Risk Prediction Models for Post-Stroke Dementia

Predicting post-stroke cognitive impairment using electronic health record data

BACKGROUND

Secondary prevention interventions to reduce post-stroke cognitive impairment (PSCI) can be aided by the early identification of high-risk individuals who would benefit from risk factor modification.

AIMS

To develop and evaluate a predictive model to identify patients at increased risk of PSCI over 5 years using data easily accessible from electronic health records.

METHODS

Cohort study that included primary care patients from two academic medical centers. Patients were aged 45 years or older, without prior stroke or prevalent cognitive impairment, with primary care visits and an incident ischemic stroke between 2003 and 2016 (development/internal validation cohort) or 2010 and 2022 (external validation cohort). Predictors of PSCI were ascertained from the electronic health record. The outcome was incident dementia/cognitive impairment within 5 years and beginning 3 months following stroke, ascertained using International Classification of Diseases, Ninth/Tenth Revision (ICD-9/10) codes. For model variable selection, we considered potential predictors of PSCI and constructed 400 bootstrap samples with two-thirds of the model derivation sample. We ran 10-fold cross-validated Cox proportional hazards models using a least absolute shrinkage and selection operator (LASSO) penalty. Variables selected in >25% of samples were included.

RESULTS

The analysis included 332 incident diagnoses of PSCI in the development cohort (n = 3741), and 161 and 128 incident diagnoses in the internal (n = 1925) and external (n = 2237) validation cohorts, respectively. The C-statistic for predicting PSCI was 0.731 (95% confidence interval (CI): 0.694-0.768) in the internal validation cohort, and 0.724 (95% CI: 0.681-0.766) in the external validation cohort. A risk score based on the beta coefficients of predictors from the development cohort stratified patients into low (0-7 points), intermediate (8-11 points), and high (12-23 points) risk groups. The hazard ratios (HRs) for incident PSCI were significantly different by risk categories in internal (high, HR: 6.2, 95% CI: 4.1-9.3; Intermediate, HR: 2.7, 95% CI: 1.8-4.1) and external (high, HR: 6.1, 95% CI: 3.9-9.6; Intermediate, HR: 2.8, 95% CI: 1.9-4.3) validation cohorts.

CONCLUSION

Five-year risk of PSCI can be accurately predicted using routinely collected data. Model output can be used to risk stratify and identify individuals at increased risk for PSCI for preventive efforts.

DATA ACCESS STATEMENT

Mass General Brigham data contain protected health information and cannot be shared publicly. The data processing scripts used to perform analyses will be made available to interested researchers upon reasonable request to the corresponding author.

Domain-specific cognitive impairment 6 months after stroke: The value of early cognitive screening

BACKGROUND

Cognitive screening following stroke is widely recommended, yet few studies have considered the prognostic value of acute domain-specific function for longer-term cognitive outcome. Identifying which post-stroke cognitive impairments more commonly occur, recover, and persist, and which impairments hold prognostic value, could inform care planning, and resource allocation.

AIMS

This study aimed to determine the prevalence of domain-specific impairment acutely and at 6 months, assess the proportion of change in cognitive performance, and examine the prognostic value of acute domain-specific cognitive screening.

METHODS

A prospective stroke cohort completed the Oxford Cognitive Screen acutely (⩽2 weeks) and 6 months post-stroke. We determined the prevalence of acute and 6-month domain-specific impairment and proportion of change in performance from acute to 6 months. Hierarchical multivariable regression was used to predict global and domain-specific cognitive impairment at 6 months adjusted for demographic/vascular factors, stroke severity, and lesion volume.

RESULTS

A total of 430 stroke survivors (mean/SD age 73.9/12.5 years, 46.5% female, median/interquartile range (IQR) National Institute of Health Stroke Scale (NIHSS) 5/2-10) completed 6-month follow-up. Acutely, domain-specific impairments were highly prevalent ranging from 26.7% (n = 112) in praxis to 46.8% (n = 183) in attention. At 6 months, the proportion of domain-specific recovery was highest in praxis (n = 73, 71%) and lowest in language (n = 89, 46%) and memory (n = 82, 48%). Severity of 6-month cognitive impairment was best predicted by the addition of acute cognitive impairment (adj R2 = 0.298, p < 0.0001) over demographic and clinical factors alone (adj R2 = 0.105, p < 0.0001). Acute cognitive function was the strongest predictor of 6-month cognitive performance (p < 0.0001). Acute domain-specific impairments in memory (p < 0.0001), language (p < 0.0001), and praxis (p < 0.0001) significantly predicted overall severity of cognitive impairment at 6 months.

CONCLUSION

Post-stroke cognitive impairment is highly prevalent across all domains acutely, while impairments in language, memory, and attention predominate at 6 months. Early domain-specific screening can provide valuable prognostic information for longer-term cognitive outcomes.

Predicting post-stroke cognitive impairment using electronic health record data

Importance

Secondary prevention interventions to reduce post-stroke cognitive impairment (PSCI) can be aided by the early identification of high-risk individuals who would benefit from risk factor modification.

Objective

To develop and evaluate a predictive model to identify patients at increased risk of PSCI over 5 years using data easily accessible from electronic health records.

Design

Cohort study with patients enrolled between 2003-2016 with follow-up through 2022.

Setting

Primary care practices affiliated with two academic medical centers.

Participants

Individuals 45 years or older, without prior stroke or prevalent cognitive impairment, with primary care visits and an incident ischemic stroke between 2003-2016 (development/internal validation cohort) or 2010-2022 (external validation cohort).

Exposures

Predictors of PSCI were ascertained from the electronic health record.

Main Outcome

The outcome was incident dementia/cognitive impairment within 5 years and beginning 3 months following stroke, ascertained using ICD-9/10 codes. For model variable selection, we considered potential predictors of PSCI and constructed 400 bootstrap samples with two-thirds of the model derivation sample. We ran 10-fold cross-validated Cox proportional hazards models using a least absolute shrinkage and selection operator (LASSO) penalty. Variables selected in >25% of samples were included.

Results

The analysis included 332 incident diagnoses of PSCI in the development cohort (n=3,741), and 161 and 128 incident diagnoses in the internal (n=1,925) and external (n=2,237) validation cohorts. The c-statistic for predicting PSCI was 0.731 (95% CI: 0.694-0.768) in the internal validation cohort, and 0.724 (95% CI: 0.681-0.766) in the external validation cohort. A risk score based on the beta coefficients of predictors from the development cohort stratified patients into low (0-7 points), intermediate (8-11 points), and high (12-35 points) risk groups. The hazard ratios for incident PSCI were significantly different by risk categories in internal (High, HR: 6.2, 95% CI 4.1-9.3; Intermediate, HR 2.7, 95% CI: 1.8-4.1) and external (High, HR: 6.1, 95% CI: 3.9-9.6; Intermediate, HR 2.8, 95% CI: 1.9-4.3) validation cohorts.

Conclusions and Relevance

Five-year risk of PSCI can be accurately predicted using routinely collected data. Model output can be used to risk stratify and identify individuals at increased risk for PSCI for preventive efforts.

Understanding and Predicting Cognitive Improvement of Young Adults in Ischemic Stroke Rehabilitation Therapy

Accurate early predictions of a patient's likely cognitive improvement as a result of a stroke rehabilitation programme can assist clinicians in assembling more effective therapeutic programs. In addition, sufficient levels of explainability, which can justify these predictions, are a crucial requirement, as reported by clinicians. This article presents a machine learning (ML) prediction model targeting cognitive improvement after therapy for stroke surviving patients. The prediction model relies on electronic health records from 201 ischemic stroke surviving patients containing demographic information, cognitive assessments at admission from 24 different standardized neuropsychology tests (e.g., TMT, WAIS-III, Stroop, RAVLT, etc.), and therapy information collected during rehabilitation (72,002 entries collected between March 2007 and September 2019). The study population covered young-adult patients with a mean age of 49.51 years and only 4.47% above 65 years of age at the stroke event (no age filter applied). Twenty different classification algorithms (from Python's Scikit-learn library) are trained and evaluated, varying their hyper-parameters and the number of features received as input. Best-performing models reported Recall scores around 0.7 and F1 scores of 0.6, showing the model's ability to identify patients with poor cognitive improvement. The study includes a detailed feature importance report that helps interpret the model's inner decision workings and exposes the most influential factors in the cognitive improvement prediction. The study showed that certain therapy variables (e.g., the proportion of memory and orientation executed tasks) had an important influence on the final prediction of the cognitive improvement of patients at individual and population levels. This type of evidence can serve clinicians in adjusting the therapeutic settings (e.g., type and load of therapy activities) and selecting the one that maximizes cognitive improvement.

Risk Prediction of Cognitive Decline after Stroke

BACKGROUND AND PURPOSE

Cognitive decline is one of the major outcomes after stroke. We have developed and evaluated a risk predictive tool of post-stroke cognitive decline and assessed its clinical utility.

METHODS

In this population-based cohort, 4,783 patients with first-ever stroke from the South London Stroke Register (1995-2010) were included in developing the model. Cognitive impairment was measured using the Mini Mental State Examination (cut off 24/30) and the Abbreviated Mental Test (cut off 8/10) at 3-months and yearly thereafter. A penalised mixed-effects linear model was developed and temporal-validated in a new cohort consisted of 1,718 stroke register participants recruited from (2011-2018). Prediction errors on discrimination and calibration were assessed. The clinical utility of the model was evaluated using prognostic accuracy measurements and decision curve analysis.

RESULTS

The overall predictive model showed good accuracy, with root mean squared error of 0.12 and R2 of 73%. Good prognostic accuracy for predicting severe cognitive decline was observed AUC: (88%, 95% CI [85-90]), (89.6%, 95% CI [86-92]), (87%, 95% CI [85-91]) at 3 months, one and 5 years respectively. Average predicted recovery patterns were analysed by age, stroke subtype, Glasgow-coma scale, and left-stroke and showed variability. DECISION: curve analysis showed an increased clinical benefit, particularly at threshold probabilities of above 15% for predictive risk of cognitive impairment.

CONCLUSIONS

The derived prognostic model seems to accurately screen the risk of post-stroke cognitive decline. Such prediction could support the development of more tailored management evaluations and identify groups for further study and future trials.

The Prevalence of Cognitive Impairment on Admission to Nursing Home among Residents with and without Stroke: A Cross-Sectional Survey of Nursing Homes in Ireland

Post-stroke cognitive impairment (PSCI) is a common consequence of stroke. Epidemiological evidence indicates that, with an ageing population, stroke and PSCI are likely to increase in the coming decades. This may have considerable implications for the demand for nursing home placement. As prevalence estimates of both cognitive impairment and dementia on admission to nursing home among residents with and without stroke have not yet been compared, they were estimated and compared in this study. We performed a cross-sectional survey to establish the admission characteristics of 643 residents in 13 randomly selected nursing homes in Ireland. The survey collected data on resident's stroke and cognitive status at the time of nursing home admission. The survey found, among nursing home residents that experienced stroke prior to admission, prevalence estimates for cognitive impairment (83.8%; 95% CI = 76.9-90.6%) and dementia (66.7%; 95% CI = 57.9-75.4%) were significantly higher compared to residents that had not experienced stroke prior to admission (cognitive impairment: 56.6%; 95% CI = 52.4-60.8%; X2 (1) = 28.64; p < 0.001; dementia: 49.8%; 95% CI = 45.6-54.1%; X2 (1) = 10.47; p < 0.01). Since the prevalence of PSCI is likely to increase in the coming decades, the findings highlight an urgent need for health service planning for this increased demand for nursing home care to meet the care needs of these stroke survivors.

Longitudinal Effect of Stroke on Cognition: A Systematic Review

BACKGROUND

Stroke is associated with an increased risk of dementia; however, the impact of stroke on cognition has been found to be variable, such that stroke survivors can show decline, remain stable, or revert to baseline cognitive functioning. Knowing the natural history of cognitive impairment after stroke is important for intervention. The aim of this systematic review is to investigate the longitudinal course of cognitive function in stroke survivors.

METHODS AND RESULTS

Three electronic databases (Medline, Embase, PsycINFO) were searched using OvidSP from inception to July 15, 2016. Longitudinal studies with ≥2 time points of cognitive assessment after stroke were included. In total, 5952 articles were retrieved and 14 were included. There was a trend toward significant deterioration in cognitive test scores in stroke survivors (8 studies). Cognitive stability (3 studies) and improvement (3 studies) were also demonstrated, although follow-up time tended to be shorter in these studies. Variables associated with impairment included age, ethnicity, premorbid cognitive performance, depression, stroke location, and history of previous stroke. Associations with APOE*E4 (apolipoprotein E with the E4 allele) allele status and sex were mixed.

CONCLUSIONS

Stroke is associated with an increased risk of cognitive decline, but cognitive decline is not a consequence. Factors associated with decline, such as sociodemographic status, health-related comorbidity, stroke history, and clinical features could be used in models to predict future risk of dementia after stroke. A risk model approach could identify patients at greatest risk for timely intervention to reduce the frequency or delay the onset of poststroke cognitive impairment and dementia.