TY - GEN
T1 - Comorbidity constructs for patients with congestive heart failure and their effect on hospital outcomes of care
AU - Zikos, Dimitrios
AU - Delellis, Nailya
AU - Zimeras, Stelios
AU - Ragina, Neli
AU - Afolayan-Oloye, Olabisi I.
N1 - Funding Information:
The work has been completed in the framework of the Central Michigan University Faculty Research and Creative Endeavors (FRCE) grant, P.I: Dimitrios Zikos, Grant number: 48530, 01/01/2019 - 06/30/2020.
Publisher Copyright:
© 2019 Association for Computing Machinery.
PY - 2019/6/5
Y1 - 2019/6/5
N2 - We present a study of the cumulative effect of comorbidities on hospital length of stay (LOS), and hospital mortality, for patients with Congestive Heart Failure (CHF). This condition can be life-threatening, while a burdened disease profile significantly increases the risk for negative outcomes. Our hypothesis is that coexisting conditions often co-interact, in various ways, and these interactions can have a variable effect on outcomes; clinical decision makers should, therefore, be able to recognize these joint effects. In order to study the CHF comorbidities, we used medical claims data from CMS. Firstly, we conducted cluster analysis to find the common hospital comorbidities for CHF admissions. We then extracted the most frequent cluster: {metabolism disorders, anemia, hypertension with complications, coronary atherosclerosis, chronic kidney disease} and calculated conditional probabilities in a modular manner for all combinations within this cluster. We furthermore estimated the cumulative effect of these comorbidity combinations on the two outcomes under study in a step by step, modular manner. Results were visualized with directed acyclic graphs.
AB - We present a study of the cumulative effect of comorbidities on hospital length of stay (LOS), and hospital mortality, for patients with Congestive Heart Failure (CHF). This condition can be life-threatening, while a burdened disease profile significantly increases the risk for negative outcomes. Our hypothesis is that coexisting conditions often co-interact, in various ways, and these interactions can have a variable effect on outcomes; clinical decision makers should, therefore, be able to recognize these joint effects. In order to study the CHF comorbidities, we used medical claims data from CMS. Firstly, we conducted cluster analysis to find the common hospital comorbidities for CHF admissions. We then extracted the most frequent cluster: {metabolism disorders, anemia, hypertension with complications, coronary atherosclerosis, chronic kidney disease} and calculated conditional probabilities in a modular manner for all combinations within this cluster. We furthermore estimated the cumulative effect of these comorbidity combinations on the two outcomes under study in a step by step, modular manner. Results were visualized with directed acyclic graphs.
KW - Clustering
KW - Comorbidities
KW - Congestive heart failure
KW - Health informatics
UR - http://www.scopus.com/inward/record.url?scp=85069205082&partnerID=8YFLogxK
U2 - 10.1145/3316782.3322745
DO - 10.1145/3316782.3322745
M3 - Conference contribution
AN - SCOPUS:85069205082
T3 - ACM International Conference Proceeding Series
SP - 196
EP - 203
BT - Proceedings of the 12th ACM International Conference on PErvasive Technologies Related to Assistive Environments, PETRA 2019
PB - Association for Computing Machinery
Y2 - 5 June 2019 through 7 June 2019
ER -