How does the presence of multimorbidity in those with rheumatoid arthritis influence hospitalisations?

Problem

Rheumatoid arthritis (RA) is a chronic autoimmune disease, resulting in painful joint pathology. Despite high levels of multimorbidity (≥2 long-term conditions (LTCs)) in RA, most work has focused on comorbidity (+1 LTC) and little is known about how multimorbidity in RA impacts health outcomes, including hospitalisations. This work aims to analyse the risk of and reasons for hospitalisations in those with RA and multimorbidity using two datasets- population-based and clinical.

Approach

We used data from UK Biobank (N=502,533 participants) and the Scottish Early Rheumatoid Arthritis (SERA; N=1095) cohort to study RA and a count of up to 42 other LTCs at baseline. Both datasets were linked to Hospital Episodes Statistics data. Negative binomial regression models were used to ascertain risk of hospitalisation (controlled for both lifestyle factors (smoking status, frequency of alcohol intake, and body mass index (BMI)) and demographic factors (sex, age and socioeconomic status)). Most common reasons for hospitalisation were calculated using primary ICD-10 codes.

Findings

N=5658 (1.1%) of participants in UK Biobank self-reported RA (mean age (SD), %F), whilst N=901 SERA (82.3%) SERA patients (mean age (SD), %F) had clinician diagnosed RA. In UK Biobank, of the N=5625 RA participants with complete LTC data, N=1424 (25.3%) reported having 0 LTCs, N=1694 (30.1%) 1 LTC, and N=2507 (44.6%) ≥2LTCs. In SERA, N=226 (25.1%) of participants had 0 LTCs, N=252 (28.0%) 1 LTC, and N=423 (46.9%) ≥2LTCs. When examining risk of hospitalisation, UK Biobank data showed that those with ≥2 LTCs and RA had a 68% increased risk of hospitalisation compared to those with RA alone (relative risk ratio (RRR) 1.68, 95% confidence intervals (CI) 1.55-1.80). In SERA, those with RA and ≥2LTCs had a 31% increased risk of hospitalisation (RRR 1.31, 95% CI 1.07-1.63) compared to those with RA alone. When examining reasons for hospitalisation, UK Biobank participants with RA and ≥2 LTCs were predominantly admitted due to RA (“M06” ICD-10 codes) and chronic kidney diseases (“N18” codes), whilst those with RA alone were admitted predominantly for RA and cancer (“C” codes). In SERA, those with RA and ≥2LTCs were predominantly admitted for respiratory disease (“J” codes), myocardial infarction (“I21” codes) and throat and chest pains (“R07” codes), whilst those with RA alone were predominantly admitted for myocardial infarction, cancer and RA.

Consequences

These findings suggest that multimorbidity should be taken into account in the management of patients with RA. In both the population based/clinical datasets, participants living with RA and ≥2 other LTCs were at increased risk of hospitalisation after controlling for demographic/lifestyle factors, although the risk was higher in the population-based cohort. There was overlap in the commonest reasons for hospitalisation, however cancer only featured for those with RA alone.