Dataset Information

Predicting emergency department visits and hospitalizations for patients with heart failure in home healthcare using a time series risk model.

ABSTRACT:

Objectives

Little is known about proactive risk assessment concerning emergency department (ED) visits and hospitalizations in patients with heart failure (HF) who receive home healthcare (HHC) services. This study developed a time series risk model for predicting ED visits and hospitalizations in patients with HF using longitudinal electronic health record data. We also explored which data sources yield the best-performing models over various time windows.

Materials and methods

We used data collected from 9362 patients from a large HHC agency. We iteratively developed risk models using both structured (eg, standard assessment tools, vital signs, visit characteristics) and unstructured data (eg, clinical notes). Seven specific sets of variables included: (1) the Outcome and Assessment Information Set, (2) vital signs, (3) visit characteristics, (4) rule-based natural language processing-derived variables, (5) term frequency-inverse document frequency variables, (6) Bio-Clinical Bidirectional Encoder Representations from Transformers variables, and (7) topic modeling. Risk models were developed for 18 time windows (1-15, 30, 45, and 60 days) before an ED visit or hospitalization. Risk prediction performances were compared using recall, precision, accuracy, F1, and area under the receiver operating curve (AUC).

Results

The best-performing model was built using a combination of all 7 sets of variables and the time window of 4 days before an ED visit or hospitalization (AUC = 0.89 and F1 = 0.69).

Discussion and conclusion

This prediction model suggests that HHC clinicians can identify patients with HF at risk for visiting the ED or hospitalization within 4 days before the event, allowing for earlier targeted interventions.

SUBMITTER: Chae S

PROVIDER: S-EPMC10531127 | biostudies-literature | 2023 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Predicting emergency department visits and hospitalizations for patients with heart failure in home healthcare using a time series risk model.

Chae Sena S Davoudi Anahita A Song Jiyoun J Evans Lauren L Hobensack Mollie M Bowles Kathryn H KH McDonald Margaret V MV Barrón Yolanda Y Rossetti Sarah Collins SC Cato Kenrick K Sridharan Sridevi S Topaz Maxim M

Journal of the American Medical Informatics Association : JAMIA 20230901 10

<h4>Objectives</h4>Little is known about proactive risk assessment concerning emergency department (ED) visits and hospitalizations in patients with heart failure (HF) who receive home healthcare (HHC) services. This study developed a time series risk model for predicting ED visits and hospitalizations in patients with HF using longitudinal electronic health record data. We also explored which data sources yield the best-performing models over various time windows.<h4>Materials and methods</h4>W ...[more]

PMID: 37433577

Similar Datasets

Project description:ImportanceSevere maternal morbidity (SMM) can have long-term health consequences for the affected mother. The association between SMM and future maternal mental health conditions has not been well studied.ObjectiveTo assess the association between SMM in the first recorded birth and the risk of hospitalization or emergency department (ED) visits for a mental health condition over a 13-year period.Design, setting, and participantsThis population-based retrospective cohort study used data from postpartum individuals aged 18 to 55 years with a first hospital delivery between 2008 and 2021 in 11 provinces and territories in Canada, except Québec. Data were analyzed from January to June 2023.ExposureSMM, defined as a composite of conditions, such as septic shock, severe preeclampsia or eclampsia, severe hemorrhage with intervention, or other complications, occurring after 20 weeks' gestation and up to 42 days after a first delivery.Main outcomes and measuresThe main outcome was a hospitalization or ED visit for a mental health condition, including mood and anxiety disorders, substance use, schizophrenia, and other psychotic disorder, or suicidality or self-harm event, arising at least 43 days after the first birth hospitalization. Cox regression models generated hazard ratios with 95% CIs, adjusted for baseline maternal comorbidities, maternal age at delivery, income quintile, type of residence, hospital type, and delivery year.ResultsOf 2 026 594 individuals with a first hospital delivery, 1 579 392 individuals (mean [SD] age, 30.0 [5.4] years) had complete ED and hospital records and were included in analyses; among these, 35 825 individuals (2.3%) had SMM. Compared with individuals without SMM, those with SMM were older (mean [SD] age, 29.9 [5.4] years vs 30.7 [6.0] years), were more likely to deliver in a teaching tertiary care hospital (40.8% vs 51.1%), and to have preexisting conditions (eg, ≥2 conditions: 1.2% vs 5.3%), gestational diabetes (8.2% vs 11.7%), stillbirth (0.5% vs 1.6%), preterm birth (7.7% vs 25.0%), or cesarean delivery (31.0% vs 54.3%). After a median (IQR) duration of 2.6 (1.3-6.4) years, 1287 (96.1 per 10 000) individuals with SMM had a mental health hospitalization or ED visit, compared with 41 779 (73.2 per 10 000) individuals without SMM (adjusted hazard ratio, 1.26 [95% CI, 1.19-1.34]).Conclusions and relevanceIn this cohort study of postpartum individuals with and without SMM in pregnancy and delivery, there was an increased risk of mental health hospitalizations or ED visits up to 13 years after a delivery complicated by SMM. Enhanced surveillance and provision of postpartum mental health resources may be especially important after SMM.

Project description:ImportanceCannabis use disorders (CUD) are associated with adverse health effects, including mental disorders and motor vehicle collision-related injuries. However, little is known about whether CUDs are associated with increased mortality risk.ObjectiveTo examine whether individuals receiving incident hospital-based care (an emergency department visit or hospitalization) for a CUD is associated with increased risk of death.Design, setting, and participantsThis population-based retrospective cohort study included all individuals aged 15 to 105 years living in Ontario, Canada, between 2006 and 2021 (n = 11 622 571 individuals). Overall and cause-specific mortality were compared between individuals with incident hospital-based CUD care and age- and sex-matched members of the general population or individuals with hospital-based care for other substance use disorders using cause-specific hazard models adjusted for comorbid mental health, substance use, and chronic health conditions. Statistical analysis was performed from September to December 2024.ExposureIncident hospital-based CUD care.Main outcomes and measuresOverall and cause-specific mortality identified using vital statistics.ResultsThe matched analysis included 527 972 individuals (mean [SD] age, 29.9 [13.6] years; 330 034 [62.5%] female) with a median (IQR) follow-up of 5 (3-9) years; 106 994 had incident CUD. Within 5 years of incident hospital-based CUD care, 3770 individuals (3.5%) died compared with 3770 (0.6%) of matched general population members. After adjusting for comorbid conditions, individuals with incident hospital-based CUD care were at increased risk of death relative to the general population (adjusted hazard ratio [aHR], 2.79 [95% CI, 2.62-2.97]). Individuals with hospital-based CUD care were at increased risk of all investigated types of death and particularly elevated risk of death by suicide (aHR, 9.70 [95% CI, 6.04-15.57]), trauma (aHR, 4.55 [95% CI, 3.55-5.82]), opioid poisoning (aHR, 5.03 [95% CI, 2.86-8.84]), other drug poisonings (aHR, 4.56 [95% CI, 3.11-6.68]), and lung cancer (aHR, 3.81 [95% CI, 2.39-6.07]) relative to the general population. Compared with an individual with hospital-based care for CUD, individuals with hospital-based care for alcohol (aHR, 1.30 [95% CI, 1.26-1.34]), stimulants (aHR, 1.69 [95% CI, 1.62-1.75]), and opioids (aHR, 2.19 [95% CI, 2.10-2.27]) were at relatively increased risk of death within 5 years.Conclusions and relevanceIn this cohort study of all residents of Ontario, Canada, individuals with incident hospital-based CUD care were at markedly increased risk of death compared with the general population. These findings suggest important clinical and policy implications, given global trends toward cannabis legalization and market commercialization accompanied by increasing cannabis use and CUDs.

Project description:Vaso-occlusive episodes (VOEs) are a hallmark of sickle cell disease (SCD), and account for >90% of health care encounters for this patient population. The Cooperative Study of Sickle Cell Disease, a large study enrolling >3000 patients, showed that the majority of SCD patients (80%) experienced 0-3 major pain crises/year. Only a small minority (~5%) experienced ≥6 VOEs/year. Our study sought to further understand this difference in VOE frequency between SCD patients. We analyzed 25 patients (13M/12F, mean age of 28.8) with ≥6 ED visits or hospitalizations/year (high utilizers), and compared these with 9 patients (6M/3F, mean age of 37.6) who had ≤2 ED visits or hospitalizations/year (low utilizers). All subjects were given a demographic survey along with questionnaires for depression, anxiety, and Health Locus of Control. Each subject then underwent quantitative sensory testing (QST) with three different modalities: pressure pain sensitivity, heat and cold sensitivity, and Von Frey monofilament testing. Laboratory and clinical data were collected through subjects' medical records. CBC and chemistry analysis showed high utilizers had higher WBC (p<0.01), ANC (p<0.01), total bilirubin (p = 0.02), and lower MCV (p = 0.03). Opioid use (morphine equivalents) over the past 6 months was significantly higher in the high utilizer group (12125.7 mg vs 2423.1 mg, p = 0.005). QST results showed lower pressure pain threshold at the ulna (224.4 KPa vs 338.9 KPa, p = 0.04) in the high utilizer group. High utilizers also had higher anxiety (9.0 vs 4.6, p = 0.04) and depression scores (10.0 vs 6.0, p = 0.051). While the low utilizer group had higher education levels with more associate and bachelor degrees (p = 0.009), there was no difference in income or employment. These data show that many biological and psychosocial factors contribute to high health care utilization in SCD. A multi-disciplinary and multi-faceted approach will be required to address this complex problem.

Project description:BackgroundEmergency department (ED) visits show a high volatility over time. Therefore, EDs are likely to be crowded at peak-volume moments. ED crowding is a widely reported problem with negative consequences for patients as well as staff. Previous studies on the predictive value of weather variables on ED visits show conflicting results. Also, no such studies were performed in the Netherlands. Therefore, we evaluated prediction models for the number of ED visits in our large the Netherlands teaching hospital based on calendar and weather variables as potential predictors.MethodsData on all ED visits from June 2016 until December 31, 2019, were extracted. The 2016-2018 data were used as training set, the 2019 data as test set. Weather data were extracted from three publicly available datasets from the Royal Netherlands Meteorological Institute. Weather observations in proximity of the hospital were used to predict the weather in the hospital's catchment area by applying the inverse distance weighting interpolation method. The predictability of daily ED visits was examined by creating linear prediction models using stepwise selection; the mean absolute percentage error (MAPE) was used as measurement of fit.ResultsThe number of daily ED visits shows a positive time trend and a large impact of calendar events (higher on Mondays and Fridays, lower on Saturdays and Sundays, higher at special times such as carnival, lower in holidays falling on Monday through Saturday, and summer vacation). The weather itself was a better predictor than weather volatility, but only showed a small effect; the calendar-only prediction model had very similar coefficients to the calendar+weather model for the days of the week, time trend, and special time periods (both MAPE's were 8.7%).ConclusionsBecause of this similar performance, and the inaccuracy caused by weather forecasts, we decided the calendar-only model would be most useful in our hospital; it can probably be transferred for use in EDs of the same size and in a similar region. However, the variability in ED visits is considerable. Therefore, one should always anticipate potential unforeseen spikes and dips in ED visits that are not shown by the model.

Project description:AimsTo identify clusters of risk factors in home health care and determine if the clusters are associated with hospitalizations or emergency department visits.DesignA retrospective cohort study.MethodsThis study included 61,454 patients pertaining to 79,079 episodes receiving home health care between 2015 and 2017 from one of the largest home health care organizations in the United States. Potential risk factors were extracted from structured data and unstructured clinical notes analysed by natural language processing. A K-means cluster analysis was conducted. Kaplan-Meier analysis was conducted to identify the association between clusters and hospitalizations or emergency department visits during home health care.ResultsA total of 11.6% of home health episodes resulted in hospitalizations or emergency department visits. Risk factors formed three clusters. Cluster 1 is characterized by a combination of risk factors related to "impaired physical comfort with pain," defined as situations where patients may experience increased pain. Cluster 2 is characterized by "high comorbidity burden" defined as multiple comorbidities or other risks for hospitalization (e.g., prior falls). Cluster 3 is characterized by "impaired cognitive/psychological and skin integrity" including dementia or skin ulcer. Compared to Cluster 1, the risk of hospitalizations or emergency department visits increased by 1.95 times for Cluster 2 and by 2.12 times for Cluster 3 (all p < .001).ConclusionRisk factors were clustered into three types describing distinct characteristics for hospitalizations or emergency department visits. Different combinations of risk factors affected the likelihood of these negative outcomes.ImpactCluster-based risk prediction models could be integrated into early warning systems to identify patients at risk for hospitalizations or emergency department visits leading to more timely, patient-centred care, ultimately preventing these events.Patient or public contributionThere was no involvement of patients in developing the research question, determining the outcome measures, or implementing the study.

Dataset Information

Predicting emergency department visits and hospitalizations for patients with heart failure in home healthcare using a time series risk model.

Objectives

Materials and methods

Results

Discussion and conclusion

Publications

Predicting emergency department visits and hospitalizations for patients with heart failure in home healthcare using a time series risk model.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets