Real-World Effectiveness of Nirsevimab Against Respiratory Syncytial Virus: A Test-Negative Case-Control Study

Study Design and Study Population

The effectiveness of nirsevimab against medically attended RSV infection was estimated using the test-negative case-control study design. The study population included all patients who were born after October 1st, 2022, were tested for RSV due to a suspected acute respiratory infection (ARI)[10], and received care in a facility affiliated with the Yale New Haven Health System (YNHHS) between October 1^st 2023 and May 9^th 2024. The YNHHS is the largest health system in Connecticut, and consists of five integrated hospital networks, 30 emergency or urgent care centers, and over 130 outpatient clinics from Westchester County, New York, to Rhode Island, all integrated using a single electronic health record (EHR) system.

Patients were excluded from these analyses if they were not age-eligible for nirsevimab when it became available on October 1^st 2023[11], or if they resided outside Connecticut, New York, or Rhode Island. The geographic restriction was implemented to ensure that immunization records could be verified through state immunization registries, which are directly integrated with the YNHHS EHR. Infants were considered eligible for nirsevimab if they were born during the season (after October 1st, 2023), or if they were under 8 months old and entering their first RSV season, or if they were between 8 and 12 months old with a risk factor for severe RSV (see Figure 1 for the detailed inclusion process and Table S1 for the definitions for risk factors).

• Download figure
• Open in new tab

Figure 1. Selection of RSV test records

Data Sources and Study Definitions

For all the infants who met the eligibility criteria, reviews of medical records and state immunization registry searches were conducted to capture information on patient characteristics, immunization history, and potential confounders. Relevant clinical and laboratory data associated with each patient’s RSV test, such as chief complaints, problem lists, encounter diagnoses, and presence of any other acute or chronic diseases were abstracted by trained investigators from the EHR. The clinical outcomes following hospitalization were also recorded, such as hospital and ICU length of stay, and maximum respiratory support needed during hospitalization. Patient characteristics including age, race and ethnicity, gestational age, birth weight, and type of insurance were also abstracted.

Cases were defined as infants with a medically attended RSV infection confirmed by nasopharyngeal polymerase chain reaction (PCR). Controls were infants with ARI who tested negative for RSV. For a given patient, if there was more than one positive test during the study period, only the first was included in the study. If an eligible control had more than one negative test within 14 days, the first negative test was selected for this period. The primary exposure of interest in our study was the nirsevimab immunization status. Only documented immunization dates were included in the analysis. Infants were classified as “immunized” if they received a dose of nirsevimab prior to their RSV test.

Statistical analysis

The characteristics of the study population were summarized using frequency distributions and measures of central tendency. Univariate analyses were performed to compare RSV-positive patients with negative controls, and unimmunized with immunized infants. Covariate balance between groups was assessed to detect potential confounders using standardized mean differences (SMD), with absolute SMDs of less than 0.20 indicating adequate balance.

For our primary analysis, the effectiveness of nirsevimab against medically attended RSV infection was quantified using all eligible patients in our study population. Effectiveness was calculated as one minus the odds ratio (OR) of immunization with nirsevimab among cases and controls, using logistic regression. Adjusted effectiveness estimates were derived using multivariable models that controlled for age, calendar month, and the presence of at least one risk factor. Non-collinear potential confounders were selected for the final models through backward selection based on the Akaike information criterion (AIC). Model formulas and AIC values are provided in the supplementary materials (Figure S1, Table S2).

For our secondary aims, separate models were fitted to analyze the data based on clinical setting (inpatient vs outpatient), disease severity, nirsevimab dosage used, and time from immunization. For the severity analysis, patients were considered to have severe disease if they were hospitalized within 14 days of RSV testing and required either transfer to a pediatric ICU or high levels of respiratory support during hospitalization, such as high-flow nasal cannula (≥2 liters per minute), continuous or bilevel positive airway pressure, or invasive mechanical ventilation. The extent to which the effectiveness of nirsevimab decreased over time was quantified using logistic regression within a Bayesian framework. The parameter representing the effectiveness of immunization was time-varying at bi-weekly intervals of time since immunization. These models used weakly-informative prior distributions and imposed a monotonic structure on the regression coefficients that represent nirsevimab’s effectiveness for increasing time since immunization. Medians and 95% credible intervals were calculated from the collected posterior samples, and convergence was evaluated using trace plots (Figure S6). A comprehensive description of the Bayesian models is provided in the supplementary material.

The effectiveness of nirsevimab was also assessed using broader endpoints, including all-cause lower respiratory tract infection (LRTI) and all-cause LRTI hospitalization across the entire respiratory season, with additional stratification by early (October to January) and late (February to May) periods.

Sensitivity analyses

Several sensitivity analyses were conducted to assess the robustness of our findings. First, we assessed for differences in effectiveness when employing different exposure and outcome definitions. Specifically, we explored restricting our analysis to only medical visits where encounter diagnoses indicating lower respiratory tract infection (LRTI) were recorded as either a primary or secondary diagnosis (Table S1). We also explored restricting controls to only those who tested positive for other respiratory viruses (i.e., influenza, adenovirus, rhinovirus, parainfluenza). In terms of exposure, we explored defining patients as “immunized” if they received nirsevimab ≥7 days prior to RSV testing, as was done in earlier reports to account for the mean RSV incubation period and the time required to reach peak antibody concentration[9]. Second, we assessed whether excluding infants whose mothers received the maternal RSV vaccine or those who were born during the previous RSV season would significantly alter the results. Last, we repeated our analysis using the hepatitis B vaccine as a “sham” exposure, as previously described [12, 13]. Since the hepatitis B vaccine is recommended to be given to all newborns but does not affect the risk of RSV infection, we expect that in the absence of bias, the proportions of cases and of controls who were immunized with the hepatitis B vaccine will not be significantly different.

Further details on study definitions and statistical analysis are provided in the Supplemental Methods. All analyses were conducted in R, version 4.3.1[14]. The code for the analysis can be found at https://github.com/Hanmeng-Xu/RSV_mAb_VE.

The institutional review board (IRB) at the Yale University School of Medicine approved the study (HIC:2000036550). The funders of this study had no role in study design; collection, analysis, and interpretation of data; writing the report; and the decision to submit the report for publication.

Study population

Between October 1, 2023, and May 9, 2024, a total of 3,992 RSV tests were performed within the YNHH system on infants born after October 1, 2023. Of these, 3,090 met our eligibility criteria and were included in the analysis (Figure 1). The analytic sample consisted of 680 RSV-positive and 2410 RSV-negative patients. Most RSV tests occurred during emergency department or urgent care clinic visits (n= 2,505, 81.1%) in the months of December and January (Figure S2). The median age at the time of RSV testing was 6.7 months (IQR 3.6 to 9.7 months). RSV-positive cases were slightly younger than RSV-negative controls (6.1 vs. 6.9 months, p = 0.007) and had a lower proportion of prematurity (11.2% vs. 14.2%, p = 0.045). However, other demographic and clinical factors, including sex, race/ethnicity, insurance type, and prevalence of comorbidities, were comparable between the two groups (Table 1).

The overall uptake of nirsevimab in the study sample was 10.7% (330/3,090). Among those who received nirsevimab before RSV testing, 71.2% (n = 235) received the 50 mg dose, while 28.8% (n = 95) received the 100 mg dose. Correlates of immunization are detailed in Table S3.

Overall, 24.4% (166/680) of RSV-positive cases resulted in hospitalization. Among those hospitalized, 58.4% (97/166) required more than two liters of respiratory support, and 13.8% (23/166) required admission to the ICU (Table S4).

Effectiveness of nirsevimab

The adjusted effectiveness of nirsevimab against any medically attended RSV infection was 68.4% (95% CI: 50.3-80.8%). Effectiveness was 61.6% (95% CI: 35.6-78.6%) for preventing RSV-associated outpatient visits, 80.5% (95% CI: 52.0-93.5%) for preventing hospital admissions, and 84.6% (95% CI: 58.7-95.6%) for preventing severe RSV (Figure 2). These findings aligned with existing clinical trials and observational data (Figure S3). The effectiveness estimates for the 100 mg and 50 mg doses were comparable, with overlapping confidence intervals (Figure S4).

• Download figure
• Open in new tab

Figure 2. Effectiveness of nirsevimab against medically attended RSV, by clinical setting and severity.

Adjusted models controlled for age, calendar month, and the presence of at least one risk factor (see Table S2).

Nirsevimab effectiveness against medically attended RSV infection decreased from 79% at 2 weeks post-immunization to 55% at 14 weeks post-immunization. A similar pattern of waning effectiveness was observed across different outcomes (Figure 3); our results were comparable to those derived from clinical trial data (Figure S5).

• Download figure
• Open in new tab

Figure 3. Effectiveness of nirsevimab by time since immunization.

The dots represent the median estimates of effectiveness of nirsevimab in preventing the various clinical outcomes, and the error bars indicate the 95% credible intervals of the estimates.

Protective effectiveness was observed against all-cause LRTI and all-cause LRTI hospitalization during the peak months of the RSV season (49.4% [95% CI: 10.7-72.9%] and 79.1% [95% CI: 27.6-94.9%] during November and December, respectively), comparable to results from other studies (Figure 4). Given that RSV was the predominant virus during these months (RSV positivity rate 39.3%), the above estimate largely reflects the effectiveness of nirsevimab against RSV. In contrast, between February and May, when the RSV positivity dropped below 3.9%, there was no effectiveness of nirsevimab against all-cause LRTI and all-cause LRTI hospitalizations (Figure 4).

• Download figure
• Open in new tab

Figure 4. Forest plot of nirsevimab effectiveness against (A) all-cause LRTI and (B) all-cause LRTI hospitalization, stratified by time.

Current study estimates are shown in blue. For comparison, estimates from three previous studies are also included. Pre-licensure clinical trial estimates are shown in gold, and other post-licensure study estimates are in black. Estimates were stratified by time (full season[30, 34, 35], peak months[5, 21], off-peak months) Only the estimate for age group 3-12 months was shown for Levy et al. 2024[21].

All sensitivity analyses generated consistent results, with less than 10% differences in point estimates of effectiveness (Table S5). As expected, the proportions of cases and controls that received the hepatitis B vaccine were nearly identical (54% and 57%, respectively, p=0.18), and the corresponding effectiveness of the hepatitis B vaccine against RSV was not statistically significant (Table S5).

Conclusion

This study confirms that nirsevimab is highly effective at preventing RSV-associated outpatient visits, hospitalizations, and severe disease requiring ICU admission or high-flow oxygen. Its effectiveness persisted for at least three months post-vaccination, consistent with results from randomized trials. These findings reinforce the benefits of RSV immunoprophylaxis and support US guidelines recommending nirsevimab for all infants entering their first RSV season.