Intervention to Promote Communication About Goals of Care for Hospitalized Patients With Serious Illness: A Randomized Clinical Trial.

Importance: Discussions about goals of care are important for high-quality palliative care yet are often lacking for hospitalized older patients with serious illness. Objective: To evaluate a communication-priming intervention to promote goals-of-care discussions between clinicians and hospitalized older patients with serious illness. Design, Setting, and Participants: A pragmatic, randomized clinical trial of a clinician-facing communication-priming intervention vs usual care was conducted at 3 US hospitals within 1 health care system, including a university, county, and community hospital. Eligible hospitalized patients were aged 55 years or older with any of the chronic illnesses used by the Dartmouth Atlas project to study end-of-life care or were aged 80 years or older. Patients with documented goals-of-care discussions or a palliative care consultation between hospital admission and eligibility screening were excluded. Randomization occurred between April 2020 and March 2021 and was stratified by study site and history of dementia. Intervention: Physicians and advance practice clinicians who were treating the patients randomized to the intervention received a 1-page, patient-specific intervention (Jumpstart Guide) to prompt and guide goals-of-care discussions. Main Outcomes and Measures: The primary outcome was the proportion of patients with electronic health record-documented goals-of-care discussions within 30 days. There was also an evaluation of whether the effect of the intervention varied by age, sex, history of dementia, minoritized race or ethnicity, or study site. Results: Of 3918 patients screened, 2512 were enrolled (mean age, 71.7 [SD, 10.8] years and 42% were women) and randomized (1255 to the intervention group and 1257 to the usual care group). The patients were American Indian or Alaska Native (1.8%), Asian (12%), Black (13%), Hispanic (6%), Native Hawaiian or Pacific Islander (0.5%), non-Hispanic (93%), and White (70%). The proportion of patients with electronic health record-documented goals-of-care discussions within 30 days was 34.5% (433 of 1255 patients) in the intervention group vs 30.4% (382 of 1257 patients) in the usual care group (hospital- and dementia-adjusted difference, 4.1% [95% CI, 0.4% to 7.8%]). The analyses of the treatment effect modifiers suggested that the intervention had a larger effect size among patients with minoritized race or ethnicity. Among 803 patients with minoritized race or ethnicity, the hospital- and dementia-adjusted proportion with goals-of-care discussions was 10.2% (95% CI, 4.0% to 16.5%) higher in the intervention group than in the usual care group. Among 1641 non-Hispanic White patients, the adjusted proportion with goals-of-care discussions was 1.6% (95% CI, -3.0% to 6.2%) higher in the intervention group than in the usual care group. There was no evidence of differential treatment effects of the intervention on the primary outcome by age, sex, history of dementia, or study site. Conclusions and Relevance: Among hospitalized older adults with serious illness, a pragmatic clinician-facing communication-priming intervention significantly improved documentation of goals-of-care discussions in the electronic health record, with a greater effect size in racially or ethnically minoritized patients. Trial Registration: ClinicalTrials.gov Identifier: NCT04281784.

Clinician Recognition of the Acute Respiratory Distress Syndrome: Risk Factors for Under-Recognition and Trends Over Time.

OBJECTIVES: The acute respiratory distress syndrome is common in critically ill patients. Recognition is crucial because acute respiratory distress syndrome is associated with a high mortality rate, and low tidal volume ventilation improves mortality. However, acute respiratory distress syndrome often goes unrecognized. Risk factors for under-recognition and trends over time have not been fully described. DESIGN: Retrospective chart review of patients with acute respiratory distress syndrome from a prospective cohort study of critically ill patients. For each patient's ICU stay, we searched the chart for terms that indicated that acute respiratory distress syndrome was diagnosed, in the differential diagnosis, or treated with low tidal volume ventilation. SETTING: ICUs at a tertiary hospital at the University of California, San Francisco between 2008 and 2016. PATIENTS: Critically ill patients with acute respiratory distress syndrome. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Acute respiratory distress syndrome was recognized in 70% of patients, and recognition increased from 60% in 2008-2009 to 92% in 2016 (p = 0.004). Use of tidal volumes less than 6.5 mL/kg also increased (p < 0.001) from 20% to 92%. Increased acute respiratory distress syndrome severity (p = 0.01) and vasopressor use (p = 0.04) were associated with greater recognition. Clinician diagnosis of acute respiratory distress syndrome and inclusion of acute respiratory distress syndrome in the differential diagnosis were associated with tidal volumes less than 6.5 mL/kg (51% use of tidal volume ≤ 6.5 mL/kg if acute respiratory distress syndrome recognized vs 15% if not recognized; p = 0.002). Diagnosing acute respiratory distress syndrome was associated with lower tidal volume in multivariate analysis. CONCLUSIONS: Although acute respiratory distress syndrome recognition and low tidal volume ventilation use have increased over time, they remain less than universal. Clinician recognition of acute respiratory distress syndrome is associated with both systemic and respiratory severity of illness and is also associated with use of low tidal volume ventilation.

The Association of Physician Orders for Life-Sustaining Treatment With Intensity of Treatment Among Patients Presenting to the Emergency Department.

STUDY OBJECTIVE: Physician Orders for Life-Sustaining Treatment (POLST) forms are intended to help prevent the provision of unwanted medical interventions among patients with advanced illness or frailty who are approaching the end of life. We seek to evaluate how POLST form completion, treatment limitations, or both influence intensity of treatment among patients who present to the emergency department (ED). METHODS: This was a retrospective cohort study of adults who presented to the ED at an academic medical center in Oregon between April 2015 and October 2016. POLST form completion and treatment limitations were the main exposures. Primary outcome was hospital admission; secondary outcomes included ICU admission and a composite measure of aggressive treatment. RESULTS: A total of 26,128 patients were included; 1,769 (6.8%) had completed POLST forms. Among patients with POLST, 52.1% had full treatment orders, and 6.4% had their forms accessed before admission. POLST form completion was not associated with hospital admission (adjusted odds ratio [aOR]=0.97; 95% confidence interval [CI] 0.84 to 1.12), ICU admission (aOR=0.82; 95% CI 0.55 to 1.22), or aggressive treatment (aOR=1.06; 95% CI 0.75 to 1.51). Compared with POLST forms with full treatment orders, those with treatment limitations were not associated with hospital admission (aOR=1.12; 95% CI 0.92 to 1.37) or aggressive treatment (aOR=0.87; 95% CI 0.5 to 1.52), but were associated with lower odds of ICU admission (aOR=0.31; 95% CI 0.16 to 0.61). CONCLUSION: Among patients presenting to the ED with POLST, the majority of POLST forms had orders for full treatment and were not accessed by emergency providers. These findings may partially explain why we found no association of POLST with treatment intensity. However, treatment limitations on POLST forms were associated with reduced odds of ICU admission. Implementation and accessibility of POLST forms are crucial when considering their effect on the provision of treatment consistent with patients' preferences.

Association of Physician Orders for Life-Sustaining Treatment With ICU Admission Among Patients Hospitalized Near the End of Life.

Importance: Patients with chronic illness frequently use Physician Orders for Life-Sustaining Treatment (POLST) to document treatment limitations. Objectives: To evaluate the association between POLST order for medical interventions and intensive care unit (ICU) admission for patients hospitalized near the end of life. Design, Setting, and Participants: Retrospective cohort study of patients with POLSTs and with chronic illness who died between January 1, 2010, and December 31, 2017, and were hospitalized 6 months or less before death in a 2-hospital academic health care system. Exposures: POLST order for medical interventions ("comfort measures only" vs "limited additional interventions" vs "full treatment"), age, race/ethnicity, education, days from POLST completion to admission, histories of cancer or dementia, and admission for traumatic injury. Main Outcomes and Measures: The primary outcome was the association between POLST order and ICU admission during the last hospitalization of life; the secondary outcome was receipt of a composite of 4 life-sustaining treatments: mechanical ventilation, vasopressors, dialysis, and cardiopulmonary resuscitation. For evaluating factors associated with POLST-discordant care, the outcome was ICU admission contrary to POLST order for medical interventions during the last hospitalization of life. Results: Among 1818 decedents (mean age, 70.8 [SD, 14.7] years; 41% women), 401 (22%) had POLST orders for comfort measures only, 761 (42%) had orders for limited additional interventions, and 656 (36%) had orders for full treatment. ICU admissions occurred in 31% (95% CI, 26%-35%) of patients with comfort-only orders, 46% (95% CI, 42%-49%) with limited-interventions orders, and 62% (95% CI, 58%-66%) with full-treatment orders. One or more life-sustaining treatments were delivered to 14% (95% CI, 11%-17%) of patients with comfort-only orders and to 20% (95% CI, 17%-23%) of patients with limited-interventions orders. Compared with patients with full-treatment POLSTs, those with comfort-only and limited-interventions orders were significantly less likely to receive ICU admission (comfort only: 123/401 [31%] vs 406/656 [62%], aRR, 0.53 [95% CI, 0.45-0.62]; limited interventions: 349/761 [46%] vs 406/656 [62%], aRR, 0.79 [95% CI, 0.71-0.87]). Across patients with comfort-only and limited-interventions POLSTs, 38% (95% CI, 35%-40%) received POLST-discordant care. Patients with cancer were significantly less likely to receive POLST-discordant care than those without cancer (comfort only: 41/181 [23%] vs 80/220 [36%], aRR, 0.60 [95% CI, 0.43-0.85]; limited interventions: 100/321 [31%] vs 215/440 [49%], aRR, 0.63 [95% CI, 0.51-0.78]). Patients with dementia and comfort-only orders were significantly less likely to receive POLST-discordant care than those without dementia (23/111 [21%] vs 98/290 [34%], aRR, 0.44 [95% CI, 0.29-0.67]). Patients admitted for traumatic injury were significantly more likely to receive POLST-discordant care (comfort only: 29/64 [45%] vs 92/337 [27%], aRR, 1.52 [95% CI, 1.08-2.14]; limited interventions: 51/91 [56%] vs 264/670 [39%], aRR, 1.36 [95% CI, 1.09-1.68]). In patients with limited-interventions orders, older age was significantly associated with less POLST-discordant care (aRR, 0.93 per 10 years [95% CI, 0.88-1.00]). Conclusions and Relevance: Among patients with POLSTs and with chronic life-limiting illness who were hospitalized within 6 months of death, treatment-limiting POLSTs were significantly associated with lower rates of ICU admission compared with full-treatment POLSTs. However, 38% of patients with treatment-limiting POLSTs received intensive care that was potentially discordant with their POLST.

Novel Risk Factors for Posttraumatic Stress Disorder Symptoms in Family Members of Acute Respiratory Distress Syndrome Survivors.

OBJECTIVES: Family members of ICU survivors report long-term psychologic symptoms of posttraumatic stress disorder. We describe patient- and family-member risk factors for posttraumatic stress disorder symptoms among family members of survivors of acute respiratory distress syndrome. DESIGN: Prospective cohort study of family members of acute respiratory distress syndrome survivors. SETTING: Single tertiary care center in Seattle, Washington. SUBJECTS: From 2010 to 2015, we assembled an inception cohort of adult acute respiratory distress syndrome survivors who identified family members involved in ICU and post-ICU care. One-hundred sixty-two family members enrolled in the study, corresponding to 120 patients. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Family members were assessed for self-reported psychologic symptoms 6 months after patient discharge using the Posttraumatic Stress Disorder Checklist-Civilian Version, the Patient Health Questionnaire 9-item depression scale, and the Generalized Anxiety Disorder 7-item scale. The primary outcome was posttraumatic stress disorder symptoms, and secondary outcomes were symptoms of depression and anxiety. We used clustered multivariable logistic regression to identify patient- and family-member risk factors for psychologic symptoms. Posttraumatic stress disorder symptoms were present in 31% (95% CI, 24-39%) of family participants. Family member risk factors for posttraumatic stress disorder symptoms included preexisting mental health disorders (adjusted odds ratio, 3.22; 95% CI, 1.42-7.31), recent personal experience of serious physical illness (adjusted odds ratio, 3.07; 95% CI, 1.40-6.75), and female gender (adjusted odds ratio, 5.18; 95% CI, 1.74-15.4). Family members of previously healthy patients (Charlson index of zero) had higher frequency of posttraumatic stress disorder symptoms (adjusted odds ratio, 2.25; 95% CI, 1.06-4.77). Markers of patient illness severity were not associated with family posttraumatic stress disorder symptoms. CONCLUSIONS: The prevalence of long-term posttraumatic stress disorder symptoms among family members of acute respiratory distress syndrome survivors is high. Family members with preexisting mental health disorders, recent experiences of serious physical illness, and family members of previously healthy patients are at increased risk for posttraumatic stress disorder symptoms.

Factors Associated with Costly Hospital Care among Patients with Dementia and Acute Respiratory Failure.

Rationale: Understanding contributors to costly and potentially burdensome care for patients with dementia is of interest to healthcare systems and may facilitate efforts to promote goal-concordant care. Objective: To identify risk factors, in particular whether an early goals-of-care discussion (GOCD) took place, for high-cost hospitalization among patients with dementia and acute respiratory failure. Methods: We conducted an electronic health record-based retrospective cohort study of 298 adults with dementia hospitalized with respiratory failure (receiving ⩾48 h of mechanical ventilation) within an academic healthcare system. We collected demographic and clinical characteristics, including clinical markers of advanced dementia (weight loss, pressure ulcers, hypernatremia, mobility limitations) and intensive care unit (ICU) service (medical, surgical, neurologic). We ascertained whether a GOCD was documented within 48 hours of ICU admission. We used logistic regression to identify patient characteristics associated with high-cost hospitalization measured using the hospital system accounting database and defined as total cost in the top third of the sample (⩾$145,000). We examined a path model that included hospital length of stay as a final mediator between exposure variables and high-cost hospitalization. Results: Patients in the sample had a median age of 71 (IQR, 62-79) years. Approximately half (49%) were admitted to a medical ICU, 29% to a surgical ICU, and 22% to a neurologic ICU. More than half (59%) had a clinical indicator of advanced dementia. A minority (31%) had a GOCD documented within 48 hours of ICU admission; those who did had a 50% lower risk of a high-cost hospitalization (risk ratio, 0.50; 95% confidence interval, 0.2-0.8). Older age, limited English proficiency, and nursing home residence were associated with a lower likelihood of high-cost hospitalization, whereas greater comorbidity burden and admission to a surgical or neurologic ICU compared with a medical ICU were associated with a higher likelihood of high-cost hospitalization. Conclusions: Early GOCDs for patients with dementia and respiratory failure may promote high-value care by ensuring aggressive and costly life support interventions are aligned with patients' goals. Future work should focus on increasing early palliative care delivery for patients with dementia and respiratory failure, in particular in surgical and neurologic ICU settings.

Mechanical Ventilation in Older Adults With Dementia: Opportunities to Promote Goal-Concordant Care.

CONTEXT: Recent studies show increasing use of mechanical ventilation among people living with dementia. There are concerns that this trend may not be driven by patient preferences. OBJECTIVES: To better understand decision-making regarding mechanical ventilation in people living with dementia. METHODS: This was an electronic health record-based retrospective cohort study of older adults with dementia (n = 295) hospitalized at one of two teaching hospitals between 2015 and 2019 who were supported with mechanical ventilation (n = 191) or died without mechanical ventilation (n = 104). Multivariable logistic regression was used to examine associations between patient characteristics and mechanical ventilation use. RESULTS: The median age was 78 years (IQR 71-86), 41% were female, 28% resided in a nursing home, and 58% had clinical markers of advanced dementia (dehydration, weight loss, mobility limitations, or pressure ulcers). Among patients supported with mechanical ventilation, 70% were intubated within 24 hours of presentation, including 31% intubated before hospital arrival. Younger age, higher illness acuity, and absence of a treatment-limiting Physician Orders for Life-Sustaining Treatment document were associated with mechanical ventilation use; nursing home residence and clinical markers of advanced dementia were not. Most patients (89%) had a documented goals of care discussion (GOCD) during hospitalization. CONCLUSION: Future efforts to promote goal-concordant care surrounding mechanical ventilation use for people living with dementia should involve identifying barriers to goal-concordant care in pre-hospital settings, assessing the timeliness of in-hospital GOCD, and developing strategies for in-the-moment crisis communication across settings.

Assessment of Natural Language Processing of Electronic Health Records to Measure Goals-of-Care Discussions as a Clinical Trial Outcome.

Importance: Many clinical trial outcomes are documented in free-text electronic health records (EHRs), making manual data collection costly and infeasible at scale. Natural language processing (NLP) is a promising approach for measuring such outcomes efficiently, but ignoring NLP-related misclassification may lead to underpowered studies. Objective: To evaluate the performance, feasibility, and power implications of using NLP to measure the primary outcome of EHR-documented goals-of-care discussions in a pragmatic randomized clinical trial of a communication intervention. Design, Setting, and Participants: This diagnostic study compared the performance, feasibility, and power implications of measuring EHR-documented goals-of-care discussions using 3 approaches: (1) deep-learning NLP, (2) NLP-screened human abstraction (manual verification of NLP-positive records), and (3) conventional manual abstraction. The study included hospitalized patients aged 55 years or older with serious illness enrolled between April 23, 2020, and March 26, 2021, in a pragmatic randomized clinical trial of a communication intervention in a multihospital US academic health system. Main Outcomes and Measures: Main outcomes were natural language processing performance characteristics, human abstractor-hours, and misclassification-adjusted statistical power of methods of measuring clinician-documented goals-of-care discussions. Performance of NLP was evaluated with receiver operating characteristic (ROC) curves and precision-recall (PR) analyses and examined the effects of misclassification on power using mathematical substitution and Monte Carlo simulation. Results: A total of 2512 trial participants (mean [SD] age, 71.7 [10.8] years; 1456 [58%] female) amassed 44 324 clinical notes during 30-day follow-up. In a validation sample of 159 participants, deep-learning NLP trained on a separate training data set identified patients with documented goals-of-care discussions with moderate accuracy (maximal F1 score, 0.82; area under the ROC curve, 0.924; area under the PR curve, 0.879). Manual abstraction of the outcome from the trial data set would require an estimated 2000 abstractor-hours and would power the trial to detect a risk difference of 5.4% (assuming 33.5% control-arm prevalence, 80% power, and 2-sided α = .05). Measuring the outcome by NLP alone would power the trial to detect a risk difference of 7.6%. Measuring the outcome by NLP-screened human abstraction would require 34.3 abstractor-hours to achieve estimated sensitivity of 92.6% and would power the trial to detect a risk difference of 5.7%. Monte Carlo simulations corroborated misclassification-adjusted power calculations. Conclusions and Relevance: In this diagnostic study, deep-learning NLP and NLP-screened human abstraction had favorable characteristics for measuring an EHR outcome at scale. Adjusted power calculations accurately quantified power loss from NLP-related misclassification, suggesting that incorporation of this approach into the design of studies using NLP would be beneficial.

Predictors of Documented Goals-of-Care Discussion for Hospitalized Patients With Chronic Illness.

CONTEXT: Goals-of-care discussions are important for patient-centered care among hospitalized patients with serious illness. However, there are little data on the occurrence, predictors, and timing of these discussions. OBJECTIVES: To examine the occurrence, predictors, and timing of electronic health record (EHR)-documented goals-of-care discussions for hospitalized patients. METHODS: This retrospective cohort study used natural language processing (NLP) to examine EHR-documented goals-of-care discussions for adults with chronic life-limiting illness or age ≥80 hospitalized 2015-2019. The primary outcome was NLP-identified documentation of a goals-of-care discussion during the index hospitalization. We used multivariable logistic regression to evaluate associations with baseline characteristics. RESULTS: Of 16,262 consecutive, eligible patients without missing data, 5,918 (36.4%) had a documented goals-of-care discussion during hospitalization; approximately 57% of these discussions occurred within 24 hours of admission. In multivariable analysis, documented goals-of-care discussions were more common for women (OR=1.26, 95%CI 1.18-1.36), older patients (OR=1.04 per year, 95%CI 1.03-1.04), and patients with more comorbidities (OR=1.11 per Deyo-Charlson point, 95%CI 1.10-1.13), cancer (OR=1.88, 95%CI 1.72-2.06), dementia (OR=2.60, 95%CI 2.29-2.94), higher acute illness severity (OR=1.12 per National Early Warning Score point, 95%CI 1.11-1.14), or prior advance care planning documents (OR=1.18, 95%CI 1.08-1.30). Documentation of these discussions was less common for racially or ethnically minoritized patients (OR=0.823, 95%CI 0.75-0.90). CONCLUSION: Among hospitalized patients with serious illness, documented goals-of-care discussions identified by NLP were more common among patients with older age and increased burden of acute or chronic illness, and less common among racially or ethnically minoritized patients. This suggests important disparities in goals-of-care discussions.

Improving communication about goals of care for hospitalized patients with serious illness: Study protocol for two complementary randomized trials.

BACKGROUND: Although goals-of-care discussions are important for high-quality palliative care, this communication is often lacking for hospitalized older patients with serious illness. Electronic health records (EHR) provide an opportunity to identify patients who might benefit from these discussions and promote their occurrence, yet prior interventions using the EHR for this purpose are limited. We designed two complementary yet independent randomized trials to examine effectiveness of a communication-priming intervention (Jumpstart) for hospitalized older adults with serious illness. METHODS: We report the protocol for these 2 randomized trials. Trial 1 has two arms, usual care and a clinician-facing Jumpstart, and is a pragmatic trial assessing outcomes with the EHR only (n = 2000). Trial 2 has three arms: usual care, clinician-facing Jumpstart, and clinician- and patient-facing (bi-directional) Jumpstart (n = 600). We hypothesize the clinician-facing Jumpstart will improve outcomes over usual care and the bi-directional Jumpstart will improve outcomes over the clinician-facing Jumpstart and usual care. We use a hybrid effectiveness-implementation design to examine implementation barriers and facilitators. OUTCOMES: For both trials, the primary outcome is EHR documentation of a goals-of-care discussion within 30 days of randomization; additional outcomes include intensity of end-of-life care. Trial 2 also examines patient- or family-reported outcomes assessed by surveys targeting 3-5 days and 4-8 weeks after randomization including quality of goals-of-care communication, receipt of goal-concordant care, and psychological symptoms. CONCLUSIONS: This novel study incorporates two complementary randomized trials and a hybrid effectiveness-implementation approach to improve the quality and value of care for hospitalized older adults with serious illness. CLINICAL TRIALS REGISTRATION: STUDY00007031-A and STUDY00007031-B.

Efficacy of a Communication-Priming Intervention on Documented Goals-of-Care Discussions in Hospitalized Patients With Serious Illness: A Randomized Clinical Trial.

Importance: High-quality goals-of-care communication is critical to delivering goal-concordant, patient-centered care to hospitalized patients with chronic life-limiting illness. However, implementation and documentation of goals-of-care discussions remain important shortcomings in many health systems. Objective: To evaluate the efficacy, feasibility, and acceptability of a patient-facing and clinician-facing communication-priming intervention to promote goals-of-care communication for patients hospitalized with serious illness. Design, Setting, and Participants: This randomized clinical trial enrolled patients from November 6, 2018, to February 18, 2020. The setting was 2 hospitals in an academic health care system in Seattle, Washington. Participants included hospitalized adults with chronic life-limiting illness, aged 65 years or older and with markers of frailty, or aged 80 years or older. Data analysis was performed from August 2020 to August 2021. Intervention: Patients were randomized to usual care with baseline questionnaires (control) vs the Jumpstart communication-priming intervention. Patients or surrogates in the intervention group and their clinicians received patient-specific Jumpstart Guides populated with data from questionnaires and the electronic health records (EHRs) that were designed to prompt and guide a goals-of-care discussion. Main Outcomes and Measures: The primary outcome was EHR documentation of a goals-of-care discussion between randomization and hospital discharge. Additional outcomes included patient-reported or surrogate-reported goals-of-care discussions, patient-reported or surrogate-reported quality of communication, and intervention feasibility and acceptability. Results: Of 428 eligible patients, this study enrolled 150 patients (35% enrollment rate; mean [SD] age, 59.2 [13.6] years; 66 women [44%]; 132 [88%] by patient consent and 18 [12%] by surrogate consent). Seventy-five patients each were randomized to the intervention and control groups. Compared with the control group, the cumulative incidence of EHR-documented goals-of-care discussions between randomization and hospital discharge was higher in the intervention group (16 of 75 patients [21%] vs 6 of 75 patients [8%]; risk difference, 13% [95% CI, 2%-24%]; risk ratio, 2.67 [95% CI, 1.10-6.44]; P = .04). Patient-reported or surrogate-reported goals-of-care discussions did not differ significantly between groups (30 of 66 patients [45%] vs 36 of 66 patients [55%]), although the intrarater consistency of patient and surrogate reports was poor. Patient-rated or surrogate-rated quality of communication did not differ significantly between groups. The intervention was feasible and acceptable to patients, surrogates, and clinicians. Conclusions and Relevance: In this randomized clinical trial, a patient-facing and clinician-facing communication priming intervention for seriously ill, hospitalized patients promoted EHR-documented goals-of-care discussions before discharge with good feasibility and acceptability. Communication-priming interventions should be reexamined in a larger randomized clinical trial to better understand their effectiveness in the inpatient setting. Trial Registration: ClinicalTrials.gov Identifier: NCT03746392.

Mixed-methods evaluation of three natural language processing modeling approaches for measuring documented goals-of-care discussions in the electronic health record.

CONTEXT: Documented goals-of-care discussions are an important quality metric for patients with serious illness. Natural language processing (NLP) is a promising approach for identifying goals-of-care discussions in the electronic health record (EHR). OBJECTIVES: To compare three NLP modeling approaches for identifying EHR documentation of goals-of-care discussions and generate hypotheses about differences in performance. METHODS: We conducted a mixed-methods study to evaluate performance and misclassification for three NLP featurization approaches modeled with regularized logistic regression: bag-of-words (BOW), rule-based, and a hybrid approach. From a prospective cohort of 150 patients hospitalized with serious illness over 2018 to 2020, we collected 4391 inpatient EHR notes; 99 (2.3%) contained documented goals-of-care discussions. We used leave-one-out cross-validation to estimate performance by comparing pooled NLP predictions to human abstractors with receiver-operating-characteristic (ROC) and precision-recall (PR) analyses. We qualitatively examined a purposive sample of 70 NLP-misclassified notes using content analysis to identify linguistic features that allowed us to generate hypotheses underpinning misclassification. RESULTS: All three modeling approaches discriminated between notes with and without goals-of-care discussions (AUCROC: BOW, 0.907; rule-based, 0.948; hybrid, 0.965). Precision and recall were only moderate (precision at 70% recall: BOW, 16.2%; rule-based, 50.4%; hybrid, 49.3%; AUCPR: BOW, 0.505; rule-based, 0.579; hybrid, 0.599). Qualitative analysis revealed patterns underlying performance differences between BOW and rule-based approaches. CONCLUSION: NLP holds promise for identifying EHR-documented goals-of-care discussions. However, the rarity of goals-of-care content in EHR data limits performance. Our findings highlight opportunities to optimize NLP modeling approaches, and support further exploration of different NLP approaches to identify goals-of-care discussions.

The influence of POLST on treatment intensity at the end of life: A systematic review.

BACKGROUND: Despite its widespread implementation, it is unclear whether Physician Orders for Life-Sustaining Treatment (POLST) are safe and improve the delivery of care that patients desire. We sought to systematically review the influence of POLST on treatment intensity among patients with serious illness and/or frailty. METHODS: We performed a systematic review of POLST and similar programs using MEDLINE, EMBASE, CINAHL, Cochrane Central Register of Controlled Trials, Cochrane Database for Systematic Reviews, and PsycINFO, from inception through February 28, 2020. We included adults with serious illness and/or frailty with life expectancy <1 year. Primary outcomes included place of death and receipt of high-intensity treatment (i.e., hospitalization in the last 30- and 90-days of life, ICU admission in the last 30-days of life, and number of care setting transitions in last week of life). RESULTS: Among 104,554 patients across 20 observational studies, 27,090 had POLST. No randomized controlled trials were identified. The mean age of POLST users was 78.7 years, 55.3% were female, and 93.0% were white. The majority of POLST users (55.3%) had orders for comfort measures only. Most studies showed that, compared to full treatment orders on POLST, treatment limitations were associated with decreased in-hospital death and receipt of high-intensity treatment, particularly in pre-hospital settings. However, in the acute care setting, a sizable number of patients likely received POLST-discordant care. The overall strength of evidence was moderate based on eight retrospective cohort studies of good quality that showed a consistent, similar direction of outcomes with moderate-to-large effect sizes. CONCLUSION: We found moderate strength of evidence that treatment limitations on POLST may reduce treatment intensity among patients with serious illness. However, the evidence base is limited and demonstrates potential unintended consequences of POLST. We identify several important knowledge gaps that should be addressed to help maximize benefits and minimize risks of POLST.

Identifying Goals of Care Conversations in the Electronic Health Record Using Natural Language Processing and Machine Learning.

CONTEXT: Goals-of-care discussions are an important quality metric in palliative care. However, goals-of-care discussions are often documented as free text in diverse locations. It is difficult to identify these discussions in the electronic health record (EHR) efficiently. OBJECTIVES: To develop, train, and test an automated approach to identifying goals-of-care discussions in the EHR, using natural language processing (NLP) and machine learning (ML). METHODS: From the electronic health records of an academic health system, we collected a purposive sample of 3183 EHR notes (1435 inpatient notes and 1748 outpatient notes) from 1426 patients with serious illness over 2008-2016, and manually reviewed each note for documentation of goals-of-care discussions. Separately, we developed a program to identify notes containing documentation of goals-of-care discussions using NLP and supervised ML. We estimated the performance characteristics of the NLP/ML program across 100 pairs of randomly partitioned training and test sets. We repeated these methods for inpatient-only and outpatient-only subsets. RESULTS: Of 3183 notes, 689 contained documentation of goals-of-care discussions. The mean sensitivity of the NLP/ML program was 82.3% (SD 3.2%), and the mean specificity was 97.4% (SD 0.7%). NLP/ML results had a median positive likelihood ratio of 32.2 (IQR 27.5-39.2) and a median negative likelihood ratio of 0.18 (IQR 0.16-0.20). Performance was better in inpatient-only samples than outpatient-only samples. CONCLUSION: Using NLP and ML techniques, we developed a novel approach to identifying goals-of-care discussions in the EHR. NLP and ML represent a potential approach toward measuring goals-of-care discussions as a research outcome and quality metric.

Conflicting Orders in Physician Orders for Life-Sustaining Treatment Forms.

BACKGROUND/OBJECTIVES: Many older persons with chronic illness use Physician Orders for Life-Sustaining Treatment (POLST) to document portable medical orders for emergency care. However, some POLSTs contain combinations of orders that do not translate into a cohesive care plan (eg, cardiopulmonary resuscitation [CPR] without intensive care, or intensive care without antibiotics). This study characterizes the prevalence and predictors of POLSTs with conflicting orders. DESIGN: Retrospective cohort study. SETTING: Large academic health system. PARTICIPANTS: A total of 3,123 POLST users with chronic life-limiting illness who died between 2010 and 2015 (mean age = 69.7 years). MEASUREMENTS: In a retrospective review of all POLSTs in participants' electronic health records, we describe the prevalence of POLSTs with conflicting orders for cardiac arrest and medical interventions, and use clustered logistic regression to evaluate potential predictors of conflicting orders. We also examine the prevalence of conflicts between POLST orders for antibiotics and artificial nutrition with orders for cardiac arrest or medical interventions. RESULTS: Among 3,924 complete POLSTs belonging to 3,123 decedents, 209 (5.3%) POLSTs contained orders to "attempt CPR" paired with orders for "limited interventions" or "comfort measures only"; 745/3169 (23.5%) POLSTs paired orders to restrict antibiotics with orders to deliver non-comfort-only care; and, 170/3098 (5.5%) POLSTs paired orders to withhold artificial nutrition with orders to deliver CPR or intensive care. Among POLSTs with orders to avoid intensive care, orders to attempt CPR were more likely to be present in POLSTs completed earlier in the patient's illness course (adjusted odds ratio = 1.27 per twofold increase in days from POLST to death; 95% confidence interval = 1.18-1.36; P < .001). CONCLUSION: Although most POLSTs are actionable by clinicians, 5% had conflicting orders for cardiac arrest and medical interventions, and 24% had one or more conflicts between orders for cardiac arrest, medical interventions, antibiotics, and artificial nutrition. These conflicting orders make implementation of POLST challenging for clinicians in acute care settings.

Pragmatic methods to avoid intensive care unit admission when it does not align with patient and family goals.

For patients with chronic, life-limiting illnesses, admission to the intensive care unit (ICU) near the end of life might not improve patient outcomes or be consistent with patient and family values, goals, and preferences. In this context, advance care planning and palliative care interventions designed to clarify patients' values, goals, and preferences have the potential to reduce provision of high-intensity interventions that are unwanted or non-beneficial. In this Series paper, we have assessed interventions that are effective at helping patients with chronic, life-limiting illnesses to avoid an unwanted ICU admission. The evidence found was largely from observational studies, with considerable heterogeneity in populations, methods, and types of interventions. Results from randomised trials of interventions to improve communication about goals of care are scarce, of variable quality, and mixed. Although observational studies show that advance care planning and palliative care interventions are associated with a reduced number of ICU admissions at the end of life, causality has not been well established. Using the available evidence we suggest recommendations to help to avoid ICU admission when it does not align with patient and family values, goals, and preferences and conclude with future directions for research.

Toward Understanding the Relationship Between Prioritized Values and Preferences for Cardiopulmonary Resuscitation Among Seriously Ill Adults.

CONTEXT: Prioritizing among potentially conflicting end-of-life values may help patients discriminate among treatments and allow clinicians to align treatments with values. OBJECTIVES: To investigate end-of-life values that patients prioritize when facing explicit trade-offs and identify predictors of patients whose values and treatment preferences seem inconsistent. METHODS: Analysis of surveys from a multi-center cluster-randomized trial of patients with serious illness. Respondents prioritized end-of-life values and identified cardiopulmonary resuscitation (CPR) preferences in two health states. RESULTS: Of 535 patients, 60% prioritized relief of discomfort over extending life, 17% prioritized extending life over relief of discomfort, and 23% were unsure. Patients prioritizing extending life were most likely to prefer CPR, with 93% preferring CPR in current health and 67% preferring CPR if dependent on others, compared with 69% and 21%, respectively, for patients prioritizing relief of discomfort, and 78% and 33%, respectively, for patients unsure of their prioritized value (P < 0.001 for all comparisons). Among patients prioritizing relief of discomfort, preference for CPR in current health was less likely among older patients (odds ratio 0.958 per year; 95% CI 0.935, 0.981) and more likely with better self-perceived health (odds ratio 1.402 per level of health; 95% CI 1.090, 1.804). CONCLUSION: Clinicians face challenges as they clarify patient values and align treatments with values. Patients' values predicted CPR preferences, but a substantial proportion of patients expressed CPR preferences that appeared potentially inconsistent with their primary value. Clinicians should question assumptions about relationships between values and CPR preferences. Further research is needed to identify ways to use values to guide treatment decisions.

Predictors of Advance Care Planning Documentation in Patients With Underlying Chronic Illness Who Died of Traumatic Injury.

CONTEXT: Advance care planning (ACP) is difficult in the setting of a life-threatening trauma but may be equally important in this context, especially with increasing numbers of trauma victims being elderly or having multimorbidity. OBJECTIVES: Identify predictors of absent ACP documentation in the electronic health records of patients with underlying chronic illness who died of traumatic injury. METHODS: We used death records and electronic health records to identify decedents with chronic life-limiting illness who died of traumatic injury between 2010 and 2015 and to evaluate factors associated with documentation of living wills, durable powers of attorney, or physician orders for life-sustaining treatment. RESULTS: Only 22% of decedents had ACP documentation at time of injury. Among those without preinjury ACP documentation, 4% completed ACP documentation after injury. In multipredictor analyses, patients were less likely to have ACP documentation at the time of injury if they were younger (P < 0.001), had fewer chronic illnesses (P = 0.002), and had fewer nonsurgical hospitalizations (P = 0.042) in the year before injury. Among patients without ACP documentation before injury, those with fewer postinjury nonsurgical hospitalizations were less likely to complete ACP documentation after injury (P = 0.019). CONCLUSIONS: Our findings suggest that patient characteristics play an important role in the completion of ACP among patients with chronic life-limiting illness and who died from sudden severe injury. Interventions to improve ACP completion by patients with serious chronic conditions have the potential for increasing goal-concordant care in the event of traumatic injury.