WEEZTHMA

Objectif(s) de la recherche et intérêt pour la santé publique

Finalité de l'étude

Objectifs poursuivis

Domaines médicaux investigués

Bénéfices attendus

This is a population-based prospective cohort design study using data from the SNDS (SNIIRAM+ PMSI+ Causes médicales de décès) collected between September 2023 and January 2030 in the pediatric population.
Study outcomes of hospitalization for wheezing episodes among infants immunized with nirsevimab (exposed group) vs non-immunized (non-exposed group) during the second year of life with other outcomes assessed from birth through 5 years of age, as the infants reach this age and their data become available. This study seeks to assess the effectiveness of passive immunization with nirsevimab during the first six months of life on hospitalization for wheezing episodes during the second year of life. Immunization was determined by nirsevimab uptake between September 15, 2023 and January 31, 2024 in maternity units. A single dose of 50 mg, for infants weighing < 5 kgs, or 100 mg, for infants weighing ≥ 5 kgs, administered intramuscularly was recommended. Information on nirsevimab administration in maternity units was searched in the PMSI database using the UCD codes. Information on nirsevimab dispensing was searched in the DCIR database using ATC/CIP13 codes.

Primary Objective: assess the effectiveness of passive immunization with nirsevimab during the first six months of life on hospitalizations for wheezing episodes during the second year of life.
Secondary Objectives:
For infants who had received passive immunization with nirsevimab during the first six months of life, we will assess
1. The effectiveness of passive immunization with nirsevimab on recurrent wheezing during the second year of life
2. The effectiveness of passive immunization with nirsevimab on the use of short acting beta-agonists bronchodilators during the second year of life
3. The effectiveness of passive immunization with nirsevimab on the use oral corticosteroids for respiratory events during the second year of life
4. The effectiveness of passive immunization with nirsevimab on the use of maintenance asthma therapy during the second year of life

For infants who had received passive immunization with nirsevimab during the first year of life, we will assess
5. The effectiveness of passive immunization with nirsevimab on hospitalization for wheezing episodes between 2 and 6 years old.
6. The effectiveness of passive immunization with nirsevimab on the occurrence of recurrent wheezing between 2 and 6 years old.
7. The effectiveness of passive immunization with nirsevimab on the use of short acting beta-agonists bronchodilators between 2 and 6 years old.
8. The effectiveness of passive immunization with nirsevimab on the use oral corticosteroids for respiratory events between 2 and 6 years old.
9. The effectiveness of passive immunization with nirsevimab on the use of maintenance asthma therapy between 2 and 6 years old.
10. The effectiveness of passive immunization with nirsevimab on the occurrence of emergency wards visits for wheezing between 2 and 6 years old

Using all participants’ socio-demographic characteristics and comorbidities listed above, a propensity score using the inverse probability of treatment weighting (IPTW) method will be used to reduce potential indication bias from nirsevimab receipt. First, the probability of receiving nirsevimab will be estimated for each child using a logistic regression model based on all baseline covariates. These probabilities will be used as propensity scores.
Then, to account for an initial covariate imbalance between exposed and non-exposed groups, a pseudo-randomized population will be created by attributing a stabilized weight to each patient derived from the inverse probability of receiving treatment (here nirsevimab) using the previously estimated propensity scores.
Covariates will be considered to be well balanced if the standardized mean difference is < 0·1. To illustrate the distribution of weights after IPTW in each group and to reflect the balance achieved through weighting, the Effective Sample Size (ESS) will be used.
Finally, we will compute the average treatment effect by use of a proportional hazards cox regression model on the pseudorandomized population, with death as a competing risk. Hazard ratios (HRs) and 95% CIs with robust SEs will be reported.
Several sensitivity analyses will be performed, including an analysis based on a nearest-neighbor propensity score matching as an alternative to the IPTW approach, within a caliper of 0·2 of the standard deviation of the logit of the propensity score. Unlike IPTW, this analysis restricts the population to members with similar propensity scores, thus emulating the populations of a randomized clinical trial.