Dataset Information

Use of Electronic Health Record Data to Estimate the Probability of Alcohol Withdrawal Syndrome in a National Cohort of Hospitalized Veterans.

ABSTRACT:

Objectives

Inpatient alcohol withdrawal syndrome (AWS) is common and early treatment improves outcomes, but no prior study has used electronic health record (EHR) data, available at admission, to predict the probability of inpatient AWS. This study estimated the probability of inpatient AWS using prior-year EHR data, hypothesizing that documented alcohol use disorder (AUD) and AWS would be strongly associated with inpatient AWS while exploring associations with other patient characteristics.

Methods

The study investigated patients hospitalized ≥24 hours on medical services in the Veterans Health Administration during 2013 using EHR data extracted from the Veterans Health Administration Corporate Data Warehouse. ICD-9-CM diagnosis code, demographic, and healthcare utilization data documented in the year before admission defined prior-year AUD, AWS, and other factors associated with inpatient AWS. The primary outcome, inpatient AWS, was defined by inpatient ICD-9-CM codes.

Results

The unadjusted probability of AWS was 5.0% (95% CI 4.5%-5.4%) among 209,151 medical inpatients overall, 26.4% (95% CI 24.4%-28.4%) among those with prior-year AUD, and 62.5% (95% CI 35.2%-39.7%) among those with prior-year AWS. Of those with AWS, 86% had documented prior-year AUD and/or AWS. Other patient characteristics associated with increased probability of inpatient AWS (P < 0.001) were: male sex, single relationship status, homelessness, seizure, and cirrhosis.

Conclusions

Although inpatient providers often use history to predict AWS, this is the first study in hospitalized patients to inform and validate this practice, showing that prior-year diagnosis of AUD and/or AWS in particular, can identify the majority of inpatients who should be monitored for AWS.

SUBMITTER: Steel TL

PROVIDER: S-EPMC8262104 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Importance:Although severe alcohol withdrawal syndrome (SAWS) is associated with substantial morbidity and mortality, most at-risk patients will not develop this syndrome. Predicting its occurrence is important because the mortality rate is high when untreated. Objective:To assess the accuracy and predictive value of symptoms and signs for identifying hospitalized patients at risk of SAWS, defined as delirium tremens, withdrawal seizure, or clinically diagnosed severe withdrawal. Data Sources:MEDLINE and EMBASE (1946-January 2018) were searched for articles investigating symptoms and signs predictive of SAWS in adults. Reference lists of retrieved articles were also searched. Study Selection:Original studies that were included compared symptoms, signs, and risk assessment tools among patients who developed SAWS and patients who did not. Data Extraction and Synthesis:Data were extracted and used to calculate likelihood ratios (LRs), sensitivity, and specificity. A meta-analysis was performed to calculate summary LR. Results:Of 530 identified studies, 14 high-quality studies that included 71?295 patients and 1355 relevant cases of SAWS (1051 cases), seizure (53 cases), or delirium tremens (251 cases) were analyzed. A history of delirium tremens (LR, 2.9 [95% CI 1.7-5.2]) and baseline systolic blood pressure 140 mm Hg or higher (LR, 1.7 [95% CI, 1.3-2.3) were associated with an increased likelihood of SAWS. No single symptom or sign was associated with exclusion of SAWS. Six high-quality studies evaluated combinations of clinical findings and were useful for identifying patients in acute care facilities at high risk of developing SAWS. Of these combinations, the Prediction of Alcohol Withdrawal Severity Scale (PAWSS) was most useful, with an LR of 174 (95% CI, 43-696; specificity, 0.93) when patients had 4 or more individual findings and an LR of 0.07 (95% CI, 0.02-0.26; sensitivity, 0.99) when there were 3 or fewer findings. Conclusions and Relevance:Assessment tools that use a combination of symptoms and signs are useful for identifying patients at risk of developing severe alcohol withdrawal syndrome. Most studies of these tools were not fully validated, limiting their generalizability.

Project description:BackgroundAlcohol withdrawal syndrome (AWS) occurs when alcohol-dependent individuals abruptly reduce or stop drinking. Hospitalized alcohol-dependent patients are at risk. Hospitals need a validated screening tool to assess withdrawal risk, but no validated tools are currently available.ObjectiveTo examine the admission Alcohol Use Disorders Identification Test-(Piccinelli) Consumption (AUDIT-PC) ability to predict the subsequent development of AWS among hospitalized medical-surgical patients admitted to a non-intensive care setting.DesignRetrospective case–control study of patients discharged from the hospital with a diagnosis of AWS. All patients with AWS were classified as presenting with AWS or developing AWS later during admission. Patients admitted to an intensive care setting and those missing AUDIT-PC scores were excluded from analysis. A hierarchical (by hospital unit) logistic regression was performed and receiver-operating characteristics were examined on those developing AWS after admission and randomly selected controls. Because those diagnosing AWS were not blinded to the AUDIT-PC scores, a sensitivity analysis was performed.ParticipantsThe study cohort included all patients age ≥18 years admitted to any medical or surgical units in a single health care system from 6 October 2009 to 7 October 2010.Key resultsAfter exclusions, 414 patients were identified with AWS. The 223 (53.9 %) who developed AWS after admission were compared to 466 randomly selected controls without AWS. An AUDIT-PC score ≥4 at admission provides 91.0 % sensitivity and 89.7 % specificity (AUC=0.95; 95 % CI, 0.94–0.97) for AWS, and maximizes the correct classification while resulting in 17 false positives for every true positive identified. Performance remained excellent on sensitivity analysis (AUC=0.92; 95 % CI, 0.90–0.93). Increasing AUDIT-PC scores were associated with an increased risk of AWS (OR=1.68, 95 % CI 1.55–1.82, p<0.001).ConclusionsThe admission AUDIT-PC score is an excellent discriminator of AWS and could be an important component of future clinical prediction rules. Calibration and further validation on a large prospectivecohort is indicated.

Project description:ImportanceCurrent methods for identifying hospitalized patients at increased risk of delirium require nurse-administered questionnaires with moderate accuracy.ObjectiveTo develop and validate a machine learning model that predicts incident delirium risk based on electronic health data available on admission.Design, setting, and participantsRetrospective cohort study evaluating 5 machine learning algorithms to predict delirium using 796 clinical variables identified by an expert panel as relevant to delirium prediction and consistently available in electronic health records within 24 hours of admission. The training set comprised 14 227 adult patients with non-intensive care unit hospital stays and no delirium on admission who were discharged between January 1, 2016, and August 31, 2017, from UCSF Health, a large academic health institution. The test set comprised 3996 patients with hospital stays who were discharged between August 1, 2017, and November 30, 2017.ExposuresPatient demographic characteristics, diagnoses, nursing records, laboratory results, and medications available in electronic health records during hospitalization.Main outcomes and measuresDelirium was defined as a positive Nursing Delirium Screening Scale or Confusion Assessment Method for the Intensive Care Unit score. Models were assessed using the area under the receiver operating characteristic curve (AUC) and compared against the 4-point scoring system AWOL (age >79 years, failure to spell world backward, disorientation to place, and higher nurse-rated illness severity), a validated delirium risk-assessment tool routinely administered in this cohort.ResultsThe training set included 14 227 patients (5113 [35.9%] aged >64 years; 7335 [51.6%] female; 687 [4.8%] with delirium), and the test set included 3996 patients (1491 [37.3%] aged >64 years; 1966 [49.2%] female; 191 [4.8%] with delirium). In total, the analysis included 18 223 hospital admissions (6604 [36.2%] aged >64 years; 9301 [51.0%] female; 878 [4.8%] with delirium). The AWOL system achieved a baseline AUC of 0.678. The gradient boosting machine model performed best, with an AUC of 0.855. Setting specificity at 90%, the model had a 59.7% (95% CI, 52.4%-66.7%) sensitivity, 23.1% (95% CI, 20.5%-25.9%) positive predictive value, 97.8% (95% CI, 97.4%-98.1%) negative predictive value, and a number needed to screen of 4.8. Penalized logistic regression and random forest models also performed well, with AUCs of 0.854 and 0.848, respectively.Conclusions and relevanceMachine learning can be used to estimate hospital-acquired delirium risk using electronic health record data available within 24 hours of hospital admission. Such a model may allow more precise targeting of delirium prevention resources without increasing the burden on health care professionals.