Twenty-four prescribing cascades, ranked by how often they actually happen

Design

Population-level retrospective cohort study using prescription sequence symmetry analysis

Population

2,297,942 community-dwelling adults aged 66 and over in Ontario, Canada

Primary outcome

Prevalence, incidence and adjusted sequence ratio for 65 candidate prescribing cascades

Effect

24 cascades met all three criteria; highest incidence iron to laxative 11.9% (95% CI 11.7–12.1); strongest sequence ratio corticosteroid to antipsychotic 2.55 (2.47–2.64)

Prescribing cascades — treating a drug's side effect with another drug — are familiar as a concept and vague as a clinical target. This study made them concrete. Starting from 65 cascades agreed by international Delphi consensus, it tested each against administrative data for 2,297,942 community-dwelling adults aged 66 and over in Ontario, keeping only those that were common at population level, common in individuals, and showed a real temporal sequence.

Twenty-four survived. The highest-incidence were iron supplement to laxative (11.9%, 95% CI 11.7–12.1), statin to pain reliever (10.9%, 10.7–11.0), and cholinesterase inhibitor to sleep agent (10.3%, 9.9–10.7). The strongest temporal associations, measured as adjusted sequence ratios, were corticosteroid to antipsychotic (2.55, 2.47–2.64), laxative to antidiarrhoeal (2.53, 2.43–2.65), and cholinesterase inhibitor to antiemetic (2.24, 1.90–2.65). Cardiovascular drugs were the most common class starting a cascade.

These are not exotic. They are the prescriptions a general clinic writes every day, and each one names a specific pair worth pausing over. The practical use is as a short list to recognise: when the second drug in one of these pairs is about to be written, the question is whether the first drug should come down instead.

• Before prescribing a laxative to a patient on iron, review the iron — dose, formulation, and whether it is still needed.
• Treat new musculoskeletal pain in a patient recently started on a statin as a statin question first.
• A new sleep agent or antiemetic in a patient on a cholinesterase inhibitor should prompt review of that drug, not an addition to it.
• Corticosteroid to antipsychotic had the strongest temporal signal of all — expect neuropsychiatric effects from steroids and plan the taper rather than medicating the agitation.
• Iron-induced constipation is particularly relevant in India, where iron supplementation is widespread and often long-term.

Why it matters

"Watch for prescribing cascades" is advice nobody can act on; a ranked list of 24 specific drug pairs is.

Don't overread it

This is administrative data — a sequence ratio shows a pattern, not that any particular prescription was wrong.

The statistics, in plain English

An adjusted sequence ratio compares how often drug B follows drug A with how often it precedes it. A ratio above 1 means the sequence is asymmetric, which is what you would expect if A were causing the problem B treats — but asymmetry is consistent with confounding by indication too, since the same underlying illness can prompt both drugs in a predictable order. This design shows patterns worth interrogating; it does not prove any individual cascade was inappropriate.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.