Artificial Intelligence-Based Models for Prediction of Mortality in ICU Patients: A Scoping Review.

BACKGROUND AND OBJECTIVE: Healthcare professionals may be able to anticipate more accurately a patient's timing of death and assess their possibility of recovery by implementing a real-time clinical decision support system. Using such a tool, the healthcare system can better understand a patient's condition and make more informed judgements about distributing limited resources. This scoping review aimed to analyze various death prediction AI (Artificial Intelligence) algorithms that have been used in ICU (Intensive Care Unit) patient populations. METHODS: The search strategy of this study involved keyword combinations of outcome and patient setting such as mortality, survival, ICU, terminal care. These terms were used to perform database searches in MEDLINE, Embase, and PubMed up to July 2022. The variables, characteristics, and performance of the identified predictive models were summarized. The accuracy of the models was compared using their Area Under the Curve (AUC) values. RESULTS: Databases search yielded an initial pool of 8271 articles. A two-step screening process was then applied: first, titles and abstracts were reviewed for relevance, reducing the pool to 429 articles. Next, a full-text review was conducted, further narrowing down the selection to 400 key studies. Out of 400 studies on different tools or models for prediction of mortality in ICUs, 16 papers focused on AI-based models which were ultimately included in this study that have deployed different AI-based and machine learning models to make a prediction about negative patient outcome. The accuracy and performance of the different models varied depending on the patient populations and medical conditions. It was found that AI models compared with traditional tools like SAP3 or APACHE IV score were more accurate in death prediction, with some models achieving an AUC of up to 92.9%. The overall mortality rate ranged from 5% to more than 60% in different studies. CONCLUSION: We found that AI-based models exhibit varying performance across different patient populations. To enhance the accuracy of mortality prediction, we recommend customizing models for specific patient groups and medical contexts. By doing so, healthcare professionals may more effectively assess mortality risk and tailor treatments accordingly. Additionally, incorporating additional variables-such as genetic information-into new models can further improve their accuracy.

Effects of mindfulness-based interventions on cognition in people with multiple sclerosis: a systematic review and meta-analysis of randomized controlled trials.

Introduction: Cognitive impairment affects up to 65% of people with multiple sclerosis (PwMS), undermining functional independence and quality of life. The objective of this study is to synthesize existing randomized controlled trial (RCT) evidence on the effects of Mindfulness-based interventions (MBIs) on cognitive function in PwMS. Methods: A systematic literature search was conducted to identify RCTs assessing MBIs effects on cognitive functioning in PwMS. Using pre-defined criteria, two independent reviewers screened titles, abstracts, and extracted data from included studies. Meta-analysis was performed, where possible, using a random effects model. Narrative synthesis was undertaken. Preferred Reporting Items for Systematic Reviews and Meta-analysis guidance was followed. PROSPERO_ID:(CRD42021286429). Results: Twelve eligible RCTs were identified, n=700 PwMS. MBIs included both standardized and tailored interventions, in-person and virtually. A variety of measures of cognitive functioning were reported. Five studies (n=254 PwMS) were included in meta-analysis; pooled results suggested MBIs effectively improved scores on the Paced Auditory Serial Addition Test (PASAT)-2 (SMD=0.38; 95% CI 0.06-0.71; I2 63%; p=0.02), whereas improvements were of borderline significance on the PASAT-3 (SMD=0.32; 95% CI -0.01-0.64; I2 65%; p=0.06), and, although trending to positive, were statistically insignificant on the Perceived Deficits Questionnaire (SMD=0.34; 95 CI -0.05-0.74; I2 0%; p=0.09) and Symbol Digits Modality Test (SMD=0.25; 95% CI -0.15-0.66; I2 0%; p=0.21). Conclusion: Preliminary findings in meta-analysis are inconsistent but suggest potential benefits from MBI training on cognitive functioning in PwMS. High quality RCTs are necessary to test more definitively the impact of MBIs on cognitive functioning in PwMS. Systematic review registration: PROSPERO, identifier CRD42021286429.

The Changing Prevalence of Pressure Injury among Ontarians with SCI/D at Rehabilitation Admission: Opportunities for Improvement.

BACKGROUND: Despite preventability, 20-50% of patients with acute spinal cord injury/disease (SCI/D) develop hospital-acquired pressure injuries (PIs). The Spinal Cord Injury Implementation and Evaluation Quality Care Consortium (SCI IEQCC) aimed to mitigate PI risk through patient-reported daily skin checks alongside usual care. METHODS: This quality improvement initiative utilized an interrupted time series design, encompassing adults ≥ 18 years admitted for inpatient rehabilitation across five Ontario sites from 2020 to 2023. Patient demographics, etiology, and impairment data were obtained from a national registry, while participating sites gathered data on PI onset, location, and severity. Run charts depicted temporal trends, and statistical analyses, including chi-square and logistic regression, compared patients with and without PIs. RESULTS: Data from 1767 discharged SCI/D patients revealed that 26% had ≥1 PI, with 59% being prevalent and 41% incident. Most severe PIs (stages III and IV and unstageable) were acquired prior to admission. Process indicator fidelity was reasonable at 68%. Patients with PIs experienced longer hospital stays, lower Functional Independence Measure (FIM) changes, and FIM efficiency during rehabilitation. CONCLUSIONS: PI prevalence is increasing, particularly sacral injuries at admission, while incident cases have decreased since 2021 due to regular skin checks. This trend calls for proactive health system interventions to reduce costs and improve patient outcomes.

Psychosocial Factors Associated with Prolonged Length of Stay in Acquired Brain Injury Rehabilitation: A Retrospective Cohort Study.

OBJECTIVE: In a climate of rising healthcare costs and increasing pressure to reduce inpatient length of stay, hospitals must balance their role as care providers with that as resource stewards. There is a need to understand what factors are associated with patients staying beyond rehabilitation length of stay targets. The aim of this study was to determine psychosocial patient factors that are identifiable on admission that influence length of stay targets in acquired brain injury rehabilitation. METHODS: A retrospective case series of 167 inpatients with acquired brain injury was conducted at an urban, academic rehabilitation hospital. A total of 29 factors were used for data analysis. Logistic and multiple linear regression analysis was utilized to determine if any patient factors were associated with patients exceeding their length of stay targets. RESULTS: Premorbid communal living status (e.g. group home) was associated with an odds ratio of 14.67 of exceeding length of stay target. Patients who did not drive prior to their admission had an odds ratio of 2.63 of exceeding their length of stay target. CONCLUSION: Premorbid communal living and premorbid non-driving status are predictors of patients with acquired brain injuries exceeding target rehabilitation length of stay. These findings may help acquired brain injury rehabilitation programmes plan for the needs of and advocate for patients.

Effect of mobile application types on stroke rehabilitation: a systematic review.

BACKGROUND: Stroke is a significant contributor of worldwide disability and morbidity with substantial economic consequences. Rehabilitation is a vital component of stroke recovery, but inpatient stroke rehabilitation programs can struggle to meet the recommended hours of therapy per day outlined by the Canadian Stroke Best Practices and American Heart Association. Mobile applications (apps) are an emerging technology which may help bridge this deficit, however this area is understudied. The purpose of this study is to review the effect of mobile apps for stroke rehabilitation on stroke impairments and functional outcomes. Specifically, this paper will delve into the impact of varying mobile app types on stroke rehabilitation. METHODS: This systematic review included 29 studies: 11 randomized control trials and 18 quasi-experimental studies. Data extrapolation mapped 5 mobile app types (therapy apps, education apps, rehab videos, reminders, and a combination of rehab videos with reminders) to stroke deficits (motor paresis, aphasia, neglect), adherence to exercise, activities of daily living (ADLs), quality of life, secondary stroke prevention, and depression and anxiety. RESULTS: There were multiple studies supporting the use of therapy apps for motor paresis or aphasia, rehab videos for exercise adherence, and reminders for exercise adherence. For permutations involving other app types with stroke deficits or functional outcomes (adherence to exercise, ADLs, quality of life, secondary stroke prevention, depression and anxiety), the results were either non-significant or limited by a paucity of studies. CONCLUSION: Mobile apps demonstrate potential to assist with stroke recovery and augment face to face rehabilitation, however, development of a mobile app should be carefully planned when targeting specific stroke deficits or functional outcomes. This study found that mobile app types which mimicked principles of effective face-to-face therapy (massed practice, task-specific practice, goal-oriented practice, multisensory stimulation, rhythmic cueing, feedback, social interaction, and constraint-induced therapy) and education (interactivity, feedback, repetition, practice exercises, social learning) had the greatest benefits. Protocol registration PROPSERO (ID CRD42021186534). Registered 21 February 2021.

Convergent validity and responsiveness of The Standing and Walking Assessment Tool (SWAT) among individuals with non-traumatic spinal cord injury.

Aim: This study aimed to (1) describe the use of the Standing and Walking Assessment Tool (SWAT) among individuals with non-traumatic spinal cord injury or disease (NT-SCI/D); (2) evaluate the convergent validity of SWAT for use among inpatients with NT-SCI/D; (3) describe SWAT responsiveness; and (4) explore the relationship between hours of walking therapy and SWAT change. Methods: A quality improvement project was conducted at the University Health Network between 2019 and 2022. Participants' demographics and impairments data, rehabilitation length of stay, and FIM scores were obtained from the National Rehabilitation Reporting System. The walking measure data were collected by therapists as part of routine practice. Hours of part- or whole-gait practice were abstracted from medical records. To determine convergent validity, Spearman's correlation coefficients were calculated between SWAT stages (admission and discharge) and the walking measures. The change in SWAT levels was calculated to determine responsiveness. Spearman's correlation coefficient was calculated between SWAT change and hours of walking therapy. Results: Among adult NT-SCI/D participants with potential walking capacity (SWAT≥1B), the majority were classified as American Spinal Injury Association (ASIA) Impairment Scale D (AIS D) at admission. The SWAT category of 1C (N = 100, 18%) was the most frequent at admission. The most frequent SWAT stage at discharge was 3C among participants with NT-SCI/D, with positive conversions in SWAT stages from admission to discharge (N = 276, 33%). The mean change in SWAT score was 3 for participants with T-SCI and NT-SCI/D. Moderate correlations between SWAT stages and walking measures were observed. The correlation of hours of gait therapy with the SWAT change (admission to discharge) was 0.44 (p < 0001). Conclusion: The SWAT has sufficient convergent validity and responsiveness for describing standing and walking recovery and communicating/monitoring rehabilitation progress among patients with NT-SCI/D.

Using Risk Scores to Estimate Lower Extremity Fragility Fracture Risk among Individuals with Chronic Spinal Cord Injury: A Preliminary Model.

Objectives: To develop SCI-FX, a risk score to estimate 5-year lower extremity fragility fracture risk among patients living with chronic spinal cord injury (cSCI). Methods: Adults with traumatic cSCI (n = 90) participated in a 2-year prospective longitudinal cohort study describing bone mineral density (BMD) change and fracture incidence conducted at the Lyndhurst Centre (University Health Network), University of Waterloo, and Physical Disability Rehabilitation Institute of Québec City. Prior publication and clinical intuition were used to identify fragility fracture risk factors including prior fragility fracture, years post-injury, motor complete injury (AIS A/B), benzodiazepine use, opioid use, and parental osteoporosis. We conducted bivariate analyses to identify variables associated with fracture. Multiple logistic regressions were performed using fragility fracture incidence as the dependent variable and all variables from the univariate analyses with a highly liberal p value at 0.2. Using the odds ratios (ORs) from the multiple logistic regression model, a point system for fragility fracture risk score was developed, and the odds of fracture for each point was estimated. Results: All initial variables, with the exception of benzodiazepine exposure, were included in the final model. Conclusion: We identified a simple preliminary model for clinicians to estimate 5-year fracture risk among patients with cSCI based on their total score.

Protocol for a single group, mixed methods study investigating the efficacy of photovoice to improve self-efficacy related to balance and falls for spinal cord injury.

INTRODUCTION: Many individuals living with spinal cord injury (SCI) experience falls and a fear of falling, both of which can impact participation in daily activities and quality of life. A single group, convergent mixed methods study will be conducted to examine the effects of a photovoice intervention on falls self-efficacy among individuals living with chronic SCI. Secondary objectives include examining the effects of photovoice on fear of falling, participation and quality of life and exploring participants' experiences and perceptions of the photovoice intervention through qualitative interviews. METHODS AND ANALYSIS: Adults with SCI (n=40) will be divided into groups according to their mobility status (ie, those who ambulate and those who primarily use a wheelchair). The study will be conducted virtually over three consecutive phases, totalling 30 weeks. Each group will self-report falls for 12 weeks prior to and following the intervention (phases 1 and 3, respectively). The 6-week photovoice intervention (phase 2) will be comprised of two photo assignments, two individual interviews with a researcher and a peer mentor, and four group meetings. Participants will discuss these photos at the interviews and group meetings. Standardised questionnaires of falls self-efficacy, fear of falling, participation and life satisfaction will be administered at four time points (ie, beginning of each phase and the end of phase 3). Questionnaire scores will be examined over time using repeated-measures analysis of variance. A semistructured interview will be completed at the end of phase 3 to gain feedback on the photovoice intervention. Qualitative data will be analysed using reflexive thematic analysis. ETHICS AND DISSEMINATION: Ethics approval was obtained prior to study enrolment. Findings will be shared through peer-reviewed scientific publications and participant-directed knowledge translation activities. TRIAL REGISTRATION NUMBER: NCT04864262.

Examination of the Relationships Among Social Networks and Loneliness on Health and Life Satisfaction in People with Spinal Cord Injury/Dysfunction.

OBJECTIVE: To examine the associations among social networks and loneliness on health and life satisfaction in adults with chronic spinal cord injury/dysfunction (SCI/D). DESIGN: Cross-sectional telephone survey study. SETTING: Tertiary spinal cord injury rehabilitation center in Ontario, Canada. PARTICIPANTS: Community-dwelling adults with chronic SCI/D (N=170). INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: The main outcome measures were the Short-Form 36 to assess health and the Life Satisfaction-11 to assess life satisfaction. RESULTS: A hierarchical regression model predicting physical health accounted for 24% (P<.001) of the variance. The only social network variable to significantly contribute to the model was having a higher proportion of network members living in one's household (P<.05). A model predicting mental health accounted for 44% (P<.001) of the variance, with having a higher proportion of network members living in one's household (P<.05) and lower feelings of loneliness (P<.001) associated with better mental health. Finally, the model predicting life satisfaction accounted for 62% (P<.001) of the variance, with lower greater levels of social network intimacy (P<.01) and lower feelings of loneliness (P<.001) being significant predictors. CONCLUSIONS: These findings highlight the importance of having access to network members in one's home for better physical and mental health after SCI/D as well as the negative association between loneliness and mental health and life satisfaction. There is a need for approaches to ensure that people with SCI/D in the community feel supported to mitigate feelings of loneliness to optimize their health and wellbeing.

Evaluating Intrinsic Fall Risk Factors After Incomplete Spinal Cord Injury: Distinguishing Fallers From Nonfallers.

OBJECTIVE: To determine whether performance on measures of lower extremity muscle strength, sensory function, postural control, gait speed, and balance self-efficacy could distinguish fallers from nonfallers among ambulatory individuals with spinal cord injury or disease (SCI/D). DESIGN: Prospective cohort study. SETTING: Community. PARTICIPANTS: Individuals (N=26; 6 female, aged 58.9±18.2y) with motor incomplete SCI/D (American Spinal Injury Association Impairment Scale rating C [n=5] or D [n=21]) participated. Participants were 7.5±9.1 years post injury. Seventeen participants experienced traumatic causes of spinal cord injury. MAIN OUTCOME MEASURES: Participants completed laboratory-based and clinical measures of postural control, gait speed, balance self-efficacy, and lower extremity strength, as well as proprioception and cutaneous pressure sensitivity. Participants were then followed for up to 1 year to track falls using a survey. The survey queried the circumstances and consequences of each fall. If a participant's number of falls equaled or exceeded the median number of falls experience by all participants, they were classified a faller. RESULTS: Median follow-up duration was 362 days and median time to first fall was 60.5 days. Fifteen participants were classified as fallers. Most falls occurred during the morning or afternoon (81%), at home (75%), and while walking (47%). The following laboratory-based and clinical measures distinguished fallers from nonfallers (P<.05): measures of lower extremity strength, cutaneous pressure sensitivity, walking speed, and center of pressure velocity in the mediolateral direction. CONCLUSIONS: There are laboratory-based and clinical measures that can prospectively distinguish fallers from nonfallers among ambulatory individuals with spinal cord injury. These findings may assist clinicians when evaluating their patients' fall risk.