When should a prescribing cascade trigger an alert?

Agreement: I Agree Body: Dear Editor Rochon and colleagues identified 24 potentially inappropriate prescribing cascades (PIPCs) that could inform prescribing review and clinical decision support.[1] Their findings raise a practical question for implementation: not only which cascades should trigger attention, but when. The primary analysis allowed up to one year between initiation of the two drugs, with a 180-day sensitivity analysis.[1] This is appropriate for population-level signal detection.

BMJ

By AnonymousSeptember 17, 2026South Africa

Agreement: I AgreeBody:

Dear Editor

Rochon and colleagues identified 24 potentially inappropriate prescribing cascades (PIPCs) that could inform prescribing review and clinical decision support.[1] Their findings raise a practical question for implementation: not only which cascades should trigger attention, but when.

The primary analysis allowed up to one year between initiation of the two drugs, with a 180-day sensitivity analysis.[1] This is appropriate for population-level signal detection. Importantly, however, the authors also note that the timing of individual cascades varies: loop-diuretic dispensing after calcium-channel blockers may occur within 30 days, increased loop-diuretic use after gabapentinoids within 90 days, whereas beta-blocker-to-antidepressant cascades may take several months.[1]

The analytic window used to detect a population signal should not necessarily be assumed to translate directly into the optimal look-back period for clinical decision support. For some cascades, the elapsed time since initiation of the first drug could help determine whether an alert should fire when the subsequent drug is prescribed. For others with a slower or broader latency, longitudinal monitoring or scheduled medication review may be more useful than an interruptive alert. The timing and workflow fit of alerts are recognised determinants of useful clinical decision support.[2]

Before these signals are operationalised in prescriber- or pharmacist-facing systems, cascade-specific time-to-subsequent-drug distributions would therefore be informative. Reporting medians, interquartile ranges, and cumulative proportions at clinically relevant intervals could help define appropriate look-back periods and, importantly, the type of intervention most suited to each cascade.

The study identifies which prescribing cascades warrant attention at the population level. Understanding when they emerge may help determine how that attention is delivered in practice.

References
1.Rochon PA, Austin PC, Gurwitz JH, et al. Exploring high priority potentially inappropriate prescribing cascades in older adults: population level retrospective cohort study. BMJ. 2026;394:e100499. doi:10.1136/bmj-2026-100499.
2.Olakotan OO, Yusof MM. The appropriateness of clinical decision support systems alerts in supporting clinical workflows: a systematic review. Health Informatics J. 2021;27(2):14604582211007536. doi:10.1177/14604582211007536..

No competing Interests: YesThe following competing Interests: Electronic Publication Date: Thursday, September 17, 2026 - 02:10AI use: No, I have not used AIHighwire Comment Subject:

Exploring high priority potentially inappropriate prescribing cascades in older adults: population level retrospective cohort study