Economic evaluations of Haemophilus influenzae type b (Hib) vaccine: a systematic review

Abstract

Aims: The World Health Organization (WHO) recommends the use of Haemophilus influenzae type b (Hib) conjugate vaccines, but China and Thailand have not used Hib vaccination in their national immunization programs. This systematic review aimed to update published economic evaluations of Hib vaccinations and to determine factors that potentially affected their cost-effectiveness.

Methods: Searches were performed from the inception until December 2015 using 13 databases: CAB direct; CEA registry; EconLit; EMBASE; E-library; NHSEED; PAHO; POPLINE; PubMed; Redalyc project; RePEc; SciELO; and WHOLIS. Reference lists of relevant studies and grey literature were also searched. Full economic evaluations of Hib vaccination with results of costs and outcomes were included. The WHO checklist was used to evaluate the quality of the included studies. Data from eligible studies were extracted using a standardized data collection form.

Results: Out of 830 articles, 27 were included. Almost half of the studies (12/27) were conducted in high-income countries. Twelve studies (12/27) investigated the Hib vaccine as an addition to the existing vaccination program. Most studies (17/27) examined a 3-dose schedule of Hib vaccine. Nineteen studies (19/27) reported the model used, where all were decision tree models. Most of the studies (23/27) demonstrated an economic value of Hib vaccination programs, key influential parameters being incidence rates of Hib disease and vaccine price.

Conclusions: Hib vaccination programs are mostly found to be cost-effective across geographic regions and country income levels, and Hib vaccination is recommended for inclusion into all national immunization programs. The findings are expected to support policy-makers for making decisions on allocating limited resources of the Hib vaccination program effectively.

Keywords:

• Haemophilus influenzae type b (Hib)
• vaccine
• cost-effectiveness analysis
• economic evaluation
• systematic review

Introduction

Haemophilus influenzae type b (Hib) is a gram-negative coccobacillus which infects infants in particular¹. It transmits via the respiratory tract from infected to susceptible individuals, mostly children aged less than 5 years. Hib may cause not only meningitis, severe pneumonia, etc., but also epiglottitis, etc., although the latter are rarer^2,3. Nowadays, Hib is estimated to be responsible globally for 200,000–700,000 deaths per year⁴. Prior to vaccination, the worldwide incidence rate of Hib diseases estimated was 370 per 100,000 children per year in children aged 0–4 years, resulting in 2.2 million cases per year. The number of deaths was over 520,000, or the case fatality rate was ∼23%⁵. The burden of Hib diseases is higher in resource-poor countries than in other countries⁶.

Hib conjugated vaccine has been available for the prevention of Hib infections since the late 1980s^7,8. Because of its high efficacy and good tolerability profile, this vaccine is recommended by WHO to be included in their Expanded Program on Immunization (EPI) throughout the world since 1998⁴. Likewise, the Global Alliance for Vaccines and Immunization (GAVI) began to offer Hib conjugated vaccine to resource-poor nations in 2001^9,10. However, the global coverage ranged from 21% in the western Pacific regions to 90% in the Americas (for 2014)⁸, while China and Thailand have not used Hib vaccine in their national immunization programs¹¹.

A previous systematic review of Hib vaccination programs¹² demonstrated wide variations of the economic evaluations resulting from differences in epidemiologic and health system settings and differences in methodology and the models used between studies. Another review by Griffiths and Miners³ attempted to identify and evaluate the cost-effectiveness of Hib vaccination, particularly in low- and middle-income countries, but the paucity of studies from resource-poor countries and their poor quality prevented the drawing of definitive conclusions. According to their Global Vaccine Action Plan 2011–2020, WHO aims to facilitate policy decisions by providing economic evidence of vaccination. Therefore, WHO commissioned the present review to provide an updated summary of economic evaluation of Hib vaccination programs. To provide this, we sought to conduct a comprehensive review of the full economic evaluations published from a myriad of countries, and to determine factors affecting the cost-effectiveness of the vaccine. The findings are expected to support the ability of policy-makers to make rational decisions on the effective allocation of limited resources.

Methods

Search strategy

Searching was performed from the inception to December 2015 using 13 independent databases: CAB direct; CEA registry; EconLit; EMBASE; E-library; Medline; NHSEED; PAHO; POPLINE; Redalyc project; RePEc; SciELO; and WHOLIS. Reference lists of relevant published studies and grey literature, i.e. conference posters, study abstracts, etc., were also searched. The search terms were Haemophilus influenzae type b AND (vaccine* OR vaccination* OR vaccination program* OR immunization OR immunization program*) AND (cost* OR cost analysis OR economic* OR economic evaluation OR pharmacoeconomic*). All search terms applied are presented in Supplementary Appendix Table A1.

Procedure

Four investigators (BC, MH, WP, and NC) independently reviewed the titles and abstracts. Full economic evaluations of the Hib vaccinations with the results of cost and health outcomes reported were included. Data from all eligible articles were extracted using a standardized data collection form by three investigators (BC, MH, and WP). Disagreements were resolved by discussion with SK and NC. Gross domestic product (GDP) per capita in United States Dollars (USD) of 2015 was obtained from the World Bank¹³. Affiliation was identified by the first listed institutional affiliation of the first author. Funding source was determined by any support for the study stated in the acknowledgments or declarations. If any co-author was employed by a vaccine company, this would be considered as a funding source of study. A checklist for appraising the quality of economic evaluations of the immunization program by WHO¹⁴ was used to estimate the quality of included studies. They were assessed based on specific methodology and reporting of good practices for conducting the economic evaluation study of the immunization program, i.e. identification of study question; expression of the study perspective, time horizon and discount rate; methods and structure of the model used for data analysis; the assumption behind the calculation of costs; sensitivity analysis performing as well as the justification of study conclusions.

Currency conversions

Local currencies were first converted into USD using the stated year of currency conversion, or (if not available) article publication year. Consequently, they were inflated to USD values in 2015 using the US Consumer price index for all urban consumers¹⁵. Since the vaccine purchase costs and GDP per capita thresholds were set internationally, they were all converted into USD using the US exchange rates¹⁶ on January 1st of the base year, and not inflated.

Results

Study selection

The initial search yielded 1,419 articles, of which 589 duplicates were removed. The remaining 830 articles were screened by title and abstract, of which 772 articles were excluded because of their irrelevance to Hib and the study design, leaving 58 studies being assessed for eligibility. A further 31 articles were excluded for the following reasons: conference abstract (n = 4), non-English language (n = 4), non-full economic evaluation (n = 8), and non-economic evaluation (n = 15). This yielded a total of 27 relevant studies in this review, compared to 17 articles in the previous systematic review³ (Supplementary Appendix Table A2). A PRISMA flow diagram is shown in Figure 1.

Figure 1. A PRISMA flow diagram describing the study selection process.

Study characteristics

Country and funding

Among all included studies, almost half of the studies (12/27: 44%)^8,17–27 were conducted in high income countries (HICs), seven studies (7/27: 26%)^28–34 in lower-middle income countries (LMICs), three studies (3/27: 11%)^35–37 in upper-middle income countries (UMICs), and two studies (2/27: 7%)^38,39 in low income countries (LICs). The remaining three studies (3/27: 11%)^9,40,41 were conducted in multiple countries. In 18 studies^{8,9,17–23,27,28,30,31,33,35,36,40,41} were reported as funded by non-profit organization (5/18: 30%)^8,9,23,33,40, government (3/18: 17%)^17,21,22, industry (3/18: 17%)^18–20, and academia (1/18: 6%)³⁶. The remaining six studies (6/18: 33%)^{27,28,30,31,35,41} were funded by two types of sponsor (Table 1).

Table 1. General characteristics of included studies.

Reference	Published year	Studied year	Affiliations of the first author	Country of the first author	Countries of study	Regions	Income groups	Funding
Akumu et al.^28	2007	2004	Kenya Medical Research Institute	Kenya	Kenya	Sub-Saharan Africa	LMICs	1. Kenya Medical Research Institute 2. WHO
Broughton^29	2007	2005	Columbia University	US	Indonesia	East Asia and Pacific	LMICs	—
Clark et al.^30	2013	2010	Department of health services research and policy, London School of Hygiene and Tropical Medicine	UK	India	South Asia	LMICs	1. GSK 2. Merck 3. Pfizer 4. GAVI Alliance
Duke^39	2002	2002	Gorona Hospital	Papua New Guinea	Papua New Guinea	East Asia and Pacific	LICs	—
Gargano et al.^38	2015	2013	Centers for Disease Control and Prevention	US	Somalia	Sub-Saharan Africa	LICs	—
Garpenholt et al.^27	1998	1995	Linkoping University	Sweden	Sweden	Europe and Central Asia	HICs	1. The Research Committee of the Orebro Country Council 2. Pasteur Merieux/MSD, Copenhagen
Gessner et al.^31	2008	2007	Agence de Medecine Preventive, France	Indonesia	Indonesia	East Asia and Pacific	LMICs	1. The office of Health and Nutrition, USA 2. SanofiPasteur, France
Ginsberg et al.^17	1993	1991	University-Hadassah School of Public Health	Israel	Israel	Middle East and North Africa	HICs	The neighborhood family health centers
Griffiths et al.^40	2011	2009	Hib Initiative, London School of Hygiene and Tropical Medicine	UK	1. Uzbekistan 2. Belarus	Europe and Central Asia	1. LMICs 2. UMICs	WHO
Griffiths et al.^9	2013	2010	London school of hygiene and tropical medicine	UK	Country groups 1. GAVI (Africa) 2. GAVI (Asia) 3. LMICs 4. UMICs	Multi-region	1. LICs 2. LMICs 3. UMICs	The GAVI alliance
Gupta et al.^32	2013	2009	School of Public Health, Post Graduate Institute of Medical Education and Research	India	India	South Asia	LMICs	—
Harris et al.^18	1994	1990	Department of Public Health, The University of Western Australia	Australia	Australia	East Asia and Pacific	HICs	MSD, Australia
Hay and Daum^19	1987	1984	Stanford University	US	US	North America	HICs	Rescon, Inc
Hay and Daum^20	1990	1988	Stanford University	US	US	North America	HICs	Connaught laboratories, Inc.
Hussey et al.^35	1995	1992	University of Cape Town	South Africa	South Africa	Sub-Saharan Africa	UMICs	1. Medical Research Council 2. Lederle Laboratories
Le et al.^33	2015	2013	University of Texas School of Public Health	US	Vietnam	East Asia and Pacific	LMICs	The GAVI alliance
Limcangco et al.^34	2001	1998	College of Pharmacy, University of the Philippines	Philippines	Philippines	East Asia	LMICs	—
Livartowski et al.^21	1996	1993	Departement de pediatric, Faculte de medecine, Universite Paris	France	France	Europe and Central Asia	HICs	French government
McIntyre et al.^22	1994	1991	Department of Pediatrics, Westmead Hospital	Australia	Australia	East Asia and Pacific	HICs	Department of Health, Housing and Community Services, and the Epidemiology and Health Services Evaluation Branch of the New South Wales
Moradi-Lakeh et al.^36	2012	2008	Tehran University of Medical Sciences	Iran	Iran	Middle East and North Africa	UMICs	Iran University of Medical Sciences
Muangchan and Warinsatian^37	2011	2006	National Vaccine Committee Office, Department of Disease Control, Ministry of Public Health	Thailand	Thailand	East Asia and Pacific	UMICs	—
Niessen et al.^41	2009	2005	Department of International Health, Johns Hopkins Bloomberg School of Public Health	US	WHO country regions 1. African 2. Americas 3. Eastern Mediterranean 4. South-East Asia 5. Western Pacific	Multi-region	1. LICs 2. LMICs 3. UMICs 4. HICs	1. Netherland Environmental Assessment Agency 2. WHO 3. The United Nations Children’s Fund
Platonov et al.^23	2006	2001	Central Institute of Epidemiology, Russian Federation	Russia	Russia	Europe and Central Asia	HICs	WHO
Pokorn et al.^24	2001	1999	University Medical Center Ljubljana	Slovenia	Slovenia	Europe and Central Asia	HICs	—
Shin et al.^8	2008	2003	Seoul National University, School of Public Health	Korea	Korea	East Asia and Pacific	HICs	WHO
Trollfors^25	1994	1993	University of Goteborg	Sweden	Sweden	Europe and Central Asia	HICs	—
Zhou et al.^26	2002	2000	US Department of Health and Human Services	US	US	North America	HICs	—

Abbreviations. GAVI, Global Alliance for Vaccines and Immunization; GSK, GlaxoSmithKline; HICs, High-Income Countries; LICs, Low-Income Countries; LMICs, Lower Middle-Income Countries; MSD, Merck Sharp and Dohme; UMICs, Upper Middle-Income Countries; UK, United Kingdom; US, United States; WHO; World Health Organization.

Type of economic analysis

Out of 27 studies, 13^{8,9,19,20,24,25,27,30,34,35,37,39,41} performed one analysis of economic evaluation (cost-effectiveness analysis, CEA; cost-utility analysis, CUA; cost-benefit analysis, CBA), while 14 studies^{17,18,21–23,26,28,29,31–33,36,38,40} performed multiple analyses of economic evaluation. When only one analysis was performed, there were nine CBA (9/13: 69%)^{8,19,20,24,25,27,34,35,37}, three CUA (3/13: 23%)^9,30,41, and one CEA (1/13: 8%)³⁹. Among studies that performed multiple analyses, there were six CEA and CUA studies (6/14: 43%)^{18,22,28,32,33,38}, four CBA, CEA, and CUA studies (4/14: 29%)^23,26,31,36, three CBA and CUA studies (3/14: 21%)^29,36,40, and one CBA and CEA study (1/14: 7%)¹⁷ (Table 2).

Table 2. Economic evaluation-related characteristics of included studies.

References	Target vaccinated populations	Interventions	Comparators	Study types	Perspectives	Time horizons	Discount rates	Thresholds	Sensitivity analyses
Akumu et al.^28	Children <5 years	Combined vaccination (DTP-Hib-HB)	Without Hib vaccination (DTP-HB)	1. CEA 2. CUA	Healthcare payer	Lifetime	1. Cost 3% 2. Benefit 3%	1 time of GDP	One-way
Broughton^29	Children <1 year	Combined vaccination (DTP-Hib-HB)	Without Hib vaccination (DTP-HB)	1. CBA 2. CUA	Societal	Lifetime	1. Cost 3% 2. Benefit 3%	—	1. One-way 2. Two-way
Clark et al.^30	Children <5 years	Single Hib vaccination	Without vaccination	CUA	1. Societal 2. Healthcare payer	Lifetime	1. Cost 3% 2. Benefit 3%	1–3 times of GDP	One-way
Duke^39	—	1. Single Hib vaccination 2. Single Hib vaccination	1. Chloramphenicol 2. Ceftriaxone	CEA	Societal	Lifetime	1. Cost 5% 2. Benefit 5%	—	One-way
Gargano et al.^38	Children <2 years	1. Single Hib vaccination 2. Combined vaccination (Hib-PCV10)	Without vaccination	1. CEA 2. CUA	Healthcare payer	Lifetime	—	1–3 times of GDP	One-way
Garpenholt et al.^27	Children <5 years	Single Hib vaccination	Without vaccination	CBA	Societal	Lifetime	1. Cost 5% 2. Benefit 5%	—	One-way
Gessner et al.^31	Children <2 years	1. Single Hib vaccination 2. Combined vaccination (DTP-Hib-HB)	Without vaccination	1. CBA 2. CEA 3. CUA	1. Societal 2. Healthcare payer	Lifetime	1. Cost 3% 2. Benefit 3%	1 time of GNI	One-way
Ginsberg et al.^17	Children <5 years	Single Hib vaccination	Without vaccination	1. CBA 2. CEA	Societal	Lifetime	1. Cost 5% 2. Benefit 5%	—	—
Griffiths et al.^40	Children <5 years	Single Hib vaccination	Without vaccination	1. CBA 2. CUA	Societal	Lifetime	1. Cost 3% 2. Benefit 3%	1–3 times of GDP	One-way
Griffiths et al.^9	Children <5 years	Combined vaccination (DTP-HB-Hib)	Without Hib vaccination (DTP-HB)	CUA	1. Societal 2. Healthcare payer	Lifetime	1. Cost 3% 2. Benefit 3%	1 time of GNI	PSA
Gupta et al.^32	Children <5 years	Combined vaccination (DTP-Hib-HB)	Without Hib vaccination (DTP-HB)	1. CEA 2. CUA	1. Societal 2. Healthcare payer	Lifetime	Cost 3%	3 times of GDP	One-way
Harris et al.^18	Children <1 year	Single Hib vaccination	Without vaccination	1. CEA 2. CUA	1. Healthcare payer 2. Others^a	Lifetime	1. Cost 5% 2. Benefit 5%	1 time of GDP	One-way
Hay and Daum^19	1. Children at 18 months 2. Children at 24 months	1. Single Hib vaccination 2. Combined vaccination	Without vaccination	CBA	Societal	60 years	1. Cost 3% 2. Benefit 3%	—	One-way
Hay and Daum^20	1. Children at 18 months 2. Children at 24 months	1. Single Hib vaccination 2. Combined vaccination	Without vaccination	CBA	Societal	60 years	1. Cost 3% 2. Benefit 3%	—	One-way
Hussey et al.^35	Children <1 year	Combined vaccination (DTP-Hib)	Without Hib vaccination (DTP)	CBA	Societal	Lifetime	1. Cost 2% 2. Benefit 2%	—	One-way
Le et al.^33	Children <5 years	Combined vaccination (DTP-Hib-HB)	Without Hib vaccination (DTP-HB)	1. CEA 2. CUA	1. Societal 2. Healthcare payer	Lifetime	1. Cost 6.9% 2. Benefit 5%	3 times of GDP	1. One-way 2. PSA
Limcangco et al.^34	Children <1 year	Combined vaccination (DTP-Hib)	Without Hib vaccination (DTP)	CBA	1. Societal 2. Healthcare payer	Lifetime	Cost 5%	—	One-way
Livartowski et al.^21	Children <1 year	Combined vaccination (DTP-Hib)	Without Hib vaccination (DTP)	1. CBA 2. CUA	1. Healthcare payer 2. Others^a	Lifetime	1. Cost 6% 2. Benefit 6%	—	One-way
McIntyre et al.^22	1. Children <6 months 2. Children <12 months 3. Children <18 months	Combined vaccination (DTP-Hib)	Without vaccination	1. CEA 2. CUA	Other^a	Lifetime	1. Cost 5% 2. Benefit 5%	—	One-way
Moradi-Lakeh et al.^36	Children <1 year	Combined vaccination (DTP-Hib)	Without Hib vaccination (DTP)	1. CBA 2. CEA 3. CUA	Societal	Lifetime	1. Cost 3% 2. Benefit 3%	3 times of GDP	1. One-way 2. Two-way
Muangchana and Warinsatian^37	Children <5 years	Combined vaccination (DTP-Hib)	Without Hib vaccination (DTP)	CBA	Societal	Lifetime	1. Cost 3% 2. Benefit 3%	—	1. One-way 2. Multi-way
Niessen et al.^41	Children <5 years	Combined vaccination (DTP-Hib-HB)	Without Hib vaccination (DTP-HB)	CUA	Healthcare payer	10 years	Cost 3%	—	—
Platonov et al.^23	Children <5 years	Combined vaccination (DTP-Hib)	Without Hib vaccination (DTP)	1. CBA 2. CEA 3. CUA	Societal	Lifetime	Cost 3%	3 times of GDP	1. One-way 2. PSA
Pokorn et al.^24	Children <1 year	Single Hib vaccination	Without vaccination	CBA	1. Societal 2. Healthcare payer	Lifetime	1. Cost 5% 2. Benefit 5%	—	One-way
Shin et al.^8	Children <1 year	Single Hib vaccination	Without vaccination	CBA	Societal	Lifetime	1. Cost 5% 2. Benefit 5%	—	One-way
Trollfors^25	Children <1 year	Single Hib vaccination	Without vaccination	CBA	Societal	Lifetime	—	—	—
Zhou et al.^26	Children <1 year	Combined vaccination (DTP-Hib)	Without vaccination	1. CBA 2. CEA 3. CUA	1. Societal 2. Other^a	50 years	1. Cost 3% 2. Benefit 3%	—	One-way

Abbreviations. CBA, Cost-Benefit Analysis; CEA, Cost-Effectiveness Analysis; CUA, Cost-Utility Analysis; DALY, Disability-adjusted life year; DTP, Diphtheria-Tetanus-Pertussis; GDP, Gross Domestic Product; GNI, Gross National Income; HB, Hepatitis B; Hib, Haemophilus influenzae type b; PCV7, 7-valent Pneumococcal Conjugated Vaccine; PSA, Probabilistic Sensitivity Analysis; QALY, Quality-adjusted life year.

^a Other perspective was divided into two definitions: (1) direct cost perspective including direct medical and direct non-medical costs in the analysis was used in three studies^18,22,26; and (2) Patients’ families perspective (co-payments) was used in one study²¹.

Population characteristics and comparators

Thirteen studies (13/27: 48%)^{8,9,17,23,26–28,30,32,33,37,40,41} investigated the cost-effectiveness analysis of Hib vaccinations of children younger than 5 years, while eight studies (8/27: 30%)^{18,21,24,25,29,34–36} used children younger than 1 year. Five studies (5/27: 19%)^{17,19,20,31,38} sub-grouped children by age to determine cost-effectiveness of Hib vaccination. One study (1/27: 4%)³⁹ provided no detail about the target vaccinated population. For comparisons, 12 studies (12/27: 44%)^{9,21,23,28,29,32–37,41} compared the cost-effectiveness of Hib vaccination as an adjunct to existing vaccination programs, whereas eight studies (8/27: 30%)^{8,17,18,24,25,27,30,40} compared Hib vaccinated populations with those receiving no vaccination against anything (Table 2).

Perspective and time horizon

Thirteen studies (13/27: 48%)^{8,17,19,20,23,25,27,29,35–37,39,40} undertook analyses by societal perspectives, three (3/27: 11%)^28,38,41 by healthcare payer, while seven (7/27: 26%)^{18,30–34,40} analysed both perspectives. For time horizons, most studies (23/27: 85%)^{8,9,17,18,21–25,27–40} used whole lifetime, while others used 60 years (2/27: 7%)^19,20, 50 years (1/27: 4%)²⁶, and 10 years (1/27: 4%)⁴¹ (Table 2).

Modeling structure and manifestation of Hib disease syndrome

Nineteen studies^{8,9,18–20,22,24,26,29–34,36–40} employed a static transmission model, specifically, a decision tree for analyses; none employed a dynamic transmission model. One study²⁸ showed 100% elimination of overt Hib symptoms, while another study²⁹ assumed a high proportion of vaccinated individuals might lead to high apparent efficacy through the herd protection effect. For study of specific Hib-mediated syndromes, meningitis was incorporated in all but one study (26/27: 96%)^{8,9,17–37,39–41}, pneumonia 15/27 studies (55%)^{8,9,26,28–33,35–38,40,41}, and epiglottitis 9/27 studies (33%)^{18–22,25–27,36} (Supplementary Appendix Table A3).

Dosage regimens, coverage/wastage rates, and costs of vaccine

Nearly all of the studies (26 studies)^{8,9,17–26,28–41} reported dosage regimens of Hib vaccine. The majority of studies (17/26: 65%)^{8,18,22,24,25,28–37,39,41} used a 3-dose schedule. Four studies (4/26: 15%)^9,17,26,40 used 3- and 4-dose schedules, two studies (2/26: 8%)^21,23 used a 4-dose schedule, two studies (2/26: 8%)^19,20 used a 1-dose schedule, and one study (1/26: 4%)³⁸ used a 2-dose schedule. Vaccination coverage rates were reported in 22 studies^{8,9,18–21,23,24,26–30,32–38,40,41}, irrespective of dose regimen, and were 60% or greater (range = 60%^19,20 to 100%²¹). Fourteen studies^{8,9,18,23,24,26,28,30,33,36–38,40,41} reported vaccine wastage rate ranged from 3%^18,24 to 27%³⁰. In addition, 26 studies^{8,9,17–25,27–41} reported costs of vaccine ranging from USD 0.3³¹ to 22.5^25,27, while 12 studies^{8,18–20,22–24,27,34–36,38} reported costs of vaccine delivery ranging from USD 0.26³⁴ to 20^19,20 (Supplementary Appendix Table A4).

Important epidemiological parameters

We categorized two important epidemiological parameters—the effect of vaccination and the incidence rates of Hib disease in the country under study were categorized according to: (1) study originating locally in the home country (local study) or (2) study conducted remotely from an outside country (non-local study). All study designs of the sources were also categorized based on the hierarchy of study designs⁴²—systematic review and meta-analyses of randomized controlled trials; randomized controlled trial (RCT); observational studies (cohort, cross-sectional and case-control studies); case report or case series; and expert opinion.

Protection by vaccination: Vaccine protection effect was described in 21 studies^{8,9,17–23,25,27,29,30,32–39}. The effect against overall overt Hib disease ranged from 40%³⁹ to 100%^17,21,35, and for meningitis 89%²⁹ to 95%³⁹, and pneumonia 5%³³ to 95%³⁷ (Supplementary Appendix Table A5). Out of 21 studies, 20^{8,17–23,25,27,29,30,32–39} reported study origin: 15 studies (15/20: 75%)^{17,18,21–23,25,27,30,32–35,37–39} were non-local, of which seven (33%)^{17,18,22,25,34,38,39} were RCTs, four were case reports or case series (19%)^23,27,33,37, and three were observational studies (14%)^21,30,35. One study (5%)³² applied both RCT and observational design. In contrast, only four studies were locally generated (20%)^19,20,29,36 and another (5%)⁸ contained both local and non-local data (Supplementary Appendix Tables A6 and A7).

Incidence rate: The incidence rate of Hib disease was described in 27 studies^8,9,17–41. Among studies investigating the target vaccinated population in children younger than 1 year, annual incidences of meningitis ranged from 17²¹ to 167¹⁷ cases per 100,000 children, while those of pneumonia ranged from 589⁹ to 2,930³³ cases per 100,000 children. In studies investigating the target vaccinated population in children younger than 5 years of age, annual incidences of meningitis ranged from 3²⁷ to 71^28,41 cases per 100,000 children, while those of pneumonia ranged from 63²⁶–2,717³² cases per 100,000 children (Supplementary Appendix Table A8).

There were 22 studies reporting incidence, of which 15 (68%)^{17,18,21–23,25–29,32,34,35,37,39} were local studies, and 10/15 (67%)^{17,18,22,23,25,27,28,32,34,39} were observational. Two studies (2/15: 13%)^26,37 were both observational and case report or case series. Others were both RCT and observational²⁹, case report or case series²¹, and expert opinion³⁵. On the other hand, there were four non-local studies (18%)^9,38,40,41. Three studies (3/22: 14%)^8,30,33 were both local and non-local (Supplementary Appendix Tables A9 and A10).

Sensitivity analysis

Among all included studies, 24^8,9,18–24,^26–40 conducted sensitivity analyses. Of these, most studies (18/24: 75%)^{8,18–22,24,26–28,30–32,34,35,38–40} performed one-way sensitivity analysis, three studies (3/24: 13%)^29,36,37 performed both one- and two-way sensitivity analyses, two studies (2/24: 8%)^23,33 performed one-way sensitivity analysis and probabilistic sensitivity analysis, and one study (1/24: 4%)⁹ performed probabilistic sensitivity analysis (Table 2).

Quality of included studies

Overall, the included studies were of high quality, except four^17,25,28,41 that lacked methodological rigor, specifically ambiguity in the economic model and assumptions behind the calculation of costs in conducting the economic evaluation.

All studies clearly identified the study question, intervention(s), comparator(s), study perspective(s), and time horizon. Most studies reported discount rates (25/27: 93%)^{8,9,17–24,26–37,39–41}, as a high proportion of these studies used 3% for discounting on both cost and benefit in accordance with the local guidelines, and performed sensitivity analyses (24/27: 89%)^{8,9,18–24,26–40} to assess the robustness of the study findings. According to the WHO guidance on the economic evaluation of immunization programs¹⁴, the model should be transparent, in that the structure and all assumptions are clearly described. Only 20 studies (20/27: 74%)^{8,9,18–20,22,24,26,27,29–34,36–40} clearly described methods and structure of the model used for data analysis. Most of the studies (24/27: 89%)^{8,9,17–24,26,27,29–31,33–41} clearly described the measurements and the assumption behind the calculation of costs. However, none of the included studies estimated vaccine efficacy based on a systematic review of the literature, as recommended by WHO as a good practice for estimating the effects of vaccine in conducting economic evaluation of the immunization program. Adverse events from vaccine immunization were reported and included in the analysis among six studies (6/27; 22%)^{18–20,24,26,35}, thus leading to a variety of cost-effectiveness findings of their analyses. All studies clearly justified the conclusions of the study based on the study results and data reported.

Study results

Vaccination

Among all included studies, 23/27 (85%)^{9,17–22,25–36,38–41} demonstrated that Hib vaccination was either cost-saving or cost-effective. Four studies (4/27, 15%)^8,23,24,27 showed that vaccination was not cost-effective in their main analyses, except when the societal perspective was used²⁴, or analysis incorporated intangible benefits, i.e. avoiding the pain and suffering caused by the disease³⁷ (Table 3). Interestingly, three of them^8,23,24 were conducted in HICs, whereas one³⁷ was conducted in UMICs. Among the 12 HICs studies, nine (9/12: 75%)^{17–22,25–27} were cost-effective compared to LICs, LMICs, and UMICs. Likewise, out of 12 studies conducted in LICs, LMICs, and UMICs, 11 (11/12: 92%)^{28–36,38,39} were cost-effective. To be cost-effective, the vaccine price in Korea needed to be reduced from USD 16.86 to 12.97 per dose, and from USD 5 to 0.5 for Russia. Results of all economic evaluation studies are shown in Tables 3–6.

Table 3. Results of benefit/cost ratios and net savings from cost-benefit analysis studies.

Reference (year)	Benefit/cost ratios	Net savings (USD)	Results
Hay and Daum^19 (1987)
Single Hib vaccination
UV	2.04	129,836,430	Cost-saving
UV18	1.09	12,251,550
Combined vaccination
UV + RH	2.92	243,672,710
UV + RH + RD	2.90	253,495,990
UV + RD	2.06	139,657,710
UV + DC + RD + RH	1.91	202,648,530
UV + DC + RH	1.89	192,975,510
UV + DC + RD	1.41	90,200,700
UV + DC	1.37	80,527,680
Hay and Daum^20 (1990)
Single Hib vaccination
PRP	1.20	32,693,520	Cost-saving
Combined vaccination
(PRP-D or HbOC)+RH	4.41	557,013,990
(PRP-D or HbOC)+RH + RD	4.32	569,281,580
(PRP-D or HbOC)+RD	3.53	427,237,670
PRP + RH + RD	2.16	198,471,180
PRP + RH	2.14	185,087,640
PRP + RD	1.27	46,077,070
Ginsberg et al.^17 (1993)
Single Hib vaccination
3-dose schedule	1.26	8,592,851.21	Cost-saving
4-dose schedule	1.45	8,947,447.44
Trollfors^25 (1994)	—	—	Cost-saving
Hussey et al.^35 (1995)
Human capital approach	1.29	1,496,957.90	Cost-saving
Subsistence/human capital approach	1.43	2,187,617.31
Livartowski et al.^21 (1996)
Healthcare payer perspective	—	3,041,929,410	Cost-saving
Other perspective^b	—	274,370,104
Garpenholt et al.^27 (1998)	—	28,526,027.1	Cost-saving
Limcangco et al.^34 (2001)
Healthcare payer perspective	—	1,386,778.16	Cost-saving
Societal perspective	—	9,067,395.73
Pokorn et al.^24 (2001)
Healthcare payer perspective	0.99	12,001.08	Cost-saving
Societal perspective	1.39	–196,548.94	Not-cost saving
Zhou et al.^26 (2002)
Societal perspective	5.3	2,874,656,340	Cost-saving
Other perspective^b	3.4	1,308,102,630
Platonov et al.^23 (2006)	0.06	–1,870,033.35	Not cost-saving
Broughton^29 (2007)
UNICEF Price	—	10,995,258.68	Cost-saving
GAVI Price	—	4,490,337.43
Gessner et al.^31 (2008)
Healthcare payer perspective
Combined vaccination (DTP-Hib-HB)	—	19,433,057.5	Cost-saving
Single Hib vaccination	—	768,659.73
Societal perspective
Combined vaccination (DTP-Hib-HB)	—	27,320,592.5
Single Hib vaccination	—	—
Shin et al.^8 (2008)	0.77	–8,616,584.69	Not cost-saving
Muangchana and Warinsatian^37 (2011)	—	79,946,210.3	Cost-saving^a
Griffiths et al.^40 (2011)
Belarus	—	—	Cost-saving
Uzbekistan	—	6,827,563.82
Moradi-Lakeh et al.^36 (2012)	2.14	9,897,926.49	Cost-saving

Abbreviations. DC, Vaccination of all day-care attendees; DTP, Diphtheria-Tetanus-Pertussis; GAVI, Global Alliance for Vaccines and Immunization; HB, Hepatitis B; HbOC, Haemophilus influenzae type b oligosaccharide-mutant diphtheria toxin conjugate vaccine; Hib, Haemophilus influenzae type b; PRP, Haemophilus influenzae type b polysaccharide vaccine; PRP-D, Haemophilus influenzae type b polysaccharide-diphtheria toxoid conjugate vaccine; RD, Rifampin prophylaxis of appropriate day-care contacts; RH, Rifampin prophylaxis of appropriate household contacts; UNICEF, United Nations Children's Emergency Fund; USD, United States Dollar; UV, 24-month Universal vaccination; UV18, 18-month Universal vaccination.

Net savings were inflated to USD value at the year 2015 using the USD Consumer price index (CPI) for all urban consumers (CPI-U).

^a Hib vaccination program was cost-effective only if intangible benefits, i.e. avoiding the pain and suffering caused by the disease, were incorporated in the model analysis.

^b Other perspective was divided into two definitions: (1) direct cost perspective including direct medical and direct non-medical costs in the analysis was used in three studies^18,22,26; (2) Patients‘families perspective (co-payments) was used in one study²¹.

Table 4. Incremental cost effectiveness ratios (cost per case averted, cost per death averted, and cost per life-year gained).

Reference (year)	Incremental cost effectiveness ratios, USD
	Cost per case averted	Cost per death averted	Cost per life-year gained
Ginsberg et al.^17 (1993)
Single Hib vaccination (4-dose schedule)	—	1,512,593.07	19,807.10
McIntyre et al.^22 (1994)
Combined Hib vaccination (DTP-Hib)
1-dose schedule	—	—	1,717.59
2-dose schedule	—	—	10,519.59
3-dose schedule	—	—	9,451.10
Livartowski et al.^21 (1996)
Payer perspective	—	—	16,128.64
Other perspective^a	—	—	30,642.36
Duke^39 (2002)	—	1,602.07	—
Platonov et al.^23 (2006)	91,118.69 ± 18,844.97	888,442.24 ± 308,005.71	—
Akumu et al.^28 (2007)	336.27	1,501.90	—
Gessner et al.^31 (2008)
Societal perspective
Combined vaccination	—	3,648.84	—
Single Hib vaccination	—	5,073.17	—
Moradi-Lakeh et al.^36 (2012)	7,435.16	86,828.70	—
Gupta et al.^32 (2013)
Healthcare payer perspective	43.09	9,732.04	331.44
Societal perspective	127.05	28,728.80	977.73
Gargano et al.^38 (2015)
Single Hib vaccination
Hib (one-dose schedule)	670.26	7,130.44	—
Hib (two-dose schedule)	986.54	10,945.98	—
Combined vaccination
Hib with PCV10 (one-dose schedule)	445.64	4,740.86	—
Hib with PCV10 (two-dose schedule)	740.98	4,740.86	—
Le et al.^33 (2015)
Healthcare payer perspective	6,954	29,449	—
Societal perspective	6,252	26,476	—

All costs were presented in United States Dollars (USD), and were inflated to the value at 2015 using the USD Consumer price index (CPI) for all urban

consumers (CPI-U).

Abbreviations. Hib, Haemophilus influenzae type b; PCV10, 10-valent Pneumococcal Conjugated Vaccine; USD, United States Dollar.

^a Other perspective was defined as: (1) direct cost perspective including direct medical and direct non-medical costs in the analysis was used in three studies²²; (2) Patients‘families perspective (co-payments) was used in one study²¹.

Table 5. Incremental cost-effectiveness ratios (cost per QALY gained).

Reference (year)	Cost per QALY gained, USD
Harris et al.^18 (1994)
Healthcare payer perspective	242,728.45
Other perspective^a	6,462.06
McIntyre et al.^22 (1994)
Combined Hib vaccination (DTP-Hib)
1-dose schedule	2,142.20
2-dose schedule	15,898.59
3-dose schedule	12,059.68
Livartowski et al.^21 (1996)
Payer perspective	10,154.83
Other perspective^a	19,291.79
Zhou et al.^26 (2002)
Societal perspective	6,536.00
Other perspective^a	8,000.44

All cost were presented in United States Dollars (USD), and were inflated to the value at 2015 using the Consumer price index (CPI) for all urban consumers (CPI-U).

Abbreviations. Hib, Haemophilus influenzae type b; QALY, Quality-adjusted life year; USD, United States Dollar.

^a Other perspective was defined as: (1) direct cost perspective including direct medical and direct non-medical costs in the analysis was used in three studies^18,22,26; (2) Patients‘ families perspective (co-payments) was used in one study²¹.

Table 6. Incremental cost-effectiveness ratios (cost per DALY averted).

Reference (year)	Cost per DALY averted, USD
Platonov et al.^23 (2006)	14,510.10 ± 3,869.09
Akumu et al.^28 (2007)	47.68
Broughton^29 (2007)	98.68
Gessner et al.^31 (2008)
Societal perspective
Combined vaccination	84.59
Single Hib vaccination	116.60
Niessen et al.^41 (2009)
Low scenario estimate
SEAR B	192.96
SEAR D	288.84
WPR B	139.56
AFR D	32.77
AFR E	55.83
EMR D	42.48
AMR D	116.51
AMR B	157.77
High scenario estimate
SEAR B	493.94
SEAR D	355.59
WPR B	740.30
AFR D	83.74
AFR E	145.63
EMR D	108.01
AMR D	297.33
AMR B	406.56
Moradi-Lakeh et al.^36 (2012)	5,495.87
Gupta et al.^32 (2013)
Healthcare payer perspective	904.82
Societal perspective	306.03
Clark et al.^30 (2013)
Healthcare payer perspective	410.87
Societal perspective	359.78
Griffiths et al.^9 (2013)
Healthcare payer perspective
GAVI (Africa)	27.17
GAVI (Asia)	38.04
LMICs	119.57
UMICs	492.39
Societal perspective
GAVI (Africa)	Cost-saving
GAVI (Asia)	10.87
LMICs	98.91
UMICs	401.09
Gargano et al.^38 (2015)	—
Single Hib vaccination	140.27
Hib (one-dose schedule)	206.47
Hib (two-dose schedule)
Combined vaccination
Hib with PCV10 (one-dose schedule)	93.27
Hib with PCV10 (two-dose schedule)	155.08
Le et al.^33 (2015)
Healthcare payer perspective	1,393.88
Societal perspective	1,252.45

All cost were presented in United States Dollars (USD), and were inflated to the value at year 2015 using the Consumer price index (CPI) for all urban consumers (CPI-U).

Abbreviations. AFR, WHO African Region; AMR, WHO Region of the Americas; EMR, WHO Eastern Mediterranean Region; SEAR, WHO South-East Asia Region; WPR, WHO Western; WHO regions were subdivided based on child and adult mortality:

A, very low child and very low adult mortality; B, low child and low adult mortality; C, low child and high adult mortality; D, high child and high adult mortality; E, high child and very high adult mortality. DALY, Disability-adjusted life year; GAVI, Global Alliance for Vaccines and Immunization; Hib, Haemophilus influenzae type b; LMICs, Lower Middle-Income Countries; PCV10, 10-valent Pneumococcal Conjugated Vaccine; UMICs, Upper Middle-Income Countries; USD, United States Dollars.

Parameters influencing cost-effectiveness

Our review demonstrated that the incidence rate of Hib disease was the most influential determinant of cost-effectiveness findings in 11/27 studies (41%)^{8,9,20–24,31,32,37,39}, followed by vaccine price (6/27: 22%)^{18,20,22,23,32,37}, discount rate (4/27: 15%)^26,33,35,36, vaccine efficacy (3/27: 11%)^20,22,38, and vaccine coverage rate (3/27: 11%)^19,27,30. As described in Supplementary Appendix Table A8, incidence rates of Hib disease varied widely, e.g. for meningitis, 13.7 cases per 100,000 children in Thailand³⁷, but 157 cases per 100,000 children in Sweden²⁵, in keeping with previous studies^43,44 comparing Asia, where the incidence was 1/10 of that in North America and Europe. Several reasons may explain this disparity: difficulty in detection or a sub-optimal microbiologic capacity in identifying Hib disease in clinical specimens⁴⁵, therefore distorting the economic evaluation.

The second influence on cost-effectiveness analysis is vaccine price. Thus, the vaccine price in Thailand was USD 9.98, accounting for 80% of the overall vaccination cost³⁷. This was the highest price among LICs, LMICs, and UMICs, and reducing its cost by 92% creates cost-effectiveness.

Likewise, a study from Russia²³ demonstrated that the vaccine price had to reduce ∼90% of the current price, in order for the Hib vaccine to be cost-effective. Notably, in a study with cost-effective results, i.e. Gupta et al.³² demonstrated that vaccine price was found to be the main determinant of its cost-effectiveness. In this review, the vaccine price was found to range from USD 0.3³¹ in LMICs (Indonesia) to 22.5²⁵ in HICs (Sweden). Interestingly, the average vaccine price in HICs was ∼3-times higher than that among LICs, LMICs, and UMICs. In LICs, LMICs, and UMICs, vaccine prices ranged from USD 0.3³¹ to 9.98^30,37 (Average = USD 3.13), while in HICs it ranged from USD 4.32³⁹ to 22.5²⁵ (Average = USD 11.59).

Discussion

Our systematic review shows that the economic value of Hib vaccination program has been demonstrated by the majority of studies. The incidence rate of Hib disease and vaccine price were important determinants of vaccination implementation. Compared with a previous systematic review, more studies were captured by including more databases, search terms, and a longer search period.

To date, only China and Thailand have no national Hib vaccination program. In China, the high cost of vaccine might be an important concern, while other possible reasons include the limited awareness among policy-makers, healthcare providers, as well as clinicians regarding Hib vaccine, the inability of the national immunization program to include additional vaccine, etc.⁴⁶. In addition, it is revealed that, in China, due to the lack of economic evaluation of Hib vaccine conducted, this has deterred informed decision-making among local policy-makers. As for Thailand, it was found that the Hib vaccination program was cost-effective only if intangible benefits, i.e. avoiding the pain and suffering caused by the disease, were incorporated in the model analysis. As a result, the vaccine is not included in Thailand’s national immunization program. Indeed, the inclusion of Hib vaccine in the national immunization program of these countries is likely to be challenging for policy-makers, since many other factors need to be considered, including competing priorities, rather than taking into account only the cost-effectiveness findings.

Among the influential parameters, vaccine efficacy was estimated RCT data derived from non-local studies, while incidence rates were estimated based on local observational studies. Although RCT minimizes various biases and confounders, it restricts participant characteristics, assigned interventions, and outcomes, hence limiting the generalizability. Furthermore, using RCT data from a non-local study is potentially misleading. However, due to the limited local data available, caution should be exercised when non-local data are used. On the other hand, using data from a local observational study is recommended when estimating incidence rates. In addition, it is important to ensure quality of the observational study to generate an accurate estimate based on real world evidence under the specified local context.

Ideally, a dynamic transmission model in an economic evaluation of Hib vaccination should be used. This is to capture the full impact of the vaccine, which includes not only the direct effect of vaccine by reducing the probability of individuals developing the disease, but also the indirect herd effects that lower transmission to others. However, this is limited when a static transmission model is used, as it assumes a constant risk of infection over time. Therefore, only the direct effect of the vaccine will be captured in the study model. Nevertheless, the use of a dynamic transmission model requires a complex technical approach, and specific data are needed. Although the herd immunity effect was examined in sensitivity or scenario analysis in two studies, the probability of an individual becoming infected and the number of infectious individuals in the population were not considered. This might be due to the difficulty in acquiring relevant input parameters and the lack of technical expertise in conducting the study.

The use of the so-called fixed GDP based thresholds of 1× and 3× GDP per capita⁴⁷ in the decision-making process may demonstrate a small impact for country-level decision-making⁴⁸. It is timely that countries should develop a context-specific process for decision-making that is supported by legislation, has stakeholder buy-in, as well as being transparent, consistent, and fair. In addition, rather than considerations based solely on a single cost-effectiveness threshold value, cost-effectiveness information should be combined with other methods, i.e. budget impact and feasibility considerations, in the decision-making process^48,49.

To the best of our knowledge, this is the most up to date systematic review of economic evaluations of Hib vaccination. A number of strengths in our review should be highlighted. First of all, we conducted a comprehensive search through databases, reference lists of relevant published studies and grey literature, in order to ensure that all economic evaluations under specified inclusion criteria were included. Second, the source of important epidemiological parameters was identified, since the variation of these values influence the cost-effectiveness, and, hence, underpins the reliability and validity of the study results. Additionally, we identified influential parameters that potentially affected the economic evaluation of Hib vaccination in order to provide researchers with insights when conducting the future economic evaluation, especially for Hib vaccination programs. Furthermore, we appraised the quality of included studies using a checklist for appraising the quality of economic evaluations of the immunization program developed by WHO. Therefore, we believe that our results are valid and provide useful information for policy-makers when considering implementation of Hib vaccination.

Several limitations in this review should be acknowledged. First of all, the findings of this review are limited by our search strategies, including the keywords of searching, databases searched, the time period of searching, and the inclusion criteria. Second, we included only studies that were published in English, despite the search terms not being limited by language restriction when performing the search, but because of the lack of language experts involved in our study. Furthermore, our findings of the cost-effectiveness of Hib vaccination being either cost-effective or not might not be applicable to other settings, since it was country-specific. Moreover, many factors need to be considered apart from the economic standpoint, i.e. the availability of healthcare workforces, budget constraints, political issues, etc. However, we believe that the value of this review is two-fold: to describe the descriptive information of economic evaluations of Hib vaccination and present the result of cost-effectiveness analyses, as well as to provide healthcare providers, researchers, and policy-makers with the trend of economic evaluation, methods of conducting the study, and crucial input parameters that were recommended for the future economic evaluations conducted, especially for Hib vaccination programs.

Conclusion

Hib vaccination programs are mostly found to be cost-effective across geographic regions and national income levels, and are recommended for inclusion into all national immunization programs. The incidence rate of Hib disease and vaccine price are important concerns for the vaccine implementation. Our findings are expected to support policy-makers in making decisions on allocating the limited resources of the Hib vaccination program effectively.

Transparency

Declaration of funding

This study is funded by the World Health Organization (WHO), Geneva, Switzerland.

Declaration of financial/other relationships

The authors have no competing interests.

Acknowledgments

We would like to thank Professor C. Norman Scholfield and Huey Yi Chong, who assisted in English language editing and revised the manuscript.