraxella catarrhalis, and Streptococcus pyogenes [36,37]. Similarly, an early formulation of PCV10 was shown in an RCT to reduce AOM episodes with NTHi isolated as the offending pathogen [56]; however, real-world [57,58] and post-licensure clinical trial data [59–61] have not shown effectiveness with this vaccine against OM due to NTHi.

1.3. Limitations of current PCV cost-effectiveness analysis for OM

Cost-effectiveness analyses (CEAs) of PCVs have traditionally evaluated only direct health gains, healthcare costs, prevented productivity losses among vaccine recipients, and prevented productivity losses in the community (Figure 2) [62–65]. For OM, most CEAs of PCVs have estimated AOM cases prevented and direct health system costs avoided, based on assumptions of PCV effectiveness against vaccine- and non-vaccine serotypes or against all-cause AOM [63,64,66–68]. Furthermore, most studies have not accounted for PCV effects on severe OM or its sequelae. The few CEAs that included long-term consequences of OM as a model input only considered hearing loss, captured as a quality-adjusted life year decrement [69–71]. Finally, many CEAs adopt a healthcare payer, rather than a societal perspective, and omit longer-term improvements in child health, child and caregiver quality of life (QoL), educational outcomes, and the impact of reduced antimicrobial resistance (AMR) from PCV use [63–65].

Figure 2. Only the tip of the iceberg is considered in CEAs of PCV impact on OM. Abbreviations: AMR, antimicrobial resistance; CSOM, chronic suppurative otitis media; OM, otitis media; PCV, pneumococcal conjugate vaccine; QoL, quality of life.

Although PCVs have reduced mortality and morbidity globally, the overall level of PCV uptake and access could be improved when benchmarked against other pediatric vaccines [72]. Revealing their full value, by including both narrow and broader benefits in CEAs, may help address this. Guidelines from the International Society for Pharmacoeconomics and Outcomes Research (ISPOR), WHO, and the European Vaccine Economics Community all advocate that broader benefits now be included when performing CEAs of vaccines, where appropriate, and that this should consider narrow and broad cost offsets resulting from reduced healthcare utilization [73–75]. Given these guideline updates, this review explores the broader burden of OM on children, caregivers, and society to better inform future economic evaluations of PCVs and to help convey the full value of PCVs to decision makers.

2. Health-related outcomes and costs of OM in children

2.1. Hearing loss and child development

Frequent or severe attacks of OM increase the risk of hearing loss in children. Children with OM tend to have average hearing thresholds of more than 20 dB [76–80], suggesting at least mild hearing loss by conventional classifications [81]. More severe hearing loss is associated with recurrent OM and CSOM [8,10,82–84], cholesteatoma (a benign skin growth behind the ear drum usually caused by recurrent OM) [80,85], and OM in children with cleft lip and palate [86]. In a cross-sectional study of Nigerian schoolchildren, mild-to-moderate conductive hearing loss was detected in around two-thirds of CSOM cases, with children from lower socioeconomic backgrounds being most vulnerable [10]. Hearing loss may last for several weeks or months before an OM episode resolves; occasionally, hearing loss can continue into adulthood [76,77,87].

Few studies have determined the prevalence of permanent hearing loss caused by OM. One estimate suggests OM-associated permanent hearing impairment (defined as permanent hearing loss for best ear >25 dB) had a global prevalence of 30.8 per 10,000 in 2005 [5], the prevalence being relatively low in higher-income countries (typically <4 per 10,000) but much higher in some lower-income regions, for example South Asia (97.0 per 10,000) and Sub-Saharan Africa (30–35 per 10,000). In South Asia, by 5 years of age, six out of 1,000 children were estimated to have permanent hearing impairment caused by sequelae of OM, whereas this occurred in less than one out of 10,000 children in higher-income regions of North America, Europe, and Australasia.

OM-related hearing loss can affect speech and language development, leading to impaired social interaction and learning in children [10,82,88,89]. Even temporary hearing loss during a critical period of a child’s learning may have long-lasting effects on the development of the brain’s speech and language centers [90–92]. OM can cause children to have considerable difficulties perceiving teachers and hearing peers in noisy classrooms, which can cause fatigue from the increased effort needed to listen [93]. Together, impaired hearing and communication can impede learning and necessitate extra educational support and healthcare resources, such as speech therapy. If left untreated, these long-term sequelae of persistent or severe OM might ultimately limit future career prospects [72].

2.2. Quality of life in children

While OM-related reduction in QoL has been included in CEAs, it is usually captured as a quality-adjusted life year decrement [66,94–97]. Other types of patient burden are excluded, such as absenteeism from school or nursery and the long-term impact on QoL in adults who suffered from severe or recurrent OM as children. Evidence suggests that recurrent and severe OM can harm children’s QoL by negatively affecting auditory processing and communication, school readiness, social competence, psychosocial wellbeing, movement and balance, and sleep (Figure 3) [98–105]. Health-related QoL in children with OM is usually evaluated using the Otitis Media-6 (OM-6) questionnaire, which includes six questions that are completed by caregivers. Studies using the OM-6 have shown that OM moderately reduces overall QoL [99–102], and one prospective cross-sectional study reported a significantly higher OM-6 score (i.e. lower QoL) in children with recurrent or chronic OM than in healthy age-matched controls [100]. Studies using the OM-6 also suggest that OM affects short-term aspects of QoL to a greater extent than long-term aspects [103–105]. Studies aimed at assessing the burden of OM on QoL may, however, be biased toward the immediate effects of AOM over longer-term effects like speech impairment, hearing loss, and emotional distress.

Figure 3. Impacts of recurrent and chronic otitis media (OM) on children’s and parent/caregivers’ quality of life.

Longer-term effects might progress even after resolution of OM and may not be captured during the study period. A recent systematic review found that few studies have examined the impact of OM on QoL in children and that their methodological quality and outcome measurements vary [98]. Additionally, a 2019 World Development Report estimated that over 2.4 million disability-adjusted life-years were lost due to otitis media [106]. Given both long-term sequelae and death are more frequent with CSOM than AOM, most life-years lost and life-years with a disability result from chronic infections.

2.3. Healthcare costs and resource utilization

There is substantial health resource utilization associated with more complicated OM cases, which is rarely taken into account by CEAs of PCVs. A cost-of-illness study conducted in Belgium found that public healthcare payer costs for long-term complications and sequelae of AOM ranged from €119 to €7,957 per patient per year [107]. For the 25 patients included, the total cost amounted to €250,000 (third-party plus patient perspective costs) over an 11-year follow-up period after the first AOM event. In addition, OM is a leading reason for pediatric ambulatory visits. In 1997–1999, just before the introduction of PCV7, OM-related visits accounted for one in every 10 ambulatory visits in children in the US [44]. Patients with chronic OM also have a higher risk of developing comorbidities that deplete additional health system resources, such as sudden sensorineural hearing loss [108], and nosocomial or other infections, for example following a tympanostomy [109].

Evidence of healthcare resource utilization from low-to-middle-income countries (LMICs) is less available. Generally, CSOM cases are less likely to be seen in these countries due to limited access to health services and scarcity of ear, nose, and throat specialists [110,111].

3. Health-related outcomes and costs of infant OM for caregivers

3.1. Quality of life in caregivers

Recurrent and severe OM affect not only a child’s QoL, but also that of their parents and caregivers (Figure 3) [112,113]. Caregiver QoL is rarely included in CEAs of PCVs. Episodes of OM in children often force changes in their parents’ daily activities by adding medical responsibilities, for example, attending clinical appointments and managing the episode at home [99,103,114–117]. Associated absenteeism from work can lead to lost earnings, amplifying their distress [118]. OM and its sequelae can also interfere with parents’ sleep and create emotional distress, especially anxiety [115,116]. For example, beyond the immediate distressing symptoms of OM, parents may be concerned about their child’s behavior and learning. Further, chronic OM can lead parents to doubt treatment efficacy and their own competency to manage the illness, experience feelings of stigmatization and social isolation, and become weary in coping with illness management and treatment [116]. Because parents are often poorly informed about OM, they may also feel uncertain about how best to help their child [119].

3.2. Societal productivity loss

OM, especially chronic and recurrent OM, can result in negative long-term consequences for society and the economy (Figure 4). Caring for a child with OM may result in missed workdays, lower productivity, and income loss. The high incidence of OM translates this into a large societal burden through lost productivity and decreased QoL [120]. In an online interview study conducted in 12 countries with parents whose children experienced an AOM episode (N = 1438), 73% reported absence from work or having to rearrange working hours to care for their child [113]. Of those who took leave, 67% stayed at home for 2–7 days. A US model-based analysis estimated the indirect cost per AOM episode in the 3 months from diagnosis was approximately $1,200 (in 1999 USD) due to caregiver work loss and transportation [118]. These indirect costs accounted for nearly 90% of the 3-month total cost of AOM treatment. In a retrospective analysis of the Canadian pediatric PCV program, $990 million (in Canadian dollars) of indirect costs were saved over a 10-year period alone, with these indirect cost savings attributed to saved hours of lost productivity among caregivers [121]. In addition to these substantial societal benefits, the study estimated approximately 3.7 million AOM episodes were averted and $627 million of direct OM-related costs were saved. The societal burden is typically higher in LMICs than in developed countries [5,122]. For example, in Malaysia, caregivers missed an average of 21 hours from work to care for their child with AOM and experienced reduced productivity at work [103]. Associated income losses may cause particular hardship in LMICs, where absenteeism, consultations, and treatment expenses can take up a considerable proportion of household income [103,111].

Figure 4. PCVs help prevent long-term consequences of otitis media. Abbreviations: AOM, acute otitis media; OM, otitis media; PCV, pneumococcal conjugate vaccine; QoL, quality of life.

4. Community or health system externalities of OM prevention

4.1. Antimicrobial resistance

As the most common condition leading to the prescription of antibacterial agents, AOM is a key contributor to global AMR [2,11,123]. Clinicians in many countries continue to prescribe broad-spectrum antibiotics to manage mild, initial AOM episodes, even though most clinical practice guidelines recommend watchful waiting [11]. AMR may be of particular concern in developing countries, where there are often no guidelines for antibacterial treatment of OM and antibiotics may be easily purchased without mandatory medical prescriptions [124,125].

Studies show that pneumococci, especially serotypes 19A, 15A, and 35B, have become increasingly resistant to antibiotics [18,42]. A systematic review and meta-analysis by Mather et al. [126] found that 15% of AOM cultures were resistant to amoxicillin, an antibiotic commonly used to treat OM. If these trends are compounded over time and geographic regions, OM could become increasingly resistant to antibiotics, leaving fewer treatment options [18,127].

Pneumococcal vaccination early in life helps reduce AMR by preventing initial OM episodes, in turn reducing cases of chronic and recurrent OM that require antibiotic treatment [37] (Figure 4). Real-world studies show reduced rates of antimicrobial-resistant vaccine-type pneumococcal infections and lower rates of antibiotic prescription since the introduction of PCV7, PCV10, and PCV13 [14,18,47–49,127], yet these impacts are consistently excluded from CEAs.

4.2. Macroeconomics

CEAs of PCVs and other vaccines in pediatric immunization programs rarely consider positive impacts from vaccination on the macroeconomy. By directly improving child health and providing indirect economic benefits, these vaccines help stimulate changes in household behavior (e.g. through increasing household savings), allow public sector budgets to be allocated elsewhere, and increase gross domestic product, especially in LMICs [72,128]. Improvements in population health through vaccination increase foreign investment and can improve workforce productivity [129]. For example, in China, vaccinations given before 15 years of age have been shown to provide long-term benefits on non-health outcomes such as education and cognitive skills, with vaccinated individuals enjoying around one more year of schooling, and performing substantially better on several cognitive tests later in life [130].

4.3. Social equity

For children living in impoverished communities and countries, there is often an excess of vaccine-preventable mortality and morbidity [131]. Widely used vaccines, including PCVs, may improve social equity (e.g. fairness in the distribution of health within a population) provided that vaccination coverage among the poorest of the population is sufficient and vaccines are accessible [132,133]. The use of PCVs in New Zealand has been associated with reduced ethnic and socioeconomic disparities in OM hospitalizations [134]. Similarly, in a US study, socioeconomic and racial disparities in the incidence of invasive pneumococcal disease were found to be lower after the introduction of PCV13 than during the pre-PCV7 era [133]. To continue improving social equity, attention is needed to ensure that PCV programs target disadvantaged individuals and communities, and that these programs are distributed to children of all socioeconomic backgrounds equally.

5. Conclusion

National governments and decision makers worldwide face a common dilemma regarding the efficient appropriation of limited healthcare resources for the maximal benefit of the population. Interventions that target disease prevention, especially vaccination, are preferred to waiting until a disease manifests and treatment is needed. Given the high prevalence of OM, PCVs have substantial positive externalities compared with other health technologies because the benefits are realized across the entire population.

OM causes substantial morbidity, economic losses, and harm to QoL globally, yet many broader impacts of OM on children, their families, and society are overlooked, some of which last far beyond the acute disease phase. Recurrent and chronic forms of OM can result in long-lasting hearing loss, which impairs children at a crucial age in their speech and language development and impacts their QoL and educational achievement. Caregivers also suffer due to emotional distress, income loss, and weariness from managing and treating the illness. Because it is so prevalent, OM additionally burdens society through absenteeism and productivity losses, and can contribute to increased rates of AMR due to inappropriate or excessive antibiotic use – leaving physicians with fewer treatment options for bacterial infections. These broader impacts are especially prominent in developing countries, which have higher prevalence of recurrent OM, CSOM, and resulting sequelae, and fewer health resources to manage them.

Infant immunization with PCVs helps prevent initial episodes of OM and, therefore, breaks the chain of events leading to long-term economic, clinical, humanistic, and societal consequences of chronic and recurrent OM. Conventional CEAs of PCVs have not fully accounted for this effect, meaning that the true value of PCVs to society has been underestimated. Decision makers, therefore, may not be informed on the vaccine’s broader societal benefits, which can lead to undervaluation of PCV programs, suboptimal allocation of resources to PCVs in NIPs, exclusion of PCVs from NIPs, and low PCV uptake rates [62,128,135,136]. Understanding and incorporating broader effects of OM as an outcome of economic evaluations, and highlighting the need for capturing these outcomes, will be important for improving access to PCVs. In parallel, efforts are needed, especially in LMICs, to collect more local country health data on broader societal impacts, as well as sero-epidemiologic data, to optimize future economic analyses.

6. Expert opinion

To help inform future health technology decisions, PCV economic evaluations must evolve to include broader value elements, including societal health impacts and costs of OM among patients, caregivers, and the larger population. Guidelines from ISPOR for the economic evaluation of vaccines recommend that CEAs capture a disease’s long-term effects [73], which is especially relevant for modeling the impact of PCVs on OM. Recent guidelines also advise using a societal perspective as the preferred base case in CEAs [74,75], or reporting outcomes from both payer and societal perspectives [73]. Within the societal perspective, estimating effects on productivity is recommended to account for improvements in children’s future capabilities (e.g. development, education, and work), reductions in AMR, and averted productivity loss of caregivers, to name a few. Additionally, the guidelines recognize that macroeconomic effects of vaccines can be analyzed, although it is likely to require alternative or complementary methodologies to those used in traditional health technology assessments. For example, fiscal health modeling could be used to estimate how a vaccine relates to changes in tax revenues and transfer costs attributable to morbidity and mortality [73].

More recently, the Office of Health Economics published a narrative for Broad Recognition of Value in Vaccines Engagement (BRAVE) [137] in which five priority gaps were identified as being either excluded or not consistently considered in vaccine health technology assessments: broad cost-offsets at the community-level; effects on carers’ health; transmission value; effects on AMR; and macroeconomic effects (Figure 5). For evaluations of PCVs against OM, bridging these priority gaps will require decision makers to recognize the importance of broader vaccine impacts, along with the development of improved metrics, models, and case studies for quantifying the burden of OM on child development, carers’ health, and the macroeconomy. Additionally, the ongoing coronavirus disease 2019 (COVID-19) pandemic provides further learnings on the broader societal effects of vaccination [137,138] that may encourage policymakers to incorporate the broader long-term and societal impacts of PCVs in economic evaluations in the future. Nonetheless, benefits from PCVs depend on other aspects, such as local epidemiological, socioeconomic, environmental, and health system factors, which may ultimately require additional investment in the health sector (e.g. investment into other public health interventions and vaccines) and other sectors (e.g. education and infrastructure) [136]. Therefore, policymakers will need to ensure that resource allocation is based on an intervention’s full value, including broader societal benefits, and the opportunity costs that exist for society by including the new intervention. As argued by others [136], economic evaluations of vaccination should be considered alongside these measures rather than individually.

Figure 5. Applying the five BRAVE priority gaps to recognizing the broad value of PCVs against OM. Abbreviations: AMR, antimicrobial resistance; CEA, cost-effectiveness analysis; OM, otitis media; PCV, pneumococcal conjugate vaccine; QoL, quality of life.

Currently, there are data limitations for including broader societal impacts associated with OM prevention in CEAs of PCVs because these impacts have rarely been quantified. More country-level evidence on the broader impacts of OM and vaccination is especially needed in LMICs to defend future investments in vaccines and immunization programs [135]. A challenge in building this evidence will be the collection of broad cost-offsets and population health data. However, innovative analyses of historical observational datasets could provide evidence to inform modeling assumptions – for example, data on the national incidence and prevalence of recurrent OM and CSOM over time. Studies linking changes in health to long-term behavioral and developmental outcomes have been suggested, although such studies should incorporate analysis techniques that reduce potential biases [135]. Prospective follow-up studies are also needed to measure broader impacts of OM that go beyond the initial infection, which may provide valuable missing long-term clinical and economic data from patients and caregivers to inform CEAs. Moreover, web-based surveys could be conducted to examine how OM affects parents/caregivers QoL, out-of-pocket expenses, and productivity losses. Researchers will also need to consider whether ‘double counting’ may arise by including both productivity loss and QoL reduction, given the influence income reduction may have on utility scores. Acknowledging this, although income loss may somewhat affect utility scores, this does not offset the substantial societal cost due to a patient’s or caregiver’s productivity loss. Another priority is to generate high-quality evidence on the magnitude of PCV impact on AMR, which would likely require reinforced national surveillance of resistant infections [137]. In the short-term, economic evaluations should acknowledge these data limitations and make informed assumptions regarding uncertain input parameters. For example, input parameters or ranges recommended by an expert panel could be used. Finally, the information needs of stakeholders from different sectors (including health and finance sectors, and external donors) should be collected to guide the incorporation of broader benefits into future economic evaluations and assist with their effective communication to decision makers.

Another challenge is that there are few active laboratory-based and population-based surveillance systems for OM. Consequently, the causal pathogen(s) and/or specific strain causing most OM cases is often unknown at a population level and incidence can vary across studies within a nation or region. Many economic analyses use modeling assumptions to overcome these data barriers. For example, studies will assume that a proportion of all OM is due to pneumococcus or that the underlying pneumococcal serotypes causing OM are the same as the invasive pneumococcal disease serotype distribution. Additional clinical studies and surveillance activities are needed to inform these assumptions and model the full burden of OM more accurately.

In the absence of data on vaccine effectiveness against broader OM outcomes, any estimates of these outcomes must also consider confounding factors. For example, various other factors may affect child speech and language development, AMR, and workforce productivity. Similarly, other interventions beside PCVs, such as influenza vaccines [139], prevent a small proportion of OM outcomes and should be separated in model-based economic evaluations.

Both PCV13 and PCV10 have shown greater reductions than PCV7 in vaccine-type pneumococcal carriage, OM episodes, and OM-associated healthcare utilization through broader serotype coverage [13–18,40–42,44,45,54,55]. However, countries with established PCV NIPs have observed increased carriage and disease from non-vaccine serotypes that were initially less prevalent across all age groups [18,35,140], including some serotypes with reduced antimicrobial susceptibility [127,141]. The potential licensure of new higher valent PCVs in the next few years will extend coverage to additional serotypes that have begun to emerge globally (Figure 1) [142]. Specifically, a 15-valent PCV (PCV15; Merck) that includes serotypes 22F and 33F is expected to become available. However, as a new vaccine, its impact on OM will need to be ascertained through clinical trial and real-world studies. Additionally, a next-generation 20-valent PCV (PCV20; Pfizer) is anticipated to be licensed for pediatric use to offer even broader coverage of pneumococcal serotypes by extending PCV13 coverage to serotypes 8, 10A, 11A, 12F, 15B, 22F, and 33F. Future economic evaluations of these and other higher-valent vaccines will need to incorporate learnings from established PCV NIPs, for example national age-specific serotype trends for PCV and non-PCV serotypes. Higher-valent PCVs are expected to continue the decline in pneumococcal carriage, disease, death, and AMR. However, the magnitude of the decline will depend on their uptake in NIPs, which will in turn depend on providing decision makers with economic evaluations that communicate the full value of the vaccine program to society.

Declaration of interest

All authors are employees of and may hold stock or stock options in Pfizer Inc. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Author contributions

All authors substantially contributed to the conception and design of the review article and interpreting the relevant literature, and all were involved in writing the review article or revised it for intellectual content.

Article highlights

• Otitis media (OM) is one of the most common diseases in young children worldwide and a leading reason for physician visits and antibiotic use.

• Most cases of OM resolve spontaneously, but recurrent and severe forms can have substantial long-term impacts such as hearing loss, impaired language and educational development, and lower patient and caregiver quality of life.

• Pneumococcal conjugate vaccines (PCVs) have contributed to a decline in OM incidence and antimicrobial-resistant vaccine serotypes.

• Economic evaluations of PCV pediatric national immunization programs traditionally measure only the direct, short-term health benefits and healthcare cost savings associated with acute OM, and therefore undervalue the full benefits of PCVs.

• Several health and economic bodies have called for the broader impact of vaccines to be recognized and incorporated into cost-effectiveness analyses.

• Including the full burden of OM in economic evaluations is important to communicate the full value of PCVs for informed decision making.

Reviewer disclosures

A reviewer has disclosed that they have previously received a grant from MSD to support analysis of our data on pneumococcal serotypes. Peer reviewers on this manuscript have no other relevant financial or other relationships to disclose.

Acknowledgments

Medical writing was provided by Dr. Jonathan Pitt (Evidera, Paris, France) and funded by Pfizer Inc.

Additional information

Funding

This manuscript was funded by Pfizer Inc.

Notes

1. Though other manufacturers are developing PCV13 and PCV10 vaccines, in this article, PCV13 refers to the 13-valent PCV (Prevnar/Prevenar 13®) marketed by Pfizer, and PCV10 refers to the 10-valent PCV (Synflorix®) marketed by GlaxoSmithKline, unless stated otherwise.