Dataset Information

Deep significance clustering: a novel approach for identifying risk-stratified and predictive patient subgroups.

ABSTRACT:

Objective

Deep significance clustering (DICE) is a self-supervised learning framework. DICE identifies clinically similar and risk-stratified subgroups that neither unsupervised clustering algorithms nor supervised risk prediction algorithms alone are guaranteed to generate.

Materials and methods

Enabled by an optimization process that enforces statistical significance between the outcome and subgroup membership, DICE jointly trains 3 components, representation learning, clustering, and outcome prediction while providing interpretability to the deep representations. DICE also allows unseen patients to be predicted into trained subgroups for population-level risk stratification. We evaluated DICE using electronic health record datasets derived from 2 urban hospitals. Outcomes and patient cohorts used include discharge disposition to home among heart failure (HF) patients and acute kidney injury among COVID-19 (Cov-AKI) patients, respectively.

Results

Compared to baseline approaches including principal component analysis, DICE demonstrated superior performance in the cluster purity metrics: Silhouette score (0.48 for HF, 0.51 for Cov-AKI), Calinski-Harabasz index (212 for HF, 254 for Cov-AKI), and Davies-Bouldin index (0.86 for HF, 0.66 for Cov-AKI), and prediction metric: area under the Receiver operating characteristic (ROC) curve (0.83 for HF, 0.78 for Cov-AKI). Clinical evaluation of DICE-generated subgroups revealed more meaningful distributions of member characteristics across subgroups, and higher risk ratios between subgroups. Furthermore, DICE-generated subgroup membership alone was moderately predictive of outcomes.

Discussion

DICE addresses a gap in current machine learning approaches where predicted risk may not lead directly to actionable clinical steps.

Conclusion

DICE demonstrated the potential to apply in heterogeneous populations, where having the same quantitative risk does not equate with having a similar clinical profile.

SUBMITTER: Huang Y

PROVIDER: S-EPMC8500061 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:OBJECTIVES:Identifying subgroups of ICU patients with similar clinical needs and trajectories may provide a framework for more efficient ICU care through the design of care platforms tailored around patients' shared needs. However, objective methods for identifying these ICU patient subgroups are lacking. We used a machine learning approach to empirically identify ICU patient subgroups through clustering analysis and evaluate whether these groups might represent appropriate targets for care redesign efforts. DESIGN:We performed clustering analysis using data from patients' hospital stays to retrospectively identify patient subgroups from a large, heterogeneous ICU population. SETTING:Kaiser Permanente Northern California, a healthcare delivery system serving 3.9 million members. PATIENTS:ICU patients 18 years old or older with an ICU admission between January 1, 2012, and December 31, 2012, at one of 21 Kaiser Permanente Northern California hospitals. INTERVENTIONS:None. MEASUREMENTS AND MAIN RESULTS:We used clustering analysis to identify putative clusters among 5,000 patients randomly selected from 24,884 ICU patients. To assess cluster validity, we evaluated the distribution and frequency of patient characteristics and the need for invasive therapies. We then applied a classifier built from the sample cohort to the remaining 19,884 patients to compare the derivation and validation clusters. Clustering analysis successfully identified six clinically recognizable subgroups that differed significantly in all baseline characteristics and clinical trajectories, despite sharing common diagnoses. In the validation cohort, the proportion of patients assigned to each cluster was similar and demonstrated significant differences across clusters for all variables. CONCLUSIONS:A machine learning approach revealed important differences between empirically derived subgroups of ICU patients that are not typically revealed by admitting diagnosis or severity of illness alone. Similar data-driven approaches may provide a framework for future organizational innovations in ICU care tailored around patients' shared needs.

Project description:ObjectiveMany healthcare systems employ population-based risk scores to prospectively identify patients at high risk of poor outcomes, but it is unclear whether single point-in-time scores adequately represent future risk. We sought to identify and characterize latent subgroups of high-risk patients based on risk score trajectories.Study designObservational study of 7289 patients discharged from Veterans Health Administration (VA) hospitals during a 1-week period in November 2012 and categorized in the top 5th percentile of risk for hospitalization.MethodsUsing VA administrative data, we calculated weekly risk scores using the validated Care Assessment Needs model, reflecting the predicted probability of hospitalization. We applied the non-parametric k-means algorithm to identify latent subgroups of patients based on the trajectory of patients' hospitalization probability over a 2-year period. We then compared baseline sociodemographic characteristics, comorbidities, health service use, and social instability markers between identified latent subgroups.ResultsThe best-fitting model identified two subgroups: moderately high and persistently high risk. The moderately high subgroup included 65% of patients and was characterized by moderate subgroup-level hospitalization probability decreasing from 0.22 to 0.10 between weeks 1 and 66, then remaining constant through the study end. The persistently high subgroup, comprising the remaining 35% of patients, had a subgroup-level probability increasing from 0.38 to 0.41 between weeks 1 and 52, and declining to 0.30 at study end. Persistently high-risk patients were older, had higher prevalence of social instability and comorbidities, and used more health services.ConclusionsOn average, one third of patients initially identified as high risk stayed at very high risk over a 2-year follow-up period, while risk for the other two thirds decreased to a moderately high level. This suggests that multiple approaches may be needed to address high-risk patient needs longitudinally or intermittently.

Project description:Background Patient-reported outcome measures (PROMs) are increasingly being used to assess the effectiveness of elective total knee arthroplasty (TKA). However, little is known about how PROMs scores change over time in these patients. The aim of this study was to identify the trajectories of quality of life and joint functioning, and their associated demographic and clinical features in patients undergoing elective TKA. Methods A prospective, cohort study was conducted, in which PROMs questionnaires (Euro Quality 5 Dimensions 3L, EQ-5D-3L, and Knee injury and Osteoarthritis Outcome Score Patient Satisfaction, KOOS-PS) were administered to patients at a single center undergoing elective TKA before surgery, and at 6 and 12 months after surgery. Latent class growth mixture models were used to analyze the patterns of change in PROMs scores over time. Multinomial logistic regression was used to investigate the association between patient characteristics and PROMs trajectories. Results A total of 564 patients were included in the study. The analysis highlighted differential patterns of improvement after TKA. Three distinct PROMs trajectories were identified for each PROMs questionnaire, with one trajectory indicating the most favorable outcome. Female gender appears to be associated with a presentation to surgery with worse perceived quality of life and joint function than males, but also more rapid improvement after surgery. Having an ASA score greater than 3 is instead associated with a worse functional recovery after TKA. Conclusion The results suggest three main PROMs trajectories in patients undergoing elective TKA. Most patients reported improved quality of life and joint functioning at 6 months, which then stabilized. However, other subgroups showed more varied trajectories. Further research is needed to confirm these findings and to explore the potential clinical implications of these results. Supplementary Information The online version contains supplementary material available at 10.1186/s12891-023-06373-2.

Project description:A growing number of studies have explored the process of pre-implantation embryonic development of multiple mammalian species. However, the conservation and variation among different species in their developmental programming are poorly defined due to the lack of effective computational methods for detecting co-regularized genes that are conserved across species. The most sophisticated method to date for identifying conserved co-regulated genes is a two-step approach. This approach first identifies gene clusters for each species by a cluster analysis of gene expression data, and subsequently computes the overlaps of clusters identified from different species to reveal common subgroups. This approach is ineffective to deal with the noise in the expression data introduced by the complicated procedures in quantifying gene expression. Furthermore, due to the sequential nature of the approach, the gene clusters identified in the first step may have little overlap among different species in the second step, thus difficult to detect conserved co-regulated genes.We propose a cross-species bi-clustering approach which first denoises the gene expression data of each species into a data matrix. The rows of the data matrices of different species represent the same set of genes that are characterized by their expression patterns over the developmental stages of each species as columns. A novel bi-clustering method is then developed to cluster genes into subgroups by a joint sparse rank-one factorization of all the data matrices. This method decomposes a data matrix into a product of a column vector and a row vector where the column vector is a consistent indicator across the matrices (species) to identify the same gene cluster and the row vector specifies for each species the developmental stages that the clustered genes co-regulate. Efficient optimization algorithm has been developed with convergence analysis. This approach was first validated on synthetic data and compared to the two-step method and several recent joint clustering methods. We then applied this approach to two real world datasets of gene expression during the pre-implantation embryonic development of the human and mouse. Co-regulated genes consistent between the human and mouse were identified, offering insights into conserved functions, as well as similarities and differences in genome activation timing between the human and mouse embryos.The R package containing the implementation of the proposed method in C?++?is available at: https://github.com/JavonSun/mvbc.git and also at the R platform https://www.r-project.org/jinbo@engr.uconn.edu.