Dataset Information

Identifying Latent Subgroups of High-Risk Patients Using Risk Score Trajectories.

ABSTRACT: OBJECTIVE:Many 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 DESIGN:Observational 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. METHODS:Using 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. RESULTS:The 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. CONCLUSIONS:On 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.

SUBMITTER: Wong ES

PROVIDER: S-EPMC6258600 | biostudies-literature | 2018 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Identifying Latent Subgroups of High-Risk Patients Using Risk Score Trajectories.

Wong Edwin S ES Yoon Jean J Piegari Rebecca I RI Rosland Ann-Marie M AM Fihn Stephan D SD Chang Evelyn T ET

Journal of general internal medicine 20180917 12

<h4>Objective</h4>Many 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.<h4>Study design</h4>Observational study of 7289 patients discharged from Veterans Health Administration (VA) hospitals during a 1-week period in November 20 ...[more]

PMID: 30225769

Similar Datasets

Project description:INTRODUCTION:Segmentation of the high-need, high-cost (HNHC) population is required for reorganizing care to accommodate person-centered, integrated care delivery. Therefore, we aimed to identify and characterize relevant subgroups of the HNHC population in primary care by using demographic, biomedical, and socioeconomic patient characteristics. METHODS:This was a retrospective cohort study within a Dutch primary care group, with a follow-up period from September 1, 2014 to August 31, 2017. Chronically ill patients were included in the HNHC population if they belonged to the top 10% of care utilizers and/or suffered from multimorbidity and had an above-average care utilization. In a latent class analysis, forty-one patient characteristics were initially used as potential indicators of heterogeneity in HNHC patients' needs. RESULTS:Patient data from 12 602 HNHC patients was used. A 4-class model was considered statistically and clinically superior. The classes were named according to the characteristics that were most dominantly present and distinctive between the classes (i.e. mainly age, household position, and source of income). Class 1 ('older adults living with partner') included 39.3% of patients, class 2 ('older adults living alone') included 25.5% of patients, class 3 ('middle-aged, employed adults with family') included 23.3% of patients, and class 4 ('middle-aged adults with social welfare dependency') included 11.9% of patients. Diabetes was the most common condition in all classes; the second most prevalent condition differed between osteoarthritis in class 1 (21.7%) and 2 (23.8%), asthma in class 3 (25.3%), and mood disorders in class 4 (23.1%). Furthermore, while general practitioner (GP) care utilization increased during the follow-up period in the classes of older adults, it remained relatively stable in the middle-aged classes. CONCLUSIONS:Although the HNHC population is heterogeneous, distinct subgroups with relatively homogeneous patterns of mainly demographic and socioeconomic characteristics can be identified. This calls for tailoring care and increased attention for social determinants of health.

Project description:BackgroundJuvenile delinquents constitute a heterogeneous group, which complicates decision-making based on risk assessment. Various psychosocial factors have been used to define clinically relevant subgroups of juvenile offenders, while neurobiological variables have not yet been integrated in this context. Moreover, translation of neurobiological group differences to individual risk assessment has proven difficult. We aimed to identify clinically relevant subgroups associated with differential youth offending outcomes, based on psychosocial and neurobiological characteristics, and to test whether the resulting model can be used for risk assessment of individual cases.MethodsA group of 223 detained juveniles from juvenile justice institutions was studied. Latent class regression analysis was used to detect subgroups associated with differential offending outcome (recidivism at 12 month follow-up). As a proof of principle, it was tested in a separate group of 76 participants whether individual cases could be assigned to the identified subgroups, using a prototype 'tool' for calculating class membership.ResultsThree subgroups were identified: a 'high risk-externalizing' subgroup, a 'medium risk-adverse environment' subgroup, and a 'low risk-psychopathic traits' subgroup. Within these subgroups, both autonomic nervous system and neuroendocrinological measures added differentially to the prediction of subtypes of reoffending (no, non-violent, violent). The 'tool' for calculating class membership correctly assigned 92.1% of participants to a class and reoffending risk.ConclusionsThe LCRA approach appears to be a useful approach to integrate neurobiological and psychosocial risk factors to identify subgroups with different re-offending risk within juvenile justice institutions. This approach may be useful in the development of a biopsychosocial assessment tool and may eventually help clinicians to assign individuals to those subgroups and subsequently tailor intervention based on their re-offending risk.

Dataset Information

Identifying Latent Subgroups of High-Risk Patients Using Risk Score Trajectories.

Publications

Identifying Latent Subgroups of High-Risk Patients Using Risk Score Trajectories.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure