Objective
This study aims to explore the association between inpatient gout flareârelated variables and the length of stay (LOS) in hospitalized people with comorbid gout.
Methods
Using data from the Aotearoa/New Zealand national data collections, this cohort study included adults with comorbid gout who were admitted to publicly funded hospitals during 2017 for reasons other than gout. The primary outcome was LOS. Association between 20 variables and the LOS was explored using two generalized linear models. Directed acyclic graph (DAG) was constructed to evaluate the causal relationship between preâadmission urate lowering therapy (ULT) and LOS.
Results
The cohort included 36 047 admissions. We identified five variables associated with shorter LOS (preâadmission regular urateâlowering therapy (ULT), serum urate testing, male gender, MÄori ethnicity and lowâdose aspirin) and seven variables associated with longer LOS (M3 multimorbidity index, acute admission, operation, loop diuretics, potassiumâsparing diuretics, NSAIDs, and age). Regular ULT had the strongest impact on shorter LOS (10% shorter). The model estimated an additional four days of hospitalization if the patient had multiple variables associated with longer LOS. DAG suggested a causal relationship between regular ULT and LOS under the condition that all unobserved confounders affected only ULT use, with no impact on inâhospital gout flares and/or LOS except through its influence on ULT use or as mediator of confounders that were observed.
Conclusion
We have identified a set of gout flareârelated variables found to be associated with LOS in hospitalized people with comorbid gout. Preâadmission ULT may help reduce the LOS in such patients.