Ontario data flagged 24 common prescribing-cascade patterns

The research question and why it matters

The researchers asked which previously proposed prescribing cascades are common enough, occur often enough after an initial drug, and show a strong enough temporal pattern to deserve priority for validation and prevention work in older adults.

Prescribing cascades have been described for decades, usually through individual examples, pharmacology studies or smaller datasets. An international Delphi process recently assembled 139 possible cascades and selected 65 as clinically important. This study used population data to rank that list with a common quantitative framework rather than asserting that every proposed pair is causal.

What researchers found

Twenty-four of 65 candidate cascades, or 37%, met all three criteria. The largest second-drug incidences were iron supplement followed by a laxative at 11.9%, statin followed by a pain reliever at 10.9%, and cholinesterase inhibitor followed by a sleep medicine at 10.3%. The strongest temporal signals were corticosteroid followed by an antipsychotic, with an adjusted sequence ratio of 2.55; laxative followed by an antidiarrheal, 2.53; and cholinesterase inhibitor followed by an antiemetic, 2.24.

How the research worked

The team began with 65 potential cascades selected through an international Delphi process. For each pair, they measured use of the initial drug, the incidence of the possible cascade drug among new users, and an adjusted sequence ratio comparing how often the second drug followed rather than preceded the first. A cascade was prioritized only when the initial drug prevalence was at least 5%, the second-drug incidence was at least 1%, and the lower confidence bound for the adjusted sequence ratio exceeded one.

How to interpret this design

This design can measure patterns and associations in the observed population. It cannot, by itself, prove that the exposure caused the outcome because unmeasured differences, reverse causation and selection effects may contribute.

Because the research involved people, it speaks directly to the participants and outcomes measured. It may still apply differently to people outside the eligibility criteria, age range, clinical setting, geography or follow-up period.

What it means

The study provides a ranked surveillance list for clinicians, pharmacists and researchers. It can focus chart reviews, patient conversations and prospective studies on combinations where a drug side effect might be mistaken for a new condition, while leaving individual decisions to clinical assessment.

How this fits with previous research

Prescribing cascades have been described for decades, usually through individual examples, pharmacology studies or smaller datasets. An international Delphi process recently assembled 139 possible cascades and selected 65 as clinically important. This study used population data to rank that list with a common quantitative framework rather than asserting that every proposed pair is causal.

Questions still unanswered

• Which of the 24 signals are confirmed when clinicians review symptoms, indications and patient records directly?
• How often does identifying a cascade lead to safer deprescribing or a non-drug alternative?
• Which combinations produce the greatest patient harm rather than merely the highest frequency?
• Do sex, frailty, kidney function, dementia or polypharmacy materially change particular cascade risks?
• Can electronic prescribing systems warn about likely cascades without creating excessive alerts?