Impact of Nirsevimab Immunization on Pediatric Hospitalization Rates: A Systematic Review and Meta-Analysis (2024)

A systematic review with a meta-analysis was performed to gather available evidence on the effectiveness of monoclonal antibody nirsevimab in the prevention of lower respiratory tract diseases (LRTDs) due to respiratory syncytial virus (RSV) in children and newborns (CRD42024540669). Studies reporting on real-world experience and randomized controlled trials (RCTs) were searched for in three databases (PubMed, Embase, and Scopus) until 1 May 2024. Our analysis included five RCTs, seven real-world reports, and one official report from the health authorities. Due to the cross-reporting of RCTs and the inclusion of multiple series in a single study, the meta-analysis was performed on 45,238 infants from 19 series. The meta-analysis documented a pooled immunization efficacy of 88.40% (95% confidence interval (95% CI) from 84.70 to 91.21) on the occurrence of hospital admission due to RSV, with moderate heterogeneity (I2 24.3%, 95% CI 0.0 to 56.6). Immunization efficacy decreased with the overall length of the observation time (Spearman’s r = −0.546, p = 0.016), and the risk of breakthrough infections was substantially greater in studies with observation times ≥150 days compared to studies lasting <150 days (risk ratio 2.170, 95% CI 1.860 to 2.532). However, the effect of observation time in meta-regression analysis was conflicting (β = 0.001, 95% CI −0.001 to 0.002; p = 0.092). In conclusion, the delivery of nirsevimab was quite effective in preventing hospital admissions due to LRTDs. However, further analyses of the whole RSV season are required before tailoring specific public health interventions.

Despite the high disease burden associated with RSV, available options for preventing incident cases of RSV have been limited [18,45,46].Until recently, RSV candidate vaccines ultimately failed to achieve the targeted vaccine efficacy (VE) of 70% against confirmed severe RSV disease over at least one year post vaccination [47,48].In fact, only the mAb palivizumab has been licensed and extensively used in real-world settings as a preventative treatment [49][50][51][52].Palivizumab is an F-protein-targeting mAb: despite its proven efficacy, because of the direct and indirect costs due to the need for several monthly shots (up to 5 during the RSV season) [52,53], palivizumab has only been indicated for a small subset of infants [52,54], which, more precisely, include the following: (1) children who are born at 35 weeks of gestation or less and less than 6 months of age at the onset of the RSV season; (2) children who are less than 2 years of age and required treatment for bronchopulmonary dysplasia within the last 6 months; (3) children who are less than 2 years of age and have hemodynamically significant congenital heart disease [49,50,52,55].
This unsatisfactory landscape has recently been changed by the approval of nirsevimab (Beyfortus, Astrazeneca [Södertälje, Sweden] and Sanofi [Gentilly, France]) [18,56,57], a long-acting mAb that can be delivered as a single dose for the whole of RSV season.Nirsevimab was initially approved by the European Medicine Agency (EMA) in October 2022, before then being authorized by the UK (9 November 2022), Canada (19 April 2023), and, eventually, the USA Food and Drug Administration (FDA) (17 July 2023) [58,59].Even though the introduction of nirsevimab in the immunization calendars has been hindered by the limited availability of the mAb, as well as by its complicated placement in the vaccination schedules of newborns and infants [58,60], since the inception of the winter season 2023-2024, several real-world experiences have been reported from the Northern Hemisphere [61,62].This study was, therefore, designed to systematically evaluate the efficacy and provide early estimates on the effectiveness of nirservimab in infants and Vaccines 2024, 12, 640 3 of 30 children in terms of avoiding hospital admissions due to RSV-associated LRTIs, possibly providing guidance for medical and public health professionals.

Research Concept
The outline of the present systematic review with meta-analysis was designed in accordance with the "Preferred Reporting Items for Systematic Reviews and Meta-Analysis" (PRISMA) statement [63,64] (see the PRISMA checklist in Supplementary Table S1) and registered on the PROSPERO database (https://www.crd.york.ac.uk/prospero/; accessed on 11 May 2024) with the progressive registration number CRD42024540669 [64].
Research concepts have been defined through the "PECO" strategy (i.e., Patient/ Population/Problem; Exposure; Control/Comparator; Outcome) [65,66].As summarized in Table 1, the main research question of the present study concerned whether individuals aged less than 2 years old who had received at least one dose of nirsevimab (P), upon being exposed to RSV infection during the following RSV season (E), had different rates of occurrence of LRTIs associated with RSV infection (O) compared to children of the same age group that had not been immunized with either nirsevimab or palivizumab (placebo) (C).

Research Strategy
Three databases (PubMed; EMBASE; and Scopus) were specifically searched from inception until 1 May 2024.Moreover, the official websites of national health authorities of EU countries, the USA, and Canada were searched for any report consistent with the research concept.The search strategy was based on the blueprint originally recommended by Turalde-Mapili et al. [67] for their similar study performed in 2023, and it is summarized in Table A1.

Selection Criteria
In order to be included into the present systematic review with meta-analysis, the retrieved studies should provide data on the following: (1) Immunization with nirsevimab of children aged less than 2 years; (2) Nirsevimab mAb administration (any strategy and settings); (3) Comparison of nirsevimab efficacy with placebo in either a randomized controlled trial (RCT) or real-world settings; (4) Reporting on the occurrence of LRTIs in individuals treated with nirsevimab and placebo with subsequent hospitalization.
The following exclusion criteria were then applied: (1) Immunization of children aged 2 years or more; (2) Secondary studies (i.e., systematic reviews and meta-analyses, letters, editorial comment, case reports); (3) Studies on animals (including non-human primates) or preclinical testing; (4) Outcomes other than clinical efficacy; (5) The full text was not available either through online repositories or through interlibrary loan or its main text was written in a language different from English, Italian, German, French, Spanish, or Portuguese; (6) A lack of details about the geographical setting and corresponding timeframe; (7) Reporting on the occurrence of influenza-like illnesses and/or respiratory syndromes other than LRTIs; (8) Reporting on the occurrence of LRTIs that did not include the number of cases eventually admitted to the hospital settings because of respiratory syndrome; (9) Methods other than Real-Time Quantitative Polymerase Chain Reaction (RT-qPCR) or non-RT-qPCR Nucleic Acid Amplification Tests (NAATs) were applied for the laboratory diagnosis of RSV infection.
When the retrieved studies included cross-published and/or duplicated data, only the most recent publication was included.When feasible, duplicated data were removed from both qualitative and quantitative analyses.

Selection Criteria
Articles fulfilling the inclusion criteria and not fulfilling in the exclusion criteria were title-and abstract-screened to ascertain their relevance to the research question [64,68].Only articles that, even during abstract screening, were considered consistent with the research strategy were retained and independently full-text-screened by two investigators (SC, MB).Cases of disagreements were primarily discussed between investigators, and when consensus was not reached, the chief investigator (MR) was consulted as a third arbiter.

Data Extraction
Full-text-screened studies and their supplementary materials (where available) were assessed in order to retrieve the following data: (a) Bibliographic characteristics of the study: The first author's name, year of publication, and registration number for RCTs; (b) Characteristics of the clinical study: The sample size of the study groups (recipients of nirsevimab and placebo), baseline data (demographics and comorbidities), number of countries involved in the study, and timeframe of the study.
(c) Outcome data: Definition of primary and secondary outcomes and case definition.The main outcome encompassed the reported hospital admissions due to RSV-associated LRTIs.
For studies reporting on several centers and/or patient groups, data were reported as individual series where feasible.

Quality Assessment (Risk of Bias)
The risk of bias (ROB) (i.e., the likelihood that any feature from the design or conduct of a study will lead to misleading results) [69][70][71] was assessed by means of the ROB tool from the National Toxicology Program (NTP)'s Office of Health Assessment and Translation (OHAT) (now the Health Assessment and Translation (HAT) group) [71,72].OHAT ROB provides a 4-point scale rating (from "definitely low", "probably low", and "probably high" to "definitely high") on the following potential sources of bias: participant selection (D1), confounding factors (D2), attrition/exclusion (D3), detection (D4), and selective reporting (D5), as well as other sources of bias (D6).OHAT ROB was prioritized over other instruments for ROB assessment, as it does not provide an overall rating for each study, nor does it require that studies affected by a certain degree of ROB be removed from the pooled analyses [72].estimates for the efficacy of nirsevimab were calculated for each study.The efficacy of nirsevimab (immunization efficacy, IE) was calculated in analogy with VE as the percentage reduction in disease cases in a treatment group compared to the non-treated (i.e., placebo) group.More precisely, in the present study, the IE was calculated based on the efficacy in avoiding hospital admissions due to RSV-associated LRTI cases.Mathematically, IE can, therefore, be defined as follows:

Meta-Analysis
Pooled estimates of RSV prevalence and IE were calculated through a random-effect model (REM) meta-analysis of retrieved studies and reported as point estimates with their 95% confidence intervals (95% CI).In REM, the effect of the categorical variables is defined as varying across the levels of the variable.In other words, as random effects account for variability and differences between different entities or subjects within a larger group, the REM was considered more effective at dealing with the genuine differences underlying the results of the studies, or heterogeneity [73,74].Heterogeneity, in turn, can be defined as the inconsistency of the effect between the included studies, or the percentage of total variation across included studies likely due to actual differences rather than chance [69].Heterogeneity can be quantified in percent values by means of I 2 statistic.To fit the aims of the following studies, heterogeneity was considered low for I 2 values ranging from 0 to 25%, moderate for I 2 values ranging from 26% to 50%, and substantial for I 2 values ≥50%.In order to cope with the presumptively reduced number of studies included in the meta-analyses, 95% CIs of I 2 estimates were calculated and reported [69].

Sensitivity Analysis
In order to evaluate the effect of each study on the pooled estimates, a sensitivity analysis (i.e., a study of how the uncertainty in the output of a mathematical model or system can be apportioned to different sources of uncertainty in its inputs) was performed by excluding one study at a time.

Analysis of Publication Bias
Publication bias was initially assessed through the visual analysis of the asymmetry of funnel plots.Contours of statistical significance were added in order to aid the visual interpretation of the plots.The asymmetry of funnel plot outcomes with three or more included studies was then assessed by means of Egger's test [64,75].Small-study bias was eventually assessed by means of radial plots.A p value < 0.05 was considered statistically significant for both publication and small-study bias.
As summarized in Figure 1 and detailed in Table A1, a total of 401 entries were retrieved: 60 studies from PubMed (14.96%), 145 from Scopus (36.16%), and 196 from EMBASE (48.88%).A total of 168 entries were duplicated across the searched databases (41.90%), before being removed from the screening, which eventually included a total of 233 titles and abstracts.Of them, 121 were excluded (30.17%), while 112 entries were sought for retrieval (27.93%) and then full-text-assessed.Eventually, 102 studies were removed as being inconsistent with inclusion criteria (55, 13.72%) or being nonoriginal studies (46, 11.47%), leading to the inclusion of 10 reports (2.49%) [59,61,[78][79][80][81][82][83][84][85].Moreover, through the analysis of the official websites of the public health authorities, three further entries were identified as consistent with the inclusion criteria [86][87][88].The individual characteristics of the included studies are summarized in Table 2.In brief, the pooled sample included five reports from RCTs from the clinical research on nirsevimab [78][79][80][81][82], seven studies on real-world experiences [59,61,[83][84][85]87,88], and a further report from the local health authority of Galicia [86].The observation times of the included studies (i.e., the time from the delivery of nirsevimab to the end of the registration of incident LRTIs) ranged between 57 days and 805 days, with four studies including an observation time <150 days [83][84][85]88].However, only the report from Dagan et al. [82] included data on the follow-up season following the initial delivery of nirsevimab.Six out of seven of real-world studies [59,[83][84][85]87,88] and the official report from NIRSEGAL study [86] were conducted in Western Europe, with the earliest recruitment from 25 September 2023 [59].More precisely, 4 studies and 13 case series were from Spain [59,[84][85][86], while 1 study each from France [87], Luxembourg [83], Italy [88], and the USA [61].Sample size ranged from a minimum of 288 infants for a study that only included subjects hospitalized because of RSV-LRTIs in pediatric intensive care units (PICU) to 10,259 patients from the real-world report of Ares-Gómez et al. [59] and 15,676 patients from the real-world report of López-Lacort et al. [84].The reported hospital admission rate for LRTIs in infants treated with nirsevimab ranged from 0 [88] to 1.96% The individual characteristics of the included studies are summarized in Table 2.In brief, the pooled sample included five reports from RCTs from the clinical research on nirsevimab [78][79][80][81][82], seven studies on real-world experiences [59,61,[83][84][85]87,88], and a further report from the local health authority of Galicia [86].The observation times of the included studies (i.e., the time from the delivery of nirsevimab to the end of the registration of incident LRTIs) ranged between 57 days and 805 days, with four studies including an observation time < 150 days [83][84][85]88].However, only the report from Dagan et al. [82] included data on the follow-up season following the initial delivery of nirsevimab.Six out of seven of real-world studies [59,[83][84][85]87,88] and the official report from NIRSEGAL study [86] were conducted in Western Europe, with the earliest recruitment from 25 September 2023 [59].More precisely, 4 studies and 13 case series were from Spain [59,[84][85][86], while 1 study each from France [87], Luxembourg [83], Italy [88], and the USA [61].Sample size ranged from a minimum of 288 infants for a study that only included subjects hospitalized because of RSV-LRTIs in pediatric intensive care units (PICU) to 10,259 patients from the real-world report of Ares-Gómez et al. [59] and 15,676 patients from the real-world report of López-Lacort et al. [84].The reported hospital admission rate for LRTIs in infants treated with nirsevimab ranged from 0 [88] to 1.96% [83], while it ranged between 1.61% [79,80,82] and 63.59% [61], and even 87.40%[87], for infants not treated with nirservimab.Based on the reported data, corresponding point estimates on IE for nirsevimab ranged from 62.6% (−7.3 to 86.9), reported by the early stages of RCT NCT03979313 [79,80,82], to 98.4% (72.9 to 99.9), reported by the study of Consolati et al. [88].As stated in Table A2, among retrieved RCTs, the treatment of either cases or controls with palivizumab was documented only by Simoes et al. [79], as all controls (n.290) actually received a full conventional immunization course.On the other hand, all sampled children were not exposed to a maternal immunization strategy.When dealing with real-world studies, while the studies by Lopez-Lacort et al. [84], Ernst et al. [83], and Ezpeleta et al. [85] did not document the actual exposure of sampled infants to both palivizumab and maternal immunization, the studies by Consolati et al. [88], Paireau et al. [87], and Moline et al. [61] deliberately excluded from their reports all cases managed with strategies alternative to nirsevimab.Finally, the report from the NIRSEGAL study [59,86] did not include any case managed with palivizumab, and no information was provided on maternal immunization.

Risk of Bias Assessment
The quality of the included studies is summarized in Figure 2, while their individual assessment is provided as the Table A3.Nearly all studies were characterized by high quality and a low risk of bias.The notable exception was represented by the report by Paireau et al. [87], the only report still in pre-print status at the time of this review.In fact, the authors implemented a design quite inconsistent with those of other real-world studies, as they reported on hospitalizations due to RSV-associated LRTIs, not providing the total number of doses delivered in the general population or in a specific group of infants.Moreover, the study included cases from PICU, compared to the general population used in other reports.Another partial exception was represented by the reports on the NIRSEGAL study group sourced from local health authorities, as the total number of hospital admissions due to RSV-associated LRTIs in non-treated children was not provided [86].

Risk of Bias Assessment
The quality of the included studies is summarized in Figure 2, while their individual assessment is provided as the Table A3.Nearly all studies were characterized by high quality and a low risk of bias.The notable exception was represented by the report by Paireau et al. [87], the only report still in pre-print status at the time of this review.In fact, the authors implemented a design quite inconsistent with those of other real-world studies, as they reported on hospitalizations due to RSV-associated LRTIs, not providing the total number of doses delivered in the general population or in a specific group of infants.Moreover, the study included cases from PICU, compared to the general population used in other reports.Another partial exception was represented by the reports on the NIRSEGAL study group sourced from local health authorities, as the total number of hospital admissions due to RSV-associated LRTIs in non-treated children was not provided [86].

Removal of Studies
Because of its design, as it focused on incident cases of LRTIs due to RSV in PICU, the report from Paireau et al. [87] was considered inconsistent with other reports, both RCTs and observational studies, therefore being removed from quantitative analysis.Cases from duplicated series [59,79,80,82,86] were similarly removed.The final sample, therefore, included a total of 45,238 infants (Table 3).

Removal of Studies
Because of its design, as it focused on incident cases of LRTIs due to RSV in PICU, the report from Paireau et al. [87] was considered inconsistent with other reports, both RCTs and observational studies, therefore being removed from quantitative analysis.Cases from duplicated series [59,79,80,82,86] were similarly removed.The final sample, therefore, included a total of 45,238 infants (Table 3).

Descriptive Analysis
Overall, 43,294 (95.70%) infants were reported in the first season after the delivery of nirsevimab: of them, 33,884 had been immunized (78.35%), while 9365 (21.65%) were placebos (RCTs) or controls (real-world studies) not immunized with nirsevimab.Of them, 290 were otherwise immunized with palivizumab (3.10%).Data for the follow-up season only included a total of 1944 further cases (4.30%), with a total of 6 cases of hospitalizations due to RSV-associated LRTIs (hospital admission rates of 0.15% and 0.31% for subjects immunized and not immunized, respectively) [82].
In detail, the occurrence of hospital admissions due to RSV-associated LRTIs was equal to 0.42% among infants treated with nirsevimab, compared to 7.22% among nontreated subjects, with a substantially reduced risk ratio (RR) of 0.058, with a 95% CI 0.049 to 0.070.Interestingly, when the sampled infants were assessed based on the design of the studies, the rate of hospitalization due to RSV-associated LRTIs was 2.26% in realworld studies compared to 1.08% in RCTs (RR 2.099, 95% CI 1.756 to 2.510).Similarly, an increased occurrence of hospitalization due to RSV-associated LRTIs was identified for infants sampled from studies with an observation period of 150 days or more (RR 2.170, 95% CI 1.860 to 2.532).When dealing with the timeframe of the study, no substantial differences were identified for having been a study completed before (hospitalization rate 1.53%) or after the inception of COVID-19 pandemic (1.93%, RR 1.263, 95% CI 0.969 to 1.646).Eventually, using the rate of hospitalizations due to RSV-associated LRTIs in RCTs as the reference group, studies from Spain exhibited a substantially reduced risk (RR 0.603, 95% CI 0.484 to 0.751), while those from other EU countries (RR 3.876, 95% CI 2.980 to 5.043) and, in particular, the single USA study included in the analysis (RR 54.887, 95% CI 46.132 to 65.303) were associated with an increased risk of hospitalization.
A corresponding forest plot for hospitalizations due to RSV-associated LRTIs is provided as Annex Figure A1.A total of 9365 children not immunized with nirsevimab were pooled as a sample, with a cumulative prevalence of 4.25% (95% CI 2.42 to 7.38) calculated by means of REM, including a prevalence of 2.45% (95% CI 1.61 to 10.34) for RCTs compared to 4.25% (95% CI 2.42 to 7.38) for observational studies from real-world settings, with no substantial differences between groups (chi-squared test 2.98, p = 0.084).Also, estimates for non-treated infants were affected by substantial heterogeneity (I 2 98.8%, 95% CI 98.5 to 99.0; tau 2 = 1.481,Q = 2.98, p < 0.001), even when considered by subgroups (I 2 84% for real-world observational studies vs. 92% for RCTs).
In order to assess whether observation time had any effect on the pooled estimates, a metaregression was implemented, with the subsequent calculation of the corresponding bubble plot.In the bubble plot, the observed outcomes (in this case, immunization efficacy) of the individual studies are plotted against a quantitative predictor (in this case, the observation time) and the size of the "bubble" is proportional to the weight that the studies received in the analysis (i.e., studies receiving more weight have larger points) (Figure 6).Eventually, observation time was characterized as a non-significant explanatory variable for IE (β = 0.001, 95% CI −0.001 to 0.002; p = 0.092), and the analyses were affected by low heterogeneity (I 2 = 19.0%,tau 2 = 0.055).A corresponding forest plot for hospitalizations due to RSV-associated LRTIs is provided as Annex Figure A1.A total of 9365 children not immunized with nirsevimab were pooled as a sample, with a cumulative prevalence of 4.25% (95% CI 2.42 to 7.38) calculated by means of REM, including a prevalence of 2.45% (95% CI 1.61 to 10.34) for RCTs compared to 4.25% (95% CI 2.42 to 7.38) for observational studies from real-world settings, with no substantial differences between groups (chi-squared test 2.98, p = 0.084).Also, estimates for non-treated infants were affected by substantial heterogeneity (I 2 98.8%, 95% CI 98.5 to 99.0; tau 2 = 1.481,Q = 2.98, p < 0.001), even when considered by  Pooled estimates were affected by moderate heterogeneity (I 2 24.3%, 95% CI 0.0 to 56.6; tau 2 = 0.105, Q = 23.79,p = 0.162), even though the individual subgroups were otherwise exempt (I 2 0% for RCTs, I 2 3% for real-world observational studies).
In order to assess whether observation time had any effect on the pooled estimates, a metaregression was implemented, with the subsequent calculation of the corresponding bubble plot.In the bubble plot, the observed outcomes (in this case, immunization efficacy) of the individual studies are plotted against a quantitative predictor (in this case, the observation time) and the size of the "bubble" is proportional to the weight that the studies received in the analysis (i.e., studies receiving more weight have larger points) (Figure 6).Eventually, observation time was characterized as a non-significant explanatory variable for IE (β = 0.001, 95% CI −0.001 to 0.002; p = 0.092), and the analyses were affected by low heterogeneity (I 2 = 19.0%,tau 2 = 0.055).

Sensitivity Analysis
Sensitivity analysis was performed through the removal from pooled estimates of hospital admission rates and the IE of a single study at a time.The resulting pooled estimates are reported in Figures A2-A5.Regarding hospital admission rates (Figures A3   Figure 6.Bubble plot for the metaregression of the included studies by observation days.In the regression model, the observation time was characterized as a non-significant explanatory variable for IE (β = 0.001, 95% CI −0.001 to 0.002; p = 0.092).Pooled analyses were affected by low heterogeneity (I 2 = 19.0%,tau 2 = 0.055) [59,61,78,79,[82][83][84][85]88].

Sensitivity Analysis
Sensitivity analysis was performed through the removal from pooled estimates of hospital admission rates and the IE of a single study at a time.The resulting pooled estimates are reported in Figures A2-A5.Regarding hospital admission rates (Figures A3 and A4), the removal of single studies did not affect either estimates or heterogeneity.Contrarily, the removal of the series from the Hospital Clinico Universitario Virgen de la Arrixaca from the study of López-Lacort et al. [84] (4631 patients, 33 cases of hospital admissions) led to a substantial reduction in the point value of I 2 , seemingly estimated as 0%.

Publication Bias
Publication bias was preliminarily assessed via the calculation and visual inspection of funnel plots.According to the PRISMA guidelines [64,68,77], the sample size was plotted against the effect size otherwise reported in the plot.As the size of the sample increases, individual estimates of the effect are likely to converge around the true underlying estimate [63,68,77].Funnel plots for hospital admission rates are reported in Figure A5, while the funnel plot for IE estimates is provided in Figure 7a.Estimates were scattered across the plots and highly symmetrical, ruling out the presence of publication bias due to the high share of lower prevision studies.In fact, Egger's test for funnel plot asymmetry was similarly not significant for estimates of admission rates for subjects treated with nirsevimab (t = −1.06,p value = 0.305, tau 2 = 7.218) and IE (t = −0.54,p value = 0.597, tau 2 = 1.376), while it confirmed the publication bias for prevalence studies on the admission rates for non-immunized infants (t = −2.87,p-value = 0.011, tau 2 = 58.632).Even in radial plots for the IE and hospital admissions (Figure 7b, Figure A7a), individual estimates were substantially scattered across the regression line when dealing with rates for subjects treated with nirsevimab, ruling out the small-study effect that was otherwise noticeable with estimates for prevalence in non-immunized infants (Figure A6b).

Generalizability and Implications for Daily Practice
Efficacy (i.e., the degree to which an intervention prevents a certain outcome) is calculated in RCTs under ideal and controlled circumstances, and even though a certain intervention with a documented high efficacy (e.g., COVID-19 mRNA vaccines against original Wuhan strains of SARS-CoV-2) would be expected to be also high performing (i.e., effective) in real-world settings, factors such as how the intervention is delivered underlying and heterogenous clinical conditions and the actual occurrence of the targeted condition (in this case, the RSV epidemiology) can reduce how effective the intervention is at preventing the disease [89].According to our data, the actual efficacy of nirsevimab in avoiding hospitalizations due to RSV-associated LRTIs was comparable to its real-world effectiveness, stressing how nirsevimab's potential impact on the management of RSV disease burden should be taken into consideration not only from a clinical point of view but also from a public health perspective [59,88].While palivizumab is highly effective in avoiding severe respiratory syndromes associated with RSV infections in highly selected patient groups [50,52,55,90], nirsevimab potentially provides a preventive option to be delivered to the whole of the infant population, with a likely impact on the RSV-associated disease burden [57,62,91].An often-overlocked and even forgotten disease [92,93], particularly among adults and the elderly [45,46,94], RSV has recently emerged from the pandemic as a generally acknowledged cause of medical consultations and hospitalization [25,93].The increasing notoriousness of RSV in the general population likely has several causes.On one hand, during the first months of COVID-19 pandemic, the prevention of RSV infections benefited from non-pharmaceutical interventions for respiratory pathogens primarily implemented against SARS-CoV-2 [95][96][97], but their removal led to an unprecedented increase in incidence rates reported at the global level in the following years [98][99][100][101].On the other hand, several preventative options have ultimately been made available to medical professionals and potential end users, including parents and caregivers, such as nirsevimab and two effective subunit vaccines for older adults: Abrysvo from Pfizer Inc. (Pfizer Europe MA EEIG, Brussels, Belgium) and Arexvy from GlaxoSmithKline LLC (GlaxoSmithKline Biologicals SA, Rixensart, Belgium) [48,102,103].Unfortunately, both vaccines cannot be used in infants and children [102,104,105], and even though both vaccines have been designed to be delivered in pregnant women, only Abrysvo has been approved by the Food and Drug Administration (FDA) and the European Medicine Agency (EMA) for maternal use (32 to 36 weeks gestational age of pregnancy for the FDA, 24 to 36 weeks for the EMA) [106].Even though a recent meta-analysis has stressed the potential efficacy of maternal immunization for reducing hospitalizations in offspring (RR 0.50, 95% CI 0.31 to 0.82) [107], safety issues have been raised for Arexvy, particularly regarding preterm births (relative risk 1.37, 95% CI 1.08 to 1.74, p = 0.001), forcing GSK to suspend the recruitment of new cases during a phase 3 study (RCT NCT04605159) [108].In such a setting, nirsevimab could represent a likely option for guaranteeing effective protection against LRTIs in all children, not only for children currently benefiting from the preventive treatment with palivizumab.
Still, several caveats should be mentioned.According to its design and preliminary studies, nirsevimab should guarantee around 150 days of protection against RSV [79,80,109], being particularly suitable for managing a full RSV season with a single shot compared to the monthly treatment required by palivizumab [78].Unfortunately, several real-world studies were completed well before the 150th days of the observation period [83][84][85]87,88], and not coincidentally, studies with an observation period longer than 150 days (i.e., the coverage guaranteed by a single dose of nirsevimab) were characterized by higher rate of hospitalizations due to RSV-associated LRTIs (RR 2.170, 95% CI 1.860 to 2.532) compared to the former ones.Similarly, the individual estimates for IE were negatively correlated with the length of the observation period (Spearman's r = −0.546,p = 0.016), i.e., the longer the observation time, the lower the actual estimate for the IE.Still, it should be stressed that meta-regression analysis did not confirm the actual role of observation time.
Another criticism of the generalizability of collected data is represented by the immunization strategy implemented in most real-world studies.For logical, ethical reasons, in longitudinal studies, all newborns from a certain geographical area were the potential recipients of nirsevimab [59,61,[83][84][85][86]88].As a consequence, non-treated individuals were children whose parents were either unwilling or unable to participate in this preventive intervention.In strict analogy with vaccine hesitancy, we cannot rule out that non-treated children belonged to families characterized by lower health literacy, lower socio-economic status, and some degree of social marginalization [110][111][112], which, in turn, may increase the likelihood of respiratory infections [113,114].On the other hand, because of the current epidemiology of RSV, particularly in European countries, and the different immunization strategies implemented by local health authorities, the comparisons with previous RSV seasons would be of limited significance.While the delivery of nirsevimab among newborns and infants led to substantial reductions in the total number of hospitalizations (Table A4) [83,86,88], ranging from −30.08% [83] to −77.05% [88], recently reported Spanish estimates on hospitalizations due to RSV-associated LRTIs hint at a very high rate for the RSV season in 2022 compared not only to the pandemic seasons 2021 and 2020 (606 vs. 395 and 247, respectively) but also to the pre-pandemic years 2016 (n = 551), 2017 (n = 451), and 2018 (n = 366), only being comparable to the peak year 2019 (n = 642) [115].
Finally, the IE was calculated only in longitudinal studies: even though the report from Moline et al. [61] was primarily meant to be a case-control study, we were able to pool reported data by means of its mixed design.On the contrary, the case-control report of Paireau et al. [87] was removed from pooled estimates.For both studies, an alternative and appropriate calculation of actual IE could be based on the odds ratio rather than relative risk.However, as reported in Table A5, both studies would be still associated with a high IE, i.e., 74.6% (50.7 to 86.9) for Paireau et al. [87] and 81.0%(55.1 to 92.0) for Moline et al. [61].

Limitations and Implications for Future Studies
Despite its potential significance in providing an early estimate of IE of nirsevimab against hospital admissions due to RSV-associated LRTIs, we must acknowledge several limitations of the present study.
As a preliminary step, the strict search strategy on this relatively new topic (i.e., a new mAb such as nirsevimab, commercialized in EU only since the winter season 2023) has likely guaranteed the good or very good quality of retrieved studies, but it similarly resulted in a small number of collected reports.This very same shortcoming has been identified in some similarly designed systematic review with a meta-analysis on the prevention of RSV infections [48,107,116] and will likely affect all secondary studies on RSV prevention in the upcoming years until the more extensive use of new vaccines and mAb occurs in real-world settings.
Second, despite the high or even very high quality of the retrieved studies, diverse immunization and reporting strategies, particularly from real-world studies, have reasonably introduced some degree of heterogeneity into the results, as suggested by pooled estimates on hospital admission rates, potentially affecting the collected results.For instance, while RCTs likely did not include any patient exposed to a maternal immunization strategy and deliberately excluded cases managed with palivizumab, with the notable exception of the MEDLEY subgroup of the study by Simoes et al. [79], reporting strategies from real-world studies are a lot more heterogeneous.On one hand, since the beginning of nirsevimab immunization, the delivery of palivizumab has been discontinued in the Spanish region of Galicia [59], but no information is provided from other Spanish regions.On the other hand, the studies of Consolati et al. [88], Moline et al. [61], and Paireau et al. [87] deliberately removed from their estimates infants immunized with palivizumab.Similarly, the exposure of sampled children to maternal immunization is ruled out for the aforementioned studies, while no information is provided by all of Spanish studies [59,84,85].Moreover, because of its design, we were unable to include in pooled estimates the study of Paireau et al. [87].From both a clinical and public health point of view, it was particularly frustrating as this study documented the efficacy of nirservimab in a specific and quite significant subset of patients, i.e., admission to PICU [87,117,118], the prevention of which would lead to a substantial reduction in the eventual RSV-associated disease burden in terms of both morbidity and direct and indirect costs.However, the point estimate of presumptive IE was not only in line with the general estimate but also highly comparable with that from the study of Moline et al. [61] when IE was estimated, considering the case-control design of the aforementioned studies.
Third, it is very important to stress how RCTs included in this systematic review with meta-analyses were substantially influenced by the COVID-19 pandemic in both a direct and an indirect way.According to the preliminary design of the RCT NCT03979313 (i.e., MELODY study) [80], a full trial size of 3000 participants was envisaged, but only 1500 patients were actually recruited before the World Health Organization declared the COVID-19 pandemic on 11 March 2020.As previously stressed, the implementation of lockdown measures and non-pharmaceutical interventions radically affected the global epidemiology of RSV [97,98,101,119,120], potentially impairing the eventual reliability of subsequent analyses [80,82,121,122].As a consequence, the report from Hammitt et al. [80] only included a primary analysis, and the report on the full trial size was delayed in post-pandemic settings [82].The potential consistency of pre-and post-pandemic data could, therefore, be questioned.According to our estimates, while the overall occurrence of hospital admissions due to LRTIs was not significantly higher in post-pandemic settings compared to pre-pandemic studies (RR 1.263, 95% CI 0.969 to 1.646), the risk of hospital admission due to breakthrough infections in children treated with nirservimab from the RCT NCT03979313 was greater in pre-pandemic settings compared to post-pandemic settings (0.60%, 95% CI 0.22 to 1.31 vs. 0.30%, 95% CI 0.06 to 0.86; RR 2.042, 95% CI 0.512 to 8.143).Moreover, the collected evidence included the report from Simoes et al. [79], which, in turn, provided data on the efficacy of nirsevimab from a phase 2b study (i.e., MEDLEY study) [109] that was originally designed for evaluating the safety of this mAb in a subset of infants with a high rate of premature heart or lung disease.Because of the original aim of the MEDLEY study, controls, therefore, received a full prophylaxis with palivizumab rather than placebo.Despite the rigorous design of the original RCT, its eventual consistency with other reports mostly focusing on healthy children could be then questioned.
Fourth, it should be stressed that because of the specific characteristics of RSV infection, a large share of diagnoses may be missed, even among hospitalized patients [48,123].On one hand, the collection of respiratory tract specimens could fail to provide an appropriate diagnosis in accordance with the stage of the assessed infection [124,125].Severe cases associated with LRTIs may result in false-negative results when the specimens are collected from the upper airways, as the viral infection is more likely to be active in the lower regions of the respiratory tract [113].In fact, when provided by the collected study, the proportion of RSV-associated LRTIs among the total number of hospitalized LRTIs (i.e., including both RSV-associated and non-RSV associated cases) ranged between 28.28% (95% CI 19.69 to 38.22) [78] and 100% [85] (Figure A7), suggesting that the actual proportion of RSV-associated cases may have somehow been underestimated, conversely inflating the pooled estimates for the IE.The choice of limiting our analysis on the outcome of hospital admissions due to RSV-associated LRTIs likely limited the potential heterogeneity due to the variable case definition of LRTIs and diagnostic options from RCTs and real-world settings, being otherwise recommended for the appropriate appraisal of RSV vaccines [48,105] Nonetheless, hospital admission rates for RSV-associated respiratory syndromes have been extensively documented as being highly heterogeneous, mostly due to national policies and the availability of healthcare options [19,[126][127][128].As the largest proportion of included samples from real-world settings were provided by studies from the Spanish settings [115], a cautious approach when generalizing our estimates, not only at global level but also at the regional (i.e., European) level, is, therefore, required.
Finally, when addressing the potential use of nirsevimab in real-world settings, the potential emergence of resistance should be considered, particularly in the light of the development of future vaccines.In fact, while the binding site of nirsevimab (site ø) is highly conserved due to its functional requirements [129], there is considerable evidence that even major variant substitutions within the nirsevimab binding site do not reduce the potency of this mAb [122,129].However, as preF and, in particular, its site ø are a shared target for vaccines such as RSVpreF and RSVPreF3, as well as for mRNA-1345 [130], the extensive reliance on nirsevimab for protecting infants and newborns should be associated with the accurate and continuous monitoring of new variants potentially emerging due to its selective pressure.

Conclusions
The RSV immunization of infants and children with nirsevimab has been proven to be effective at preventing hospitalizations due to LRTIs.However, due to the limited availability of complete and fully comparable follow-up for all retrieved RCT studies, the limits to the generalizability of data from real-world experience, and the potential underestimation of incident cases of LRTIs associated with RSV, further and improved reporting from ongoing real-world experiences is still recommended in order to properly address the cost-effectiveness profile of this innovative preventative option.

Figure 2 .
Figure 2. Summary of the risk of bias (ROB) estimates for observational studies [72,85].Analyses were performed according to the National Toxicology Program (NTP)'s Office of Health Assessment and Translation (OHAT) handbook and respective ROB tools including all retrieved studies (n = 13).

Figure 2 .
Figure 2. Summary of the risk of bias (ROB) estimates for observational studies [72,85].Analyses were performed according to the National Toxicology Program (NTP)'s Office of Health Assessment and Translation (OHAT) handbook and respective ROB tools including all retrieved studies (n = 13).

Figure A2 .
Figure A2.Forest plot for sensitivity analysis on hospitalization rates for respiratory syncytial virus (RSV)-associated lower respiratory tract infections in infants previously treated with nirsevimab.Sensitivity analysis was performed by excluding one study at a time and b reporting corresponding pooled estimates with heterogeneity estimates (I 2 ) [59,61,78,79,82-85,88].

Figure A2 . 31 Figure A2 .
Figure A2.Forest plot for sensitivity analysis on hospitalization rates for respiratory syncytial virus (RSV)-associated lower respiratory tract infections in infants previously treated with nirsevimab.Sensitivity analysis was performed by excluding one study at a time and b reporting corresponding pooled estimates with heterogeneity estimates (I 2 ) [59,61,78,79,82-85,88].

Figure A4 .Figure A5 .
Figure A4.Forest plot for sensitivity analysis of the immunization efficacy of nirsevimab regarding the prevention of hospital admissions due to respiratory syncytial virus (RSV)-associated lower respiratory tract infections.Sensitivity analysis was performed by excluding one study at a time and reporting corresponding pooled estimates with heterogeneity estimates (I 2 ) [59,61,78,79,82-85,88].

Figure A4 . 31 Figure A4 .Figure A5 .
Figure A4.Forest plot for sensitivity analysis of the immunization efficacy of nirsevimab regarding the prevention of hospital admissions due to respiratory syncytial virus (RSV)-associated lower respiratory tract infections.Sensitivity analysis was performed by excluding one study at a time and reporting corresponding pooled estimates with heterogeneity estimates (I 2 ) [59,61,78,79,82-85,88].

Figure A6 .
Figure A6.Radial plots for prevalence data on the occurrence of hospital admissions due to respiratory syncytial virus (RSV)-associated lower respiratory tract infections with their corresponding results for the regression test of funnel plot asymmetry (Egger's test).Subfigure (a) individuals treated with nirsevimab; subfigure (b) individuals not treated with nirsevimab [59,61,78,79,82-85,88].
Definition Population of Interest Children who had received at least one dose of nirsevimab for the prevention of respiratory syncytial virus infection Exposure exposed to RSV infection during the subsequent RSV season Control/Comparator Children who had not received nirsevimab or palivizumab for the prevention of respiratory syncytial virus infection (placebo) Outcome Occurrence of lower respiratory tract infection 2.7.1.Immunization EfficacyAs a preliminary step, the point prevalence of RSV infections among immunized and non-immunized individuals was calculated in percent values.Similarly, individual

Table A2 .
The delivery of palivizumab and maternal vaccination strategy in sampled studies.

Table A3 .
[89,119] reporting of the risk of bias (ROB) estimates for observational studies[89,119].Analyses were performed according to the National Toxicology Program (NTP)'s Office of Health Assessment and Translation (OHAT) handbook and the respective risk of bias (ROB) tool.

Table A4 .
Hospitalizations due to RSV-associated lower respiratory tract infections in 2023-2024 compared to previous RSV season(s).

Table A5 .
The estimate of immunization effectiveness for case-control studies via the calculation of the IE through odds ratio rather than relative risk.
the IE through odds ratio rather than relative risk.