Professor Clinical Pharmacology, Pharmacy, and Environmental Medicine, Department of Public Health, University of Southern Denmark. Odense C, Denmark
Background: When patients experience larger clinical transitions, such as receiving a cancer diagnosis, their pharmacotherapy should be re-evaluated. Epidemiologists might want to describe the extent of such changes. However, doing so is methodologically challenging.
Objectives: We aim to describe changes in medication use from a major clinical transition and until death and highlight the inherent methodological challenges with describing such changes, using as case medication trajectories among patients with type 2 diabetes receiving a cancer diagnosis.
Methods: All Danish patients with cancer and type 2 diabetes during 2000-2021 were matched to four patients without cancer using propensity score for cancer, based on sex, country of birth, age, duration of diabetes, comorbidities, and socioeconomic data. Within this cohort, we analysed 1) rate of incident (new) medications per 100 patient-months, and 2) the proportion of users. Commonly used medications were categorised as symptomatic (opioids, benzodiazepines and antiemetics) or preventive (antihypertensives, statins and glucose-lowering medications). To illustrate the challenges of analysing medication changes we anchored the analyses at both cohort entry and death, and further stratified the cohort by length of survival following cohort entry.
Results: In total, 41,745 patients with cancer were matched with 166,994 patients without cancer (median age 74). The 1- and 5-year mortality was 51% and 86% in the cancer group and 13% and 59% in the non-cancer group. There were substantial changes in medications usage in relation to a cancer diagnosis and death. In both situations, the use of symptomatic medications increased, and the use of preventive medication decreased. Stratification showed that the overall changes in relation to a cancer diagnosis were largely driven by patients with short length of survival ( < 2 years). In contrast, changes near death were less dependent on survival strata, and >60% used preventive medications at death, irrespective of cancer diagnosis.
Conclusions: Medication changes in relation to a cancer diagnosis were frequent and correlated to the length of survival. While medication changes were common near death, preventive medications were often used until death. The results showcase the challenges and limited clinical utility of anchoring analyses on either an event or death. While the former diluted the results by averaging changes across patients with vastly different prognoses, the latter leveraged information unavailable to clinicians treating the patients.
