An Implementation-Effectiveness Study of a Perioperative Delirium Prevention Initiative for Older Adults

Long-term mortality impact of postoperative hyperactive delirium in older hip fracture surgery patients

BACKGROUND

Postoperative hyperactive delirium is a common and serious complication in older patients undergoing surgery, but the association between delirium and mortality remains controversial. Compared to other delirium subtypes, hyperactive delirium is characterized by more overt clinical manifestations, facilitating accurate detection and evaluation. This study aimed to clarify this association by comparing long-term mortality between patients with and without postoperative hyperactive delirium, using propensity score matching for robust analysis.

METHODS

We conducted a cohort study to evaluate the association between postoperative hyperactive delirium and long-term mortality in older patients undergoing emergency hip fracture surgery. We used the Taiwan National Health Insurance Service database to identify patients aged 65 years or older who underwent emergency hip fracture surgery between 2008 and 2018. The primary outcome was all-cause mortality.

RESULTS

A total of 270,437 patients were included in the analysis, with 6,795 patients in the postoperative hyperactive delirium group and 263,642 patients in the no postoperative hyperactive delirium group. After PSM, both groups contained 6,795 patients, ensuring balanced baseline characteristics for comparison. Postoperative hyperactive delirium was an independent risk factor for all-cause death, with an adjusted hazard ratio of 1.62 (95% confidence interval, 1.51-1.74; P < 0.0001) after PSM. Subgroup analysis revealed that older patients with postoperative hyperactive delirium consistently exhibited significantly higher adjusted hazard ratios of all-cause death compared with those without postoperative hyperactive delirium, regardless of age, sex, income levels, or ASA scores. Although the difference in 5-year overall survival between groups (81.7% vs. 89.8%, P < 0.0001) was statistically significant, the high survival rates in both groups suggest a modest absolute clinical impact.

CONCLUSION

Postoperative hyperactive delirium is an independent risk factor for long-term mortality in older patients undergoing emergency hip fracture surgery. While the statistical association is evident, it is important to carefully consider the modest absolute difference in survival rates and its implications for clinical application.

Strategies to Limit Benzodiazepine Use in Anesthesia for Older Adults: A Randomized Clinical Trial

Importance

Despite guidelines recommending avoidance of benzodiazepine administration to older patients, many of them now receive benzodiazepines as a part of anesthesia care. The effectiveness of clinician- and patient-facing interventions to discourage such use remains insufficiently characterized.

Objective

To evaluate the effect of clinician peer comparison, patient informational mail, or a combination of these interventions compared with usual care on the rate of perioperative benzodiazepine administration to older patients.

Design, Setting, and Participants

This 2 × 2 factorial, stepped-wedge, cluster randomized clinical trial of a corporate quality improvement initiative was conducted between August 8, 2022, and May 28, 2023, across 415 hospitals, surgery centers, and physician offices in 8 US states served by anesthesia clinicians from a national anesthesia practice. Participants were adults aged 65 years or older who underwent an elective surgical or endoscopic procedure with general anesthesia. Data analyses followed the intention-to-treat principle.

Intervention

Patients were randomly assigned to 1 of 4 groups-clinician peer comparison (wherein clinicians received feedback regarding their performance compared with other clinicians in the practice), patient informational mail (wherein patients received an informational letter encouraging them to have a discussion regarding medication selection with their clinician on the day of surgery), both interventions, or usual care (no intervention).

Main Outcomes and Measures

Rate of benzodiazepine administration during anesthesia care and patient satisfaction with anesthesia care (measured by the Anesthesia Patient Satisfaction Questionnaire, version 2).

Results

Among the 509 269 enrolled participants (255 871 females [50.2%]; mean [SD] age, 74 [7] years), 81 363 (16.0%) were assigned to clinician peer comparison, 98 520 (19.3%) to patient informational mail, 169 712 (33.3%) to both interventions, and 159 674 (31.4%) to usual care. Among patients who received benzodiazepine during anesthesia care, 24.5% were in the usual care group compared with 19.7% in the clinician peer comparison group, 20.0% in the patient informational mail group, and 19.7% in the combination group. After adjustment for time, none of the study interventions were associated with lower odds of benzodiazepine administration compared with usual care (odds ratio [OR], 1.02 [95% CI, 0.98-1.07]; P = .35 for clinician peer comparison; OR, 1.01 [95% CI, 0.96-1.05]; P = .81 for patient informational mail; and OR, 1.11 [95% CI, 1.05-1.16]; P < .001 for combined interventions). Satisfaction scores were high in all groups and did not vary by treatment assignment.

Conclusions and Relevance

This randomized clinical trial found that clinician peer comparison, patient informational mail, or a combination of both interventions did not reduce benzodiazepine administration to older patients compared with usual care; patient satisfaction remained high throughout the study. Overall, the findings suggest a need to explore other patient-targeted interventions to improve anesthesia care.

Trial Registration

Clinicaltrials.gov Identifier: NCT05436392.

Feasibility and effectiveness of an evidence-based intervention bundle to improve peri-operative care of older adults: A quality improvement study

BACKGROUND

Maintaining functional status is an essential therapeutic goal in peri-operative care of older adults. Although several recommendations concerning peri-operative management are available, pragmatic approaches to their sustainable implementation are lacking.

OBJECTIVE

Multiple evidence-based recommendations for peri-operative care of older adults were bundled into a multicomponent intervention and assessed for feasibility and effectiveness.

DESIGN

A quality improvement study with before-after comparison using a hybrid implementation-effectiveness approach.

SETTING

University Medical Centre. The trial was conducted from 2017 to 2020, follow-up was 1 week, 1 month and 6 months after surgery.

PATIENTS

Patients at least 65 years old were scheduled for surgery; excluded: intercranial, ophthalmological, and emergency surgery; planned postoperative stay in the intensive care unit. A total of 720 patients were eligible; 278 patients were recruited, 95 (34) of whom were female.

INTERVENTIONS

The intervention consisted of a set of recommended measures for peri-operative management of older patients, including pre-operative assessment and measures to manage frailty, malnutrition, polypharmacy, cognitive impairment and delirium. Patients were sequentially assigned to three groups: control (no change from usual care), transition (to implementation of the intervention) and intervention (fully implemented).

MAIN OUTCOME MEASURES

Feasibility was assessed by determining the level of implementation, and barriers were identified by conducting qualitative interviews with the medical staff. Intervention effectiveness was estimated by Instrumental Activities of Daily Living (IADL; Lawton and Brody, primary clinical outcome) 1 and six months postoperatively. The secondary outcomes included postoperative complications, cognitive performance, quality of life and length of hospital stay.

RESULTS

The implementation rate was 77%. Pre-operative IADL was 9.9/10 (range 8 to 10; SD 0.4) and 9.7/10 (6 to 10, 0.8) for the control and intervention groups, respectively. There was no statistically significant difference between the groups in IADL (0.07, 95% CI -0.23 to 0.36, P = 0.66) and 6 months after surgery (0.01, 95% CI -0.29 to 0.31, P = 0.95).

CONCLUSION

The implementation of evidence-based interventions to improve peri-operative care of older patients showed good feasibility in clinical routine but did not improve patients' functional status, which was already at a high level pre-operatively.

TRIAL REGISTRATION

ClinicalTrials.gov, Identifier: NCT03325413.

Association between postoperative hyperactive delirium and major complications in elderly patients undergoing emergency hip fracture surgery: A large-scale cohort study

AIM

This cohort study aimed to explore the connection between postoperative hyperactive delirium and major complications in elderly patients undergoing emergency hip fracture surgery.

METHODS

Elderly patients aged 65 years and older undergoing emergency hip fracture surgery were included in the study. The presence of postoperative hyperactive delirium was assessed, and logistic regression analysis, following propensity score matching, was conducted to investigate the association between postoperative hyperactive delirium and major complications occurring 30 and 90 days post-surgery. The analysis controlled for potential confounding factors.

RESULTS

After propensity score matching, the analysis included 13 590 patients, equally distributed with 6795 in each group. The group experiencing postoperative hyperactive delirium exhibited a significantly elevated risk of 30-day postoperative complications, including acute renal failure, pneumonia, septicemia, and stroke, with adjusted odds ratios ranging from 1.64 to 2.39. Furthermore, this group displayed notably higher rates of 90-day postoperative complications, encompassing mortality, acute renal failure, pneumonia, septicemia, and stroke, with a significantly increased incidence of mortality within 90 days.

CONCLUSION

Postoperative hyperactive delirium in elderly patients undergoing emergency hip fracture surgery is significantly linked to an increased risk of major complications at both 30 and 90 days post-surgery. These findings underscore the critical importance of delirium prevention and management in this patient population, offering the potential to reduce the occurrence of postoperative complications. Geriatr Gerontol Int 2024; 24: 730-736.

Prevalence and characteristics of preoperative patients with depression

BACKGROUND

Within the perioperative period, depression-related diagnoses are associated with postoperative complications. We developed a perioperative depression screening programme to assess disease prevalence and feasibility for intervention.

METHODS

Adult patients in multiple surgical departments at a single academic centre were screened for depression via the electronic health record patient portal or preoperative anaesthesia clinic before surgery, using the Patient Health Questionnaire (PHQ)-2 and -8. We utilised a broad method, screening all patients, and a focused method, only screening patients with a history of depression. Logistic regression was used to identify characteristics associated with clinically significant depression (PHQ-8 ≥10). Symptomatic patients were administered a brief psychoeducational intervention and referred for mental health services.

RESULTS

A total of 3735 patients were identified by the broad and focused screens, of whom 2940 (79%) returned PHQ-2 data and were included in analysis. The broad screen (N=1216) found 46 (4%) patients who reported symptoms of moderate or greater severity. The focused screen (N=1724) found 242 (14%) patients with symptoms of moderate or greater severity and over all higher rates of depression across the symptom severity scale. Using the total screened pool, logistic regression identified a history of depression as the strongest associated patient characteristic variable but this did not capture most cases. Finally, we found that 66% of patients who were contacted about mental health services accepted referrals or sought outside care.

CONCLUSIONS

At least 4% of preoperative patients have clinically significant symptoms of depression, most of whom do not have a chart history of depression.

Development and validation of delirium prediction models for noncardiac surgery patients

STUDY OBJECTIVE

Postoperative delirium is associated with morbidity and mortality, and its incidence varies widely. Using known predisposing and precipitating factors, we sought to develop postoperative delirium prediction models for noncardiac surgical patients.

DESIGN

Retrospective prediction model study.

SETTING

Major quaternary medical center.

PATIENTS

Our January 2016 to June 2020 training dataset included 51,677 patients of whom 2795 patients had delirium. Our July 2020 to January 2022 validation dataset included 14,438 patients of whom 912 patients had delirium.

INTERVENTIONS

None.

MEASUREMENTS

We trained and validated two static prediction models and one dynamic delirium prediction model. For the static models, we used random survival forests and traditional Cox proportional hazard models to predict postoperative delirium from preoperative variables, or from a combination of preoperative and intraoperative variables. We also used landmark modeling to dynamically predict postoperative delirium using preoperative, intraoperative, and postoperative variables before onset of delirium.

MAIN RESULTS

In the validation analyses, the static random forest model had a c-statistic of 0.81 (95% CI: 0.79, 0.82) and a Brier score of 0.04 with preoperative variables only, and a c-statistic of 0.86 (95% CI: 0.84, 0.87) and a Brier score of 0.04 when preoperative and intraoperative variables were combined. The corresponding Cox models had similar discrimination metrics with slightly better calibration. The dynamic model - using all available data, i.e., preoperative, intraoperative and postoperative data - had an overall c-index of 0.84 (95% CI: 0.83, 0.85).

CONCLUSIONS

Using preoperative and intraoperative variables, simple static models performed as well as a dynamic delirium prediction model that also included postoperative variables. Baseline predisposing factors thus appear to contribute far more to delirium after noncardiac surgery than intraoperative or postoperative variables. Improved postoperative data capture may help improve delirium prediction and should be evaluated in future studies.

Current Trends for Delirium Screening within the Emergency Department

Delirium is an acute neurological disorder that involves attention and cognition. It is associated with a high risk of morbidity and mortality among older people (>65 years old). In the context of the Emergency Department (ED), it is frequently experienced by patients but often not recognized. Literature studies have identified some screening instruments for an initial evaluation of delirium. Most of these tools have not been validated yet in the context of emergencies, but, in other settings, they were very useful for assessing and maximizing the recognition of this condition among older patients. We conducted a review of the literature, including randomized control trials, clinical and observational studies, and research studies published in recent years, confirming that most of the screening tools for delirium used in the intensive care unit (ICU) or the geriatric department have not been tested in the ED, and the ideal timing and form of the delirium assessment process for older adults have not been defined yet. The aim of our review is to summarize the updated evidence about the screening tools for delirium in the context of the ED, due to the fact that overcrowding of the ED and the stressful condition of emergency situations (that contribute to the onset of delirium) could expose older patients to a high risk of complications and mortality if delirium is not promptly recognized. In conclusion, we support the evidence that delirium is a current and real condition that emergency physicians have to face daily, and we are aware that more research is needed to explore this field in order to improve the overall outcomes of older patients admitted to the ED.

Evaluation of a behavioural intervention to reduce perioperative midazolam administration to older adults

Background

Older patients commonly receive benzodiazepines during anaesthesia despite guidelines recommending avoidance. Interventions to reduce perioperative benzodiazepine use are not well studied. We hypothesized an automated electronic medical record alert targeting anaesthesia providers would reduce administration of benzodiazepines to older adults undergoing general anaesthesia.

Methods

We conducted a retrospective study of adults who underwent surgery at 5 hospitals within one US academic health system. One of the hospitals received an intervention consisting of provider education and an automated electronic medical record alert discouraging benzodiazepine administration to patients aged 70 years or older. We used difference-in-differences analysis to compare patterns of midazolam use 12-months before and after intervention at the intervention hospital, using the 4 non-intervention hospitals as contemporaneous comparators.

Results

The primary analysis sample included 20,347 cases among patients aged 70 and older. At the intervention hospital, midazolam was administered in 454/4,240 (10.7%) cases pre-alert versus 250/3,750 (6.7%) post-alert (p<0.001). At comparator hospitals, respective rates were 3,186/6,366 (50.0%) versus 2,935/5,991 (49.0%) (p=0.24). After adjustment, the intervention was associated with a 3.2 percentage point (p.p.) reduction in the percentage of cases with midazolam administration (95% CI: (-5.2, -1.1); p=0.002). Midazolam dose was unaffected (adjusted mean difference -0.01 mg, 95% CI: (-0.20, 0.18); p=0.90). In 76,735 cases among patients aged 18-69, the percentage of cases with midazolam administration decreased by 6.9 p. p. (95% CI: (-8.0, -5.7); p<0.001).

Conclusion

Provider-facing alerts in the intraoperative electronic medical record, coupled with education, can reduce midazolam administration to older patients presenting for surgery but may affect care of younger patients.

Machine learning-based prediction model for postoperative delirium in non-cardiac surgery

BACKGROUND

Postoperative delirium is a common complication that is distressing. This study aimed to demonstrate a prediction model for delirium.

METHODS

Among 203,374undergoing non-cardiac surgery between January 2011 and June 2019 at Samsung Medical Center, 2,865 (1.4%) were diagnosed with postoperative delirium. After comparing performances of machine learning algorithms, we chose variables for a prediction model based on an extreme gradient boosting algorithm. Using the top five variables, we generated a prediction model for delirium and conducted an external validation. The Kaplan-Meier and Cox survival analyses were used to analyse the difference of delirium occurrence in patients classified as a prediction model.

RESULTS

The top five variables selected for the postoperative delirium prediction model were age, operation duration, physical status classification, male sex, and surgical risk. An optimal probability threshold in this model was estimated to be 0.02. The area under the receiver operating characteristic (AUROC) curve was 0.870 with a 95% confidence interval of 0.855-0.885, and the sensitivity and specificity of the model were 0.76 and 0.84, respectively. In an external validation, the AUROC was 0.867 (0.845-0.877). In the survival analysis, delirium occurred more frequently in the group of patients predicted as delirium using an internal validation dataset (p < 0.001).

CONCLUSION

Based on machine learning techniques, we analyzed a prediction model of delirium in patients who underwent non-cardiac surgery. Screening for delirium based on the prediction model could improve postoperative care. The working model is provided online and is available for further verification among other populations.

TRIAL REGISTRATION

KCT 0006363.

Anesthetic Practice Trends and Perceptions Toward Postoperative Delirium: A Mixed-Methods Analysis

BACKGROUND

Delirium is the most common postoperative complication in older adults, though anesthesiologist awareness of delirium prevention guidelines-and associated practice trends-remains unknown.

METHODS

This was a convergent mixed-methods study, which simultaneously analyzed quantitative and qualitative data to determine delirium guideline awareness among anesthesiologists and practice patterns based on guideline recommendations. Quantitative data were abstracted from the Multicenter Perioperative Outcomes Group database for noncardiac surgery patients (2009-2020) aged 65 years and older. Linear trends were reported for select guideline-based delirium prevention recommendations via regression modeling. Anesthesiologists (n = 40) from a major academic center without a structured delirium reduction program on hospital wards were then surveyed regarding knowledge and practices with respect to postoperative delirium. For qualitative data, 3 focus groups were held to further discuss guideline awareness and identify challenges with delirium prevention.

RESULTS

Quantitative results demonstrated a significant decline in the proportion of cases with midazolam between 2009 and 2020, with the largest decrease observed with urologic surgeries (-3.9%/y; 95% confidence interval [CI], -4.2 to -3.6; P < .001). Use of regional anesthesia increased over this period, particularly with gynecologic surgeries (+2.3%/y; 95% CI, 1.2-3.4; P = .001). Anesthesiologist survey results revealed variable guideline awareness, as 21 of 39 (54%) respondents reported being aware of guidelines for anesthetic management of older adults. Importantly, unawareness of delirium management guidelines was the most frequently cited challenge (17 of 37, 46%) when caring for older adults. Finally, focus group participants were largely unaware of postoperative delirium guidelines. However, participants conveyed key barriers to delirium identification and prevention, including the unclear pathophysiology, nonmodifiable risk factors, and system-based hospital challenges. Participants also expressed a desire for decision-support systems, integrated within the perioperative workflow, that provide evidence-based recommendations for reducing delirium risk.

CONCLUSIONS

Perioperative practice trends are indicative of an improving environment for postoperative delirium. However, delirium guideline awareness remains variable among anesthesiologists, and key barriers continue to exist for identifying and preventing postoperative delirium.

Potentially Inappropriate Medication Administration Is Associated With Adverse Postoperative Outcomes in Older Surgical Patients: A Retrospective Cohort Study

BACKGROUND

The American Geriatrics Society (AGS) Beers Criteria is an explicit list of potentially inappropriate medications (PIMs) best avoided in adults ≥65 years of age. Cognitively impaired and frail surgical patients often experience poor outcomes after surgery, but the impacts of PIMs on these patients are unclear. Our objective was to assess whether perioperative PIM administration was associated with poor outcomes in geriatric surgical patients. We then evaluated the association between PIM administration and postoperative outcomes in subgroups of patients who were frail or cognitively impaired.

METHODS

We performed a retrospective cohort study of patients ≥65 years of age who underwent elective inpatient surgery at a large academic medical center from February 2018 to January 2020. Edmonton Frail Scale and Mini-Cog screening tools were administered to all patients at their preoperative clinic visit. A Mini-Cog score of 0 to 2 was considered cognitive impairment, and frailty was defined by an Edmonton Frail Scale score of ≥8. Patients were divided into 2 groups depending on whether they received at least 1 PIM (PIM+), based on the 2019 AGS Beers Criteria, in the perioperative period or none (PIM-). We assessed the association of preoperative frailty, cognitive impairment, and perioperative PIM administration with the length of hospital stay and discharge disposition using multiple regression analyses adjusted for age, sex, ASA physical status, and intensive care unit (ICU) admission.

RESULTS

Of the 1627 included patients (mean age, 73.7 years), 69.3% (n = 1128) received at least 1 PIM. A total of 12.7% of patients were frail, and 11.1% of patients were cognitively impaired; 64% of the frail patients and 58% of the cognitively impaired patients received at least 1 PIM. Perioperative PIM administration was associated with longer hospital stay after surgery (PIM-, 3.56 ± 5.2 vs PIM+, 4.93 ± 5.66 days; P < .001; 95% confidence interval [CI], 0.360-0.546). Frail patients who received PIMs had an average length of stay (LOS) that was nearly 2 days longer than frail patients who did not receive PIMs (PIM-, 4.48 ± 5.04 vs PIM+, 6.33 ± 5.89 days; P = .02). Multiple regression analysis revealed no significant association between PIM administration and proportion of patients discharged to a care facility (PIM+, 26.3% vs PIM-, 28.7%; P = .87; 95% CI, -0.046 to 0.054).

CONCLUSIONS

Perioperative PIM administration was common in older surgical patients, including cognitively impaired and frail patients. PIM administration was associated with an increased hospital LOS, particularly in frail patients. There was no association found between PIM administration and discharge disposition.

Cognitive impairment and frailty screening in older surgical patients: a rural tertiary care centre experience

Introduction

Despite a clear association between cognitive impairment and physical frailty and poor postoperative outcomes in older adults, preoperative rates are rarely assessed. We sought to implement a preoperative cognitive impairment and frailty screening programme to meet the unique needs of our rural academic centre.

Methods

Through stakeholder interviews, we identified five primary drivers underlying screening implementation: staff education, technology infrastructure, workload impact, screening value and patient–provider communication. Based on these findings, we implemented cognitive dysfunction (AD8, Mini-Cog) and frailty (Clinical Frailty Scale) screening in our preoperative care clinic and select surgical clinics.

Results

In the preoperative care clinic, many of our patients scored positive for clinical frailty (428 of 1231, 35%) and for cognitive impairment (264 of 1781, 14.8%). In our surgical clinics, 27% (35 of 131) and 9% (12 of 131) scored positive for clinical frailty and cognitive impairment, respectively. Compliance to screening improved from 48% to 86% 1 year later.

Conclusion

We qualitatively analysed stakeholder feedback to drive the successful implementation of a preoperative cognitive impairment and frailty screening programme in our rural tertiary care centre. Preliminary data suggest that a clinically significant proportion of older adults screen positive for preoperative cognitive impairment and frailty and would benefit from tailored inpatient care.

Postoperative delirium prediction using machine learning models and preoperative electronic health record data

Background

Accurate, pragmatic risk stratification for postoperative delirium (POD) is necessary to target preventative resources toward high-risk patients. Machine learning (ML) offers a novel approach to leveraging electronic health record (EHR) data for POD prediction. We sought to develop and internally validate a ML-derived POD risk prediction model using preoperative risk features, and to compare its performance to models developed with traditional logistic regression.

Methods

This was a retrospective analysis of preoperative EHR data from 24,885 adults undergoing a procedure requiring anesthesia care, recovering in the main post-anesthesia care unit, and staying in the hospital at least overnight between December 2016 and December 2019 at either of two hospitals in a tertiary care health system. One hundred fifteen preoperative risk features including demographics, comorbidities, nursing assessments, surgery type, and other preoperative EHR data were used to predict postoperative delirium (POD), defined as any instance of Nursing Delirium Screening Scale ≥2 or positive Confusion Assessment Method for the Intensive Care Unit within the first 7 postoperative days. Two ML models (Neural Network and XGBoost), two traditional logistic regression models (“clinician-guided” and “ML hybrid”), and a previously described delirium risk stratification tool (AWOL-S) were evaluated using the area under the receiver operating characteristic curve (AUC-ROC), sensitivity, specificity, positive likelihood ratio, and positive predictive value. Model calibration was assessed with a calibration curve. Patients with no POD assessments charted or at least 20% of input variables missing were excluded.

Results

POD incidence was 5.3%. The AUC-ROC for Neural Net was 0.841 [95% CI 0. 816–0.863] and for XGBoost was 0.851 [95% CI 0.827–0.874], which was significantly better than the clinician-guided (AUC-ROC 0.763 [0.734–0.793], p < 0.001) and ML hybrid (AUC-ROC 0.824 [0.800–0.849], p < 0.001) regression models and AWOL-S (AUC-ROC 0.762 [95% CI 0.713–0.812], p < 0.001). Neural Net, XGBoost, and ML hybrid models demonstrated excellent calibration, while calibration of the clinician-guided and AWOL-S models was moderate; they tended to overestimate delirium risk in those already at highest risk.

Conclusion

Using pragmatically collected EHR data, two ML models predicted POD in a broad perioperative population with high discrimination. Optimal application of the models would provide automated, real-time delirium risk stratification to improve perioperative management of surgical patients at risk for POD.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12871-021-01543-y.

Derivation, Validation, Sustained Performance, and Clinical Impact of an Electronic Medical Record-Based Perioperative Delirium Risk Stratification Tool

BACKGROUND

Postoperative delirium is an important problem for surgical inpatients and was the target of a multidisciplinary quality improvement project at our institution. We developed and tested a semiautomated delirium risk stratification instrument, Age, WORLD backwards, Orientation, iLlness severity, Surgery-specific risk (AWOL-S), in 3 independent cohorts from our tertiary care hospital and describe its performance characteristics and impact on clinical care.

METHODS

The risk stratification instrument was derived with elective surgical patients who were admitted at least overnight and received at least 1 postoperative delirium screen (Nursing Delirium Screening Scale [NuDESC] or Confusion Assessment Method for the Intensive Care Unit [CAM-ICU]) and preoperative cognitive screening tests (orientation to place and ability to spell WORLD backward). Using data pragmatically collected between December 7, 2016, and June 15, 2017, we derived a logistic regression model predicting probability of delirium in the first 7 postoperative hospital days. A priori predictors included age, cognitive screening, illness severity or American Society of Anesthesiologists physical status, and surgical delirium risk. We applied model odds ratios to 2 subsequent cohorts ("validation" and "sustained performance") and assessed performance using area under the receiver operator characteristic curves (AUC-ROC). A post hoc sensitivity analysis assessed performance in emergency and preadmitted patients. Finally, we retrospectively evaluated the use of benzodiazepines and anticholinergic medications in patients who screened at high risk for delirium.

RESULTS

The logistic regression model used to derive odds ratios for the risk prediction tool included 2091 patients. Model AUC-ROC was 0.71 (0.67-0.75), compared with 0.65 (0.58-0.72) in the validation (n = 908) and 0.75 (0.71-0.78) in the sustained performance (n = 3168) cohorts. Sensitivity was approximately 75% in the derivation and sustained performance cohorts; specificity was approximately 59%. The AUC-ROC for emergency and preadmitted patients was 0.71 (0.67-0.75; n = 1301). After AWOL-S was implemented clinically, patients at high risk for delirium (n = 3630) had 21% (3%-36%) lower relative risk of receiving an anticholinergic medication perioperatively after controlling for secular trends.

CONCLUSIONS

The AWOL-S delirium risk stratification tool has moderate accuracy for delirium prediction in a cohort of elective surgical patients, and performance is largely unchanged in emergent/preadmitted surgical patients. Using AWOL-S risk stratification as a part of a multidisciplinary delirium reduction intervention was associated with significantly lower rates of perioperative anticholinergic but not benzodiazepine, medications in those at high risk for delirium. AWOL-S offers a feasible starting point for electronic medical record-based postoperative delirium risk stratification and may serve as a useful paradigm for other institutions.