Background: National estimates of drug discontinuation for deprescribing targets in older adults are limited, partly due to challenges distinguishing planned deprescribing from poor adherence. Focusing on individuals with multidose dispensing (MDD), characterized by high adherence by design, may yield realistic discontinuation rates.

Aims: To estimate the rates of discontinuation for chronically used drugs targeted for deprescribing among older adults, and to describe reinitiation among users of MDD and standard dispensing (non-MDD).

Methods: In this nationwide cohort study, Swedish adults aged ≥ 75 were identified from national registers. At baseline (1 January 2021), chronic users of seven drug classes were defined. We estimated the 12-month cumulative incidence of discontinuation (defined as no new dispensing during the treatment episode of the prior dispensing plus a 180-day grace period) and the proportion of patients restarting therapy within 180 days after discontinuation.

Results: We identified 162 518 chronic users: benzodiazepines (n = 69 511), PPIs (n = 43 973), antidepressants (n = 41 577), statins (n = 36 085), cholinesterase inhibitors (n = 6408), bisphosphonates (n = 5801) and antipsychotics (n = 4380). Discontinuation rates were low (8.3–51.5 per 1000 person-years), and non-MDD users had higher discontinuation and reinitiation rates across all drugs.

Conclusion: Discontinuation among Swedish older adults is infrequent. Irregular dispensing is likely misclassified as deprescribing, and MDD users may better reflect true discontinuation in routinely collected data.