Objective:

To clarify aspects of the study assessing oseltamivir chemoprophylaxis in nursing homes during influenza outbreaks.

Approach:

• Data Source: The study utilized electronic medication administration record (eMAR) data to assess oseltamivir chemoprophylaxis rather than clinician orders.
• Treatment Assignment: Treatment strategies were assigned at the nursing home level, with an emphasis on the differences in characteristics influencing hospitalization risk.
• Bias Mitigation: A randomize-censor-weight approach was employed to mitigate bias from measured determinants of adherence to NH-level strategies.

Key Findings:

• Substantial unmeasured oseltamivir use in the nonintensive group is unlikely.
• Censoring after the 2-day grace period resulted in a substantial reduction in the intensive-strategy risk set.
• Inverse-probability-of-censoring weighting was used to address potential bias from informative censoring.

Interpretation:

The results support rapid, extensive, staff-administered oseltamivir chemoprophylaxis during influenza outbreaks in nursing homes.

Limitations:

• Residual bias due to unmeasured NH-level or outbreak-level factors remains possible.
• Censoring may generate selection bias, particularly when the treatment strategy is uncommon.

Conclusion:

Future studies should replicate this approach and specify all components of a target trial, ideally using eMAR data.

Sources:

• Original Study
• Additional Reference
• Further Reading
• Study Replication