The cost effectiveness of palivizumab: a systematic review of the evidence

Abstract

Objective:

Palivizumab is a prophylactic therapy shown to reduce the number of respiratory syncytial virus (RSV)-related hospitalizations but has a high acquisition cost. The objective was to systematically examine the cost effectiveness of palivizumab in defined infant groups and identify important cost and outcome determinants.

Methods:

Literature searches of MedLine, the Cost-Effectiveness Analysis registry and the UK NHS Economic Evaluation Database (NHS EED) were conducted to identify economic evaluations of palivizumab compared to no prophylactic treatment for RSV prevention in any infant population. Study quality was evaluated using Quality of Health Economic Studies (QHES) criteria and results converted to 2009 CAN$ for comparison.

Results:

A total of 23 articles meeting inclusion criteria were identified, including 11 cost-utility analyses (CUAs) and 12 cost-effectiveness analyses (CEAs). Quality of individual analyses was fairly high (range 60–100, median 86). Results ranged from cost dominance for prophylaxis to $3,365,769/QALY depending on population, outcome measures, and input parameters. Base-case and sensitivity-analysis mortality rates varied between studies and influenced results.

Conclusions:

RSV prophylaxis with palivizumab is cost effective in specific groups of high-risk infants, especially those with multiple environmental risk factors. Cost-effectiveness estimates vary between populations and settings and are more positive in those at highest risk for RSV hospitalization.

Limitations:

Direct comparison of the published reports was limited by restriction to English language articles and the varied methodologies, input measures, and populations across the studies reviewed. Although reported currencies were converted to a common unit for comparison, this does not completely account for monetary and inflation differences.

Key words:

• Cost effectiveness analysis
• Palivizumab
• Respiratory syncytial virus

Introduction

Respiratory syncytial virus (RSV) infection occurs in nearly 50% of infants during the first year of life and almost all children have had RSV-related upper or lower respiratory tract disease (LRTI) by 2 years of age1,2. Approximately half the children who develop RSV infection during their first year become re-infected1. In the majority of cases, affected infants are healthy but 7.7% require subsequent medical attention from family practitioners and 2.6% require assistance in the emergency department3. A total of 0.5–2% of all infants with RSV infection are hospitalized with bronchiolitis and pneumonia and the severity of illness is often determined by preterm gestational age (GA), low birth weight and existing underlying medical disorders such as airway anomalies, cardiac and chronic lung disease and neuromuscular impairments4–7. Environmental risk factors such as number of siblings, crowding in the household and day care attendance have also been found to increase the risk of RSV infection and hospitalization8. In Canada, RSV hospitalizations number 5800–12,000 annually while in the US between 1997–2000 there were 311,000 admissions for bronchiolitis4,9,10. The burden of illness is substantial world-wide. In Kenya, RSV accounts for 85,000 LRTI cases annually11, in the Netherlands there are 2000 annual RSV hospitalizations12 while in Liverpool UK, the RSV hospitalization rate was 19.7% over the 1998–2000 RSV seasons13. The overall RSV related mortality rate is 1–4% and has decreased significantly because of state of the art pediatric intensive care management6,7,14–16. However, the burden of illness has a significant impact on health care resources both within the hospital and in the community following patient discharge17–19. Of equal importance are family-related costs relative to the care of these children and the potential impact on the child’s health related quality of life and long term respiratory sequelae such as wheezing, asthma and pulmonary function20–24.

The cost incurred for RSV hospitalization in Canadian children 1–4 years of age between 1993 and 1994 was CAN$18.5 million, with 62% of the direct costs being for hospital services4. Similarly, in the US from 1997–2000, the average annual direct costs for hospital care totaled US$750 million and in 2002 were estimated at US$1.1 billion10,25. Importantly, costs vary significantly in infants with risk factors and those with pre-existing medical disorders. In one study, costs increased from US$3777 for infants without risk factors to $7376 for prematurity, $11,335 for those with congenital heart disease (CHD) and $13,241 for infants with congenital anomalies18. Apart from direct costs, RSV illness imposes a significant financial burden on families. The average out of pocket expenses range from $214–295 per child to $644 for premature infants and encompass costs for travel, lost work days and family doctor consultation visits26,27.

Palivizumab, a genetically engineered, humanized monoclonal antibody initially became available through the special access program and was approved for RSV prophylaxis in Canada in 2002. The efficacy and safety of the product was proven in infants less than 35 weeks’ gestation and those with chronic lung disease <2 years of age through a large scale, multicenter, randomized, double-blind, placebo-controlled trial15. A similar randomized trial conducted in infants with hemodynamically significant CHD demonstrated an overall 45% relative reduction in RSV hospitalizations (p = 0.003)16. This set the stage for universal position statements by pediatric committees on the use of palivizumab prophylaxis in high-risk populations and the strategy rapidly gained approval28–30.

As with all biologics, the acquisition costs for palivizumab are substantial. To address value for money, there have been several cost-effectiveness and cost-utility analyses on the use of palivizumab prophylaxis in high-risk populations based on societal and payers’ perspectives. However, there has been a wide variability in results and conclusions across studies. The primary objective of this study was to examine the cost effectiveness of palivizumab prophylaxis and the factors that affect it by conducting a systematic review of the scientific literature and existing databases and critically evaluating the results and determinants of outcomes.

Methods

Literature search

A systematic literature search was conducted to identify studies examining the cost effectiveness of palivizumab versus no prophylaxis for RSV in children. A MEDLINE (Ovid) search was performed from 1950 to February 2010 using the following criteria: (‘respiratory syncytial virus’ OR ‘rsv’) AND (palivizumab OR monoclonal antibody$) AND (‘costs and cost analysis’ OR econom$ OR ‘cost-benefit analysis’ OR cost effective$ OR cost$ OR pharmacoeconomic$ OR budget$). Searches of the Cost-Effectiveness Analysis registry and the UK National Health Service (NHS) Economic Evaluation Database (NHS EED) were also performed, as well as examinations of reference lists of the review articles retrieved. The search was limited to English language studies.

Cost-utility analyses (CUAs), cost-effectiveness analyses (CEAs), and cost-benefit analyses (CBAs) were considered for inclusion where palivizumab prophylaxis was compared to no prophylactic treatment for RSV prevention in any infant population. Cost comparison studies and cost analyses were excluded.

Analysis

Studies were assessed for quality using the guidelines for economic submissions to the British Medical Journal31 and the Consensus Health Economic Criteria (CHEC) list32 and given a quality score (QS) on a scale of 0–100 based on the Quality of Health Economic Studies (QHES) Instrument33. Results were converted to 2009 CAN$ to facilitate comparison using the Bank of Canada’s online currency converter34 and inflation calculator35. Where currency or price year was not stated, country and year of study publication were used.

Results

A total of 23 articles meeting inclusion criteria were identified for review (Figure 1); 11 of these were CUAs measuring outcomes in incremental cost-effectiveness ratios (ICERs) based on cost per quality-adjusted life year (QALY) gained and 12 were CEAs measuring outcomes in ICERs per life-year gained (LYG) or hospital admission prevented (HAP). Characteristics of these studies are found in Table 1. Table 2 compiles the base-case results of each study. QHES scores are given in Table 3.

Figure 1.  Search strategy for articles included in the review.

Table 1.  Characteristics of cost-effectiveness analyses of palivizumab versus no prophylaxis.

Paper	Country	Population	Study design	Time horizon	Perspective
Chirico et al.	Italy	High-risk preterm (<36 weeks’ GA) infants	CUA	Lifetime	Payer
ElHassan et al.	USA	Preterm (26–32 weeks’ GA) infants without CLD	CUA	8 years	Societal
Lanctôt et al.	Canada	Preterm (32–35 weeks’ GA) infants	CUA	Lifetime	Payer (societal secondary)
Nuijten et al.	UK	Preterm (≤35 weeks’ GA) infants, infants with CHD or BPD	CUA	Lifetime	Payer (societal secondary)
Nuijten et al.	Netherlands	Preterm (32–35 weeks’ GA) infants, infants with BPD or CHD	CUA	Lifetime	Societal (payer secondary)
Nuijten et al.	Germany	Infants with CHD	CUA	Lifetime	Societal (payer secondary)
Nuijten and Wittenberg	Spain	Preterm infants ≤32 weeks’ GA, infants <6 months of age	CUA	Lifetime	Payer (societal secondary)
Resch et al.	Austria	Preterm (≤35 weeks’ GA) infants, infants with CHD or BPD	CUA	Lifetime	Payer (societal secondary)
Tam et al.	Canada	Term Inuit infants	CUA	Lifetime	Societal
Wang et al.	UK	Preterm infants (≤35 weeks’ GA) with and without BPD; infants with CHD	CUA	Lifetime	Payer (societal secondary)
Yount and Mahle	USA	Infants with CHD	CUA	Lifetime	Societal
Banerji et al.	Canada	Infants of all GAs in the Canadian Eastern Arctic	CEA	1 year	Payer
Chan et al.	Malaysia	Preterm or those with CHD or CLD	CEA	Not stated	Payer
Fariña et al.	Argentina	Infants ≤35 weeks’ GA and <6 months of age; ≤28 weeks’ GA and <1 year of age; or ≤2 years of age with BPD	CEA	Not stated	Societal
Joffe et al.	USA	Preterm infants (<36 weeks’ GA)	CEA	1 year	Societal
Lofland et al.	USA	Preterm infants; infants with CHD, BPD, or compromised immune function	CEA	Not stated	Payer
Rackham et al.	UK	Infants <1 year of age with CHD	CEA	Not stated	Payer
Reeve et al.	Australia	Infants of birth weight <2500 g: all; Indigenous; from multiple births	CEA	Not stated	Payer
Rodriguez et al.	Argentina	Preterm infants with and without BPD and siblings	CEA	Not stated	Payer
Roeckl-Wiedmann et al.	Germany	Preterm (≤35 weeks’) infants with a variety of risk factors	CEA	1 year	Societal
Simpson et al.	UK	Preterm (≤35 weeks’ GA) infants <6mos at start of RSV season; infants <2 years of age with BPD	CEA	Not stated	Payer
Stevens et al.	USA	Preterm infants (≤32 weeks’ GA)	CEA	1 year	Payer
Vogel et al.	New Zealand	Preterm infants (<32 weeks’ GA) with CLD	CEA	1 year	Payer

Table 2.  Results of palivizumab cost-effectiveness analyses.

Paper	Country	Population	Currency & year	Discount rate	Hospitalization rate (%)		Cost-effectiveness threshold	Base-case ICERs	Base-case ICERs (2009 CAD$)	Sensitivity analysis range	Concluded cost effective?
					Palivizumab	No prophylaxis
Cost-utility analyses
Chirico et al.	Italy	<33 weeks’ GA	2007 €	3%	2.0; 1.3 ICU	10.3; 3.0 ICU	€50,000	€9380.00/QALY	$14,881/QALY	€2665–21,477/QALY	Yes
		33–35 weeks’ GA			1.5; 1.3 ICU	9.8; 3.0 ICU		€14,937.32/QALY	$23,967/QALY
		BPD			5.6; 1.3 ICU	18.4; 3.0 ICU		€2,731.81/QALY	$4331/QALY
ElHassan et al.	USA	26–32 weeks’ GA, no CLD	2002 USD	3%	0.011–0.052	0.064–0.206	$200,000/QALY	$675,780–1,855,000/QALY	$1,223,156–$3,365,769/QALY	$103,053–1,855,000/QALY	No
Lanctôt et al.	Canada	32–35 weeks’ GA, no CLD	2007 CAN$	5%	1.4–11.4	6.3–51.2	CAD$50 K, 75 K, 100 K	$13,029–30,618/QALY	$13,354–$31,383/QALY	$6,618–50,701/QALY	Yes
Nuijten et al.	UK	≤35 weeks’ GA	2003 £	3.5%	4.8	10.6	£25,000	£14,883/QALY	$41,658/QALY
		BPD						£20,953/QALY	$58,648/QALY
		CHD						£6664/QALY	$18,653/QALY	£5508–58685/QALY	Yes
Nuijten et al.	Netherlands	32–35 weeks’ GA BPD CHD	2006 €	Economic outcomes 4.0%; clinical outcomes 1.5%	1.8 7.9 5.3	8.1 12.8 9.7	Not specified	€18,563/QALY €23,461/QALY €7,067/QALY	$26,862/QALY $33,962/QALY $10,279/QALY	Dominant to €34,856/QALY	Yes
Nuijten et al.	Germany	CHD	2006 €	5%	5.3	9.7	€30,000/QALY	€9,529/QALY	$13,860/QALY	€933–4089/QALY	Yes
Nuijten and Wittenberg	Spain	≤32 weeks’ GA, <6 months of age at the start of the RSV season	2006 €	3%	3.95	13.25	€30,000/QALY	€12,814/QALY	$18,638/QALY	Dominant to €19,752/QALY	Yes
Resch et al.	Austria	≤35 weeks’ GA	2006 €	5%	4.8	10.6	€30,000	€20,704/QALY	$30,114/QALY	€2244–41,707/QALY	Yes
		BPD						€31,867/QALY	$46,351/QALY
		CHD						€11,390/QALY	$16,566/QALY
Tam et al.	Canada	Term Inuit infants	2007 CAN$	5%	1.8	10	CAD$50 K, 75 K, 100 K	$39,435/QALY	$40,421/QALY	$21,737–32,709/QALY	Yes
Wang et al.	UK	≤35 weeks’ GA without CLD	2006 £	3.5%	0.018	0.081	£30,000/QALY	£454,100/QALY	$1,038,231/QALY	£24,100–3,905,500/QALY	No
		CLD			0.079	0.128		£63,800/QALY	$134,160/QALY
		CHD			0.0532	0.0972		£79,800/QALY	$168,608/QALY
Yount and Mahle	USA	CHD	2002USD	3%	5.3	9.7	Not specified	$114,337/QALY	$207,458/QALY	$60,233–382,675/QALY	No
Cost-effectiveness analyses
Banerji et al.	Canada	All GAs, Canadian Eastern Arctic	2007 CAN$	Not specified	3.65	16.6	Not specified	$39,440/HAP	$40,426/HAP	None	Yes
Chan et al.	Malaysia	<2 years of age with CLD; ≤32 weeks’ GA; 32–25 weeks’ GA with multiple risk factors	1997 USD	Not specified	4.8	10.6	Not specified	$2921.13/HAP	$4987/HAP	None	No
Fariña et al.	Argentina	Preterm or BPD	2000 USD	Not specified	10.7	23.8	Not specified	$15,358/HAP	$26,474/HAP	$2171–$48,630/HAP	No
Joffe et al.	USA	<36 weeks’ GA	1995 USD	3%	0.54–11.1	1.2–24.6	Not specified	$33,000–1,200,000/LYG	$60,495–$2,198,512/LYG	$1500–150,000/HAP	No
Lofland et al.	USA	Preterm, CHD, BPD, or compromised immune function	Not stated	Not specified	5 (RSV incidence rate)	10–38 (incidence rate)	Not specified	$0 (cost savings) – $79,706/HAP	$0–$121,445/HAP	None	Not stated
Rackham et al.	UK	<1 year of age with CHD	2002 USD	Not specified	7	13	Not specified	$34,100/HAP	$61,872/HAP	None	Not stated
Reeve et al.	Australia	Birth weight <2500 g	AUS $ Year not given	Not specified	0.44	2.2	Not specified	$88,547/HAP	$79,333/HAP	None	No
		Indigenous and birth weight <2500 g						$73,294/HAP	$65,667/HAP
		Birth weight <2500 g, multiple birth						$69,861/HAP	$62,592/HAP
Rodriguez et al.	Argentina	Preterm:	2007 USD	Not specified	1–21	5–36	Not specified			None	Yes
		With BPD, without siblings						$43,027/HAP	$51,375/HAP
		With BPD, with siblings						$21,152/HAP	$25,256/HAP
		Without BPD, without siblings						$89,902/HAP	$107,345/HAP
		Without BPD, with siblings						$13,198/HAP	$15,758/HAP
Roeckl-Wiedmann et al.	Germany	≤35 weeks’ GA, variety of risk factors	€ Year not given	None	1.53–24.2	3.4–53.9	Not specified	€6639–204,684/HAP	$12,052–$371,569/HAP	€493–75,993 (3 highest risk groups)	No
Simpson et al.	UK	≤35 weeks’ GA and <6 months of age; <2 years of age with BPD	2000 £	1.5%	4.8	10.6	£30,000/LYG	£96,000/LYG	$270,916/LYG	£36,000–72,000/HAP £80,000–£161,000/LYG	No
Stevens et al.	USA	≤32 weeks’ GA	USD Year not given	Not specified	3.42	7.6	Not specified	$37,612/HAP	$70,183/HAP	None	No
Vogel et al.	New Zealand	<32 weeks’ GA	2000 NZD	Not specified	1.8–14.0	8.2–22.9	Not specified	NZ$28,000–166,700/HAP	$25,084–$149,337/HAP	NZ$13,000–NZ$223,000	No

ICER, incremental cost-effectiveness ratio; GA, gestational age; BPD, bronchopulmonary dysplasia; CLD, chronic lung disease; CHD, congenital heart disease; QALY, quality-adjusted life year; LYG, life-year gained; HAP, hospital admission prevented.

Table 3.  Quality of health economic studies (QHES) scores.

Paper	QHES score	Paper	QHES score
Chirico et al.	86	Chan et al.	75
ElHassan et al.	84	Fariña et al.	67
Lanctôt et al.	94	Joffe et al.	90
Nuijten et al. (UK)	94	Lofland et al.	76
Nuijten et al. (NL)	94	Rackham et al.	68
Nuijten et al. (GER)	94	Reeve et al.	67
Nuijten and Wittenberg	91	Rodriguez et al.	60
Resch et al.	91	Roeckl-Wiedmann et al.	89
Tam et al.	100	Simpson et al.	76
Wang et al.	94	Stevens et al.	75
Yount and Mahle	84	Vogel et al.	93
Banerji et al.	78

Cost-utility analysis (CUAs)

The quality of CUAs was fairly high (from 84 to 100 with a median score of 94). The studies clearly described their study population, perspective, and time horizon. Efficacy of palivizumab was comprehensively established, and both results and consequences were succinctly documented. All but one study36 used a lifetime time horizon, and all discounted future costs and benefits and performed sensitivity analyses. Only three studies36–38 explicitly discussed magnitude and direction of potential biases. There was variability in input parameters and costs included in each analysis.

Mortality rate and inclusion of sequelae such as asthma and wheezing were notable sources of variability in the cost-effectiveness results. Three analyses – two by Nuijten et al. (UK 200839, Netherlands 200940) and one by Resch et al.41 – were conducted using similar decision analytical models and input parameters examining preterm infants ≤35 weeks’ GA, infants with bronchopulmonary dysplasia (BPD), and infants with hemodynamically significant congenital heart disease (CHD). The UK analysis (QS 94) reported ICERs of $41,658/QALY, $58,648/QALY, and $18,653/QALY, respectively. Resch et al. (QS 91) calculated similar values of $30,114/QALY for preterm infants, $46,351/QALY for infants with BPD, and $16,566/QALY for infants with CHD. The Netherlands analysis (QS 94) determined ICERs of $26,862/QALY for the preterm group, $33,962/QALY for BPD infants, and $10,279/QALY for subjects with CHD. A fourth study by Chirico et al.37 (QS 86) also included a subgroup of infants with BPD and calculated an ICER of $4331/QALY for that population. However, a UK cost-utility analysis performed by Wang et al. (QS 94) for the same three groups found different results42. ICERs were $1,038,231/QALY for preterm infants without CLD, $134,160/QALY for preterm infants with CLD, and $168,808/QALY for infants with CHD.

Wang et al.42 used mortality rates of 0.43% for preterm infants without BPD and 4.0% for those with BPD after RSV hospitalization compared to the Nuijten et al. and Resch et al. analyses that utilized a mortality rate of 8.11% in both groups, based on the Sampalis study19. However, the lowest ICER for the BPD group was calculated by Chirico et al. using the more conservative 4.0% rate37. In the CHD group, Nuijten et al.39,40 and Resch et al.41 applied different mortality rates in the palivizumab and no prophylaxis groups, ranging from 2.3 to 3.3% and 4.0 to 4.2% respectively, although trials had not been powered to show a significant difference in mortality between the two groups. This variation reflects diversity found in the literature and represents different measures, with the lower values including early mortality rates only during hospitalization, while the higher value comprises longer term, all-cause mortality, in confirmed and probable RSV positive hospitalizations versus matched control groups who were not hospitalized with RSV illness19. The effect of mortality on model output was demonstrated in sensitivity analyses performed by Wang et al. and Nuijten et al. (UK). Wang et al. reported ICERs ranging from $47,933/QALY for an 8.11% mortality rate to $7,768,528/QALY for a 0.05% rate, while Nuijten et al. estimated a range of ICERs from $25,846/QALY for a mortality rate of 0 in the palivizumab group to $163,588/QALY if rates are assumed to be equal in the palivizumab and control groups. Similarly, Wang et al. did not account for the potential increased risks and downstream costs of asthma in children following RSV infections in infancy. Unlike Chirico et al., both articles by Nuijten et al., and Resch et al. included asthma only in a secondary analysis, finding slightly lowered ICERs ($14,187–31,522/QALY). The impact of RSV infections in early infancy on subsequent wheezing and asthma is still contentious. However a meta-analysis of 12 clinical trials demonstrated that the association between RSV illness at a young age and subsequent asthma exists, but the effect steadily declines with increasing age into adolescence20.

CHD was determined to be the most cost-effective indication in the previously mentioned studies. An additional CHD analysis conducted by Nuijten et al.43 in Germany (QS 94) calculated an ICER of $13,860/QALY from the societal perspective. Yount and Mahle44 (QS 84) studied a similar population and calculated an ICER of $207,458/QALY for infants with CHD, significantly higher than those found in the other analyses. The low estimate of benefit from palivizumab prophylaxis was ascertained on the basis of a relatively low RSV mortality rate of 3% coupled with quality of life estimates that were derived solely from CHD rather than RSV-related effects.

Specific gestational age was also shown to affect cost-effectiveness results. This is a reflection of the variability in effect size by gestational age found in the clinical trials. Chirico et al.37 calculated an ICER of $14,881/QALY for infants <35 weeks’ GA compared to $23,967/QALY for infants 33–35 weeks’ GA. ElHassan et al. (QS 84) reported on a cohort of preterm infants (GA 26–32 weeks) without CLD, over an 8-year time horizon36. ICERs were stratified by GA and ranged from $1,223,165/QALY to $3,365,769/QALY. That analysis limited the outcome benefit of palivizumab to a conservative estimate of the risk of developing asthma, with consequent health-utility increases and hospital cost decreases; RSV mortality was not included due to limited evidence from clinical trials that it is affected by palivizumab. These assumptions largely account for the higher ICER values, but both studies indicate that cost effectiveness is higher in infants of lower GA.

The impact of indirect costs, asthma, and risk scoring was demonstrated clearly by Lanctôt et al. (QS 94). This study analyzed costs for preterm infants of GA 32–35 weeks45 and calculated ICERs of $31,383/QALY for direct medical costs not accounting for asthma, and $21,447 including asthma costs, comparable to the estimate from Chirico et al. for the 33–35-week GA group. With the inclusion of indirect costs the ICERs decreased substantially to $23,292/QALY and $13,354/QALY, respectively. Again, the higher 8.11% mortality rate19 was employed in the base case since the Sampalis et al. study is the only nested case-control trial that examined RSV-related mortality in a cohort of 33–35 GA infants post-hospitalization, over a mean follow-up duration of 2.1 years. Importantly, patient groups were further stratified by risk scores, with results showing strong evidence for cost effectiveness in infants with two or more risk factors and those who were at moderate-to-high risk of RSV hospitalization. This was also shown by Wang et al., with cost effectiveness increasing considerably in infants with siblings in day care or school42. Nuijten and Wittenberg46 (QS 91) utilized their original model and local data from Spain to study a population of preterm infants ≤32 weeks’ GA and <6 months of age at the start of the RSV season. ICERs of $18,638/QALY were determined for direct medical costs. This decreased to $9451/QALY with accounting for sequelae, and dominant values were achieved when indirect costs were included. This study further emphasizes the effects of sequelae and indirect costs. Notably, the Spanish data provided a higher effectiveness value for palivizumab using a lower mortality rate, though the higher 8.11% rate19 was also incorporated in a scenario analysis.

Tam et al.38 (QS 100) studied a specific population of term Inuit infants in the Canadian Eastern Arctic, stratified by rural or urban geographic location, reflecting distance from accessible medical care. An ICER of $40,421/QALY was determined for the whole population, ranging from cost savings for rural infants <6 months of age and high-risk rural infants <1 year of age to $153,527/QALY for infants <1 year of age in urban areas. This analysis distinctly illustrates the need for localized cost analyses in order to develop constructive guidelines for RSV prophylaxis.

The ICERs found in the CUAs range from cost dominance with palivizumab prophylaxis to $3,365,789/QALY for diverse high-risk populations based on cost, input, and outcome parameters examined. CHD is generally found to be the most cost-effective indication for palivizumab prophylaxis.

Cost-effectiveness analysis (CEAs)

The quality of CEAs varied from 60 to 93 with a median score of 75.5. The majority of publications clearly described their population and research question, but few used a lifetime time horizon and only two47,48 applied discounting to future costs and outcomes. There was some variability in costs accounted for in each study. Fewer than half (42%) used sensitivity analyses to assess the robustness of input parameters.

Two studies assessed cost effectiveness using life-years gained as an outcome measure47,48. Simpson et al. (QS 76) examined infants ≤35 weeks’ GA and under 6 months of age at the start of the RSV season and infants under 2 years of age with BPD, estimating an ICER of $270,916/LYG for the combined group. An analysis of infants ≤36 weeks’ GA by Joffe et al.47 (QS 90) examined risk groups stratified by GA, length of oxygen therapy in a neonatal intensive care unit (NICU), and month of NICU discharge. ICERs ranged from $60,495/LYG for infants 23–32 weeks’ GA who were oxygen dependent for ≥28 days and discharged between September and November to $2,198,512/LYG for infants 33–36 weeks’ GA who required oxygen for <28 days and were discharged between December and August of the following year. Both studies concluded that palivizumab was not cost effective in their study populations, though each stressed the importance of examining risk factors beyond the indications discussed. Joffe et al. used a low 1% mortality rate for infants hospitalized for RSV, and Simpson et al. assumed a 6-month course of prophylaxis rather than the more common 4–5-month course which significantly increased cost estimates.

Nine of the 11 CUA analyses also reported results in terms of LYGs; for most, these values were higher by approximately 50% than ICERs/QALY. Wang et al. found that the values were similar while Yount and Mahle calculated a slightly decreased ICER/LYG. Among all studies evaluating LYGs, results ranged from cost savings40,46 to an ICER of $2,198,512/LYG47 dependent on populations and input assumptions.

Of the ten cost-effectiveness analyses identified that measured outcomes in hospital admissions prevented (HAPs), four examined premature infants without BPD49–52. Values ranged from $15,758/HAP49 to $371,569/HAP50 and reinforced the importance of specific gestational age as well as environmental risk factors for RSV hospitalization and determinants of cost effectiveness. For infants ≤35 weeks’ GA, Roeckl-Wiedmann et al. (QS 89) found ICERS of $29,152/HAP for male infants with siblings in day care discharged between October and December and up to $235,962/HAP for male infants with no other risk factors50. Vogel et al. (QS 93) determined cost per HAP to be $19,743 for infants ≤28 weeks’ GA, $60,463 for infants 29–31 weeks’ GA, and $17,707 for infants discharged home on oxygen52. Rodriguez et al. (QS 60) calculated ICERs of $15,758/HAP for premature infants without siblings and $107,345/HAP for those with siblings49. Stevens et al. found ICERs ranging from $21,586/HAP for infants ≤28 weeks’ GA to $125,340/HAP for infants 30–32 weeks’ GA51. Each of these studies varies extensively by subgroups within the population, demonstrating the positive effect of risk factors such as home discharge on supplemental oxygen, presence of siblings in or out of day care and discharge date during the RSV season on cost effectiveness.

Two analyses also encompassed preterm infants with BPD. Roeckl-Wiedmann et al. included a subgroup of male, premature infants ≤35 weeks’ GA with BPD and siblings for which the ICER was $12,052/HAP. Vogel et al. studied infants ≤28 weeks’ GA and 29–31 weeks’ GA with BPD and found ICERs of $40,103/HAP and $149,337/HAP respectively. BPD enhanced cost effectiveness in Roeckl-Wiedmann et al. but decreased it in Vogel et al., largely because the former used an average palivizumab effectiveness rate rather than one stratified by indication. In the IMpact trial, infants with BPD had a 39% reduction in hospitalization versus 55% overall for patients in the study15. Only one CEA examined infants with CHD. Rackham et al.53 (QS 68) calculated an ICER of $61,872/HAP for infants with severe CHD, slightly higher than estimates from comparable CUAs.

Cost effectiveness was limited in studies examining cost per HAP in all high-risk groups together, further supporting the need for refined risk characterization. Chan et al.54 found a combined ICER of $4987/HAP for infants <2 years of age with BPD, infants ≤32 weeks’ GA, and infants of GA 32–35 weeks with multiple risk factors54. In Argentina, Fariña et al.55 (QS 67) focused on infants ≤35 weeks’ GA who were <6 months chronological age, infants ≤28 weeks’ GA and <1 year of age, and infants <2 years of age with BPD and calculated an ICER of $26,474/HAP55. Both of these studies concluded that palivizumab is not cost effective in these settings. Lofland et al.56 (QS 76) studied a high-risk population comprising premature infants, infants with BPD, CHD, or other congenital anomalies, and infants with compromised immune function. ICERs were reported as a function of incidence rate at two palivizumab price points, with a range from cost savings to $121,445/HAP. The higher cost and lower incidence scenarios are very similar to the other analyses and suggest limited cost effectiveness.

Banerji et al. and Reeve et al. studied unique at-risk populations57,58. Banerji et al. (QS 78) examined Inuit infants of all GAs in the Canadian Eastern Arctic. ICERs ranged from cost savings of $8321/HAP to a cost of $166,615/HAP depending on age and location (urban versus rural)57. Reeve et al. (QS 67) reviewed a population of infants with birth weights <2500 g in three risk groups: all infants, Indigenous infants, and those from multiple births. ICERs were $79,333/HAP, $65,667/HAP, and $62,592/HAP respectively58.

Cost per HAP varied from cost savings with palivizumab use to $371,569. Most concluded that palivizumab is not cost effective in large risk groups based on HAP, though none explicitly specified a threshold for acceptable results. Many further reinforced the importance of environmental risk factors49,50,54,57 in determining palivizumab cost effectiveness. Figure 2 summarizes the ICERs for outcomes in QALYs, LYGs, and HAPs in the most commonly investigated populations.

Figure 2.  Cost-effectiveness results for studies reporting outcomes for the most commonly investigated populations (white = BPD, black = CHD, horizontal lines = premature, diagonal lines = premature or BPD). Panel A: outcomes in quality-adjusted life-years (QALYs); Panel B: outcomes in life-years gained (LYGs); Panel C: outcomes in hospital admissions prevented (HAPs). ‘Premature’ population encompasses all subgroups of gestational ages ≤35 weeks studied. BPD, bronchopulmonary dysplasia; CHD, congenital heart disease; ICER, incremental cost-effectiveness ratio.

Discussion

Prophylaxis with palivizumab effectively reduces RSV hospitalization rates in high-risk infants15. Many cost analyses have been performed to quantify the economic burden of this strategy and assess cost effectiveness. Cost-effectiveness results varied widely between studies for all outcomes assessed. Input measures and costs accounted for in individual studies had a strong impact on overall costs and outcomes. Cost effectiveness was shown to be highest in a variety of high-risk subpopulations and from the perspective of society rather than the payer.

There was no significant difference in the quality of the studies that concluded palivizumab is or is not cost effective (t = 1.5, df = 19, p = 0.153). CUAs were generally more positive and of higher quality (t = 4.5, df = 19, p < 0.001) than CEAs. Three of the highest quality studies (QS ≥ 90) reported negative findings42,47,52 (Table 3). However, the majority demonstrated positive cost effectiveness and in this subset of 23 rigorous studies, the effects of discrepancies in key input assumptions strongly influenced the conclusions.

Several studies across all outcome measures reinforced the importance of environmental risk factors in the determination of palivizumab cost effectiveness38,45,48,50,55. The stronger an infant’s risk profile, the higher the probability that palivizumab will be a cost-effective option. This indicates that risk assessment and stringent application of country-specific guidelines for prophylaxis may prove beneficial, by targeting healthcare resources appropriately at predetermined high-risk infant populations.

Mortality rate was a significant source of variability between studies. The randomized placebo-controlled efficacy trials have not been powered to either provide an estimate of the mortality rate in the small number of infants hospitalized with RSV, or find a difference in mortality rate with and without palivizumab prophylaxis. The overall mortality rates in the treatment and placebo arms of the clinical trials ranged from 0.4 to 4.2 % over 5 months15,16. Only one nested cohort study is currently available that followed infants for 2 years after a confirmed or probable RSV hospitalization and found a relatively high mortality rate of 8.1%19. As a result, the mortality rates utilized in the analyses across all of the studies differed widely. Many of the studies in our review considered the effect of varying mortality rates in scenario analyses and showed that the models are sensitive to changes in this parameter. Although better data on the RSV specific mortality following palivizumab versus no palivizumab prophylaxis is essential to refining cost-effectiveness analyses, this end point may prove difficult to determine due to prohibitive sample sizes. Additionally, the dubious link between RSV infection and asthma is also variably accounted for in these analyses based on the current evidence.

With some exceptions42,44, cost effectiveness was generally higher when outcomes were measured in QALYs rather than LYGs. This indicates that the benefits of palivizumab are largely in quality of life effects rather than mortality or hospitalization costs, emphasizing the relevance of health-utility estimates in assessing palivizumab cost effectiveness. Assessing quality of life effects also helps to mitigate the uncertainty caused by lack of consensus on mortality rates by measuring outcomes differently.

The primary limitation of this review is the heterogeneity of the methodologies, costs and outcomes in the included analyses. Major cost drivers such as sequelae, as well as minor input parameters such as administrative costs, were variably accounted for across the reports. Discrepancies in the primary research on which the assumptions are based (e.g., mortality) may have also contributed to the variations identified among the studies. Although reported currencies were converted to a common unit for comparison, this does not completely account for monetary and inflation differences. As such, it is difficult to clearly determine specific reasons for overall cost disparities between countries. Finally, several non-English language articles were excluded from this review, further limiting international comparison.

Conclusion

In summary, RSV prophylaxis with palivizumab is cost effective within certain subgroups of high-risk infants. These subgroups include children with CHD, where cost effectiveness has been shown to be higher than that in preterm infants with or without BPD. Also, judicious use in the relatively small proportion of moderate-to-high risk 33–35 weeks’ GA infants based on the presence of risk factors is supported by scientific evidence59–61. In populations such as the aboriginal children where the risk of RSV hospitalization is extremely high, prophylaxis is a dominant strategy with cost savings. Environmental risk factors are also key indicators that can be used to target the most cost-effective patient groups. Country-specific studies based on local risk factors and models that accurately determine RSV hospitalization rates in defined populations will likely yield the best evidence for the cost effectiveness of palivizumab.

Transparency

Declaration of funding

No funding for this study was provided.

Declaration of financial/other interests

K.S. has disclosed that she has no relevant financial relationships. B.P. and K.L. are currently Co-Principal Investigators of a multicenter investigator initiated research study entitled ‘CARESS’ (Canadian Registry of Synagis). The Principal Investigators are supported by a grant in aid of research by Abbott Laboratories Ltd. The investigators for the CARESS study maintain control over any abstracts and presentations arising from the data and the rights to publish. B.P. and K.L. are also co-investigators on a research study funded by Abbott International entitled ‘Surveillance and cost analysis for Respiratory Syncytial Virus hospital admissions in the Arctic communities in Canada’.