Temporal patterns of multiple long-term conditions in individuals with intellectual disability living in Wales: an unsupervised clustering approach to disease trajectories.

Introduction: Identifying and understanding the co-occurrence of multiple long-term conditions (MLTCs) in individuals with intellectual disability (ID) is crucial for effective healthcare management. Individuals with ID often experience earlier onset and higher prevalence of MLTCs compared to the general population, however, the specific patterns of co-occurrence and temporal progression of these conditions remain largely unexplored. This study presents an innovative unsupervised approach for examining and characterising clusters of MLTC in individuals with ID, based on their shared disease trajectories. Methods: Using a dataset of electronic health records (EHRs) from 13,069 individuals with ID, encompassing primary and secondary care data in Wales from 2000 to 2021, this study analysed the time sequences of disease diagnoses. Significant pairwise disease associations were identified, and their temporal directionality assessed. Subsequently, an unsupervised clustering algorithm-spectral clustering-was applied to the shared disease trajectories, grouping them based on common temporal patterns. Results: The study population comprised 52.3% males and 47.7% females, with a mean of 4.5 Discussion: This study reveals complex multimorbidity patterns in individuals with ID, highlighting age and sex differences. The identified clusters provide new insights into disease progression and co-occurrence in this population. These findings can inform the development of targeted interventions and risk stratification strategies, potentially improving personalised healthcare for individuals with ID and MLTCs with the aim of improving health outcome for this vulnerable group of patients i.e. reducing frequency and length of hospital admissions and premature mortality.