Use of the moving epidemic method (MEM) to assess national surveillance data for respiratory syncytial virus (RSV) in the Netherlands, 2005 to 2017

Abstract

Background

To control respiratory syncytial virus (RSV), which causes acute respiratory infections, data and methods to assess its epidemiology are important.

Aim

We sought to describe RSV seasonality, affected age groups and RSV-type distribution over 12 consecutive seasons in the Netherlands, as well as to validate the moving epidemic method (MEM) for monitoring RSV epidemics.

Methods

We used 2005−17 laboratory surveillance data and sentinel data. For RSV seasonality evaluation, epidemic thresholds (i) at 1.2% of the cumulative number of RSV-positive patients per season and (ii) at 20 detections per week (for laboratory data) were employed. We also assessed MEM thresholds.

Results

In laboratory data RSV was reported 25,491 times (no denominator). In sentinel data 5.6% (767/13,577) of specimens tested RSV positive. Over 12 seasons, sentinel data showed percentage increases of RSV positive samples. The average epidemic length was 18.0 weeks (95% confidence intervals (CI):  16.3–19.7) and 16.5 weeks (95% CI: 14.0–18.0) for laboratory and sentinel data, respectively. Epidemics started on average in week 46 (95% CI: 45–48) and 47 (95% CI:  46–49), respectively. The peak was on average in the first week of January in both datasets. MEM showed similar results to the other methods. RSV incidence was highest in youngest (0–1 and >1–2 years) and oldest (>65–75 and > 75 years) age groups, with age distribution remaining stable over time. RSV-type dominance alternated every one or two seasons.

Conclusions

Our findings provide baseline information for immunisation advisory groups. The possibility of employing MEM to monitor RSV epidemics allows prospective, nearly real-time use of surveillance data.