Who Gets (and Who Should Get) a Serious Illness Conversation in the Hospital? An Analysis of Readmission Risk Score in an Electronic Health Record

Abstract

Serious Illness Conversations (SICs) explore patients’ prognostic awareness, hopes, and worries, and can help establish priorities for their care during and after hospitalization. While identifying patients who benefit from an SIC remains a challenge, this task may be facilitated by use of validated prediction scores available in most commercial electronic health records (EHRs), such as Epic’s Readmission Risk Score (RRS). We identified the RRS on admission for all hospital encounters from October 2018 to August 2019 and measured the area under the receiver operating characteristic (AUROC) curve to determine whether RRS could accurately discriminate post discharge 6-month mortality. For encounters with standardized SIC documentation matched in a 1:3 ratio to controls by sex and age (±5 years), we constructed a multivariable, paired logistic regression model and measured the odds of SIC documentation per every 10% absolute increase in RRS. RRS was predictive of 6-month mortality with acceptable discrimination (AUROC .71) and was significantly associated with SIC documentation (adjusted OR 1.42, 95% CI 1.24-1.63). An RRS >28% used to identify patients with post discharge 6-month mortality had a high specificity (89.0%) and negative predictive value (NPV) (97.0%), but low sensitivity (25.2%) and positive predictive value (PPV) (7.9%). RRS may serve as a practical EHR-based screen to exclude patients not requiring an SIC, thereby leaving a smaller cohort to be further evaluated for SIC needs using other validated tools and clinical assessment.

The Association of Standardized Documentation of Serious Illness Conversations With Healthcare Utilization in Hospitalized Patients: A Propensity Score Matched Cohort Analysis

“…Baseline characteristics ( Table 1 ) demonstrated that prior to matching, encounters with standardized documentation of a SIC were more frequently older, more frequently female, less frequently Hispanic, and more frequently insured by Medicare. Encounters with standardized documentation of a SIC also had higher Elixhauser/Van Walraven comorbidity scores, 29 RRS, 28 and eCART 27 scores and more likely to have had a hospitalization in the year prior to index hospitalization. Following propensity score matching, variables used in matching were balanced between groups, with SMDs of < ±.1 except for DRG weight, which had an SMD of .21.…”

“…Though inpatient palliative care consultation is typically reserved for patients with the highest mortality risk, 30,31 our ability to successfully propensity match in a ratio of 1: 3 suggests that other high-risk patients may have unmet palliative care and end-of-life needs that were not formally elicited and addressed during the hospital encounter. A variety of barriers to conducting SICs and to using the structured EHR module to document the conversation during the hospital encounter may explain this gap, including limited availability of palliative care consultants, 32 lack of standardized indications for palliative care consultation, 33 difficulty with prognostication, 28,34 limited clinician time, suboptimal training or discomfort with leading these conversations and documenting them in the institutionally-approved manner. 1 Achieving a healthcare system that values and prioritizes patient-centeredness requires that hospital-based clinicians have tools, training and resources to identify patients with elevated mortality risk, conduct these conversations and document these conversations so they can readily be carried forward in the medical record.…”

“…We used a greedy matching algorithm with caliper size .25 x standard deviation [SD] for each matching variable, without replacement, i.e., the closest match is found for each case and cannot be used in another match. We included sociodemographic, comorbidity burden, and health care utilization factors thought to be potential confounders in the propensity score (i.e., in a multivariable logistic regression model to predict receipt of a SIC): age, sex, diagnosis related group (DRG) weight, electronic Cardiac Arrest Risk Triage (eCART) score (a validated predictor of death, cardiac arrest, or ICU transfer), 27 Epic’s Risk of Unplanned Readmission (Readmission Risk Score [RRS]) 28 and Elixhauser/Van Walraven 29 comorbidity scores on admission, and a history of one or more hospitalizations in the year prior to the index hospitalization. We calculated standardized mean differences (SMDs) for covariates after matching, with < ±.1 considered indicative of balance.…”

Use of Epic Electronic Health Record System for Health Care Research: Scoping Review

“…Dillon et al 2020 [149], Nolan et al 2020 [150] Adolescent Austrian et al [151], 2018; Brooks et al [152], 2016; Carter et al [153], 2017; Chiu et al [154], 2018; Clendennen et al [155], 2015; Dowling et al [156], 2018; Flatow et al [157], 2015; Isseh et al [158], Adult 2017; Javier-DesLoges et al [159], 2018; King et al [160], 2020; Loo and Taddei [161], 2015; Makam et al [162], 2013; Mathias et al [163], 2010; McDowell et al [164], 2017; Muqri et al [165], 2022; Nikolian et al [166], 2018; Osband et al [167], 2018; Rizk et al [168], 2020; Rodriguez et al [169], 2021; Salem et al [170], 2023; Serna et al [171], 2022; Su et al [172], 2018; Vlashyn et al [173], 2023; Wagner et al [174], 2015; Zazove et al [175], 2017 Altice and Gerow [176], 2020; Bellon et al [177], 2015; Burns et al [178], 2022; Everson et al [179], 2016; Handley et al [180], 2022; Nellis et al [181], 2022; Sahu et al [182], 2022; Simon et al [183], 2023 Adult and pediatric Palestine et al [184], 2018 Not applicable Bush et al [185], 2014; Bush et al [186], 2017; Havrilla et al [187], 2022; Redd et al [188], 2014 Pediatric Beiser et al…”