3.1. Framework
Sixteen criteria were excluded from the Vaccinex framework (eradication potential, mode of transmission, image and goodwill, resistance offered by anti-vaccination groups, impact on the population of pregnant women, impact on school activities, time to development of symptoms, costs related to the production platform, productivity costs [presenteeism], generation of jobs in the country, impact on migrant population, impact on fertility, impact on the Lesbian, Gay, Bisexual, Transgender, and Intersex (LGTBI) population, impact on the female population, perception and fear, legal liability), and two criteria which were not present in Vaccinex were included (burden of disease – incidence on the outpatient setting; burden of disease – incidence on the inpatient setting). In addition, the disease prevalence criterion was renamed as “population in which the prevention strategy would be indicated”. The final framework for RSV assessment consists of 26 criteria (Table 1 and supplementary files 4 and 5).
3.2. Weights
Based on the mean weights given by the experts, the five criteria considered as most relevant to any evaluation of antibody-based preventive measures in RSV were efficacy of the preventive measure (4.6 ± 0.7 out of 5.0), severity of symptoms (4.4 ± 0.7), incidence on the inpatient setting (4.4 ± 0.9), lethality risk (4.2 ± 1.2) and serious adverse events (4.2 ± 1.3), while those of lesser relevance were innovation stimulus (2.3 ± 1.1), impact on productivity cost (2.6 ± 0.9), impact on caregivers (2.6 ± 1.0), public health awareness (2.7 ± 1.1) and cost of the disease on the patient (out of pocket expenses) [2.7 ± 0.5] (Figure 1). See supplementary file 6 for detailed weights, overall and by subgroups.
3.3. Absolute scores
The examination of the absolute scores (Figure 2 and Table 2) suggests that RSV is a disease with no effective alternative treatment and with a prevention alternative which is very limited (mean scores, availability of treatment: 4.4 ± 0.7 out of 5.0; prevention alternatives: 4.2 ± 0.7). The only antiviral treatment available (ribavirin®) can only be administered in severe cases, requires aerosolization in specific devices and its routine use is not recommended [25,26]. The only prevention alternative, palivizumab®, is only indicated for high-risk premature infants and in a hospital setting [18,27].
Furthermore, children with a disease caused by an infection by RSV are at high risk of developing comorbidities in the short, medium, and long term (mean score comorbidity risk: 3.6 ± 0.9). Additionally, the incidence of RSV is very high (mean score incidence of RSV cases: 4.0 ± 0.7), with infections occurring in 90% of children within their first 2 years of life [1,28]. This translates into a very high burden of disease, inappropriate consumption of antibiotics and healthcare resources, as reflected by its inpatient and outpatient incidence (39,690 RSV cases/100,000 children ≤1 years, 9 medical visits per infected patient [4]) vs (2,520 hospitalizations for RSV/100,000 children ≤2 years [29]), respectively (mean scores: incidence on the inpatient setting: 3.3 ± 1.3; incidence on the outpatient setting: 4.1 ± 0.8).
Moreover, it is a highly transmissible pathology (R0=4.5 [30], mean score transmissibility: 3.7 ± 1.0), with a high impact on the population of children (mean score: 4.2 ± 1.0), who present losses in quality of life after diagnosis and stress levels after hospital discharge of 39% and 79%, respectively [31,32].
Finally, bronchiolitis (one of the main diseases caused by RSV) has a rate of 82 deaths per every 100,000 children ≤2 years hospitalized [33], which is not considered as very high by the experts (mean score lethality risk: 2.1 ± 1.2). In addition, it presents moderate symptoms (mean score severity of symptoms: 3.0 ± 1.1) that usually last for three to seven days, which is regarded as not very long lasting by the MCE (mean score time of duration of acute symptoms: 2.7 ± 1.0). See supplementary file 6 and 7 for detailed absolute scores.
3.4. Relative scores
Overall, relative scores (Figure 3 and Table 3) indicate that nirsevimab was considered by the MCE as a preventive measure with clinical benefits, as it is much more effective than placebo (80% efficacy in preventing medically attended RSV-associated LRTI [34]; mean score efficacy of the preventive measure: 4.3 ± 0.5 out of 5.0), with a robust safety profile (safety outcomes associated with nirsevimab are comparable to those observed with placebo) (mean score serious adverse events: -0.1 ± 0.3).
This clinical benefit, added to the fact that nirsevimab is indicated for neonates and infants during their first RSV season (mean score population in which the prevention strategy would be indicated: 4.2 ± 1.6), is also associated with several economic and social benefits.
To assess the economic benefits, since the price of nirsevimab was unknown at the time of the study, and for the purpose of this exercise, it was benchmarked with the price of innovative vaccines in Spain. Thereby, experts granted a negative score on the monetary cost of the preventive measure, in relation to placebo (-3.7 ± 1.8).
Furthermore, there was consensus (>85% positive scores) that the implementation of nirsevimab vs. placebo would generate savings in other health system costs, such as hospitalizations, outpatient consultations, or emergency room care (mean score cost of the disease on the health system: 3.3 ± 1.5); that it would improve the labor productivity of caregivers, generating savings in costs related to absenteeism (mean score productivity cost – absenteeism: 2.6 ± 1.4); and in patients’ out-of-pocket expenses (mean score cost of the disease on the patient: 1.8 ± 1.8).
Moreover, the social benefits of using nirsevimab in comparison to placebo would translate, on the one hand, into an increase in public health awareness through greater social debate and greater parental attention in the case of introducing this preventive measure in the children childhood immunization schedule (mean score public health awareness: 2.3 ± 1.6). On the other hand, its universal recommendation would result in the elimination of health inequity (mean score impact on health inequity: 2.8 ± 1.6), contributing to avoiding RSV infections in the most disadvantaged populations, for whom RSV represents a greater burden in terms of comorbidity and mortality (see supplementary file 6 and 7 for detailed relative scores).
3.5. Final estimated value
The final estimated value (overall means, n=9) obtained in this MCDA in RSV was 0.56 ± 0.11 (0.32-0.67) for the comparison between nirsevimab and placebo (Figure 4), meaning that the introduction of nirsevimab would provide positive value in the view of the MCE.
Out of this total, 0.43 was related to the absolute criteria, whilst 0.13 was associated with the relative criteria. Individually, the criteria with the highest contribution to the final estimated value were efficacy of the preventive measure (7.5% of total value), availability of treatment (6.7%), impact on the population of children (6.5%), incidence on the outpatient setting (6.5%), and incidence of RSV cases (6.1%). See supplementary file 6 for detailed results.
3.6. Retest and sensitivity analysis
The consistency of the weights between test and retest was high, with an average intra-rater correlation coefficient (ICC) of 0.8366. Likewise, the retest scores and value estimates were very similar to the test, with average ICCs of 0.9189 and 0.9196, respectively. The retest value estimate was 4.2% higher than the test´s.
When replacing the original weights of this MCDA in RSV with the weights assigned in the Vaccinex study, the impact on outcomes was close to none, with final estimated values ranging from 0.55 to 0.56 (vs. 0.56 in the MCDA in RSV).
By excluding the outlier, the final estimated value increased 5.4%, resulting in 0.59. See supplementary file 8 for more details.
An alternative evaluation comparing nirsevimab vs. palivizumab was also undertaken in the context of RSV prevention in neonates and infants during their first RSV season, obtaining very similar results (final estimated value of 0.58). The criteria which mark the distinction between these two comparisons were the efficacy of the preventive measure (nirsevimab has an estimated value of 0.04 vs. placebo and 0.02 vs. palivizumab) and its monetary cost (nirsevimab obtained an estimated value of -0.03 vs. placebo, and 0.02 vs. palivizumab) (see supplementary file 9).