Introduction
Pediatric patients need access to safe and effective therapeutic products for the overall health and safety of the population. For new technologies to reach children, several steps must be successfully navigated. Sponsors must first decide to pursue regulatory approval and commit to conducting clinical trials that generate evidence for this purpose. Regulatory authorities then assess the safety and efficacy of the proposed new technology and determine if approval is warranted for a specific indication and population. If regulatory approval is granted, an assessment by a health technology assessment (HTA) agency often follows.
HTA evaluations are standard in Europe, Canada, South America, and Asia. HTA agencies examine the evidence package provided to them by the sponsor and evaluate the comparative cost-effectiveness of the proposed new technology vis-à-vis market alternatives. These economic evaluations aim to quantify the broader impact of a new intervention, often looking beyond safety and efficacy to domains including changes in quality of life (QOL) and other patient-reported outcomes. This allows them to estimate the cost–benefit ratio and consider how reimbursement fits in with the social values of their jurisdiction. Recommendations are then made to healthcare payers – often national governments – on whether the new intervention will generate sufficient benefit to the country as a whole to justify the cost. While the weight of HTA decisions and their legal status differ, in many places, a negative HTA opinion renders a drug essentially inaccessible to the public (Reference Kawalec, Sagan and Pilc1).
Producing the type and quality of data required for pediatric HTAs is challenging. Children have historically been excluded from clinical research due to concerns for their safety (Reference Lepola, Nelson, Matsui, Gasthuys, Allegaert, Dossche and Turner2). Compounding this, many pediatric diseases are rare, rendering the recruitment for and conduct of pediatric clinical trials challenging (Reference Aballéa, Thokagevistk and Velikanova3;Reference Agashe, O’Day, Arvin-berod, Meyer and Bramley4). Children’s dependence on parents or other caregivers, the lack of standardized methodologies for capturing QOL in children, and the possible lifetime use of a product are just a few of the factors that make quantifying the value of a pediatric intervention more complex. While recent regulatory changes have attempted to address challenges by incentivizing pediatric research (Reference Bourgeois and Hwang5;Reference Rivera and Hartzema6), adequate data generation for HTA purposes remains difficult (Reference Denburg, Giacomini, Ungar and Abelson7).
This scoping review aims to synthesize the literature on pediatric HTAs and map out the key challenges of assessing new technologies for use in children, with a particular emphasis on pharmaceutical interventions. Results are intended to identify knowledge gaps that would benefit from increased scientific inquiry and provide new insights to guide future policy change.
Methods
This review was registered on the open science framework before study initiation (Reference Hutchinson8) and conducted in accordance with the Joanna Briggs Institute Methodology for scoping reviews (Reference Peters, Godfrey, McInerney, Soares, Khalil and Parker9).
Information sources
We identified reports and reviews addressing HTAs in the pediatric domain by searching the electronic databases PubMed (NCBI), Embase (Elsevier), Web of Science Core Collection (Clarivate), and EconLit (EBSCO). The search, created and carried out by a medical librarian (PAB), included terms for technology assessment and for children or pediatrics (Supplementary Material 1). The indexed contents of the journals Health Technology Assessment and International Journal of Technology Assessment in Health Care were included in the search. Controlled vocabulary terms were included when available; no date or language limits were applied to the search. The search was last updated on January 22, 2024. Duplicate records were removed using EndNote Version 20 (10).
In addition, two authors (NH and LO) performed a gray literature search, in which they independently performed a targeted search and examination of websites of organizations relevant to HTA performance. Searches were conducted from 10/2023 to 11/2023 using the Google search engine, and relevant websites were identified using the International Network of Agencies for Health Technology Assessment member list (11).
Reference lists of included documents were screened for relevant citations, and if not previously captured in the systematic searches, they were evaluated for inclusion.
Search and eligibility criteria
We included published manuscripts, editorials and commentaries, brief reports, book chapters, conference proceedings, and published abstracts. We also included study reports, working papers, government documents, and white papers identified through our gray literature and citation searches.
Selection of sources of evidence
Article titles and abstracts from our searches were imported into Covidence, a web-based collaboration software platform designed to facilitate conducting literature reviews (12). Titles and abstracts were independently screened for full-text review by two authors (NH and LO). The same two authors independently reviewed articles in full text for inclusion. Consensus was reached in instances of discrepancy through discussion.
Evidence sources were included if they addressed pediatric HTAs, including literature reviews of HTAs and related reports of HTA decision-making. Evidence sources were excluded if they (i) were a clinical trial investigating a specific technology or an HTA of that technology, (ii) did not address the challenges of HTAs, and/or (iii) had no relevance to a pediatric population. We also excluded reports in a language other than English or French.
Data charting process and data items
Data extraction from included evidence sources was performed independently and in duplicate by two authors (NH and LO) using a pre-determined data extraction template on Covidence, which was tested for appropriateness and ease of use before deployment. Extracted data included general publication information (corresponding author, country/region, HTA agencies mentioned, regulatory authorities cited, and guidance documents mentioned), characteristics of included studies (study aim, study design, start date, end date, funding sources, and COI disclosure statement), information on data sources (description of data sources, inclusion and exclusion criteria) and potential challenges of pediatric HTAs (as described in detail below).
Synthesis of results
Once data charting was complete, data were exported from Covidence as CSV files. No quality assessment of studies was performed. Given the nature of our review, data analyses were descriptive, and results were thematically reorganized into data tables by two authors (NH and LO). Each data table served to summarize a subject area that presented particular challenges to the conduct of pediatric HTAs, discussed in order of the frequency with which they were cited: (i) QOL, health-related quality of life (HRQoL), and utility values; (ii) caregiver and family burden; (iii) data and endpoints; (iv) uncertainty; (v) population size; (vi) elicitation of external perspectives; (vii) societal values and ethics; (viii) HTA methods and processes; (ix) extrapolation; (x) population heterogeneity; (xi) price of intervention; (xii) comparator; and (xiii) costs outside the healthcare sector.
Results
Selection of sources of evidence
Electronic database searching returned 3164 records. In addition, the gray literature search added twenty-nine records; the examination of citations added eight records. After removing duplicate records, 2361 remained to be screened, of which 2215 were excluded based on abstract review. One hundred and forty-six reports were retrieved and examined for inclusion in the study based on full-text review (when available). One hundred and three reports met our inclusion criteria (Supplementary Table 1). Of these, sixty were full journal articles, twenty-three were conference abstracts, and twenty were guidelines, reports, and other documents (Figure 1).
PRISMA flow diagram.
Study characteristics
Of the 103 studies that met our inclusion criteria, most were conducted in Europe (77 (74.8 percent)) (Table 1). Most sources were published in the past 5 years (2020 – present; 66 (64.1 percent)). Over one-third of the included sources were reviews of cost-utility analyses or of HTA agency appraisals, guidance, and policy (43 (41.7 percent)). Over half (53.4 percent) of the included studies did not specify the type(s) of interventions addressed or addressed multiple intervention types; 43.7 percent focused specifically on drug and/or biologic (13) interventions.
Profile of included sources
a Categories are not mutually exclusive.
b Including 1 modeling study, 1 cross-sectional study of clinical trial data, 2 mixed methods studies, and 1 cohort study evaluating utility measures.
Challenges of conducting pediatric health technology assessments
QOL, HRQoL, and utility values
Forty-five sources discussed the assessment of QOL, HRQoL, and the generation of utility values as a key challenge of conducting pediatric HTAs (Reference Aballéa, Thokagevistk and Velikanova3;Reference Denburg, Giacomini, Ungar and Abelson7;Reference Ungar, Prosser and Burnett14–Reference Pennington57). Primary concerns included the difficulty in generating HRQoL data in children due to a lack of guidance on how this should be done (Reference Powell, Rowen, Mukuria, Keetharuth and Wailoo48), who should be asked (i.e., parents, clinicians, or children themselves) (Reference Denburg, Giacomini, Ungar and Abelson19;Reference Mpundu-Kaambwa, Bulamu and Lines25;Reference Noyes and Edwards28–Reference Fahfouhi, Thorrington and Ghabri30), or which tool should be used (50–52). Additional concerns were related to the limits of self-reporting of health in young children due to their tendency to think in extremes, resulting in evaluations on either end of a scale (Reference Bégo-Le Bagousse, Jia and Wolowacz16) or a lack of life experience, limiting their frame of reference (Reference Dewilde, Janssen, Lloyd and Shah34). Despite being frequently employed (Reference Lovett, Cooper and Lamb46;Reference Purushotham, Brown and Francmanis53), HRQoL instruments developed in adults and applied to children lacked important domains relevant to the childhood experience, for example, the ability to play outside with friends or feel confident at school (Reference Eiser and Morse56).
The translation of HRQoL data into utility values, representing patient preferences on a particular health state, ranging from 0 (death) to 1 (full health) (Reference Pennington57), raised several additional challenges. Authors highlighted the lack of guidance from HTA agencies about which techniques to use to calculate utilities in pediatric populations (Reference Bégo-Le Bagousse, Jia and Wolowacz16;Reference Hill, Rowen, Pennington, Wong and Wailoo18;Reference Sutherland, Davies and Sully37), despite evidence that adopting different techniques produced different reimbursement recommendations (Reference Meregaglia, Nicod and Drummond38) and that inadequate utility data has been cited as a reason not to reimburse a new technology (Reference Sadeuk-Benabbas, Autin, Couillerot and Clément39). Studies noted that even when child-specific HRQoL outcomes were employed, translating these into utilities often relied on data from adult studies (Reference Montgomery and Kusel15); thus, assumptions regarding the similarity between adult and childhood health experiences were embedded in the tools used to measure them (Reference Hill, Rowen, Pennington, Wong and Wailoo17). See Supplementary Table 2 for a list of utility measures discussed in the included sources.
Caregiver and family burden
Twenty-nine sources discussed the challenges of quantifying and incorporating caregiver and family burden into pediatric HTAs (Reference Aballéa, Thokagevistk and Velikanova3;Reference Ungar, Prosser and Burnett14;Reference Bégo-Le Bagousse, Jia and Wolowacz16;Reference Denburg, Giacomini, Ungar and Abelson19–Reference Conti, Gruber, Ollendorf and Neumann21;Reference Nicod, Lloyd and Morel24;Reference Cleemput and Neyt27;Reference Dawoud, Lamb and Moore29;Reference Costa and Ungar55;Reference Pennington57–Reference Richardson, Tuson, Mikudina, Pownell and Large75). Certain bodies, including the Haute Authorité de santé in France (74) and the National Institute for Health and Care Excellence (NICE) in England (Reference Ellicott, Homer, Moor and Farrington59;66;Reference Mikudina, Robertson and Upadhyaya72), have specific guidance on including caregiver burden in assessments. However, even among HTAs conducted by these agencies, in practice, only a minority of HTAs considered the effect of pediatric health interventions on caregivers (Reference Bégo-Le Bagousse, Jia and Wolowacz16;Reference Scope, Bhadhuri and Pennington58;Reference Fornaro, Drummond, Ciani and Jommi60;Reference Pennington and Wong63;Reference Ferizovic and Marshall73;Reference Richardson, Tuson, Mikudina, Pownell and Large75), although doing so was more common in pediatric HTAs than in adult HTAs (Reference Pennington57;Reference James, Gruber, Vilu and Martin71).
Sources noted a lack of standardized practices across HTAs in eliciting, measuring, and integrating caregiver effects in assessments (Reference Aballéa, Thokagevistk and Velikanova3;Reference Dawoud, Lamb and Moore29;Reference Scope, Bhadhuri and Pennington58;Reference Fornaro, Drummond, Ciani and Jommi60–Reference Lamsal, Yeh, Pullenayegum and Ungar62). Incorporating family and caregiver burden in HTAs increased cost-effectiveness by broadening the scope of who may benefit from a child’s improved health state, making it more likely that an HTA resulted in a recommendation to reimburse (Reference Dawoud, Lamb and Moore29;Reference Pennington57;Reference Scope, Bhadhuri and Pennington58;Reference Lavelle, D’Cruz and Mohit61;Reference Pennington and Wong63). However, incorporating caregiver utilities into HTA analyses could produce counterintuitive results; for example, in one model, the extension of the life of a disabled child was estimated to reduce the lifetime QOL for caregivers when compared to the quick death of the child, despite this outcome being in opposition to the hopes of caregivers themselves (64).
Data and endpoints
Twenty-eight sources raised concerns regarding the volume, quality, and data type available to inform pediatric HTAs (Reference Aballéa, Thokagevistk and Velikanova3;Reference Denburg, Giacomini, Ungar and Abelson7;Reference Conti, Gruber, Ollendorf and Neumann21;Reference Nicod, Lloyd and Morel24;Reference Moretti, Ruiz, Bonifazi, Pizzo and Kindblom33;Reference Freigofaite, Kaproulia, Verhoek and Sarri45;Reference Crossley, Chandler and Godfrey49;Reference Maas, Joos and Hiligsmann54;Reference Costa and Ungar55;Reference Fornaro, Drummond, Ciani and Jommi60;66;67;Reference Whittal, Nicod, Drummond and Facey70;Reference James, Gruber, Vilu and Martin71;Reference Gye, Goodall and De Abreu Lourenco76–Reference Crossley, Chandler and Godfrey89). A lack of high-quality pediatric data was highlighted as an obstacle to accurately measuring the added benefit of new pediatric interventions (Reference Maas, Joos and Hiligsmann54) or performing any pharmacoeconomic value assessment (Reference Gladwell, Lee, Tate, Batty and Brereton77). Proposed solutions included increasing the volume of high-quality data by compelling the industry to undertake further pediatric testing of drugs (Reference Denburg, Giacomini, Ungar and Abelson7) or expanding the definition of valuable data to include data from patient registries, claims information, electronic health records, or natural history and qualitative studies (Reference Denburg, Giacomini, Ungar and Abelson7;Reference Conti, Gruber, Ollendorf and Neumann21;Reference Åkesson, Llewellyn, Bagshaw, Kousoulakou and Larkin79;Reference Cook80). Patient advocacy groups, although often insufficiently resourced, were noted to fill an important informational void created by the limited investment by pharmaceutical companies in pediatric diseases (Reference Denburg, Giacomini, Ungar and Abelson7;Reference Nicod, Lloyd and Morel24;Reference Moretti, Ruiz, Bonifazi, Pizzo and Kindblom33;Reference Crossley, Chandler and Godfrey49;Reference Schoenmakers, Beerepoot and van den Berg84;Reference Crossley, Chandler and Godfrey89).
The widespread use of surrogate outcomes was also seen as a challenge to pediatric HTAs. While the benefit of using single-arm studies with surrogate endpoints to minimize pediatric exposure to experimental interventions was noted (Reference Walzer and Droeschel87), using surrogate endpoints that were not validated (Reference Aballéa, Thokagevistk and Velikanova3;Reference Fornaro, Drummond, Ciani and Jommi60) and that led to more significant uncertainty of long-term safety and efficacy outcomes (Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82) was of great concern. From the perspective of drug developers, the lack of consistency over which surrogate outcomes were deemed acceptable to regulatory bodies and different HTA agencies rendered preparing a single evidence package very challenging (Reference Maas, Joos and Hiligsmann54;Reference Raza, Keyzor and Shohet86).
Uncertainty
Twenty-three sources highlighted uncertainty due to the limitations in available pediatric data as a significant challenge to conducting HTAs in this population (Reference Aballéa, Thokagevistk and Velikanova3;Reference Denburg, Giacomini, Ungar and Abelson7;Reference Montgomery and Kusel15;Reference Lamb, Murray and Lovett20;Reference Conti, Gruber, Ollendorf and Neumann21;Reference Nicod, Lloyd and Morel24;Reference Moretti, Ruiz, Bonifazi, Pizzo and Kindblom33;Reference Freigofaite, Kaproulia, Verhoek and Sarri45;Reference Maas, Joos and Hiligsmann54;Reference Fornaro, Drummond, Ciani and Jommi60;Reference Whittal, Nicod, Drummond and Facey70;Reference Gye, Goodall and De Abreu Lourenco76–Reference Nicod, Whittal, Drummond and Facey78;Reference Chapman, Kumar, Whittington and Pearson81;Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82;Reference Ungar85;Reference Adkins, Nicholson, Floyd, Ratcliffe and Chevrou-Severac90–Reference Karinou, Beugre-Guyot, Bungey and Vilu95). Specifically, the lack of long-term data for treatment effects (Reference Aballéa, Thokagevistk and Velikanova3;Reference Conti, Gruber, Ollendorf and Neumann21;Reference Whittal, Nicod, Drummond and Facey70;Reference Gye, Goodall and De Abreu Lourenco76;Reference Chapman, Kumar, Whittington and Pearson81;Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82;Reference Ungar85;92;Reference Cressman and Denburg93) and safety events (Reference Denburg, Giacomini, Ungar and Abelson7;Reference Maas, Joos and Hiligsmann54;Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82) in pediatric populations was emphasized. The frequent limitations in the quality of pediatric data, emerging as it often does from small trials with heterogeneous patient populations employing surrogate outcomes and without comparator groups, among other limitations, led to uncertainty in the strength of available evidence (Reference Montgomery and Kusel15;Reference Lamb, Murray and Lovett20;Reference Nicod, Lloyd and Morel24;Reference Freigofaite, Kaproulia, Verhoek and Sarri45;Reference Whittal, Nicod, Drummond and Facey70;Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82). Recommendations were proposed to quantify the degree of uncertainty (Reference Gye, Goodall and De Abreu Lourenco76) to assist in HTA decision-making and to conduct extensive sensitivity analyses (Reference Gladwell, Lee, Tate, Batty and Brereton77) to mitigate these challenges. The adoption of supplemental processes to evaluate pediatric rare disease treatments, including conditional approval agreements and integration of subject matter and rare disease expertise in appraisal committees, were proposed to better manage uncertainty in pediatric HTAs (Reference Nicod, Whittal, Drummond and Facey78).
Population size
Twenty-two sources considered the challenges of serving small pediatric populations (Reference Aballéa, Thokagevistk and Velikanova3;Reference Agashe, O’Day, Arvin-berod, Meyer and Bramley4;Reference Denburg, Giacomini, Ungar and Abelson7;Reference Montgomery and Kusel15;Reference Conti, Gruber, Ollendorf and Neumann21;Reference Nicod, Lloyd and Morel24;Reference Moro, Schlander and Telser32;Reference Maas, Joos and Hiligsmann54;Reference Costa and Ungar55;Reference Whittal, Nicod, Drummond and Facey70;Reference Gye, Goodall and De Abreu Lourenco76;Reference Nicod, Whittal, Drummond and Facey78;Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82;Reference Schoenmakers, Beerepoot and van den Berg84;Reference Ungar85;Reference Walzer and Droeschel87;Reference Adkins, Nicholson, Floyd, Ratcliffe and Chevrou-Severac90;92;Reference Brougham, Schlander, Telser, Bakshi and Sola-Morales96–99). Rare pediatric diseases rendered evidence generation difficult, including struggles to recruit enough children to clinical trials to produce adequate statistical power to evaluate treatment effects (Reference Aballéa, Thokagevistk and Velikanova3;Reference Agashe, O’Day, Arvin-berod, Meyer and Bramley4;Reference Denburg, Giacomini, Ungar and Abelson7;Reference Conti, Gruber, Ollendorf and Neumann21;Reference Costa and Ungar55;Reference Schoenmakers, Beerepoot and van den Berg84;Reference Ungar85;Reference Walzer and Droeschel87;Reference Adkins, Nicholson, Floyd, Ratcliffe and Chevrou-Severac90;Reference Brougham, Schlander, Telser, Bakshi and Sola-Morales96–Reference Greasley, Campbell and Wall98), resulting in high evidentiary uncertainty (Reference Whittal, Nicod, Drummond and Facey70;Reference Gye, Goodall and De Abreu Lourenco76;Reference Nicod, Whittal, Drummond and Facey78;Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82). Small populations also made generating validated patient-reported outcome measures more difficult (Reference Nicod, Lloyd and Morel24). A lack of market incentives for industry sponsors to develop and evaluate interventions in pediatric populations, due to the small number of predicted end users, acted as a deterrent to additional pediatric data generation (Reference Denburg, Giacomini, Ungar and Abelson7;Reference Montgomery and Kusel15;Reference Maas, Joos and Hiligsmann54;Reference Brougham, Schlander, Telser, Bakshi and Sola-Morales96). Embedded within traditional cost-effectiveness models were assumptions that the social value of an intervention was proportionate to the number of individuals who may benefit, thus devaluing small pediatric populations (Reference Moro, Schlander and Telser32;99).
Societal values and ethics
Nineteen sources grappled with the ethical and social values relevant to HTA decision-making in children (Reference Aballéa, Thokagevistk and Velikanova3;Reference Denburg, Giacomini, Ungar and Abelson7;Reference Ungar, Prosser and Burnett14;Reference Denburg, Giacomini, Ungar and Abelson19;Reference Conti, Gruber, Ollendorf and Neumann21;Reference Moro, Schlander and Telser32;Reference Dewilde, Janssen, Lloyd and Shah34;Reference Crossley, Chandler and Godfrey49;Reference Guest, Holdgate, Patel, Rouse and Bradley51;Reference Gauvreau, Wight and Subasri65;Reference Rowen, Rivero-Arias, Devlin and Ratcliffe68;Reference Whittal, Nicod, Drummond and Facey70;Reference Mikudina, Robertson and Upadhyaya72;Reference Ungar85;Reference Brougham, Schlander, Telser, Bakshi and Sola-Morales96;Reference Boye, Matza and Feeny100–Reference Thébaut103). Emphasis was placed on children as a unique group requiring additional protections due to their lack of autonomy (Reference Ungar85) and political marginalization (Reference Denburg, Giacomini, Ungar and Abelson7;Reference Denburg, Giacomini, Ungar and Abelson19;Reference Guest, Holdgate, Patel, Rouse and Bradley51). Protectionist approaches, including a low tolerance for risk (Reference Denburg, Giacomini, Ungar and Abelson19), were noted. In addition, a favorable view of treatments that improve childhood life expectancy and quality of life, irrespective of cost-effectiveness, was observed (Reference Moro, Schlander and Telser32). The latter highlighted an important tension in pediatric HTAs, which are tasked with serving the population’s needs while also reflecting societal values that may consider quality-adjusted life years differently at various stages of life (Reference Ungar, Prosser and Burnett14;Reference Petrou102). Choosing to value life differently at one end of the spectrum versus another may be contrary to HTA policy in certain locations. For example, the Equality Act 2010 in the United Kingdom requires that NICE avoid giving special consideration to any particular population subgroup solely due to age (Reference Mikudina, Robertson and Upadhyaya72).
Elicitation of external perspectives
Seventeen sources considered the integration of external perspectives into HTAs (Reference Aballéa, Thokagevistk and Velikanova3;Reference Denburg, Giacomini, Ungar and Abelson7;Reference Ungar, Prosser and Burnett14;Reference Denburg, Giacomini, Ungar and Abelson19–Reference Conti, Gruber, Ollendorf and Neumann21;Reference Nicod, Lloyd and Morel24;Reference Moro, Schlander and Telser32;50;Reference Gauvreau, Wight and Subasri65;Reference Whittal, Nicod, Drummond and Facey70;Reference Gladwell, Lee, Tate, Batty and Brereton77;Reference Greasley, Campbell and Wall98;99;Reference Daniels and van der Wilt101;Reference Thébaut103;Reference Bush, Lenney, Spencer and Warner104). Several highlighted a need for increased pediatric expertise amongst HTA committee members (Reference Aballéa, Thokagevistk and Velikanova3;Reference Denburg, Giacomini, Ungar and Abelson7;Reference Greasley, Campbell and Wall98;Reference Bush, Lenney, Spencer and Warner104), given the unique epidemiologic, social, and developmental features of childhood health and disease (Reference Denburg, Giacomini, Ungar and Abelson7). Calls for increased physician involvement (Reference Aballéa, Thokagevistk and Velikanova3;Reference Denburg, Giacomini, Ungar and Abelson7;Reference Lamb, Murray and Lovett20;Reference Nicod, Lloyd and Morel24;50;99), family/caregiver input (Reference Ungar, Prosser and Burnett14;Reference Moro, Schlander and Telser32;Reference Gauvreau, Wight and Subasri65;99), including the integration of bereaved family members (Reference Denburg, Giacomini, Ungar and Abelson19), pediatric input (Reference Denburg, Giacomini, Ungar and Abelson7;Reference Denburg, Giacomini, Ungar and Abelson19;Reference Nicod, Lloyd and Morel24;Reference Moro, Schlander and Telser32;50;99), and the inclusion of viewpoints from the general public (Reference Conti, Gruber, Ollendorf and Neumann21;Reference Moro, Schlander and Telser32;Reference Gauvreau, Wight and Subasri65;Reference Daniels and van der Wilt101) were espoused. Overall, the integration of external perspectives was noted to provide context to HTA decision-making, expanding the scope beyond consideration of clinical and cost-effectiveness concerns.
HTA methods and processes
Seventeen sources outlined methodologies and processes relevant to the appraisal of pediatric interventions (Reference Denburg, Giacomini, Ungar and Abelson7;Reference Ungar, Prosser and Burnett14;Reference Conti, Gruber, Ollendorf and Neumann21;Reference Dawoud, Lamb and Moore29;Reference Dewilde, Janssen, Lloyd and Shah34;Reference Maas, Joos and Hiligsmann54;Reference Fornaro, Drummond, Ciani and Jommi60;Reference Gauvreau, Wight and Subasri65;Reference Whittal, Nicod, Drummond and Facey70;Reference Gye, Goodall and De Abreu Lourenco76;Reference Nicod, Whittal, Drummond and Facey78;Reference Walzer and Droeschel87;Reference Adkins, Nicholson, Floyd, Ratcliffe and Chevrou-Severac90;Reference Brougham, Schlander, Telser, Bakshi and Sola-Morales96;Reference O’Day, Meyer and Bramley105–Reference Cressman and Denburg107). While most HTA agencies do not explicitly adopt a separate pathway for evaluating interventions in pediatric populations (Reference Whittal, Nicod, Drummond and Facey70), in practice, pediatric diseases are frequently managed through rare disease pathways (Reference Nicod, Whittal, Drummond and Facey78), resulting in more bespoke assessments. Examples of tailored assessment processes included separate evaluation criteria for ultra-rare diseases and temporary reimbursement of interventions while further evidence was collected (Reference Adkins, Nicholson, Floyd, Ratcliffe and Chevrou-Severac90). Sources also noted the potential for inadvertent bias against pediatric populations due to routine HTA practices, including using short life expectancy criteria, which renders it more likely that end-of-life interventions will qualify for reimbursement but not necessarily interventions for patients whose lives have been severely truncated (Reference O’Day, Meyer and Bramley105). Discounting, where a life year gained in the present is valued more than future gains, provided another example of potential bias against pediatric illnesses, as survival benefits may take longer to accrue in children than in adults (Reference Conti, Gruber, Ollendorf and Neumann21).
Extrapolation
Fourteen sources discussed challenges in data extrapolation (Reference Aballéa, Thokagevistk and Velikanova3;Reference Denburg, Giacomini, Ungar and Abelson7;Reference Conti, Gruber, Ollendorf and Neumann21;Reference Moretti, Ruiz, Bonifazi, Pizzo and Kindblom33;Reference Freigofaite, Kaproulia, Verhoek and Sarri45;Reference Fornaro, Drummond, Ciani and Jommi60;66;Reference James, Gruber, Vilu and Martin71;Reference Gye, Goodall and De Abreu Lourenco76;Reference Greasley, Campbell and Wall98;Reference Bush, Lenney, Spencer and Warner104;Reference Ayers and Cope108–Reference Weihing, Roxlau and Bocuk110). Two types of extrapolation were discussed. First, extrapolating adult data to pediatric conditions was performed on occasion (Reference Moretti, Ruiz, Bonifazi, Pizzo and Kindblom33;Reference Freigofaite, Kaproulia, Verhoek and Sarri45;Reference James, Gruber, Vilu and Martin71;Reference Greasley, Campbell and Wall98;Reference Ernst and Rietzschel109;Reference Weihing, Roxlau and Bocuk110); the acceptability of this approach varied between HTA agencies (Reference Ernst and Rietzschel109;Reference Weihing, Roxlau and Bocuk110). Discomfort with this approach was highlighted due to the fundamental differences between adults and children, including children’s unique biology and social position (Reference Denburg, Giacomini, Ungar and Abelson7). Second, the extrapolation of short-term data to long-term outcomes was also discussed. This raised several challenges: difficulty predicting long-term outcomes from short-term pediatric data (Reference Conti, Gruber, Ollendorf and Neumann21;Reference Moretti, Ruiz, Bonifazi, Pizzo and Kindblom33), the struggle to select the appropriate time horizon for economic modeling (Reference Aballéa, Thokagevistk and Velikanova3;Reference Denburg, Giacomini, Ungar and Abelson7), and the need to adjust utility values to reflect the natural decrement in HRQoL over time (66).
Population heterogeneity
Thirteen sources outlined specific dimensions of pediatric population heterogeneity, which make evaluation of treatment benefits difficult (Reference Aballéa, Thokagevistk and Velikanova3;Reference Montgomery and Kusel15;Reference Hill, Rowen, Pennington, Wong and Wailoo17;Reference Nicod, Lloyd and Morel24;Reference Moretti, Ruiz, Bonifazi, Pizzo and Kindblom33;Reference Freigofaite, Kaproulia, Verhoek and Sarri45;Reference Crossley, Chandler and Godfrey49;Reference Whittal, Nicod, Drummond and Facey70;Reference Nicod, Whittal, Drummond and Facey78;Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82;Reference Walzer and Droeschel87;92;Reference Petrou102), including variations in age and developmental stage and variability in baseline characteristics, clinical presentation, disease course, and response to therapy between individuals (Reference Whittal, Nicod, Drummond and Facey70;Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82;Reference Walzer and Droeschel87;92). While some degree of heterogeneity is inherent in pediatric populations, broad inclusion criteria in pediatric trials that draw on a diverse international population may further amplify population differences (Reference Freigofaite, Kaproulia, Verhoek and Sarri45;Reference Crossley, Chandler and Godfrey49). Heterogeneity can impede efforts to meaningfully capture the quality of life in pediatric populations, given the inability to employ the same patient-reported outcome measures (PROMs) and preference-based scores across the entire population (Reference Aballéa, Thokagevistk and Velikanova3;Reference Montgomery and Kusel15;Reference Hill, Rowen, Pennington, Wong and Wailoo17;Reference Nicod, Lloyd and Morel24;Reference Moretti, Ruiz, Bonifazi, Pizzo and Kindblom33;Reference Petrou102).
Price of intervention
Twelve sources addressed financial considerations relevant to pediatric HTAs (Reference Agashe, O’Day, Arvin-berod, Meyer and Bramley4;Reference Denburg, Giacomini, Ungar and Abelson7;Reference Conti, Gruber, Ollendorf and Neumann21;Reference Moro, Schlander and Telser32;Reference Lloyd, Dean, Jensen, Maru and Dabbous40;Reference Whittal, Nicod, Drummond and Facey70;Reference Gye, Goodall and De Abreu Lourenco76;Reference Nicod, Whittal, Drummond and Facey78;Reference Chapman, Kumar, Whittington and Pearson81;Reference Curry, Dawson, Cork, Hollier-Hann and Richardson94–Reference Brougham, Schlander, Telser, Bakshi and Sola-Morales96). Several highlighted the lack of economic incentives driving companies to develop pediatric interventions, given the small pediatric market (Reference Denburg, Giacomini, Ungar and Abelson7;Reference Lloyd, Dean, Jensen, Maru and Dabbous40;Reference Brougham, Schlander, Telser, Bakshi and Sola-Morales96). While increased appreciation for the genomic and molecular basis for diseases may dissolve the strict separation between adult and pediatric disease populations (Reference Denburg, Giacomini, Ungar and Abelson7), thereby changing market dynamics, currently, there is a mismatch between market incentives and societal support for developing and reimbursing pediatric interventions (Reference Denburg, Giacomini, Ungar and Abelson7). Lack of competition in the rare disease space further contributes to high prices (Reference Conti, Gruber, Ollendorf and Neumann21). Proposals included innovative payment structures designed to share the risk of developing costly but potentially highly beneficial interventions (Reference Gye, Goodall and De Abreu Lourenco76;Reference Nicod, Whittal, Drummond and Facey78;Reference Chapman, Kumar, Whittington and Pearson81) and government regulation of the pediatric drug market (Reference Denburg, Giacomini, Ungar and Abelson7).
Comparator
Ten sources discussed the challenges of designing pediatric studies with an appropriate comparator (Reference Aballéa, Thokagevistk and Velikanova3;Reference Nicod, Lloyd and Morel24;Reference Maas, Joos and Hiligsmann54;Reference Fornaro, Drummond, Ciani and Jommi60;Reference Gye, Goodall and De Abreu Lourenco76;Reference Gladwell, Lee, Tate, Batty and Brereton77;Reference Walzer and Droeschel87;Reference Barchanska, Shaw, Campbell and Perez88;Reference Schoot, Otth, Frederix, Leufkens and Vassal97;Reference Tomeczkowski, Partemio, Nijhuis, Kubitz and Kavanagh111). Due to small disease populations, pediatric studies are frequently conducted without a comparator group (Reference Maas, Joos and Hiligsmann54), resulting in a lack of conventional randomized controlled trial data to support HTA decision-making (Reference Schoot, Otth, Frederix, Leufkens and Vassal97). For example, most cell and gene therapy studies in pediatric populations were single-arm trials (Reference Fornaro, Drummond, Ciani and Jommi60). However, single-arm data were also noted to be very hard to interpret due to the dynamic nature of child development, whereby comparisons to a prior baseline can be misleading (Reference Nicod, Lloyd and Morel24), and pharmacoeconomic value assessments can be difficult to perform (Reference Gladwell, Lee, Tate, Batty and Brereton77). Early discussion with HTA bodies was advised to ensure that the future data generated would meet HTA agency assessment requirements (Reference Schoot, Otth, Frederix, Leufkens and Vassal97).
Costs outside the healthcare sector
Nine sources discussed the costs outside the healthcare sector (Reference Aballéa, Thokagevistk and Velikanova3;Reference Ungar, Prosser and Burnett14;Reference Conti, Gruber, Ollendorf and Neumann21;Reference Moretti, Ruiz, Bonifazi, Pizzo and Kindblom33;Reference Costa and Ungar55;Reference Gauvreau, Wight and Subasri65;Reference Ungar85;Reference Kearns112;Reference Shaw, Omer, Walker, Petrie and Foster113). These can be substantial and include the costs of social services, special education, therapy, paid care, and reduced labor productivity due to disability (Reference Aballéa, Thokagevistk and Velikanova3;Reference Ungar, Prosser and Burnett14;Reference Conti, Gruber, Ollendorf and Neumann21;Reference Costa and Ungar55;Reference Gauvreau, Wight and Subasri65;Reference Ungar85). There is no agreement on which of these costs, if any, should be included in HTAs (Reference Moretti, Ruiz, Bonifazi, Pizzo and Kindblom33). Providing accurate costs when considering expenses beyond healthcare was noted to be very challenging (Reference Moretti, Ruiz, Bonifazi, Pizzo and Kindblom33).
Discussion
This scoping review maps out the challenges of conducting HTAs in children, drawing on 103 identified sources. Two important themes emerge from this work. The first is the unique position of children within society and the resulting strain of integrating them within a population-wide assessment system. Pediatric patients were identified as a distinct and unique group requiring protection due to their lack of autonomy and political power (Reference Denburg, Giacomini, Ungar and Abelson7;Reference Denburg, Giacomini, Ungar and Abelson19;Reference Guest, Holdgate, Patel, Rouse and Bradley51;Reference Ungar85). However, pediatric medical needs were noted to be inadequately managed within a system designed for adults. Societal values, largely supportive of the reimbursement of pediatric treatments (Reference Moro, Schlander and Telser32), conflict with market dynamics (Reference Denburg, Giacomini, Ungar and Abelson7;Reference Montgomery and Kusel15;Reference Maas, Joos and Hiligsmann54;Reference Brougham, Schlander, Telser, Bakshi and Sola-Morales96), traditional cost-effectiveness models (Reference Moro, Schlander and Telser32;99), and population-wide policies that explicitly exclude consideration of patient age to avoid discrimination on its basis (Reference Mikudina, Robertson and Upadhyaya72). Additional tools devised with adult populations in mind, such as discounting (Reference Conti, Gruber, Ollendorf and Neumann21), utilitarian cost-effectiveness models (Reference Meregaglia, Nicod and Drummond38;99), and short life expectancy criteria (Reference O’Day, Meyer and Bramley105) inadvertently led to biases against pediatric populations in reimbursement decisions. The lack of adoption by most HTAs of separate pathways to evaluate interventions in pediatric populations (Reference Whittal, Nicod, Drummond and Facey70) and the absence of pediatric expertise on HTA bodies (Reference Aballéa, Thokagevistk and Velikanova3;Reference Denburg, Giacomini, Ungar and Abelson7;Reference Greasley, Campbell and Wall98;Reference Bush, Lenney, Spencer and Warner104) further exacerbated these dynamics.
The second theme is the uncertainty that characterizes pediatric HTAs. This took on two forms. The first was uncertainty related to data constraints. Attempts to fit pediatric populations into an adult drug development, regulatory, and HTA landscape extended to expectations regarding the data package submitted to HTAs by the sponsor. Concerns regarding the lack of adequate pediatric data and the resulting uncertainty were commonly raised in the included sources. Contributing factors included the small pediatric population size (Reference Whittal, Nicod, Drummond and Facey70;Reference Gye, Goodall and De Abreu Lourenco76;Reference Nicod, Whittal, Drummond and Facey78;Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82), lack of market incentives to collect data (Reference Denburg, Giacomini, Ungar and Abelson7;Reference Montgomery and Kusel15;Reference Maas, Joos and Hiligsmann54;Reference Brougham, Schlander, Telser, Bakshi and Sola-Morales96), population heterogeneity (Reference Whittal, Nicod, Drummond and Facey70;Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82;Reference Walzer and Droeschel87;92), lack of long-term data on treatment efficacy (Reference Aballéa, Thokagevistk and Velikanova3;Reference Conti, Gruber, Ollendorf and Neumann21;Reference Whittal, Nicod, Drummond and Facey70;Reference Gye, Goodall and De Abreu Lourenco76;Reference Chapman, Kumar, Whittington and Pearson81;Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82;Reference Ungar85;92;Reference Cressman and Denburg93) and safety (Reference Denburg, Giacomini, Ungar and Abelson7;Reference Maas, Joos and Hiligsmann54;Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82), use of surrogate outcomes (Reference Clay, Kharitonova, Ansaripour, Aballea and Toumi82), and frequent lack of a comparator group in pediatric studies (Reference Agashe, O’Day, Arvin-berod, Meyer and Bramley4;Reference Maas, Joos and Hiligsmann54;Reference Fornaro, Drummond, Ciani and Jommi60).
The absence of comparator data is particularly problematic for HTA agencies, given their mandate to assess interventions against available alternatives (Reference Karres, Pino-Barrio and Benchetrit114), distinct from regulators’ focus on an individual intervention’s efficacy and safety. New efforts to centralize processes for data generation and elevate alternative sources of data include the creation of the Data Analysis and Real World Interrogation Network (115), a coordinating center established by the European Medicines Agency and the European Medicines Regulatory Network, which provides a catalog of validated real-world data on drugs and vaccines. In the United States, the FDA Rare Disease Innovation Hub (116), which seeks to integrate the lived experiences of patients and their caregivers in reviewing interventions for rare disease populations, may prove a valuable model for the engagement of the rare disease community in pharmaceutical development.
The second type of uncertainty characterizing pediatric HTA review is due to either a lack of guidance by HTAs or variations in guidance between HTA bodies in different countries. This was particularly pronounced in the assessment of QOL, where there was insufficient guidance on how to generate HRQoL data in children (Reference Powell, Rowen, Mukuria, Keetharuth and Wailoo48), the best tools to employ (50–52), and how to translate QOL and HRQoL scores into utilities (Reference Bégo-Le Bagousse, Jia and Wolowacz16;Reference Hill, Rowen, Pennington, Wong and Wailoo18;Reference Sutherland, Davies and Sully37). Similarly, there was a lack of consistency over which surrogate endpoints to use in pediatric disease trials, with divergent requirements between regulatory agencies and different HTA bodies (Reference Maas, Joos and Hiligsmann54;Reference Raza, Keyzor and Shohet86).
These challenges are not unexpected, as HTA agencies are tasked with being responsive to the needs of the population they serve. With no two populations exactly alike, national HTA bodies’ priorities and decision-making processes will differ. While final judgments regarding reimbursement will continue to vary, the EU Regulation 2021/2282 on HTA (117) has developed a process for joint clinical assessment and joint scientific consultations with technology developers to streamline the initial steps in evidence generation and HTA review. This regulation was enacted in January 2025 for oncology and advanced therapy medicinal products and will expand to include orphan medicinal products in 2028. The extent to which this resolves the tension and uncertainty in the pediatric review process will become apparent in the upcoming years.
Limitations
Our review had several limitations. First, our search of electronic databases focused on pediatric HTAs, resulting in 3164 records. We did not expand our search to include all orphan drugs, as this would have limited the feasibility of screening. Second, our search, which was conducted by a medical librarian (PAB), identified reports and reviews addressing pediatric HTAs. HTAs of specific interventions, such as those involving rehabilitation or screening programs, were not included in our results. Third, we limited our review to sources in English or French. While many HTAs offer English translations of guidelines and assessments, not all do, and this may have resulted in an over-representation of sources from English-speaking countries. HTA agencies exist globally, are active in many non-English speaking countries (11), and may face distinct challenges not captured in this review. For example, economic and political realities in Iran, in addition to a fragmented healthcare system, among other challenges, have rendered it difficult to establish a robust HTA agency (Reference Aryankhesal, Behzadifar, Bakhtiari, Azari and Behzadifar118). In Brazil, a constitutional “right to health” has been utilized to require the Ministry of Health to fund access to onasemnogene abeparvovec, an expensive gene therapy for the rare pediatric disease spinal muscle atrophy, at a cost well beyond that established by the Brazilian drug pricing authority (Reference Ivama-Brummell, Wagner, Pepe and Naci119). These are examples of unique challenges not addressed in our review. Fourth, we included a large number (23; 22.3 percent) of conference abstracts for which full-text publications were unavailable. As a scoping review aims to include all evidence relevant to the research question (Reference Peters, Godfrey, McInerney, Soares, Khalil and Parker9), we considered these sources important to include.
Conclusion
HTAs must attend to the medical demands of a nation while also being responsive to the unique needs of specific subsets of the population. The tensions that arise and the uncertainty generated in assessing pediatric interventions are difficult to manage and overcome. Increased awareness of these challenges and attention to the unique needs of pediatric patients is required. Innovative approaches, a multitude of which will be necessary to address the needs of this heterogeneous population, should be encouraged and meaningfully supported.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/S0266462325103188.
Data availability statement
Study protocol and data tables are available from the open science framework.(Reference Hutchinson8)
Funding statement
This research was informed by a meeting on the topic of Pediatric HTAs, which was funded by the Brocher Foundation. This research received no additional specific grant or other funding from any agency in the public, commercial, or not-for-profit sectors.
Competing interests
The authors declare none.
Ethics approval
This study is a scoping review of previously published data and, therefore, did not require formal ethics approval.
Open access
