Medication-related hospitalisations in patients with SLE

Patients with SLE take multiple medications. Within a large prospective longitudinal SLE cohort, we characterised medication-related hospitalisations and their preventability.

We identified consecutive admissions to our tertiary hospitals between 2015 and 2020. Two independent adjudicators evaluated if medication-related events contributed to the hospitalisation, considering (1) adverse drug events (ADEs) and (2) events from medication non-adherence, using the Leape and Bates method. We classified ADEs as potentially preventable/ameliorable if we identified modifiable factors. Logistic regressions with generalised estimating equations evaluated associations between participant characteristics and medication-related hospitalisations, accounting for repeat hospitalisations within the same participant.

We studied 68 hospitalisations among 45 participants (91% female). At first hospitalisation, the median age was 38 years (IQR 26.5-53.0) and median SLE duration was 12 years (IQR 5.5-19.5). One or more ADEs contributed to 20 (29%) hospitalisations (11/23 (48%) ADEs being preventable/ameliorable), and SLE flares associated with medication non-adherence contributed to 7 (10%) hospitalisations. Adjusting for age and sex, current prednisone use (adjusted OR (aOR) 3.7, 95% CI 1.1 to 13.0) or ≥1 current immunosuppressant (aOR 11.5, 95% CI 2.7 to 50.0), renal involvement at SLE diagnosis (aOR 6.5, 95% CI 2.7 to 15.7) and polypharmacy (≥5 medications; aOR 11.3, 95% CI 1.2 to 103.8) were associated with having an ADE-related (vs non-ADE) hospitalisation. Age at SLE diagnosis<18 years (OR 5.9, 95% CI 1.3 to 26.6) was associated with hospitalisation for a flare related to non-adherence.

Forty per cent of SLE hospitalisations were medication-related, while half were potentially preventable/ameliorable. Renal involvement, polypharmacy, prednisone and immunosuppressant use were associated with hospitalisation related to an ADE, highlighting a vulnerable group.