Objective: We compared trends in incidence rates (per 1,000 livebirths) of early-onset GBS (EOGBS), late-onset GBS (LOGBS), and percentage of GBS cases resistant to clindamycin and erythromycin under different GBS prevention strategies across countries.
Methods: We collected data on annual GBS rates (outcomes), current prevention strategy (predictor), and contextual factors (confounding variables) in a survey sent to institutions across countries. Some contextual data was supplemented from international websites and some observations for contextual variables were imputed using multiple imputation. We analysed the association between prevention strategy and trends of outcomes using linear regression models, where prevention strategy was interacted with year, and along with the contextual variables, was regressed on each outcome. We also summarised the outcome trends by world region using unadjusted regression.
Results: Overall, data covered >40 areas from 1989-2015. We found a difference in GBS trends by current prevention strategy. EOGBS decreased under screening or both screening and risk-based prevention, and increased under risk-based or no prevention. The largest increase of LOGBS was under no followed by risk-based prevention, while screening had the smallest increase. These findings were finely balanced as sensitivity analyses produced different results. There were no differences in resistance trends possibly due to a lack of power, even though the percentage of neonatal GBS cases resistant to erythromycin increased higher under screening compared to other prevention strategies.
Conclusion: Adjusting for contextual differences, trends of neonatal GBS are internationally affected by prevention strategy. However, findings should be treated cautiously as results depend on many factors.