Evaluation of algorithms to identify delirium in administrative claims and drug utilization database

Comparing LASSO and random forest models for predicting neurological dysfunction among fluoroquinolone users

BACKGROUND

Fluoroquinolones are associated with central (CNS) and peripheral (PNS) nervous system symptoms, and predicting the risk of these outcomes may have important clinical implications. Both LASSO and random forest are appealing modeling methods, yet it is not clear which method performs better for clinical risk prediction.

PURPOSE

To compare models developed using LASSO versus random forest for predicting neurological dysfunction among fluoroquinolone users.

METHODS

We developed and validated risk prediction models using claims data from a commercially insured population. The study cohort included adults dispensed an oral fluoroquinolone, and outcomes were CNS and PNS dysfunction. Model predictors included demographic variables, comorbidities and medications known to be associated with neurological symptoms, and several healthcare utilization predictors. We assessed the accuracy and calibration of these models using measures including AUC, calibration curves, and Brier scores.

RESULTS

The underlying cohort contained 16 533 (1.18%) individuals with CNS dysfunction and 46 995 (3.34%) individuals with PNS dysfunction during 120 days of follow-up. For CNS dysfunction, LASSO had an AUC of 0.81 (95% CI: 0.80, 0.82), while random forest had an AUC of 0.80 (95% CI: 0.80, 0.81). For PNS dysfunction, LASSO had an AUC of 0.75 (95% CI: 0.74, 0.76) versus an AUC of 0.73 (95% CI: 0.73, 0.74) for random forest. Both LASSO models had better calibration, with Brier scores 0.17 (LASSO) versus 0.20 (random forest) for CNS dysfunction and 0.20 (LASSO) versus 0.25 (random forest) for PNS dysfunction.

CONCLUSIONS

LASSO outperformed random forest in predicting CNS and PNS dysfunction among fluoroquinolone users, and should be considered for modeling when the cohort is modest in size, when the number of model predictors is modest, and when predictors are primarily binary.

The Utilization and Costs of Grade D USPSTF Services in Medicare, 2007-2016

BACKGROUND

Low-value care, or patient care that offers no net benefit in specific clinical scenarios, is costly and often associated with patient harm. The US Preventive Services Task Force (USPSTF) Grade D recommendations represent one of the most scientifically sound and frequently delivered groups of low-value services, but a more contemporary measurement of the utilization and spending for Grade D services beyond the small number of previously studied measures is needed.

OBJECTIVE

To estimate utilization and costs of seven USPSTF Grade D services among US Medicare beneficiaries.

DESIGN

We conducted a cross-sectional study of data from the National Ambulatory Medical Care Survey (NAMCS) from 2007 to 2016 to identify instances of Grade D services.

SETTING/PARTICIPANTS

NAMCS is a nationally representative survey of US ambulatory visits at non-federal and non-hospital-based offices that uses a multistage probability sampling design. We included all visits by Medicare enrollees, which included traditional fee-for-service, Medicare Advantage, supplemental coverage, and dual-eligible Medicare-Medicaid enrollees.

MAIN MEASURES

We measured annual utilization of seven Grade D services among adult Medicare patients, using inclusion and exclusion criteria from prior studies and the USPSTF recommendations. We calculated annual costs by multiplying annual utilization counts by mean per-unit costs of services using publicly available sources.

KEY RESULTS

During the study period, we identified 95,121 unweighted Medicare patient visits, representing approximately 2.4 billion visits. Each year, these seven Grade D services were utilized 31.1 million times for Medicare beneficiaries and cost $477,891,886. Three services-screening for asymptomatic bacteriuria, vitamin D supplements for fracture prevention, and colorectal cancer screening for adults over 85 years-comprised $322,382,772, or two-thirds of the annual costs of the Grade D services measured in this study.

CONCLUSIONS

US Medicare beneficiaries frequently received a group of rigorously defined and costly low-value preventive services. Spending on low-value preventive care concentrated among a small subset of measures, representing important opportunities to safely lower US health care spending while improving the quality of care.

Identifying patients with delirium based on unstructured clinical notes. (Preprint)

Background

Delirium in hospitalized patients is a syndrome of acute brain dysfunction. Diagnostic (International Classification of Diseases [ICD]) codes are often used in studies using electronic health records (EHRs), but they are inaccurate.

Objective

We sought to develop a more accurate method using natural language processing (NLP) to detect delirium episodes on the basis of unstructured clinical notes.

Methods

We collected 1.5 million notes from >10,000 patients from among 9 hospitals. Seven experts iteratively labeled 200,471 sentences. Using these, we trained three NLP classifiers: Support Vector Machine, Recurrent Neural Networks, and Transformer. Testing was performed using an external data set. We also evaluated associations with delirium billing (ICD) codes, medications, orders for restraints and sitters, direct assessments (Confusion Assessment Method [CAM] scores), and in-hospital mortality. F1 scores, confusion matrices, and areas under the receiver operating characteristic curve (AUCs) were used to compare NLP models. We used the φ coefficient to measure associations with other delirium indicators.

Results

The transformer NLP performed best on the following parameters: micro F1=0.978, macro F1=0.918, positive AUC=0.984, and negative AUC=0.992. NLP detections exhibited higher correlations (φ) than ICD codes with deliriogenic medications (0.194 vs 0.073 for ICD codes), restraints and sitter orders (0.358 vs 0.177), mortality (0.216 vs 0.000), and CAM scores (0.256 vs –0.028).

Conclusions

Clinical notes are an attractive alternative to ICD codes for EHR delirium studies but require automated methods. Our NLP model detects delirium with high accuracy, similar to manual chart review. Our NLP approach can provide more accurate determination of delirium for large-scale EHR-based studies regarding delirium, quality improvement, and clinical trails.

Association Between Intraoperative Arterial Hypotension and Postoperative Delirium After Noncardiac Surgery: A Retrospective Multicenter Cohort Study

BACKGROUND

It is unclear whether intraoperative arterial hypotension is associated with postoperative delirium. We hypothesized that intraoperative hypotension within a range frequently observed in clinical practice is associated with increased odds of delirium after surgery.

METHODS

Adult noncardiac surgical patients undergoing general anesthesia at 2 academic medical centers between 2005 and 2017 were included in this retrospective cohort study. The primary exposure was intraoperative hypotension, defined as the cumulative duration of an intraoperative mean arterial pressure (MAP) <55 mm Hg, categorized into and short (<15 minutes; median [interquartile range {IQR}], 2 [1-4] minutes) and prolonged (≥15 minutes; median [IQR], 21 [17-31] minutes) durations of intraoperative hypotension. The primary outcome was a new diagnosis of delirium within 30 days after surgery. In secondary analyses, we assessed the association between a MAP decrease of >30% from baseline and postoperative delirium. Multivariable logistic regression adjusted for patient- and procedure-related factors, including demographics, comorbidities, and markers of procedural severity, was used.

RESULTS

Among 316,717 included surgical patients, 2183 (0.7%) were diagnosed with delirium within 30 days after surgery; 41.7% and 2.6% of patients had a MAP <55 mm Hg for a short and a prolonged duration, respectively. A MAP <55 mm Hg was associated with postoperative delirium compared to no hypotension (short duration of MAP <55 mm Hg: adjusted odds ratio [ORadj], 1.22; 95% confidence interval [CI], 1.11-1.33; P < .001 and prolonged duration of MAP <55 mm Hg: ORadj, 1.57; 95% CI, 1.27-1.94; P < .001). Compared to a short duration of a MAP <55 mm Hg, a prolonged duration of a MAP <55 mm Hg was associated with greater odds of postoperative delirium (ORadj, 1.29; 95% CI, 1.05-1.58; P = .016). The association between intraoperative hypotension and postoperative delirium was duration-dependent (ORadj for every 10 cumulative minutes of MAP <55 mm Hg: 1.06; 95% CI, 1.02-1.09; P =.001) and magnified in patients who underwent surgeries of longer duration (P for interaction = .046; MAP <55 mm Hg versus no MAP <55 mm Hg in patients undergoing surgery of >3 hours: ORadj, 1.40; 95% CI, 1.23-1.61; P < .001). A MAP decrease of >30% from baseline was not associated with postoperative delirium compared to no hypotension, also when additionally adjusted for the cumulative duration of a MAP <55 mm Hg (short duration of MAP decrease >30%: ORadj, 1.13; 95% CI, 0.91-1.40; P = .262 and prolonged duration of MAP decrease >30%: ORadj, 1.19; 95% CI, 0.95-1.49; P = .141).

CONCLUSIONS

In patients undergoing noncardiac surgery, a MAP <55 mm Hg was associated with a duration-dependent increase in odds of postoperative delirium. This association was magnified in patients who underwent surgery of long duration.

Association of Histamine-2 Blockers and Proton-Pump Inhibitors With Delirium Development in Critically Ill Adults: A Retrospective Cohort Study

Supplemental Digital Content is available in the text.

OBJECTIVES:

Histamine-2 receptor antagonists are commonly administered for stress ulcer prophylaxis in critically ill adults and may be associated with delirium development. We aimed to determine differential associations of histamine-2 receptor antagonist or proton-pump inhibitor administration with delirium development in patients admitted to a medical ICU.

DESIGN:

Retrospective observational study using a deidentified database sourced from the University of North Carolina Health Care system. Participants were identified as having delirium utilizing an International Classification of Diseases-based algorithm. Associations among histamine-2 receptor antagonist, proton-pump inhibitor, or no medication administration and delirium were identified using relative risk. Multiple logistic regression was used to control for potential confounders including mechanical ventilation and age.

SETTING:

Academic tertiary care medical ICU in the United States.

PATIENTS:

Adults admitted to the University of North Carolina medical ICU from January 2015 to December 2019, excluding those on concurrent histamine-2 receptor antagonists and proton-pump inhibitors in the same encounter.

INTERVENTIONS:

None.

MEASUREMENTS AND MAIN RESULTS:

We identified 6,645 critically ill patients, of whom 29% (n = 1,899) received mechanical ventilation, 45% (n = 3,022) were 65 or older, and 22% (n = 1,487) died during their medical ICU encounter. Of the 6,645 patients, 31% (n = 2,057) received an histamine-2 receptor antagonist and no proton-pump inhibitors, 40% (n = 2,648) received a proton-pump inhibitor and no histamine-2 receptor antagonists, and 46% (n = 3,076) had delirium. The histamine-2 receptor antagonist group had a greater association with delirium than the proton-pump inhibitor group compared with controls receiving neither medication, after controlling for mechanical ventilation and age (risk ratio, 1.36; 1.25–1.47; p < 0.001) and (risk ratio, 1.15; 1.07–1.24; p < 0.001, respectively).

CONCLUSIONS:

Histamine-2 receptor antagonists are more strongly associated with increased delirium than proton-pump inhibitors. Prospective studies are necessary to further elucidate this association and to determine if replacement of histamine-2 receptor antagonists with proton-pump inhibitors in ICUs decreases the burden of delirium in critically ill patients.

Frailty is associated with a higher risk of developing delirium and cognitive impairment among patients with diabetic kidney disease: A longitudinal population-based cohort study

AIMS

Delirium, a form of acute brain failure, exhibits a high incidence among older adults. Recent studies have implicated frailty as an under-recognized complication of diabetes mellitus. Whether the presence of frailty increases the risk of delirium/cognitive impairment among patients with diabetic kidney disease (DKD) remains unclear.

METHODS

From the longitudinal cohort of diabetes patients (LCDP) (n = 840,000) in Taiwan, we identified adults with DKD, dividing them into those without and with different severities of frailty based on a modified FRAIL scale. Cox proportional hazard regression was utilized to examine the frailty-associated risk of delirium/cognitive impairment, identified using approaches validated by others.

RESULTS

Totally 149,145 patients with DKD (mean 61.0 years, 44.2% female) were identified, among whom 31.0%, 51.7%, 16.0% and 1.3% did not have or had 1, 2 and >2 FRAIL items at baseline. After 3.68 years, 6613 (4.4%) developed episodes of delirium/cognitive impairment. After accounting for demographic/lifestyle factors, co-morbidities, medications and interventions, patients with DKD and 1, 2 and >2 FRAIL items had a progressively higher risk of developing delirium/cognitive impairment than those without (for those with 1, 2 and >2 items, hazard ratio 1.18, 1.26 and 1.30, 95% confidence interval 1.08-1.28, 1.14-1.39 and 1.10-1.55, respectively). For every FRAIL item increase, the associated risk rose by 9%.

CONCLUSIONS

Frailty significantly increased the risk of delirium/cognitive impairment among patients with DKD. Frailty screening in these patients may assist in delirium risk stratification.

Probable Delirium and Associated Patient Characteristics in Long-Term Care and Complex Continuing Care: A Population-Based Observational Study

OBJECTIVES

To estimate the prevalence of probable delirium in long-term care (LTC) and complex continuing care (CCC) settings and to describe the resident characteristics associated with probable delirium.

DESIGN

Population-based cross-sectional study using routinely collected administrative health data.

SETTING AND PARTICIPANTS

All LTC and CCC residents in Ontario, Canada, assessed with the Resident Assessment Instrument-Minimum Dataset (RAI-MDS) assessment between July 1, 2016, and December 31, 2016 (LTC n=86,454, CCC n=10,217).

METHODS

Probable delirium was identified via the delirium Clinical Assessment Protocol on the RAI-MDS assessment, which is triggered when individuals display at least 1 of 6 delirium symptoms that are of recent onset and different from their usual functioning. RAI-MDS assessments were linked to demographic and health services utilization databases to ascertain resident demographics and health status. Multivariable logistic regression was used to identify characteristics associated with probable delirium, with adjusted odds ratios (ORs) and 95% confidence intervals (CIs) reported.

RESULTS

Delirium was probable in 3.6% of LTC residents and 16.5% of CCC patients. LTC patients displayed fewer delirium symptoms than CCC patients. The most common delirium symptom in LTC was periods of lethargy (44.6% of delirium cases); in CCC, it was mental function varying over the course of the day (63.5% of delirium cases). The odds of probable delirium varied across individual demographics and health characteristics, with increased health instability having the strongest association with the outcome in both care settings (LTC: OR 30.4, 95% CI 26.2-35.3; CCC: OR 21.0, 95% CI 16.7-26.5 for high vs low instability).

CONCLUSIONS AND IMPLICATIONS

There were differences in the presentation and burden of delirium symptoms between LTC and CCC, potentially reflecting differences in delirium severity or symptom identification. Several risk factors for probable delirium in LTC and CCC were identified that may be amenable to interventions to prevent this highly distressing condition.

Hospitalization Outcomes of Delirium in Patients Admitted to Acute Care Hospitals in Their Last Year of Life: A Population-Based Retrospective Cohort Study

CONTEXT

Delirium is a highly distressing neurocognitive disorder for patients at the end of life.

OBJECTIVES

To compare hospitalization outcomes between patients with and without delirium admitted to acute care hospitals in the last year of life.

METHODS

Using linked administrative data from ICES (previously known as the Institute for Clinical Evaluative Sciences), this population-based retrospective cohort study included adults who died in Ontario between January 1, 2014 and December 31, 2016 and were admitted to an acute care hospital in their last year of life. Delirium was identified via diagnosis codes on the hospitalization discharge record. Outcomes included lengths of stay, discharge location, and in-hospital mortality. We used multivariable generalized estimating equations to compare outcomes between patients with and without delirium.

RESULTS

Of 208,715 decedents, 9.3% experienced delirium in at least one hospitalization in the last year of life. The mean hospitalization lengths of stay was 13.8 days in patients with delirium (SD = 21.1) or 1.80 times longer (95% CI = 1.75-1.84) compared with those without delirium. Among patients discharged alive, patients with delirium were 1.32 times (95% CI = 1.27-1.38) more likely to be discharged to another institution rather than discharged home. There was no difference in in-hospital mortality between patients with and without delirium (relative risk = 1.01; 95% CI = 0.98-1.05).

CONCLUSION

In the last year of life, hospitalized patients with recorded delirium experience poorer outcomes, including longer lengths of stay and increased risk of postdischarge institution use, compared with those without delirium. These outcomes illustrate added burden for patients, their families, and the health care system, thus highlighting the need for delirium prevention and early detection in addition to informed transitional care decisions.

Modifiable factors associated with postoperative delirium after hip fracture repair: An age-stratified retrospective cohort study

BACKGROUND

Postoperative delirium in hip fracture patients is common and is associated with substantial morbidity and consumption of resources.

OBJECTIVE

Using data from the USA, we aimed to examine the relationship between postoperative delirium and (modifiable) peri-operative factors mentioned in the American Geriatrics Society Best Practice Statement on Postoperative Delirium in Older Adults, stratified by 'young old' (<80 years) and 'old-old' (≥80 years) categories.

DESIGN

Retrospective cohort study from 2006 to 2016.

SETTING

Population-based claims data from the USA.

PARTICIPANTS

Patients undergoing 505 152 hip fracture repairs between 2006 and 2016 as recorded in the Premier Healthcare Database.

MAIN OUTCOMES AND MEASURES

The main outcome was postoperative delirium; modifiable factors of interest were peri-operative opioid use (high, medium or low; <25th, 25 to 75th or >75th percentile of oral morphine equivalents), anaesthesia type (general, neuraxial, both), use of benzodiazepines (long acting, short acting, both), pethidine, nonbenzodiazepine hypnotics, ketamine, corticosteroids and gabapentinoids. Multilevel models assessed associations between these factors and postoperative delirium, in the full cohort, and separately in those aged less than 80 and at least 80 years. Odds ratios (ORs) and Bonferroni-adjusted 95% confidence intervals (95% CIs) are reported.

RESULTS

Overall, postoperative delirium incidence was 15.7% (n = 79 547). After adjustment for relevant covariates, the use of long-acting (OR 1.82, CI 1.74 to 1.89) and combined short and long-acting benzodiazepines (OR 1.56, CI 1.48 to 1.63) and ketamine (OR 1.09, CI 1.03 to 1.15), in particular, was associated with increased odds for postoperative delirium, while neuraxial anaesthesia (OR 0.91 CI 0.85 to 0.98) and opioid use (OR 0.95, CI 0.92 to 0.98 and OR 0.88, CI 0.84 to 0.92 for medium and high dose compared with low dose) were associated with lower odds; all P < 0.05. When analysing data separately by age group, effects of benzodiazepines persisted, while opioid use was only relevant in those aged less than 80 years.

CONCLUSION

We identified modifiable factors associated with postoperative delirium incidence among patients undergoing hip fracture repair surgery.

Detection of Cognitive Impairment Following Critical Illness with the Medicare Annual Wellness Visit: A Cohort Study

RATIONALE

Cognitive impairment following critical illness is common in observational studies of older intensive care unit (ICU) survivors. The rate of screening for and diagnosis of cognitive impairment in ICU survivors in non-research settings is unknown.

OBJECTIVES

To determine how often cognitive impairment was detected in older adults in the year following critical illness at an academic medical center as part of (1) the Medicare Annual Wellness Visit (AWV), and (2) as part of routine clinical care.

METHODS

This study was a retrospective cohort study conducted at an urban academic medical center. The study included 696 patients age 65 and older admitted to the medical ICU between October 1, 2016 and October 1, 2018 and discharged alive. Patients were also required to have a health system affiliated primary care provider. Patients were followed for one year. We defined cognitive impairment detected in the AWV as either an indicated diagnosis of cognitive impairment or dementia, or if the patient, family, or provider indicated memory concerns during the AWV. We modeled the incidence of AWV completion and the detection of cognitive impairment using semi-parametric additive models accounting for the competing risk of death.

RESULTS

Over one-year of follow-up, the cumulative incidence of mortality was 23.0% (95% CI 19.9% to 26.1%), with 24.7% (95% CI 21.5%-27.9%) completing the AWV. The cumulative incidence of cognitive impairment first detected through the AWV was 3.4% (95% CI, 1.8%-5.0%) at 1 year, with a higher cumulative incidence for diagnoses of cognitive impairment or dementia first indicated via encounter diagnosis codes or the EHR problem list, 5.9% (95% CI, 3.9%-7.9%).

CONCLUSION

The results of our study suggest that the currently implemented AWV is unlikely to be an adequate mechanism for detecting cognitive impairment in a high-risk population such as those recovering from critical illness.

A novel model to label delirium in an intensive care unit from clinician actions

BACKGROUND

In the intensive care unit (ICU), delirium is a common, acute, confusional state associated with high risk for short- and long-term morbidity and mortality. Machine learning (ML) has promise to address research priorities and improve delirium outcomes. However, due to clinical and billing conventions, delirium is often inconsistently or incompletely labeled in electronic health record (EHR) datasets. Here, we identify clinical actions abstracted from clinical guidelines in electronic health records (EHR) data that indicate risk of delirium among intensive care unit (ICU) patients. We develop a novel prediction model to label patients with delirium based on a large data set and assess model performance.

METHODS

EHR data on 48,451 admissions from 2001 to 2012, available through Medical Information Mart for Intensive Care-III database (MIMIC-III), was used to identify features to develop our prediction models. Five binary ML classification models (Logistic Regression; Classification and Regression Trees; Random Forests; Naïve Bayes; and Support Vector Machines) were fit and ranked by Area Under the Curve (AUC) scores. We compared our best model with two models previously proposed in the literature for goodness of fit, precision, and through biological validation.

RESULTS

Our best performing model with threshold reclassification for predicting delirium was based on a multiple logistic regression using the 31 clinical actions (AUC 0.83). Our model out performed other proposed models by biological validation on clinically meaningful, delirium-associated outcomes.

CONCLUSIONS

Hurdles in identifying accurate labels in large-scale datasets limit clinical applications of ML in delirium. We developed a novel labeling model for delirium in the ICU using a large, public data set. By using guideline-directed clinical actions independent from risk factors, treatments, and outcomes as model predictors, our classifier could be used as a delirium label for future clinically targeted models.

Development and External Validation of a Delirium Prediction Model for Hospitalized Patients With Coronavirus Disease 2019

Background

The coronavirus disease 2019 pandemic has placed unprecedented stress on health systems and has been associated with elevated risk for delirium. The convergence of pandemic resource limitation and clinical demand associated with delirium requires careful risk stratification for targeted prevention efforts.

Objectives

To develop an incident delirium predictive model among coronavirus disease 2019 patients.

Methods

We applied supervised machine learning to electronic health record data for inpatients with coronavirus disease 2019 at three hospitals to build an incident delirium diagnosis prediction model. We validated this model in three different hospitals. Both hospital cohorts included academic and community settings.

Results

Among 2907 patients across 6 hospitals, 488 (16.8%) developed delirium. Applying the predictive model in the external validation cohort of 755 patients, the c-index was 0.75 (0.71–0.79) and the lift in the top quintile was 2.1. At a sensitivity of 80%, the specificity was 56%, negative predictive value 92%, and positive predictive value 30%. Equivalent model performance was observed in subsamples stratified by age, sex, race, need for critical care and care at community vs. academic hospitals.

Conclusion

Machine learning applied to electronic health records available at the time of inpatient admission can be used to risk-stratify patients with coronavirus disease 2019 for incident delirium. Delirium is common among patients with coronavirus disease 2019, and resource constraints during a pandemic demand careful attention to the optimal application of predictive models.

Association Between Early Tracheostomy and Delirium in Older Adults in the United States

OBJECTIVES

Early tracheostomy (fewer than eight days after intubation) is associated with shorter length of stay in the intensive care unit and shorter duration of mechanical ventilation. Studies assessing the association between early tracheostomy and incidence of delirium, however, are lacking. This investigation sought to fill this gap.

DESIGN

Retrospective cross-sectional study.

SETTING

Multi-institutional acute care facilities in the United States.

PARTICIPANTS

Data were derived from the National Inpatient Sample data from 2010 to 2014. Included patients were 65 or older and underwent both intubation and tracheostomy during the hospitalization. The authors excluded patients who underwent multiple intubations or tracheostomy procedures.

INTERVENTIONS

Early tracheostomy versus non-early tracheostomy.

RESULTS

In total, 23,310 patients were included, of whom 24.8% underwent early tracheostomy. From multivariate logistic regression, early tracheostomy was associated with lower odds of having a delirium diagnosis (odds ratio [OR] 0.77, p < 0.00001) across all admission classifications. Upon subgroup analysis, early tracheostomy was associated significantly with lower odds of having delirium for patients admitted with medical (OR 0.74, p < 0.00001) and nonsurgical injury admissions (OR 0.74, p = 0.00116).

CONCLUSIONS

Early tracheostomy was associated significantly with lower odds of delirium among all patients studied. This association held true across medical and nonsurgical subgroups.

Delirium Among Hospitalized Older Adults With Acute Heart Failure Exacerbation

BACKGROUND

Delirium among older adults hospitalized with acute heart failure is associated with increased mortality. However, studies concomitantly assessing the association of delirium with both clinical and economic outcomes in this population, such as mortality, hospital cost, or length of stay, are lacking.

METHODS AND RESULTS

We conducted a retrospective observational study using National Inpatient Sample data from 2011 to 2014. Using multivariable logistic regression, we assessed the association of delirium with in-hospital mortality, then estimated the incremental hospital cost and excessive length of stay adjusting for demographic and clinical factors using multivariable generalized linear regression. The association of other medical complications on clinical and economic outcomes was also assessed. A total of 568,565 (weighted N = 2,826,131) hospitalizations of patients 65 years or older with acute heart failure from 2011 to 2014 were included in the final analysis. The reported prevalence of delirium was 4.53%. After multivariable adjustment, delirium was associated with a 2.35-fold increase in the odds of in-hospital mortality (95% confidence interval [CI] 2.23-2.47), which was lower than the odds ratio for sepsis/septicemia (5.36; 95% CI, 5.02-5.72) or respiratory failure (4.53; 95% CI, 4.38-4.69), but similar to that for acute kidney injury (2.39; 95% CI, 2.31-2.48) and higher than for non-ST elevation myocardial infarct (1.57; 95% CI, 1.46-1.68). Delirium increased the total hospital cost by $4,262 (95% CI, $4,002-4,521) and the length of stay by 1.73 days (95% CI, 1.68-1.78), which was slightly lower than, but similar to, acute kidney injury ($4,771; 95% CI, $4,644-4,897) and 1.82 days (95% CI, 1.79-1.84), and higher than non-ST elevation myocardial infarct ($1,907; 95% CI, $1,629-2,185) and 0.31 days (95% CI, 0.25-0.37).

CONCLUSIONS

Delirium was associated with increased in-hospital mortality, total hospital cost, and length of stay, and the magnitude of the effect was similar to that for acute kidney injury. Enhanced efforts to prevent delirium are needed to decrease its adverse impact on clinical and economic outcomes for hospitalized older adults with acute heart failure.

Coronary Artery Bypass Grafting Among Older Adults: Patterns, Outcomes, and Trends

INTRODUCTION

Although the numbers of older adults in the US are rapidly increasing, there is sparse recent data on the use and outcomes of coronary artery bypass grafting (CABG) among this population. We aimed to evaluate the characteristics and outcomes of older adults undergoing CABG and to measure temporal trends.

MATERIALS AND METHODS

Using data from the National Inpatient Sample (2005-2014), patients aged 85 y and older who underwent CABG were selected. Demographic, clinical, and hospital characteristics were extracted. Outcomes measured were hospital mortality, hospital length of stay, discharge home, and operative complications. Patients were grouped by 2-year increments. Differences in clinical characteristics and outcomes over time were evaluated using trend analyses.

RESULTS

There were 60,124 patients included in the cohort. The mean age was 86.8 y with majority being men (61%), white (88%), and treated in teaching hospitals (61%). Over the study period, the annual surgical volume decreased from 6689 in 2005/06 to 5150 in 2013/14. Mortality decreased from 8.5% to 5.5% (P-trend <0.001) and mean hospital length of stay decreased from 13.9 d to 12.0 d (P-trend <0.001), whereas the rate of discharge home remained stable (14.1% versus 11.6%, P-trend = 0.056). Compared with patients in 2005/06, those in 2013/14 had higher comorbidities [diabetes: 27.6% versus 17.3%; chronic kidney disease: 29.8% versus 9.2%; peripheral artery disease: 7.5% versus 6.0%; and hypertension: 83.7% versus 64.5% (all P-trend <0.001)].

CONCLUSIONS

CABG volumes are decreasing among older adults, and comorbidity burden is increasing, but outcomes are improving. These data may indicate improved preoperative optimization and better perioperative care processes.

Derivation and validation of a novel comorbidity-based delirium risk index to predict postoperative delirium using national administrative healthcare database

OBJECTIVE

To derive and validate a comorbidity-based delirium risk index (DRI) to predict postoperative delirium.

DATA SOURCE/STUDY SETTING

Data of 506 438 hip fracture repair surgeries from 2006 to 2016 were collected to derive DRI and perform internal validation from the Premier Healthcare Database, which provided billing information on 20-25 percent of hospitalizations in the USA. Additionally, data of 1 130 569 knee arthroplasty surgeries were retrieved for external validation.

STUDY DESIGN

Thirty-six commonly seen comorbidities were evaluated by logistic regression with the outcome of postoperative delirium. The hip fracture repair surgery cohort was separated into a training dataset (60 percent) and an internal validation (40 percent) dataset. The least absolute shrinkage and selection operator (LASSO) procedure was applied for variable selection, and weights were assigned to selected comorbidities to quantify corresponding risks. The newly developed DRI was then compared to the Charlson-Deyo Index for goodness-of-fit and predictive ability, using the Akaike information criterion (AIC), Bayesian information criterion (BIC), area under the ROC curve (AUC) for goodness-of-fit, and odds ratios for predictive performance. Additional internal validation was performed by splitting the data by four regions and in 4 randomly selected hospitals. External validation was conducted in patients with knee arthroplasty surgeries.

DATA COLLECTION

Hip fracture repair surgeries, knee arthroplasty surgeries, and comorbidities were identified by using ICD-9 codes. Postoperative delirium was defined by using ICD-9 codes and analyzing billing information for antipsychotics (specifically haloperidol, olanzapine, and quetiapine) typically recommended to treat delirium.

PRINCIPAL FINDINGS

The derived DRI includes 14 comorbidities and assigns comorbidities weights ranging from 1 to 6. The DRI outperformed the Charlson-Deyo Comorbidity Index with better goodness-of-fit and predictive performance.

CONCLUSIONS

Delirium risk index is a valid comorbidity index for covariate adjustment and risk prediction in the context of postoperative delirium. Future work is needed to test its performance in different patient populations and varying definitions of delirium.

The Association of Economic Outcome and Geriatric Syndromes among Older Adults with Transcatheter Aortic Valve Replacement (TAVR)

BACKGROUND

The association of geriatric syndromes and economic outcomes among patients who are undergoing transcatheter aortic valve replacement (TAVR) remains unknown.

METHODS AND RESULTS

A retrospective observational study using the National Inpatient Sample (NIS) from 2011 to 2014 was conducted with 7078 patients who were 65 years or older and underwent TAVR. The average hospital cost was US$58 703 (± SD 29 777) and length of stay (LOS) was 8.1 days (±7.20). The rates of delirium, dementia, and frailty were 8.0%, 6.1%, and 10.5%, respectively. From a multivariable generalized linear regression, delirium increased the cost by 31.5% (95% CI 25.41~37.92) and LOS by 70.3% (95% CI 60.20~83.38). Frailty increased the cost by 7.4% (95% CI 3.44~11.53) and the LOS by 22.6% (95% CI 15.15~30.55). Dementia had no significant association with either outcome. When the interactions of the geriatric syndromes were tested for association with the outcomes, delirium in the absence of dementia but presence of frailty showed the strongest association with cost (increase by 45.1%, 95% CI 26.45~66.45), and delirium in the absence of both dementia and frailty showed the strongest association with LOS (increase by 74.5%, 95% CI 62.71~87.13). When the average hospital cost and LOS were predicted using the model with interaction terms, patients with delirium and frailty (but without dementia) had the highest value (total hospital cost US$86 503 and LOS 14.9 days).

CONCLUSION

Among TAVR patients, delirium was significantly associated with increased hospital cost and LOS, and the association was significantly higher in the absence of dementia. The results of this study will be a great asset for health care providers and administrators in planning for efficient care strategy to lower health care expenditure in the hospital for older adults who underwent TAVR.

Physical Performance and Risk of Postoperative Delirium in Older Adults Undergoing Aortic Valve Replacement

Background

Delirium is a major risk factor for poor recovery after surgical aortic valve replacement (SAVR) and transcatheter aortic valve replacement (TAVR). It is unclear whether preoperative physical performance tests improve delirium prediction.

Objective

To examine whether physical performance tests can predict delirium after SAVR and TAVR, and adapt an existing delirium prediction rule for cardiac surgery, which includes Mini-Mental State Examination (MMSE), depression, prior stroke, and albumin level.

Design

Prospective cohort, 2014-2017.

Setting

Single academic center.

Subjects

A total of 187 patients undergoing SAVR (n=77) or TAVR (n=110).

Methods

The Short Physical Performance Battery (SPPB) score was calculated based on gait speed, balance, and chair stands (range: 0-12 points, lower scores indicate poor performance). Delirium was assessed using the Confusion Assessment Method. We fitted logistic regression to predict delirium using SPPB components and risk factors of delirium.

Results

Delirium occurred in 35.8% (50.7% in SAVR and 25.5% in TAVR). The risk of delirium increased for lower SPPB scores: 10-12 (28.2%), 7-9 (34.5%), 4-6 (37.5%) and 0-3 (44.1%) (p-for-trend=0.001). A model that included gait speed <0.46 meter/second (OR, 2.7; 95% CI, 1.2-6.4), chair stands time ≥11.2 seconds (OR, 3.5; 95% CI, 1.0-12.4), MMSE <24 points (OR, 2.9; 95% CI, 1.3-6.4), isolated SAVR (OR, 5.4; 95% CI, 2.1-13.8), and SAVR and coronary artery bypass grafting (OR, 15.8; 95% CI, 5.5-45.7) predicted delirium better than the existing prediction rule (C statistics: 0.71 vs 0.61; p=0.035).

Conclusion

Assessing physical performance, in addition to cognitive function, can help identify high-risk patients for delirium after SAVR and TAVR.

The occurrence and timing of delirium in acute care hospitalizations in the last year of life: A population-based retrospective cohort study

BACKGROUND

Delirium is a distressing neurocognitive disorder that is common among terminally ill individuals, although few studies have described its occurrence in the acute care setting among this population.

AIM

To describe the prevalence of delirium in patients admitted to acute care hospitals in Ontario, Canada, in their last year of life and identify factors associated with delirium.

DESIGN

Population-based retrospective cohort study using linked health administrative data. Delirium was identified through diagnosis codes on hospitalization records.

SETTING/PARTICIPANTS

Ontario decedents (1 January 2014 to 31 December 2016) admitted to an acute care hospital in their last year of life, excluding individuals age of <18 years or >105 years at admission, those not eligible for the provincial health insurance plan between their hospitalization and death dates, and non-Ontario residents.

RESULTS

Delirium was recorded as a diagnosis in 8.2% of hospitalizations. The frequency of delirium-related hospitalizations increased as death approached. Delirium prevalence was higher in patients with dementia (prevalence ratio: 1.43; 95% confidence interval: 1.36-1.50), frailty (prevalence ratio: 1.67; 95% confidence interval: 1.56-1.80), or organ failure-related cause of death (prevalence ratio: 1.23; 95% confidence interval: 1.16-1.31) and an opioid prescription (prevalence ratio: 1.17; 95% confidence interval: 1.12-1.21). Prevalence also varied by age, sex, chronic conditions, antipsychotic use, receipt of long-term care or home care, and hospitalization characteristics.

CONCLUSION

This study described the occurrence and timing of delirium in acute care hospitals in the last year of life and identified factors associated with delirium. These findings can be used to support delirium prevention and early detection in the hospital setting.

Economic Burden of Postoperative Neurocognitive Disorders Among US Medicare Patients

IMPORTANCE

Postoperative neurocognitive disorders (PNDs) after surgical procedures are common and may be associated with increased health care expenditures.

OBJECTIVE

To quantify the economic burden associated with a PND diagnosis in 1 year following surgical treatment among older patients in the United States.

DESIGN, SETTING, AND PARTICIPANTS

This retrospective cohort study used claims data from the Bundled Payments for Care Improvement Advanced Model from 4285 hospitals that submitted Medicare Fee-for-service (FFS) claims between January 2013 and December 2016. All Medicare patients aged 65 years or older who underwent an inpatient hospital admission associated with a surgical procedure, did not experience a PND before index admission, and were not undergoing dialysis or concurrently enrolled in Medicaid were included. Data were analyzed from October 2019 and May 2020.

EXPOSURES

PND, defined as an International Classification of Diseases, Ninth or Tenth Revision, diagnosis of delirium, mild cognitive impairment, or dementia within 1 year of discharge from the index surgical admission.

MAIN OUTCOMES AND MEASURES

The primary outcome was total inflation-adjusted Medicare postacute care payments within 1 year after the index surgical procedure.

RESULTS

A total of 2 380 473 patients (mean [SD] age, 75.36 (7.31) years; 1 336 736 [56.1%] women) who underwent surgical procedures were included, of whom 44 974 patients (1.9%) were diagnosed with a PND. Among all patients, most were White (2 142 157 patients [90.0%]), presenting for orthopedic surgery (1 523 782 patients [64.0%]) in urban medical centers (2 179 893 patients [91.6%]) that were private nonprofits (1 798 749 patients [75.6%]). Patients with a PND, compared with those without a PND, experienced a significantly longer hospital length of stay (mean [SD], 5.91 [6.01] days vs 4.29 [4.18] days; P < .001), were less likely to be discharged home (9947 patients [22.1%] vs 914 925 patients [39.2%]; P < .001), and had a higher incidence of mortality at 1 year after treatment (4580 patients [10.2%] vs 103 767 patients [4.4%]; P < .001). After adjusting for patient and hospital characteristics, the presence of a PND within 1 year of the index procedure was associated with an increase of $17 275 (95% CI, $17 058-$17 491) in cost in the 1-year postadmission period (P < .001).

CONCLUSIONS AND RELEVANCE

The findings of this cohort study suggest that among older Medicare patients undergoing surgical treatment, a diagnosis of a PND was associated with an increase in health care costs for up to 1 year following the surgical procedure. Given the magnitude of this cost burden, PNDs represent an appealing target for risk mitigation and improvement in value-based health care.

Long-Term Cognitive Impairment Associated With Delirium in Acute Neurological Injury

Supplemental Digital Content is available in the text.

Objectives:

To characterize the risk of long-term cognitive impairment associated with delirium in acute neurologic injury patients.

Design:

We analyzed a 10-year cohort of adult acute neurologic injury patients (stroke and traumatic brain injury) without preexisting mild cognitive impairment or dementia, utilizing administrative databases. Patients were followed for in-hospital delirium and mild cognitive impairment or dementia. We report incidence and adjusted hazard ratios for mild cognitive impairment or dementia associated with delirium. Subgroups analyzed include acute neurologic injury categories, dementia subtypes, repeated delirium exposure, and age strata.

Setting:

We used state emergency department and state inpatient databases for New York, Florida, and California. All visits are included in the databases regardless of payer status.

Patients:

We included adult patients with diagnosis of stroke and traumatic brain injury as acute neurologic injury. Patients with preexisting mild cognitive impairment or dementia were excluded.

Interventions:

None.

Measurements and Main Results:

Among 911,380 acute neurologic injury patients, 5.2% were diagnosed with delirium. Mild cognitive impairment or dementia incidence among delirium patients was approximately twice that of nondelirium patients. In adjusted models, risk of mild cognitive impairment or dementia was higher among patients with delirium (adjusted hazard ratio, 1.58). Increased risk was observed across all subgroups including patients less than or equal to 55 years old.

Conclusions:

Identification, management, and prevention of in-hospital delirium could potentially improve long-term cognitive outcomes in acute neurologic injury patients.

Early tracheostomy in acute heart failure exacerbation

BACKGROUND

The optimal timing for tracheostomy among patients with acute heart failure (AHF) exacerbation has been controversial, despite multiple studies assessing the utility of early tracheostomy. Our objective was to assess the trend of utilization and outcomes of early tracheostomy among patients with AHF exacerbation in the United States.

METHODS AND RESULTS

A retrospective cohort study using the National Inpatient Sample from 2005 to 2014 was conducted. Among those who were admitted with AHF exacerbation (n = 1,390,356), 0.26% of patients underwent tracheostomy (n = 2,571), and among them, 19.4% received early tracheostomy (n = 496). There was no significant shift in the percentage of early tracheostomy from 2008 to 2014. We used propensity score matching to compare the clinical and economic outcomes between the early tracheostomy group and late tracheostomy group. In-hospital mortality did not show any difference between the two groups (13.97% in early group vs. 18.04% in late group; p =0.163). The median total hospital cost ($53,466), total hospital length of stay (19 days), and length of stay after intubation (16 days) in the early tracheostomy group were significantly lower than in the late tracheostomy group ($73,680; 26 days; 23 days, respectively).

CONCLUSION

Early tracheostomy showed economic benefit with lower hospital costs and shorter length of stay, without a difference in in-hospital mortality compared to late tracheostomy.

Evaluation of emergency department derived delirium prediction models using a hospital-wide cohort

OBJECTIVE

Delirium is acute disorder of attention and cognition. We conducted an observational study using a hospital-wide database to validate three delirium prediction models that were developed to predict prevalent delirium within the first day of hospitalization after ED visit.

METHODS

This was a retrospective cohort study at the academic medical center to evaluate the predictive ability of three previously developed prediction models for delirium from 2014 to 2017. We included patients aged 65 years and older who were hospitalized from ED. Nurses used the Delirium Observation Screening Scale (DOSS) twice daily while hospitalized. We extracted variables to examine the three prediction models with a positive DOSS screen within the first day of admission. The predictive ability was summarized using the area under the curve (AUC).

RESULTS

We identified 2582 visits with a positive DOSS screen and 877 visits with a diagnosis of delirium from ICD9/10 codes among 12,082 encounters. The AUC of these prediction models ranged from 0.71 to 0.80 when predicting a positive DOSS screen, and 0.68 to 0.72 when predicting a ICD9/10 diagnosis of delirium. In our cohort, the delirium risk score which uses the cutoff of positive or negative predicted DOSS positive delirium with the AUC of 0.8 (p < .0001). The model demonstrated the sensitivity and the specificity of 91.2 (95% CI 90.0-92.3) and 50.3 (95% CI 49.3-51.3).

CONCLUSION

In this study, the delirium risk score had the highest predictive ability for prevalent delirium defined by a positive DOSS within the first day of hospitalization.

In-hospital outcomes and 30-day readmission rates among ischemic and hemorrhagic stroke patients with delirium

OBJECTIVE

Delirium is associated with poor outcomes among critically ill patients. However, it is not well characterized among patients with ischemic or hemorrhagic stroke (IS and HS). We provide the population-level frequency of in-hospital delirium and assess its association with in-hospital outcomes and with 30-day readmission among IS and HS patients.

METHODS

We analyzed Nationwide in-hospital and readmission data for years 2010-2015 and identified stroke patients using ICD-9 codes. Delirium was identified using validated algorithms. Outcomes were in-hospital mortality, length of stay, unfavorable discharge disposition, and 30-day readmission. We used survey design logistic regression methods to provide national estimates of proportions and 95% confidence intervals (CI) for delirium, and odds ratios (OR) for association between delirium and poor outcomes.

RESULTS

We identified 3,107,437 stroke discharges of whom 7.45% were coded to have delirium. This proportion significantly increased between 2010 (6.3%) and 2015 (8.7%) (aOR, 95% CI: 1.04, 1.03-1.05). Delirium proportion was higher among HS patients (ICH: 10.0%, SAH: 9.8%) as compared to IS patients (7.0%). Delirious stroke patients had higher in-hospital mortality (12.3% vs. 7.8%), longer in-hospital stay (11.6 days vs. 7.3 days) and a significantly greater adjusted risk of 30-day-readmission (16.7%) as compared to those without delirium (12.2%) (aRR, 95% CI: 1.13, 1.11-1.15). Upon readmission, patients with delirium at initial admission continued to have a longer length of stay (7.7 days vs. 6.6 days) and a higher in-hospital mortality (9.3% vs. 6.4%).

CONCLUSION

Delirium identified through claims data in stroke patients is independently associated with poor in-hospital outcomes both at index admission and readmission. Identification and management of delirium among stroke patients provides an opportunity to improve outcomes.

Risk factors for postoperative delirium in patients undergoing lower extremity joint arthroplasty: a retrospective population-based cohort study

BACKGROUND

With an ageing population, the demand for joint arthroplasties and the burden of postoperative delirium is likely to increase. Given the lack of large-scale data, we investigated associations between perioperative risk factors and postoperative delirium in arthroplasty surgery.

METHODS

This retrospective population-based cohort study, utilized national claims data from the all-payer Premier Healthcare database containing detailed billing information from >25% nationwide hospitalizations. Patients undergoing elective total hip/knee arthroplasty surgery (2006-2016) were included.The primary outcome was postoperative delirium, while potential risk factors included age, gender, race, insurance type, and modifiable exposures including anesthesia type, opioid prescription dose (low/medium/high), benzodiazepines, meperidine, non-benzodiazepine hypnotics, ketamine, corticosteroids, and gabapentinoids.

RESULTS

Among 1 694 795 patients' postoperative delirium was seen in 2.6% (14 785/564 226) of hip and 2.9% (32 384/1 130 569) of knee arthroplasties. Multivariable models revealed that the utilization of long acting (OR 2.10 CI 1.82 to 2.42), combined long/short acting benzodiazepines (OR 1.74 CI 1.56 to 1.94), and gabapentinoids (OR 1.26 CI 1.16 to 1.36) was associated with increased odds of postoperative delirium. Lower odds of postoperative delirium were seen for neuraxial versus general anesthesia (OR 0.81 CI 0.70 to 0.93) and with the utilization of non-steroidal anti-inflammatory drugs (OR 0.85 CI 0.79 to 0.91) as well as cyclooxygenase-2 inhibitors (OR 0.82 CI 0.77 to 0.89). Age-stratified analysis revealed lower odds with high versus low opioid dose (OR 0.86 CI 0.76 to 0.98) in patients >65 years. Findings were consistent between hip and knee arthroplasties.

CONCLUSIONS

In this large national cohort, we identified various modifiable risk factors (including anesthesia type and pharmaceutical agents) for postoperative delirium, demonstrating possible prevention pathways.

Delirium Incidence and Functional Outcomes After Transcatheter and Surgical Aortic Valve Replacement

BACKGROUND

Transcatheter aortic valve replacement (TAVR) may be associated with less delirium and allow faster recovery than surgical aortic valve replacement (SAVR).

OBJECTIVE

To examine the association of delirium and its severity with clinical and functional outcomes after SAVR and TAVR.

DESIGN

Prospective cohort study.

SETTING

An academic medical center.

PARTICIPANTS

A total of 187 patients, aged 70 years and older, undergoing SAVR (N = 77) and TAVR (N = 110) in 2014 to 2016.

MEASUREMENTS

Delirium was assessed daily using the Confusion Assessment Method (CAM), with severity measured by the CAM-Severity (CAM-S) score (range = 0-19). Outcomes were prolonged hospitalization (9 days or more); institutional discharge; and functional status, measured by ability to perform 22 daily activities and physical tasks over 12 months.

RESULTS

SAVR patients had a higher incidence of delirium than TAVR patients (50.7% vs 25.5%; P < .001), despite younger mean age (77.9 vs 83.7 years) and higher baseline Mini-Mental State Examination score (26.9 vs 24.7). SAVR patients with delirium had a shorter duration (2.2 vs 3.4 days; P = .04) with a lower mean CAM-S score (4.5 vs 5.7; P = .01) than TAVR patients with delirium. The risk of prolonged hospitalization in no, mild, and severe delirium was 18.4%, 30.8%, and 61.5% after SAVR (P for trend = .009) and 26.8%, 38.5%, and 73.3% after TAVR (P for trend = .001), respectively. The risk of institutional discharge was 42.1%, 58.3%, and 84.6% after SAVR (P for trend = .01) and 32.5%, 69.2%, and 80.0% after TAVR (P for trend <.001), respectively. Severe delirium was associated with delayed functional recovery after SAVR and persistent functional impairment after TAVR at 12 months.

CONCLUSION

Less invasive TAVR was associated with lower incidence of delirium than SAVR. Once delirium developed, TAVR patients had more severe delirium and worse functional status trajectory than SAVR patients did.

REGISTRATION

NCT01845207.

A Scoping Review to Map Empirical Evidence Regarding Key Domains and Questions in the Clinical Pathway of Delirium in Palliative Care

CONTEXT

Based on the clinical care pathway of delirium in palliative care (PC), a published analytic framework (AF) formulated research questions in key domains and recommended a scoping review to identify evidence gaps.

OBJECTIVES

To produce a literature map for key domains of the published AF: screening, prognosis and diagnosis, management, and the health-related outcomes.

METHODS

A standard scoping review framework was used by an interdisciplinary study team of nurse- and physician-delirium researchers, an information specialist, and review methodologists to conduct the review. Knowledge user engagement provided context in refining 19 AF questions. A peer-reviewed search strategy identified citations in Medline, PsycINFO, Embase, and CINAHL databases between 1980 and 2018. Two reviewers independently screened records for inclusion using explicit study eligibility criteria for the population, design, delirium diagnosis, and investigational intent.

RESULTS

Of 104 studies reporting empirical data and meeting eligibility criteria, most were conducted in patients with cancer (73.1%) and in inpatient PC units (52%). The most frequent study design was a one or more group, nonrandomized trial or cohort (67.3%). Evidence gaps were identified: delirium risk prediction; comparative effectiveness and harms of prevention, variability in delirium management across PC settings, advanced directive and substitute decision-maker input, and transition of care location; and estimating delirium reversibility. Future rigorous primary studies are required to address these gaps and preliminary concerns regarding the quality of extant literature.

CONCLUSION

Substantial evidence gaps exist, providing opportunities for future research regarding the assessment, prognosis, and management of delirium in PC settings.

Evaluation of Changes in Functional Status in the Year After Aortic Valve Replacement

IMPORTANCE

Functional status is a patient-centered outcome that is important for a meaningful gain in health-related quality of life after aortic valve replacement.

OBJECTIVE

To determine functional status trajectories in the year after transcatheter aortic valve replacement (TAVR) and surgical aortic valve replacement (SAVR).

DESIGN, SETTING, AND PARTICIPANTS

A prospective cohort study with a 12-month follow-up was conducted at a single academic center in 246 patients undergoing TAVR or SAVR for severe aortic stenosis. The study was conducted between February 1, 2014, and June 30, 2017; data analysis was performed from December 27, 2017, to May 7, 2018.

EXPOSURES

Preoperative comprehensive geriatric assessment was performed and a deficit-accumulation frailty index (CGA-FI) (range, 0-1; higher values indicate greater frailty) was calculated.

MAIN OUTCOMES AND MEASURES

Telephone interviews were conducted to assess self-reported ability to perform 22 activities and physical tasks at 1, 3, 6, 9, and 12 months after the procedure.

RESULTS

Of the 246 patients included in the study, 143 underwent TAVR (74 [51.7%] women; mean [SD] age, 84.2 [5.9] years), and 103 underwent SAVR (46 [44.7%] women; age, 78.1 [5.3] years). Five trajectories were identified based on functional status at baseline and during the follow-up: from excellent at baseline to improvement at follow-up (excellent baseline-improvement), good (high baseline-full recovery), fair (moderate baseline-minimal decline), poor (low baseline-moderate decline), and very poor (low baseline-large decline). After TAVR, the most common trajectory was fair (54 [37.8%]), followed by good (33 [23.1%]), poor (21 [14.7%]), excellent (20 [14.0%]), and very poor (12 [8.4%]) trajectories. After SAVR, the most common trajectory was good (39 [37.9%]), followed by excellent (38 [36.9%]), fair (20 [19.4%]), poor (3 [2.9%]), and very poor (1 [1.0%]) trajectories. Preoperative frailty level was associated with lower probability of functional improvement and greater probability of functional decline. After TAVR, patients with CGA-FI level of 0.20 or lower had excellent (3 [50.0%]) or good (3 [50.0%]) trajectories, whereas most patients with CGA-FI level of 0.51 or higher had poor (10 [45.5%]) or very poor (5 [22.7%]) trajectories. After SAVR, most patients with CGA-FI level of 0.20 or lower had excellent (24 [58.5%]) or good (15 [36.6%]) trajectories compared with a fair trajectory (5 [71.4%]) in those with CGA-FI levels of 0.41 to 0.50. Postoperative delirium and major complications were associated with functional decline after TAVR (delirium present vs absent: 14 [50.0%] vs 11 [13.4%]; complications present vs absent: 14 [51.9%] vs 19 [16.4%]) or lack of improvement after SAVR (delirium present vs absent: 27 [69.2%] vs 31 [81.6%]; complications present vs absent: 10 [62.5%] vs 69 [79.3%]).

CONCLUSIONS AND RELEVANCE

The findings suggest that functional decline or lack of improvement is common in older adults with severe frailty undergoing TAVR or SAVR. Although this nonrandomized study does not allow comparison of the effectiveness between TAVR and SAVR, anticipated functional trajectories may inform patient-centered decision making and perioperative care to optimize functional outcomes.

Readmission after seizure discharge in a nationally representative sample

OBJECTIVE

To determine the 30-day readmission rate after seizure-related discharge in a nationally representative sample, as well as patient, clinical, and hospital characteristics associated with readmission.

METHODS

Retrospective cohort study of adults discharged alive from a nonelective hospitalization for epilepsy or seizure, sampled from the Healthcare Cost and Utilization Project's 2014 Nationwide Readmissions Database. Descriptive statistics and logistic regression models were built to quantify and characterize nonelective readmission within 30 days.

RESULTS

A total of 139,800 admissions met inclusion criteria, of which 15,094 (10.8%) were readmitted within 30 days. Patient characteristics associated with readmission included comorbid disease burden (Elixhauser score 2: adjusted odds ratio [AOR] [95% confidence interval (CI)] 1.38 [1.21-1.57]; Elixhauser score 3: AOR 1.52 [1.34-1.73]; Elixhauser score >4: AOR 2.28 [2.01-2.58] as compared to 1) and participation in public insurance programs (Medicare: AOR 1.39 [1.26-1.54]; Medicaid: AOR 1.39 [1.26-1.54] as compared to private insurance). Adverse events (AOR 1.17 [1.05-1.30]) and prolonged length of stay, as well as nonroutine discharge (AOR 1.32 [1.23-1.42]), were also associated with increased adjusted odds of readmission. The most common primary reason for readmission was epilepsy or convulsion (17%).

CONCLUSIONS

Patients hospitalized with seizure are frequently readmitted. While readmitted patients are more likely to have multiple medical comorbidities, our study demonstrated that inpatient adverse events were also significantly associated with readmission. The most common reason for readmission was seizure or epilepsy. Together, these 2 findings suggest that a proportion of readmissions are related to modifiable care process factors and may therefore be avoidable. Further study into understanding preventable drivers of readmission in this population presents an opportunity to improve patient outcomes and health.

Longitudinal Trends and Variation in Antipsychotic Use in Older Adults After Cardiac Surgery

OBJECTIVES

To evaluate temporal trends and between-hospital variation in off-label antipsychotic medication (APM) use in older adults undergoing cardiac surgery.

DESIGN

Retrospective cohort study.

SETTING

National administrative database including 465 U.S. hospitals.

PARTICIPANTS

Individuals aged 65 and older without known indications for APMs who underwent cardiac surgery from 2004 to 2014 (N=293,212).

MEASUREMENTS

Postoperative exposure to any APMs and potentially excessive dosing were examined. Hospital-level APM prescribing intensity was defined as the proportion of individuals newly treated with APMs in the postoperative period.

RESULTS

The rate of APM use declined from 8.8% in 2004 to 6.2% in 2014 (p<.001). Use of haloperidol (parenteral 7.0% to 4.5%, p<.001; oral: 1.9% to 0.5%, p<.001), and risperidone (1.1% to 0.3%, p<.001) declined, whereas quetiapine use tripled (0.6% to 1.9%, p=.03). Hospital APM prescribing intensity varied widely, from 0.3% to 35.6%, across 465 hospitals. Treated individuals at higher-prescribing hospitals were more likely to receive APMs on the day of discharge (highest vs lowest quintile: 15.1% vs 9.6%; p<.001) and for a longer duration (4.8 vs 3.7 days; p<.001) than those at lower-prescribing hospitals. Delirium was the strongest risk factor for APM exposure (odds ratio=9.73, 95% confidence interval=9.02-10.5), whereas none of the hospital characteristics were significantly associated. The rate of potentially excessive dosing declined (60.7% to 44.9%, p<.001), and risk factors for potentially excessive dosing were similar to those for any APM exposure.

CONCLUSIONS

Our findings suggest highly variable prescribing cultures and raise concerns about inappropriate use, highlighting the need for better evidence to guide APM prescribing in hospitalized older adults after cardiac surgery.

Use of haloperidol versus atypical antipsychotics and risk of in-hospital death in patients with acute myocardial infarction: cohort study

OBJECTIVE

To compare the risk of in-hospital mortality associated with haloperidol compared with atypical antipsychotics in patients admitted to hospital with acute myocardial infarction.

DESIGN

Cohort study using a healthcare database.

SETTING

Nationwide sample of patient data from more than 700 hospitals across the United States.

PARTICIPANTS

6578 medical patients aged more than 18 years who initiated oral haloperidol or oral atypical antipsychotics (olanzapine, quetiapine, risperidone) during a hospital admission with a primary diagnosis of acute myocardial infarction between 2003 and 2014.

MAIN OUTCOME MEASURE

In-hospital mortality during seven days of follow-up from treatment initiation.

RESULTS

Among 6578 patients (mean age 75.2 years) treated with an oral antipsychotic drug, 1668 (25.4%) initiated haloperidol and 4910 (74.6%) initiated atypical antipsychotics. The mean time from admission to start of treatment (5.3 v 5.6 days) and length of stay (12.5 v 13.6 days) were similar, but the mean treatment duration was shorter in patients using haloperidol compared with those using atypical antipsychotics (2.4 v 3.9 days). 1:1 propensity score matching was used to adjust for confounding. In intention to treat analyses with the matched cohort, the absolute rate of death per 100 person days was 1.7 for haloperidol (129 deaths) and 1.1 for atypical antipsychotics (92 deaths) during seven days of follow-up from treatment initiation. The survival probability was 0.93 in patients using haloperidol and 0.94 in those using atypical antipsychotics at day 7, accounting for the loss of follow-up due to hospital discharge. The unadjusted and adjusted hazard ratios of death were 1.51 (95% confidence interval 1.22 to 1.85) and 1.50 (1.14 to 1.96), respectively. The association was strongest during the first four days of follow-up and decreased over time. By day 5, the increased risk was no longer evident (1.12, 0.79 to 1.59). In the as-treated analyses, the unadjusted and adjusted hazard ratios were 1.90 (1.43 to 2.53) and 1.93 (1.34 to 2.76), respectively.

CONCLUSION

The results suggest a small increased risk of death within seven days of initiating haloperidol compared with initiating an atypical antipsychotic in patients with acute myocardial infarction. Although residual confounding cannot be excluded, this finding deserves consideration when haloperidol is used for patients admitted to hospital with cardiac morbidity.

Genome-wide association identifies a novel locus for delirium risk

We aimed to identify common genetic variations associated with delirium through genome-wide association testing in a hospital biobank. We applied a published electronic health record-based definition of delirium to identify cases of delirium, and control individuals with no history of delirium, from a biobank spanning 2 Boston academic medical centers. Among 6035 individuals of northern European ancestry, including 421 with a history of delirium, we used logistic regression to examine genome-wide association. We identified one locus spanning multiple genes, including 3 interleukin-related genes, associated with p = 1.41e-8, and 5 other independent loci with p < 5e-7. Our results do not support previously reported candidate gene associations in delirium. Identifying common-variant associations with delirium may provide insight into the mechanisms responsible for this complex and multifactorial outcome. Using standardized claims-based phenotypes in biobanks should allow the larger scale investigations required to confirm novel loci such as the one we identify.

Postoperative delirium in total knee and hip arthroplasty patients: a study of perioperative modifiable risk factors

BACKGROUND

Postoperative delirium continues to pose major clinical difficulties. While unmodifiable factors (e.g. age and comorbidity burden) are commonly studied risk factors for delirium, the role of modifiable factors, such as anaesthesia type and commonly used perioperative medications, remains understudied. This study aims to evaluate the role of modifiable factors for delirium after hip and knee arthroplasties.

METHODS

We performed a retrospective study of 41 766 patients who underwent hip or knee arthroplasties between 2005 and 2014 at a single institution. Data were collected as part of routine patient care. Multivariable logistic regression models assessed associations between anaesthesia type and commonly used perioperative medications (opioids, benzodiazepines, and ketamine) and postoperative delirium. Odds ratios (OR) and 95% confidence intervals (CI) are reported. Various sensitivity analyses are also considered, including multiple imputation methods to address missing data.

RESULTS

Postoperative delirium occurred in 2.21% (n=922) of all patients. While patients who received neuraxial anaesthesia were at lower risk for postoperative delirium (compared with general anaesthesia; epidural OR 0.59 CI 0.38-0.93; spinal OR 0.55 CI 0.37-0.83; combined spinal/epidural OR 0.56 CI 0.40-0.80), those given intraoperative ketamine (OR 1.27 CI 1.01-1.59), opioids (OR 1.25 CI 1.09-1.44), postoperative benzodiazepines (OR 2.47 CI 2.04-2.97), and ketamine infusion (OR 10.59 CI 5.26-19.91) were at a higher risk.

CONCLUSIONS

In this cohort of hip and knee arthroplasty patients, anaesthesia type and perioperative medications were associated with increased odds for postoperative delirium. Our results support the notion that modifiable risk factors may exacerbate or attenuate risk for postoperative delirium.