Associations between early in-hospital medications and the development of delirium in patients with stroke

Prediction of stroke severity: systematic evaluation of lesion representations

OBJECTIVE

To systematically evaluate which lesion-based imaging features and methods allow for the best statistical prediction of poststroke deficits across independent datasets.

METHODS

We utilized imaging and clinical data from three independent datasets of patients experiencing acute stroke (N1 = 109, N2 = 638, N3 = 794) to statistically predict acute stroke severity (NIHSS) based on lesion volume, lesion location, and structural and functional disconnection with the lesion location using normative connectomes.

RESULTS

We found that prediction models trained on small single-center datasets could perform well using within-dataset cross-validation, but results did not generalize to independent datasets (median R2 N1 = 0.2%). Performance across independent datasets improved using large single-center training data (R2 N2 = 15.8%) and improved further using multicenter training data (R2 N3 = 24.4%). These results were consistent across lesion attributes and prediction models. Including either structural or functional disconnection in the models outperformed prediction based on volume or location alone (P < 0.001, FDR-corrected).

INTERPRETATION

We conclude that (1) prediction performance in independent datasets of patients with acute stroke cannot be inferred from cross-validated results within a dataset, as performance results obtained via these two methods differed consistently, (2) prediction performance can be improved by training on large and, importantly, multicenter datasets, and (3) structural and functional disconnection allow for improved prediction of acute stroke severity.

Exploring Antipsychotic Use for Delirium Management in Adults in Hospital, Sub-Acute Rehabilitation and Aged Care Settings: A Systematic Literature Review

BACKGROUND

International guidelines discourage antipsychotic use for delirium; however, concerns persist about their continued use in clinical practice.

OBJECTIVES

We aimed to describe the prevalence and patterns of antipsychotic use in delirium management with regard to best-practice recommendations. Primary outcomes investigated were prevalence of use, antipsychotic type, dosage and clinical indication.

METHODS

Eligibility criteria: studies of any design that examined antipsychotic use to manage delirium in adults in critical care, acute care, palliative care, rehabilitation, and aged care were included. Studies of patients in acute psychiatric care, with psychiatric illness or pre-existing antipsychotic use were excluded.

INFORMATION SOURCES

we searched five health databases on 16 August, 2023 (PubMed, CINAHL, Embase, APA PsycInfo, ProQuest Health and Medical Collection) using MeSH terms and relevant keywords, including 'delirium' and 'antipsychotic'. Risk of bias: as no included studies were randomised controlled trials, all studies were assessed for methodological quality using the Mixed Methods Appraisal Tool.

SYNTHESIS OF RESULTS

descriptive data were extracted in Covidence and synthesised in Microsoft Excel.

RESULTS

Included studies: 39 studies published between March 2004 and August 2023 from 13 countries (n = 1,359,519 patients). Most study designs were retrospective medical record audits (n = 16).

SYNTHESIS OF RESULTS

in 18 studies, participants' mean age was ≥65 years (77.79, ±5.20). Palliative care had the highest average proportion of patients with delirium managed with antipsychotics (70.87%, ±33.81%); it was lower and varied little between intensive care unit (53.53%, ±19.73%) and non-intensive care unit settings [medical, surgical and any acute care wards] (56.93%, ±26.44%) and was lowest in in-patient rehabilitation (17.8%). Seventeen different antipsychotics were reported on. In patients aged ≥65 years, haloperidol was the most frequently used and at higher than recommended mean daily doses (2.75 mg, ±2.21 mg). Other antipsychotics commonly administered were olanzapine (mean 11 mg, ±8.54 mg), quetiapine (mean 64.23 mg, ±43.20 mg) and risperidone (mean 0.97 mg, ±0.64 mg).

CONCLUSIONS

The use of antipsychotics to manage delirium is strongly discouraged in international guidelines. Antipsychotic use in delirium care is a risk for adverse health outcomes and a longer duration of delirium, especially in older people. However, this study has provided evidence that clinicians continue to use antipsychotics for delirium management, the dose, frequency and duration of which are often outside evidence-based guideline recommendations. Clinicians continue to choose antipsychotics to manage delirium symptoms to settle agitation and maintain patient and staff safety, particularly in situations where workload pressures are high. Sustained efforts are needed at the individual, team and organisational levels to educate, train and support clinicians to prioritise non-pharmacological interventions early before deciding to use antipsychotics. This could prevent delirium and avert escalation in behavioural symptoms that often lead to antipsychotic use.

Scaling behaviours of deep learning and linear algorithms for the prediction of stroke severity

Deep learning has allowed for remarkable progress in many medical scenarios. Deep learning prediction models often require 105-107 examples. It is currently unknown whether deep learning can also enhance predictions of symptoms post-stroke in real-world samples of stroke patients that are often several magnitudes smaller. Such stroke outcome predictions however could be particularly instrumental in guiding acute clinical and rehabilitation care decisions. We here compared the capacities of classically used linear and novel deep learning algorithms in their prediction of stroke severity. Our analyses relied on a total of 1430 patients assembled from the MRI-Genetics Interface Exploration collaboration and a Massachusetts General Hospital-based study. The outcome of interest was National Institutes of Health Stroke Scale-based stroke severity in the acute phase after ischaemic stroke onset, which we predict by means of MRI-derived lesion location. We automatically derived lesion segmentations from diffusion-weighted clinical MRI scans, performed spatial normalization and included a principal component analysis step, retaining 95% of the variance of the original data. We then repeatedly separated a train, validation and test set to investigate the effects of sample size; we subsampled the train set to 100, 300 and 900 and trained the algorithms to predict the stroke severity score for each sample size with regularized linear regression and an eight-layered neural network. We selected hyperparameters on the validation set. We evaluated model performance based on the explained variance (R2) in the test set. While linear regression performed significantly better for a sample size of 100 patients, deep learning started to significantly outperform linear regression when trained on 900 patients. Average prediction performance improved by ∼20% when increasing the sample size 9× [maximum for 100 patients: 0.279 ± 0.005 (R2, 95% confidence interval), 900 patients: 0.337 ± 0.006]. In summary, for sample sizes of 900 patients, deep learning showed a higher prediction performance than typically employed linear methods. These findings suggest the existence of non-linear relationships between lesion location and stroke severity that can be utilized for an improved prediction performance for larger sample sizes.