TY - GEN
T1 - Cluster and Trajectory Analysis of Multiple Long-Term Conditions in Adults with Learning Disabilities
AU - Abakasanga, Emeka
AU - Kousovista, Rania
AU - Cosma, Georgina
AU - Jun, Gyuchan Thomas
AU - Kiani, Reza
AU - Gangadharan, Satheesh
PY - 2024/8/15
Y1 - 2024/8/15
N2 - Individuals with learning disabilities (LD) are at a heightened risk of experiencing multiple long-term conditions (MLTCs) due to various factors, which can lead to increased premature mortality rates and compromised quality of life. Despite this, there is limited research employing cluster analysis to identify and categorise similar patterns of MLTCs in patients with learning disabilities. This study applies machine learning clustering algorithms to data from 13,069 adults with learning disabilities in Wales, using a 3-cluster Gaussian Mixture Model for 6,830 males and a 3-cluster BIRCH algorithm for 6,239 females. Cluster 3 for males and Cluster 1 for females represented ‘relatively healthy’ groups, characterised by predominantly younger patients with lower MLTC counts and lower hospitalization rates. However, these clusters exhibited the lowest age at mortality, 62 years for males and approximately 65 years for females, indicating a higher likelihood of preventable deaths. Subsequently, prevalent combinations of MLTCs and common disease trajectories were analysed within these clusters. Identifying distinct MLTC clusters, prevalent combinations, and trajectories provides crucial insights for optimizing care pathways, targeted interventions, and resource allocation tailored to the specific needs of individuals with learning disabilities. This ultimately aims to improve health outcomes and quality of life for this population.
AB - Individuals with learning disabilities (LD) are at a heightened risk of experiencing multiple long-term conditions (MLTCs) due to various factors, which can lead to increased premature mortality rates and compromised quality of life. Despite this, there is limited research employing cluster analysis to identify and categorise similar patterns of MLTCs in patients with learning disabilities. This study applies machine learning clustering algorithms to data from 13,069 adults with learning disabilities in Wales, using a 3-cluster Gaussian Mixture Model for 6,830 males and a 3-cluster BIRCH algorithm for 6,239 females. Cluster 3 for males and Cluster 1 for females represented ‘relatively healthy’ groups, characterised by predominantly younger patients with lower MLTC counts and lower hospitalization rates. However, these clusters exhibited the lowest age at mortality, 62 years for males and approximately 65 years for females, indicating a higher likelihood of preventable deaths. Subsequently, prevalent combinations of MLTCs and common disease trajectories were analysed within these clusters. Identifying distinct MLTC clusters, prevalent combinations, and trajectories provides crucial insights for optimizing care pathways, targeted interventions, and resource allocation tailored to the specific needs of individuals with learning disabilities. This ultimately aims to improve health outcomes and quality of life for this population.
KW - Cluster analysis
KW - Learning disability
KW - Multiple long-term conditions
KW - Trajectories
UR - http://www.scopus.com/inward/record.url?scp=85201975493&partnerID=8YFLogxK
UR - https://link.springer.com/chapter/10.1007/978-3-031-67285-9_1
U2 - 10.1007/978-3-031-67285-9_1
DO - 10.1007/978-3-031-67285-9_1
M3 - Conference publication
AN - SCOPUS:85201975493
SN - 9783031672842
T3 - Lecture Notes in Computer Science (LNCS (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics))
SP - 3
EP - 16
BT - Artificial Intelligence in Healthcare
A2 - Xie, Xianghua
A2 - Powathil, Gibin
A2 - Styles, Iain
A2 - Ceccarelli, Marco
T2 - 1st International Conference on Artificial Intelligence in Healthcare, AIiH 2024
Y2 - 4 September 2024 through 6 September 2024
ER -