Caring for Unbefriended Older Adults and Adult Orphans: A Clinician Survey

Objectives: Unbefriended older adults are those who lack the capacity to make medical decisions and do not have a completed advance directive that can guide treatment decisions or a surrogate decision maker. Adult orphans are those who retain medical decision-making capacity but are at risk of becoming unbefriended due to lack of a completed advance health care directive and lack of a surrogate decision maker. In a follow-up to the 2016 American Geriatrics Society (AGS) position statement on unbefriended older adults, we examined clinicians' experiences in caring for unbefriended older adults and adult orphans.Methods: Clinicians recruited through the AGS (N = 122) completed an online survey about their experiences with unbefriended older adults regarding the perceived frequency of contact, clinical concerns, practice strategies, and terminology; and also with adult orphans regarding the perceived frequency of contact, methods of identification, and terminology.Results: Almost all inpatient (95.9%) and outpatient (86.4%) clinicians in this sample encounter unbefriended older adults at least quarterly and 92.2% of outpatient clinicians encounter adult orphans at least quarterly. Concerns about safety (95.9%), medication self-management (90.4%), and advance care planning (86.3%) bring unbefriended older adults to outpatient clinicians' attention "sometimes" to "frequently." Prolonged hospital stays (87.7%) and delays in transitioning to end-of-life care (85.7%) bring unbefriended older adults to inpatient clinicians' attention "sometimes" to "frequently." Clinicians apply a wide range of practice strategies to these populations. Participants suggested alternative terminology to replace "unbefriended" and "adult orphan."Conclusions: This study suggests that unbefriended older adults are frequently encountered in geriatrics practice, both in the inpatient and outpatient settings, and that there is widespread awareness of adult orphans in the outpatient setting. Clinicians' awareness of both groups suggests avenues for intervention and prevention.Clinical Implications: Health care professionals in geriatric settings will likely encounter older adults in need of advocates. Clinicians, attorneys, and policymakers should collaborate to improve early detection and to meet the needs of this vulnerable population.

Factors Associated With Access to and Receipt of Liver Transplantation in Veterans With End-stage Liver Disease

IMPORTANCE

Organ scarcity means few patients with advanced liver disease undergo a transplant, making equitable distribution all the more crucial. Disparities may arise at any stage in the complex process leading up to this curative therapy.

OBJECTIVE

To examine the rate of and factors associated with referral, wait-listing, and receipt of liver allografts.

DESIGN, SETTING, AND PARTICIPANTS

This retrospective cohort study used linked data from comprehensive electronic medical records and the United Network of Organ Sharing. Adult patients with cirrhosis and a Model for End-Stage Liver Disease with addition of sodium score of at least 15 points between October 1, 2011, and December 31, 2017, were included in the study. Patients were from 129 hospitals in the integrated, US Department of Veterans Affairs health care system and were followed up through December 31, 2018. Statistical analyses were performed from April 28, 2020, to January 31, 2021.

EXPOSURES

Sociodemographic (eg, age, insurance, income), clinical (eg, liver disease etiology, severity, comorbidity), and health care facility (eg, complexity, rural or urban, presence of a liver transplant program) factors were evaluated.

MAIN OUTCOMES AND MEASURES

Referral, wait-listing, and liver transplantation.

RESULTS

Of the 34 494 patients with cirrhosis (mean [SD] age, 62 [7.7] years; 33 560 men [97.29%]; 22 509 White patients [65.25%]), 1534 (4.45%) were referred, 1035 (3.00%) were wait-listed, and 549 (1.59%) underwent a liver transplant within 3 years of meeting clinical criteria for transplantation. Patient age of 70 years or older was associated with lower rates of referral (hazard ratio [HR], 0.09; 95% CI, 0.06-0.13), wait-listing (HR, 0.07; 95% CI, 0.04-0.12), and transplant (HR, 0.08; 95% CI, 0.04-0.16). Alcohol etiology for liver cirrhosis was associated with lower rates of referral (HR, 0.38; 95% CI, 0.33-0.44), wait-listing (HR, 0.32; 95% CI, 0.27-0.38), and transplant (HR, 0.30; 95% CI, 0.23-0.37). In addition, comorbidity (none vs >1 comorbidity) was associated with lower rates of referral (HR, 0.47; 95% CI, 0.40-0.56), wait-listing (HR, 0.38; 95% CI, 0.31-0.46), and transplant (HR, 0.28; 95% CI, 0.21-0.38). African American patients were less likely to be referred (HR, 0.82; 95% CI, 0.70-0.95) and wait-listed (HR, 0.73; 95% CI, 0.61-0.88). Patients with lower annual income and those seen in facilities in the West were less likely to be referred (HR, 0.70; 95% CI, 0.53-0.93), wait-listed (HR, 0.48; 95% CI, 0.36-0.64), or undergo a transplant (HR, 0.50; 95% CI, 0.34-0.74). In a review of the medical records for 333 patients who had limited comorbidity but were not referred, organ transplant was considered as a potential option in 176 (52.85%). When documented, medical and psychosocial barriers explained most of the deficits in referral.

CONCLUSIONS AND RELEVANCE

In this cohort study, few patients with advanced liver disease received referrals, were wait-listed, or underwent a transplant. The greatest deficits occurred at the referral step. Although health systems routinely track rates and disparities for organ transplants among wait-listed patients, extending monitoring to the earlier stages may help improve equity and manage potentially modifiable barriers to transplantation.

Decrease in Mobility during the COVID-19 Pandemic and Its Association with Increase in Depression among Older Adults: A Longitudinal Remote Mobility Monitoring Using a Wearable Sensor

Background: Social isolation during COVID-19 may negatively impact older adults’ wellbeing. To assess its impact, we measured changes in physical activity and sleep among community-dwelling older adults, from pre-to post-pandemic declaration. Method: Physical activity and sleep in older adults (n = 10, age = 77.3 ± 1.9 years, female = 40%) were remotely assessed within 3-month pre-to 6-month post-pandemic declaration using a pendant-wearable system. Depression was assessed pre-and post-pandemic declaration using the Center for Epidemiologic Studies Depression scale and was compared with 48 h continuous physical activity monitoring data before and during pandemic. Results: Compared to pre-pandemic, post-pandemic time spent in standing declined by 32.7% (Cohen’s d = 0.78, p < 0.01), walking by 52.2% (d = 1.1, p < 0.01), step-counts by 55.1% (d = 1.0, p = 0.016), and postural transitions by 44.6% (d = 0.82, p = 0.017) with increase in sitting duration by 20.5% (d = 0.5, p = 0.049). Depression symptoms increased by 150% (d = 0.8, p = 0.046). Interestingly, increase in depression was significantly correlated with unbroken-prolong sitting bout (ρ = 0.677, p = 0.032), cadence (ρ = −0.70, p = 0.024), and sleep duration (ρ = −0.72, p = 0.019). Conclusion: This is one of the early longitudinal studies highlighting adverse effect of the pandemic on objectively assessed physical activity and sleep in older adults. Our observations showed need for timely intervention to mitigate hard to reverse consequences of decreased physical activity such as depression.

A Wrist-Worn Sensor-Derived Frailty Index Based on an Upper-Extremity Functional Test in Predicting Functional Mobility in Older Adults

INTRODUCTION

Preoperative frailty is an independent risk factor for postoperative complications across surgical specialties. Functional mobility such as gait, timed up and go (TUG), and 5 times sit-to-stand (5-STS) are popular preoperative frailty measurements but are not suitable for patients with severe mobility impairment. A wrist-worn sensor-derived frailty index based on an upper-extremity functional test (20-s repetitive elbow flexion-extension task; UEFI) was developed previously; however, its association with functional mobility remained unexplored. We aimed to investigate the predictive power of the UEFI in predicting functional mobility.

METHODS

We examined correlation between the UEFI and gait speed, TUG duration, and 5-STS duration in 100 older adults (≥ 65 years) using multivariate regression analysis. The UEFI was calculated using slowness, weakness, exhaustion, and flexibility of the sensor-based 20-s repetitive elbow flexion-extension task.

RESULTS

The UEFI was a significant predictor for gait speed and TUG duration and 5-STS duration (all R ≥ 0.60; all p < 0.001) with the variance (adjusted R2) of 35-37% for the dependent variables. The multivariate regression analysis revealed significant associations between the UEFI and gait speed (β = -0.84; 95% confidence interval [95% CI] = [-1.19, -0.50]; p < 0.001) and TUG duration (β = 16.2; 95% CI = [9.59, 22.8]; p < 0.001) and 5-STS duration (β = 33.3; 95% CI = [23.6, 43.2]; p < 0.001), found after accounting for confounding variables (e.g., age and fear of falling scale).

CONCLUSIONS

Our findings suggest that the UEFI can be performed with a wrist-worn sensor and has been validated with other established measures of preoperative frailty. The UEFI can be applied in a wide variety of patients, regardless of mobility limitations, in an outpatient setting.

Study of emotional distress in a comparative effectiveness trial of diabetes treatments: Rationale and design

Emotional distress, including depression and diabetes-specific distress (e.g., feeling overwhelmed by living with diabetes, feelings of failure related to diabetes self-care), is a significant and prevalent problem for patients with type 2 diabetes. Both depression and diabetes distress have been associated with metabolic/glycemic control, diabetes complications, mortality, and quality of life. Recent findings further suggest that risk for emotional distress is influenced by diabetes treatment. The GRADE Study (Glycemia Reduction Approaches in Diabetes: A Comparative Effectiveness Study) is generating prospective data that will provide a unique opportunity to examine the relationships between emotional distress, diabetes treatment, and outcomes in an experimental design. The GRADE study is a randomized clinical trial that will compare the metabolic effects of four common anti-hyperglycemic drugs when combined with metformin. This sub-study recruited a subset (n = 1739) of GRADE participants and will examine patient-level variation in baseline emotional distress as a predictor of glycemic control and other health outcomes, independent of treatment effects. The study will also provide an experimental examination of treatment regimen effects on emotional distress over time as part of the overall evaluation of comparative effectiveness. Evaluation of emotional distress using validated measures will allow us to disentangle the roles of depressive symptoms and diabetes distress, factors that share significant overlap but require distinct approaches to screening and treatment. Study findings may directly influence practice decisions regarding screening and treatment for emotional distress as part of diabetes care. ClinicalTrials.gov Identifier: NCT01794143.

Outcome Goals and Health Care Preferences of Older Adults With Multiple Chronic Conditions

IMPORTANCE

Older adults with multiple chronic conditions (MCCs) vary in their health outcome goals and the health care that they prefer to receive to achieve these goals.

OBJECTIVE

To describe the outcome goals and health care preferences of this population with MCCs.

DESIGN, SETTING, AND PARTICIPANTS

This cross-sectional study included participants in the Patient Priorities Care study who underwent health priorities identification from February 1, 2017, to August 31, 2018, in a primary care practice. Patients eligible to participate were 65 years or older, English speaking, and had at least 3 chronic conditions; in addition, they used at least 10 medications, saw at least 2 specialists, or had at least 2 emergency department visits or 1 hospitalization during the past year. Of 236 eligible patients, 163 (69%) agreed to participate in this study. Data were analyzed from August 1 to October 31, 2020.

EXPOSURES

Guided by facilitators, participants identified their core values, as many as 3 actionable and realistic outcome goals, health-related barriers to these goals, and as many as 3 helpful and 3 bothersome health care activities.

MAIN OUTCOMES AND MEASURES

Frequencies were ascertained for outcome goals and health care preferences. Preferences included health care activities (medications, health care visits, procedures, diagnostic tests, and self-management) reported as either helpful or bothersome.

RESULTS

Most of the 163 participants were White (158 [96.9%]) and women (109 [66.9%]), with a mean (SD) age of 77.6 (7.6) years. Of 459 goals, the most common encompassed meals and other activities with family and friends (111 [24.2%]), shopping (28 [6.1%]), and exercising (21 [4.6%]). Twenty individuals (12.3%) desired to live independently without specifying necessary activities. Of 312 barriers identified, the most common were pain (128 [41.0%]), fatigue (45 [14.4%]), unsteadiness (42 [13.5%]), and dyspnea (19 [6.1%]). Similar proportions of patients identified at least 1 medication that was helpful (130 [79.8%]) or bothersome (128 [78.5%]). Medications most commonly cited as helpful were pain medications, including nonopiods (36 of 55 users [65.5%]) and opioids (15 of 27 users [55.6%]); sleep medications (27 of 51 users [52.9%]); and respiratory inhalants (19 of 45 [42.2%]). Most often mentioned as bothersome were statins (25 of 97 users [25.8%]) and antidepressants (13 of 40 users [32.5%]). Thirty-two participants (19.6%) reported using too many medications. Health care visits were identified as helpful by 43 participants (26.4%); 15 (9.2%) reported too many visits. Procedures were named helpful by 38 participants (23.3%); 24 (14.7%) cited unwanted procedures. Among 48 participants with diabetes, monitoring of glucose levels was doable for 18 (37.5%) and too bothersome for 9 (18.8%).

CONCLUSIONS AND RELEVANCE

Participants identified realistic and actionable goals while varying in health care activities deemed helpful or bothersome. The goals and health care preferences of more diverse populations must be explored. Previous work suggests that clinicians can use this information in decision-making.

Implementation and Feasibility of the Re-Engineered Discharge for Surgery (RED-S) Intervention: A Pilot Study

INTRODUCTION

Adapting Project Re-Engineered Discharge (Project RED), an intervention for reducing internal medicine hospital readmissions, is a promising option for reducing colorectal surgery readmissions.

METHODS

We conducted a pilot study for the adaptation and implementation of Project RED with patients admitted for colectomy at a regional VA tertiary care center between July 2014 and January 2015. Implementation was evaluated using adherence to intervention components and results from the Survey of Healthcare Experiences of Patients. The adapted Project RED for Surgery has five components: surgical wound/ostomy-care education, scheduled follow-up appointments, medication reconciliation, an After Hospital Care Plan, and postdischarge phone calls.

RESULTS

All (n = 21) participants received postoperative wound care education, and 77% of ostomy patients received education. Follow-up appointments were scheduled for 76% with surgery clinic and 67% with primary care. Half received pharmacist-led medication reconciliation. Seventy-five percent received a postdischarge phone call. Ninety five percent of participants reported positive or satisfactory care transitions versus less than 60% of a comparison group of surgery patients from the same institution. We summarized lessons learned from this intervention study to facilitate future dissemination efforts.

CONCLUSION

The lessons learned from this pilot can guide quality improvement teams seeking to implement the Re-Engineered Discharge for Surgery intervention within their existing workflows.

Measuring Adoption of Patient Priorities-Aligned Care Using Natural Language Processing of Electronic Health Records: Development and Validation of the Model

Background

Patient Priorities Care (PPC) is a model of care that aligns health care recommendations with priorities of older adults who have multiple chronic conditions. Following identification of patient priorities, this information is documented in the patient’s electronic health record (EHR).

Objective

Our goal is to develop and validate a natural language processing (NLP) model that reliably documents when clinicians identify patient priorities (ie, values, outcome goals, and care preferences) within the EHR as a measure of PPC adoption.

Methods

This is a retrospective analysis of unstructured National Veteran Health Administration EHR free-text notes using an NLP model. The data were sourced from 778 patient notes of 658 patients from encounters with 144 social workers in the primary care setting. Each patient’s free-text clinical note was reviewed by 2 independent reviewers for the presence of PPC language such as priorities, values, and goals. We developed an NLP model that utilized statistical machine learning approaches. The performance of the NLP model in training and validation with 10-fold cross-validation is reported via accuracy, recall, and precision in comparison to the chart review.

Results

Of 778 notes, 589 (75.7%) were identified as containing PPC language (kappa=0.82, P<.001). The NLP model in the training stage had an accuracy of 0.98 (95% CI 0.98-0.99), a recall of 0.98 (95% CI 0.98-0.99), and precision of 0.98 (95% CI 0.97-1.00). The NLP model in the validation stage had an accuracy of 0.92 (95% CI 0.90-0.94), recall of 0.84 (95% CI 0.79-0.89), and precision of 0.84 (95% CI 0.77-0.91). In contrast, an approach using simple search terms for PPC only had a precision of 0.757.

Conclusions

An automated NLP model can reliably measure with high precision, recall, and accuracy when clinicians document patient priorities as a key step in the adoption of PPC.

A Telehealth-supported, Integrated care with CHWs, and MEdication-access (TIME) Program for Diabetes Improves HbA1c: a Randomized Clinical Trial

BACKGROUND

Many individuals with diabetes live in low- or middle-income settings. Glycemic control is challenging, particularly in resource-limited areas that face numerous healthcare barriers.

OBJECTIVE

To compare HbA1c outcomes for individuals randomized to TIME, a Telehealth-supported, Integrated care with CHWs (Community Health Workers), and MEdication-access program (intervention) versus usual care (wait-list control).

DESIGN

Randomized clinical trial.

PARTICIPANTS

Low-income Latino(a) adults with type 2 diabetes.

INTERVENTIONS

TIME consisted of (1) CHW-participant telehealth communication via mobile health (mHealth) for 12 months, (2) CHW-led monthly group visits for 6 months, and (3) weekly CHW-physician diabetes training and support via telehealth (video conferencing).

MAIN MEASURES

Investigators compared TIME versus control participant baseline to month 6 changes of HbA1c (primary outcome), blood pressure, body mass index (BMI), weight, and adherence to seven American Diabetes Association (ADA) standards of care. CHW assistance in identifying barriers to healthcare in the intervention group were measured at the end of mHealth communication (12 months).

KEY RESULTS

A total of 89 individuals participated. TIME individuals compared to control participants had significant HbA1c decreases (9.02 to 7.59% (- 1.43%) vs. 8.71 to 8.26% (- 0.45%), respectively, p = 0.002), blood pressure changes (systolic: - 6.89 mmHg vs. 0.03 mmHg, p = 0.023; diastolic: - 3.36 mmHg vs. 0.2 mmHg, respectively, p = 0.046), and ADA guideline adherence (p < 0.001) from baseline to month 6. At month 6, more TIME than control participants achieved > 0.50% HbA1c reductions (88.57% vs. 43.75%, p < 0.001). BMI and weight changes were not significant between groups. Many (54.6%) TIME participants experienced > 1 barrier to care, of whom 91.7% had medication issues. CHWs identified the majority (87.5%) of barriers.

CONCLUSIONS

TIME participants resulted in improved outcomes including HbA1c. CHWs are uniquely positioned to identify barriers to care particularly related to medications that may have gone unrecognized otherwise. Larger trials are needed to determine the scalability and sustainability of the intervention.

CLINICAL TRIAL

NCT03394456, accessed at https://clinicaltrials.gov/ct2/show/NCT03394456.

Implementing Health Care Quality Measures in Electronic Health Records: A Conceptual Model

BACKGROUND/OBJECTIVES

There is significant literature on the development and validation of quality measures, but comparably less on their implementation into learning health systems. Electronic Health Records (EHRs) have made vast amounts of data available for quality improvement purposes. In this paper we describe a conceptual model for EHR implementation of quality measures.

DESIGN

The model involves five steps: (1) select a measure; (2) define measure criteria; (3) validate criteria and measurement process; (4) improve recording of measure-related activity; and (5) engage quality improvement processes. The model was used to develop and implement a quality measure in the Home-Based Medical Care (HBMC) setting.

SETTING

Harris Health House Call Program (HHHC) provides primary medical and palliative care for homebound patients in Houston.

PARTICIPANTS

Four-hundred twenty-four primary care patients followed in the HHHC.

MEASUREMENT

Completion rate of the 9-item Patient Health Questionnaire (PHQ-9) within the Electronic Health Record of newly enrolled HHHC patients.

RESULTS

Use of the conceptual model to guide implementation of a quality measure of depression screening in a HMBC practice was successful. Additional components of early leadership and clinician buy-in were required, as well as strong relationships with IT to ease implementation and limit disruptions in clinicians' work-flow.

CONCLUSION

This conceptual model was feasible for guiding implementation of a quality measure for depression care of HBMC patients, and it can guide broader implementation of EHR-based quality measures in the future.

Developing Leaders and Scholars in Health Care Improvement: The VA Quality Scholars Program Competencies

Despite the need for leaders in health care improvement across health professions, there are no standards for the knowledge and skills that should be achieved through advanced interprofessional health care improvement training. Existing health care improvement training competencies focus on foundational knowledge expected of all trainees or for specific career pathways. Health care improvement leaders fill multiple roles within organizations and promote interprofessional improvement practice. The diverse skill set required of modern health care improvement leaders necessitates the development of training competencies specifically for fellowships in applied health care improvement. The authors describe the development of the revised national Veterans Affairs Quality Scholars (VAQS) Program competencies. The VAQS Program is an interprofessional, postdoctoral training program whose mission is to develop leaders and scholars to improve health care. An interprofessional committee of VAQS faculty reviewed and revised the competencies over 4 months beginning in fall 2018. The first draft was developed using 111 competencies submitted by 11 VAQS training sites and a review of published competencies. The final version included 22 competencies spanning 5 domains: interprofessional collaboration and teamwork, improvement and implementation science, organization and system leadership, methodological skills and analytic techniques for improvement and research, and teaching and coaching. Once attained, the VAQS competencies will guide the skill development that interprofessional health care improvement leaders need to participate in and lead health care improvement scholarship and implementation. These broad competencies are relevant to advanced training programs that develop health care improvement leaders and scholars and may be used by employers to understand the knowledge and skills expected of individuals who complete advanced fellowships in applied health care improvement.

Reduction of in-hospital cardiac arrest with sequential deployment of rapid response team and medical emergency team to the emergency department and acute care wards

Purpose

This study aimed to determine if sequential deployment of a nurse-led Rapid Response Team (RRT) and an intensivist-led Medical Emergency Team (MET) for critically ill patients in the Emergency Department (ED) and acute care wards improved hospital-wide cardiac arrest rates.

Methods

In this single-center, retrospective observational cohort study, we compared the cardiac arrest rates per 1000 patient-days during two time periods. Our hospital instituted a nurse-led RRT in 2012 and added an intensivist-led MET in 2014. We compared the cardiac arrest rates during the nurse-led RRT period and the combined RRT-MET period. With the sequential approach, nurse-led RRT evaluated and managed rapid response calls in acute care wards and if required escalated care and co-managed with an intensivist-led MET. We specifically compared the rates of pulseless electrical activity (PEA) in the two periods. We also looked at the cardiac arrest rates in the ED as RRT-MET co-managed patients with the ED team.

Results

Hospital-wide cardiac arrests decreased from 2.2 events per 1000 patient-days in the nurse-led RRT period to 0.8 events per 1000 patient-days in the combined RRT and MET period (p-value = 0.001). Hospital-wide PEA arrests and shockable rhythms both decreased significantly. PEA rhythms significantly decreased in acute care wards and the ED.

Conclusion

Implementing an intensivist-led MET-RRT significantly decreased the overall cardiac arrest rate relative to the rate under a nurse-led RRT model. Additional MET capabilities and early initiation of advanced, time-sensitive therapies likely had the most impact.

Development, Validation, and Evaluation of a Simple Machine Learning Model to Predict Cirrhosis Mortality

Importance

Machine-learning algorithms offer better predictive accuracy than traditional prognostic models but are too complex and opaque for clinical use.

Objective

To compare different machine learning methods in predicting overall mortality in cirrhosis and to use machine learning to select easily scored clinical variables for a novel cirrhosis prognostic model.

Design, Setting, and Participants

This prognostic study used a retrospective cohort of adult patients with cirrhosis or its complications seen in 130 hospitals and affiliated ambulatory clinics in the integrated, national Veterans Affairs health care system from October 1, 2011, to September 30, 2015. Patients were followed up through December 31, 2018. Data were analyzed from October 1, 2017, to May 31, 2020.

Exposures

Potential predictors included demographic characteristics; liver disease etiology, severity, and complications; use of health care resources; comorbid conditions; and comprehensive laboratory and medication data. Patients were randomly selected for model development (66.7%) and validation (33.3%). Three different statistical and machine learning methods were evaluated: gradient descent boosting, logistic regression with least absolute shrinkage and selection operator (LASSO) regularization, and logistic regression with LASSO constrained to select no more than 10 predictors (partial pathway model). Predictor inclusion and model performance were evaluated in a 5-fold cross-validation. Last, the predictors identified in the most parsimonious (the partial path) model were refit using maximum-likelihood estimation (Cirrhosis Mortality Model [CiMM]), and its predictive performance was compared with that of the widely used Model for End Stage Liver Disease with sodium (MELD-Na) score.

Main Outcomes and Measures

All-cause mortality.

Results

Of the 107 939 patients with cirrhosis (mean [SD] age, 62.7 [9.6] years; 96.6% male; 66.3% white, 18.4% African American), the annual mortality rate ranged from 8.8% to 15.3%. In total, 32.7% of patients died within 3 years, and 46.2% died within 5 years after the index date. Models predicting 1-year mortality had good discrimination for the gradient descent boosting (area under the receiver operating characteristics curve [AUC], 0.81; 95% CI, 0.80-0.82), logistic regression with LASSO regularization (AUC, 0.78; 95% CI, 0.77-0.79), and the partial path logistic model (AUC, 0.78; 95% CI, 0.76-0.78). All models showed good calibration. The final CiMM model with machine learning-derived clinical variables offered significantly better discrimination than the MELD-Na score, with AUCs of 0.78 (95% CI, 0.77-0.79) vs 0.67 (95% CI, 0.66-0.68) for 1-year mortality, respectively (DeLong z = 17.00; P < .001).

Conclusions and Relevance

In this study, simple machine learning techniques performed as well as the more advanced ensemble gradient boosting. Using the clinical variables identified from simple machine learning in a cirrhosis mortality model produced a new score more transparent than machine learning and more predictive than the MELD-Na score.

Using a Logic Model to Design and Evaluate a Quality Improvement Leadership Course

Strong leadership is an essential factor in the success of quality improvement (QI) initiatives that generate and sustain improvements in patient outcomes. Notably, there is a rising need for frontline clinicians, who are often charged with leading QI efforts, to receive training in blended QI and leadership methods and skills. The Leading Healthcare Improvement (LHI) course is a longitudinal leadership course embedded within the Department of Veterans Affairs Quality Scholars (VAQS) program, a multisite interprofessional QI fellowship program. The LHI course was developed to provide frontline clinicians who are emerging QI leaders with the skills to lead and advance improvement efforts at their institutions. It consists of eight 60-minute online sessions and was implemented and delivered to a cohort of interprofessional fellows at 9 sites during the 2017-2018 academic year.This article describes the use of a logic model as a framework to guide the planning, implementation, and evaluation of the LHI course. The authors developed 5 logic model components: inputs, activities, outputs, short-term outcomes, and long-term outcomes. They defined the short-term outcomes using feedback from fellows and an evaluation of the fellows' abstract submissions to the VAQS Summer Institute. Submissions were reviewed to identify how fellows applied the LHI course concepts to QI projects at their respective sites. The authors also collected preliminary impact data from fellows to determine long-term outcomes.Finally, they used the logic model to inform changes to the LHI course based on the evaluation data they collected and developed plans to measure the impact of the course on learners, patients, and the health care system. The authors conclude with lessons learned to guide others who are implementing similar QI efforts.

Feasibility of Clinicians Aligning Health Care with Patient Priorities in Geriatrics Ambulatory Care

BACKGROUND/OBJECTIVES

Aligning healthcare decisions with patients' priorities may improve care for older adults with multiple chronic conditions (MCCs). We conducted a pilot study to assess the feasibility of identifying patient priorities in routine geriatrics care and to compare clinicians' recommendations for patients who did or did not have their priorities identified.

DESIGN

Retrospective chart review.

SETTING

Veterans Administration Medical Center Geriatrics Clinic.

PARTICIPANTS

Older adults with MCCs receiving Patient Priorities Care (PPC; n = 35) were matched with patients receiving usual care (UC; n = 35). Both PPC and UC patients were cared for by three primary care providers (PCPs) in an ambulatory geriatric clinic.

INTERVENTION

In the PPC group, a clinician facilitator met with each patient to identify their healthcare priorities and transmitted patients' priorities in the electronic health record (EHR). Trained PCPs then sought to align healthcare decisions with patients' priorities. In the UC group, patients received usual care from the same PCPs.

MEASUREMENTS

We matched patients by clinician seen, patient's age, number of active conditions, medications, hospitalizations, functional status, and prior hospitalizations. EHRs were reviewed to identify care decisions including medications added or stopped, referrals and consults added or avoided, referrals to community services and supports, self-management activities added or avoided, and total number of changes to care. Mean differences in recommended care between PPC and UC patients from the same PCPs were examined.

RESULTS

Clinician facilitators could identify patient priorities during routine clinic encounters. Compared with patients in the UC group, those in the PPC group had, on average, fewer medications added (P = .05), more referrals to community services and supports (P = .03), and more priorities-aligned self-management tasks added (P = .005).

CONCLUSION

These findings support the feasibility of identifying and documenting patient priorities during routine encounters. Results also suggest that clinicians use patient priorities in recommending care.

Validation of the MEDSAIL Tool to Screen for Capacity for Safe and Independent Living Among Nursing Home Residents

OBJECTIVES

Capacity for safe and independent living (SAIL) refers to an individual's ability to solve problems associated with everyday life and perform activities necessary for living independently. Little guidance exists on the assessment of capacity for SAIL among nursing home residents. As a result, capacity for SAIL is not fully considered in the development of discharge plans to ensure safety and independence in the community. We reasoned that this problem could be addressed with the Making and Executing Decisions for Safe and Independent Living (MEDSAIL) tool, developed to screen for capacity for SAIL among community-dwelling older adults. In this report, we describe findings on the validity of the MEDSAIL when used with nursing home residents.

DESIGN

Prospective cross-sectional pilot study.

SETTING AND PARTICIPANTS

Twenty-four residents of a Veterans Health Affairs Community Living Center (CLC; nursing home); exclusion criteria were cognitive impairment too severe to complete the protocol, diagnosis of serious mental illness or developmental disability, inability to hear, or inability to communicate verbally.

METHODS

Participants completed 2 assessments: the MEDSAIL interview administered by a research assistant and the criterion standard capacity interview administered by a geriatric psychiatrist. We examined internal consistency, divergent validity, and criterion-based validity.

RESULTS

Five of 7 MEDSAIL scenarios approximated acceptable levels of internal consistency (α >0.70). MEDSAIL scores were highly positively correlated with criterion standard capacity determination (0.88, P = .001), and the Wilcoxon rank-sum test statistic for the 2 assessments was also statistically significant (P = .001).

CONCLUSIONS AND IMPLICATIONS

MEDSAIL has promise as a user-friendly brief screening tool for use by nursing home staff to understand resident capacity for SAIL. This information can be used in the development of discharge plans to keep the resident safe and independent in the community. In addition, tailoring the MEDSAIL scenarios specifically to the nursing home setting may further enhance the tool's validity and utility in this new application.

1157 Comparison Of Polysomnography Total Sleep Time In Veterans With A Dementia Diagnosis, Incipient Dementia, And No Dementia

Introduction

Neuroprotection, early diagnosis, and behavioral intervention are national priorities for dementia research. Sleep duration is emerging as an important potential remediable risk factor. In this study, we examined the total sleep time derived from overnight polysomnography (PSG) studies in veterans with a current dementia diagnosis at the time of PSG study (dementia), future diagnosis of dementia following the PSG study (incipient dementia), and no diagnosis of dementia at any time point (no dementia) over a 19-year period.

Methods

We identified 69,847 PSG sleep studies using CPT code 95810 and all-cause dementia diagnosis using ICD 9/10 codes (e.g., F03.90) from 2000-19 in the US Department of Veteran Affairs (VA) national database. To be included patients must have ≥ 1 VA visits in 12 months leading up to PSG. Dementia diagnosis must be documented on two separate visits between 12 months prior to 6 months following PSG for current dementia group and anytime after the PSG for incipient dementia. We used natural language processing to extract TST values from the patient free-text notes. Analysis of variance was used to compare PSG TST of the three groups.

Results

Patients had a mean age of 55.4±13.8 at the time of PSG study, 91.5% were male, and 64% were Caucasian. TST of dementia patients (N=1,031) was m=257±110m (d=0.33, p&lt;.05), incipient dementia (N=1,875) was m=253±116m (d=0.35, p&lt;.05) versus no dementia (61,871) m=292±104mins.

Conclusion

Patients with a diagnosis of dementia at the time of PSG study and patients that went on to receive a diagnosis following their PSG study had a significantly lower total sleep time compared to patients that have never received a dementia diagnosis. This is an important study that compares sleep duration during overnight PSG studies and dementia diagnosis across 19 years in the largest integrated healthcare system in the US.

Support

This material is based upon work supported in part by the Department of Veteran Affairs, Veterans Health Administration, Office of Research and Development, and the Center for Innovations in Quality, Effectiveness and Safety (CIN 13-413). Dr. Nowakowski is also supported by a National Institutes of Health (NIH) Grant (R01NR018342).

1156 Polysomnographic Total Sleep Time: A Novel Biomarker For Dementia

Introduction

Neuroprotection, early diagnosis, and behavioral intervention are national priorities for dementia research. Sleep duration is emerging as an important potential remediable risk factor. In this study, we examined whether total sleep time (TST) derived from attended overnight polysomnography (PSG) studies is associated with an increased prevalence of dementia diagnosis and determined the optimal cut-point.

Methods

We identified 69,847 PSG sleep studies using CPT code 95810 from 2000-19 in the US Department of Veteran Affairs (VA) national database of patient care. We used natural language processing to verify PSG reports and extract TST values from the patient free-text notes. We examined a TST of 240-420 minutes in 10-minute increments using a run chart (time series) approach to determine the optimal cut-point for determining greater odds of dementia.

Results

Patients had a mean age of 55.4±13.8, 91.5% were male, and 64% were Caucasian. PSG studies revealed a mean TST of 310.6±79.5 minutes. The run chart time series analysis revealing &lt; 360 minutes being the optimal cut-point for increased odds of dementia (OR: 1.64, 95% CI: 1.36-1.99, p&lt;.05).

Conclusion

Lower TST predicted higher prevalence of dementia diagnosis. TST of 360 minutes may serve as the optimal cut-point to determine greater odds of dementia. This is an important study examining PSG sleep duration and the prevalence of dementia across 19 years in the largest integrated healthcare system in the US. TST may function as a potential biomarker for developing dementia.

Support

This material is based upon work supported in part by the Department of Veteran Affairs, Veterans Health Administration, Office of Research and Development, and the Center for Innovations in Quality, Effectiveness and Safety (CIN 13-413). Dr. Nowakowski is also supported by a National Institutes of Health (NIH) Grant (R01NR018342).

1180 The Use Of Natural Language Processing To Extract Data From Psg Sleep Study Reports Using National Vha Electronic Medical Record Data

Introduction

In 2007, Congress asked the Department of Veteran Affairs to pay closer attention to the incidence of sleep disorders among veterans. We aimed to use natural language processing (NLP), a method that applies algorithms to understand the meaning and structure of sentences within Electronic Health Record (EHR) patient free-text notes, to identify the number of attended polysomnography (PSG) studies conducted in the Veterans Health Administration (VHA) and to evaluate the performance of NLP in extracting sleep data from the notes.

Methods

We identified 481,115 sleep studies using CPT code 95810 from 2000-19 in the national VHA. We used rule-based regular expression method (phrases: “sleep stage” and “arousal index”) to identify attended PSG reports in the patient free-text notes in the EHR, of which 69,847 records met the rule-based criteria. We randomly selected 178 notes to compare the accuracy of the algorithm in mining sleep parameters: total sleep time (TST), sleep efficiency (SE) and sleep onset latency (SOL) compared to human manual chart review.

Results

The number of documented PSG studies increased each year from 963 in 2000 to 14,209 in 2018. System performance of NLP compared to manually annotated reference standard in detecting sleep parameters was 83% for TST, 87% for SE, and 81% for SOL (accuracy benchmark ≥ 80%).

Conclusion

This study showed that NLP is a useful technique to mine EHR and extract data from patients’ free-text notes. Reasons that NLP is not 100% accurate included, the note authors used different phrasing (e.g., “recording duration”) which the NLP algorithm did not detect/extract or authors omitting sleep continuity variables from the notes. Nevertheless, this automated strategy to identify and extract sleep data can serve as an effective tool in large health care systems to be used for research and evaluation to improve sleep medicine patient care and outcomes.

Support

This material is based upon work supported in part by the Department of Veteran Affairs, Veterans Health Administration, Office of Research and Development, and the Center for Innovations in Quality, Effectiveness and Safety (CIN 13-413). Dr. Nowakowski is also supported by a National Institutes of Health (NIH) Grant (R01NR018342).

Hip Fracture Risk among Hemodialysis-Dependent Patients Prescribed Opioids and Gabapentinoids

BACKGROUND

Despite opioids' known association with hip fracture risk in the general population, they are commonly prescribed to patients with ESKD. Whether use of opioids or gabapentinoids (also used to treat pain in patients with ESKD) contributes to hip fracture risk in patients with ESKD on hemodialysis remains unknown.

METHODS

In a case-control study nested within the US Renal Data System, we identified all hip fracture events recorded among patients dependent on hemodialysis from January 2009 through September 2015. Eligible cases were risk-set matched on index date with ten eligible controls. We required >1 year of Medicare Parts A and B coverage and >3 years of part D coverage to study cumulative longer-term exposure. To examine new, short-term exposure, we selected individuals with >18 months of Part D coverage and no prior opioid or gabapentinoid use between 18 and 7 months before index. We used conditional logistic regression to estimate unadjusted and multivariable-adjusted odds ratios (ORs) and 95% confidence intervals (95% CI).

RESULTS

For the longer-term analyses, we identified 4912 first-time hip fracture cases and 49,120 controls. Opioid use was associated with increased hip fracture risk (adjusted OR, 1.39; 95% CI, 1.26 to 1.53). Subgroups of low, moderate, and high use yielded adjusted ORs of 1.33 (95% CI, 1.20 to 1.47), 1.53 (95% CI, 1.36 to 1.72), and 1.66 (95% CI, 1.45 to 1.90), respectively. The association with hip fractures was also elevated with new, short-term use (adjusted OR, 1.38; 95% CI, 1.25 to 1.52). There were no associations between gabapentinoid use and hip fracture.

CONCLUSIONS

Among patients dependent on hemodialysis in the United States, both short-term and longer-term use of opioid analgesics were associated with hip fracture events.