Predictive modeling of ambulatory outcomes after spinal cord injury using machine learning.

STUDY DESIGN: Retrospective multi-site cohort study. OBJECTIVES: To develop an accurate machine learning predictive model using predictor variables from the acute rehabilitation period to determine ambulatory status in spinal cord injury (SCI) one year post injury. SETTING: Model SCI System (SCIMS) database between January 2000 and May 2019. METHODS: Retrospective cohort study using data that were previously collected as part of the SCI Model System (SCIMS) database. A total of 4523 patients were analyzed comparing traditional models (van Middendorp and Hicks) compared to machine learning algorithms including Elastic Net Penalized Logistic Regression (ENPLR), Gradient Boosted Machine (GBM), and Artificial Neural Networks (ANN). RESULTS: Compared with GBM and ANN, ENPLR was determined to be the preferred model based on predictive accuracy metrics, calibration, and variable selection. The primary metric to judge discrimination was the area under the receiver operating characteristic curve (AUC). When compared to the van Middendorp all patients (0.916), ASIA A and D (0.951) and ASIA B and C (0.775) and Hicks all patients (0.89), ASIA A and D (0.934) and ASIA B and C (0.775), ENPLR demonstrated improved AUC for all patients (0.931), ASIA A and D (0.965) ASIA B and C (0.803). CONCLUSIONS: Utilizing artificial intelligence and machine learning methods are feasible for accurately classifying outcomes in SCI and may provide improved sensitivity in identifying which individuals are less likely to ambulate and may benefit from augmentative strategies, such as neuromodulation. Future directions should include the use of additional variables to further refine these models.

Effect of physical therapy timing on patient-reported outcomes for individuals with acute low back pain: A systematic review with meta analysis of randomized controlled trials.

OBJECTIVE: The purpose of this systematic review with meta-analysis was to investigate the effect of early physical therapy (PT) for the management of acute low back pain (LBP) on patient-reported outcomes of pain and disability, compared to delayed PT or non-PT care. LITERATURE SURVEY: Randomized controlled trials in three electronic databases (MEDLINE, CINAHL, Embase) were searched from inception to June 12, 2020, and updated on September 23, 2021. METHODOLOGY: Eligible participants were individuals with acute low back pain. The intervention was early PT compared to delayed PT or non-PT care. Primary outcomes included the patient-reported outcomes of pain and disability. The following information was extracted from included articles: demographic data, sample size, selection criteria, PT interventions, and pain and disability outcomes. Data were extracted following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Methodological quality was assessed using the Physiotherapy Evidence Database (PEDro) Scale. Random effects models were used for the meta-analysis. SYNTHESIS: Seven of 391 articles met the eligibility criteria and were included in the meta-analysis. Random effects meta-analysis comparing early PT to non-PT care for acute LBP indicated a significant reduction in pain (standard mean difference [SMD] = 0.43, 95% confidence interval [CI]: -0.69 to -0.17) and disability (SMD = 0.36, 95% CI: -0.57 to -0.16) in the short term. Early PT compared to delayed PT did not result in improvement in short-term pain (SMD = -0.24, 95% CI: -0.52 to 0.04) or disability (SMD = 0.28, 95% CI: -0.56 to 0.01), or long-term pain (SMD = 0.21, 95% CI: -0.15 to 0.57) or disability (SMD = 0.14, 95% CI: -0.15 to 0.42). CONCLUSIONS: This systematic review and meta-analysis suggest early PT versus non-PT care is associated with statistically significant reductions in short-term pain and disability (up to 6 weeks) with small effect sizes. The results indicate a nonsignificant trend favoring a small benefit of early PT over delayed PT for outcomes at short-term follow-up but no effect at long-term follow-up (6 months or greater).

Intrapulmonary and Intracardiac Shunts in Adult COVID-19 Versus Non-COVID Acute Respiratory Distress Syndrome ICU Patients Using Echocardiography and Contrast Bubble Studies (COVID-Shunt Study): A Prospective, Observational Cohort Study.

OBJECTIVES: Studies have suggested intrapulmonary shunts may contribute to hypoxemia in COVID-19 acute respiratory distress syndrome (ARDS) with worse associated outcomes. We evaluated the presence of right-to-left (R-L) shunts in COVID-19 and non-COVID ARDS patients using a comprehensive hypoxemia workup for shunt etiology and associations with mortality. DESIGN: Prospective, observational cohort study. SETTING: Four tertiary hospitals in Edmonton, Alberta, Canada. PATIENTS: Adult critically ill, mechanically ventilated, ICU patients admitted with COVID-19 or non-COVID (November 16, 2020, to September 1, 2021). INTERVENTIONS: Agitated-saline bubble studies with transthoracic echocardiography/transcranial Doppler ± transesophageal echocardiography assessed for R-L shunts presence. MEASUREMENTS AND MAIN RESULTS: Primary outcomes were shunt frequency and association with hospital mortality. Logistic regression analysis was used for adjustment. The study enrolled 226 patients (182 COVID-19 vs 42 non-COVID). Median age was 58 years (interquartile range [IQR], 47-67 yr) and Acute Physiology and Chronic Health Evaluation II scores of 30 (IQR, 21-36). In COVID-19 patients, the frequency of R-L shunt was 31 of 182 COVID patients (17.0%) versus 10 of 44 non-COVID patients (22.7%), with no difference detected in shunt rates (risk difference [RD], -5.7%; 95% CI, -18.4 to 7.0; p = 0.38). In the COVID-19 group, hospital mortality was higher for those with R-L shunt compared with those without (54.8% vs 35.8%; RD, 19.0%; 95% CI, 0.1-37.9; p = 0.05). This did not persist at 90-day mortality nor after adjustment with regression. CONCLUSIONS: There was no evidence of increased R-L shunt rates in COVID-19 compared with non-COVID controls. R-L shunt was associated with increased in-hospital mortality for COVID-19 patients, but this did not persist at 90-day mortality or after adjusting using logistic regression.

Electroencephalogram patterns in critical care: A primer for acute care doctors.

Electroencephalograms are commonly ordered by acute care doctors but not always understood. Other reviews have covered when and how to perform electroencephalograms. This primer has a different, unique, and complementary goal. We review basic electroencephalogram interpretation and terminology for nonexperts. Our goal is to encourage common understanding, facilitate inter specialty collaboration, dispel common misunderstandings, and inform the current and future use of this precious resource. This primer is categorically not to replace the expert neurologist or technician. Quite the contrary, it should help explain how nuanced electroencephalogram can be, and why indiscriminate electroencephalogram is inappropriate. Some might argue not to teach nonexperts lest they overestimate their abilities or reach. We humbly submit that it is even more inappropriate to not know the basics of a test that is ordered frequently and resource intensive. We cover the characteristics of the "normal" electroencephalogram, electroencephalogram slowing, periodic epileptiform discharges (and its subtypes), burst suppression, and electrographic seizures (and its subtypes). Alongside characteristic electroencephalogram findings, we provide clinical pearls. These should further explain what the reporter is communicating and whether additional testing is beneficial. Along with teaching the basics and whetting the appetite of the general clinician, this resource could increase mutual understanding and mutual appreciation between those who order electroencephalograms and those who interpret them. While there is more to electroencephalogram than can be delivered via a single concise primer, it offers a multidisciplinary starting point for those interested in the present and future of this commonly ordered test.

Ibuprofen use is associated with reduced C-reactive protein and interleukin-6 levels in chronic spinal cord injury.

Objective: To assess the association between ibuprofen use and the systemic inflammatory biomarkers C-reactive protein (CRP) and interleukin-6 (IL-6) in chronic Spinal Cord Injury (SCI).Study design: Prospective cohort study.Setting: Community dwelling individuals with SCI.Participants: 338 (278 male, 60 female) community dwelling individuals with chronic SCI (≥1-year post-injury).Interventions: None.Main outcome measures: CRP and IL-6 levels were quantified by ultra-sensitive ELISA assay. General linear models were used to assess associations between various clinical and demographic factors and CRP and IL-6 levels.Results: There were 50 active ibuprofen users and 288 non-users. After adjusting for clinical and demographic factors, ibuprofen users had significantly lower CRP levels (2.3 mg/L versus 3.5 mg/L, P = 0.04) and IL-6 levels (3.2 pg/ml versus 4.0 pg/ml, P = 0.04) compared to nonusers.Conclusions: Our study suggests that self-reported ibuprofen use may be negatively associated with CRP and IL-6 levels in chronic SCI after adjusting for known confounding factors, and suggests ibuprofen use may be an important, potential variable to consider in future studies focused on systemic inflammation in SCI. Future prospective studies require assessing frequency, duration, and dosage-dependent effects of ibuprofen on systemic markers of inflammation in chronic SCI. These findings may support future clinical trials to determine safety and efficacy of ibuprofen treatment for various outcomes in chronic SCI.

Study Registration for the Field of Prevention Science: Considering Options and Paths Forward.

The practice of prospectively registering the details of intervention studies in a public database or registry is gaining momentum across disciplines as a strategy for increasing the transparency, credibility, and accessibility of study findings. In this article, we consider five registries that may be relevant for registration of intervention studies in the field of prevention science: ClinicalTrials.gov, the American Economic Association Registry of Randomized Controlled Trials (AEA RCT Registry), the Open Science Framework Preregistration (OSF Preregistration), the Registry for International Development Impact Evaluations (RIDIE), and the Registry of Efficacy and Effectiveness Studies (REES). We examine the five registries in terms of substantive focus, study designs, and contents of registry entries. We consider two paths forward for prospective registration of intervention studies in the field of prevention science: Path A: register all studies in ClinicalTrials.gov and Path B: allow individual researchers to select the registry with the "best fit." Lastly, we consider how the field might begin to establish norms around registration.

Extreme Delta Brush in Anti-NMDAR Encephalitis Correlates With Poor Functional Outcome and Death.

Objective: To characterize EEG findings in anti-NMDAR encephalitis patients looking for the proportion of EEGs that were abnormal, presence of extreme delta brush (EDB), and to relate EEG findings to clinical outcomes (Glasgow Outcome Scale (GOS) at 6 months, need for ICU admission, and death). Methods: This retrospective cohort single center study included patients with anti-NMDAR encephalitis who had ≥1 EEGs obtained from 2014 to 2021. EEGs were retrospectively analyzed by 2 reviewers. Clinical outcomes of interest were extracted through hospital and clinic chart review. Results: Twenty-one patients with anti-NMDAR encephalitis were included. Sixty-four EEGs were analyzed. Four EEGs (6.3%) were within normal limits. Focal or generalized slowing (without EDB) was seen on 44 EEGs (68.8%). EDB was seen on 16 EEGs (25.0%) in 9 of 21 patients (42.9%). The presence of EDB was significantly associated with need for ICU admission (p = 0.02), poorer outcome at 6 months as per the GOS (p = 0.002), and with death (p=0.02). EDB was present on ≥1 EEG of every patient who died. Conclusions: The presence of EDB on EEG in anti-NMDAR encephalitis patients is associated with increased need for ICU admission, worse functional outcomes at 6 months, and risk of death.

Anti-N-methyl-d-aspartate receptor encephalitis: A primer for acute care healthcare professionals.

This primer summarizes the diagnosis, treatment, complications, and prognosis of anti-N-methyl-d-aspartate receptor encephalitis for healthcare professionals, especially those in acute care specialities. Anti-N-methyl-d-aspartate receptor encephalitis is an immune-mediated encephalitis that is classically paraneoplastic and associated with ovarian teratomas in young women. Other less common neoplastic triggers include testicular cancers, Hodgkin lymphoma, lung and breast cancers. It may also be triggered by infection, occurring as a para-infectious phenomenon, seen most commonly after herpes simplex-1 encephalitis. Presentation varies but typically consists of behavioural and cognitive manifestations, seizures, dysautonomia, movement disorders, central hypoventilation, and coma, necessitating intensive care unit admission. Diagnosis of anti-N-methyl-d-aspartate receptor encephalitis requires high clinical suspicion plus ancillary testing, the most sensitive being cerebrospinal fluid analysis for anti-N-methyl-d-aspartate receptor antibodies. Imaging in search of an ovarian teratoma should be exhaustive and tumours need to be surgically treated. Treatment should be expeditious with pulsed steroids and either plasma exchange or intravenous immunoglobulin. Second-line treatments include intravenous rituximab, cyclophosphamide, azathioprine, and intrathecal methotrexate. Most patients recover to be functionally independent, but the in-hospital course can be months long followed by extensive rehabilitation. Given the lengthy course of illness, we explain why education and debriefing are important for staff, and where families can obtain additional help.

Trends in the Physical Medicine and Rehabilitation Match: Analysis of NRMP Data from 2007 to 2018.

INTRODUCTION: We provide an updated analysis of data about U.S. Physical Medicine and Rehabilitation (PM&R) residency program applicants collected by the National Resident Matching Program (NRMP). OBJECTIVE: Analyze trends within NRMP data for PM&R residency match rates, compare matched to unmatched applicants, and compare PM&R applicants to other medical specialties. DESIGN: Secondary analysis of NRMP data. SETTING: NRMP data set. PARTICIPANTS: Residency program applicants who participated in the NRMP Match, 2007 to 2018. INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: Number of applicants, match rates, difference in characteristics including rank order list (ROL), U.S. Medical Licensing Examination (USMLE) Step 1 and Step 2 Clinical Knowledge (CK) scores, publications, Alpha Omega Alpha (AOA) status, PhD degree, and experiences in research, volunteer, and work. RESULTS: Number of applicants and residency positions increased from 2007 to 2018. Length of ROL increased and was longer for matched compared to unmatched applicants, with maximum mean difference of 7.4 in 2016 (95% confidence interval [CI] 5.6-9.2). Matched U.S. Allopathic Seniors had higher USMLE scores compared to unmatched, with a mean difference of 12.7 for Step 1 (95% CI 8.3-17.0) and 12.6 (95% CI 8.6-16.6) for Step 2 CK (P < .001). Number of publications and volunteer experiences were higher for matched U.S. Allopathic Seniors (0.64, 95% CI 0.09-1.2 and 1.5, 95% CI 0.65-2.3, respectively). PM&R USMLE Step 1 and 2 CK scores increased at a significantly faster rate than for all other specialties, with estimated rate differences of 0.46 (95% CI 0.21-0.71) and 0.69 (95% CI 0.45-0.93) points per year, respectively. CONCLUSIONS: PM&R residency has become more competitive. USMLE Step 1 and 2 CK scores have outpaced the inflation of scores in other specialties. ROL length has increased, suggesting more ranked programs to successfully match. These analyses update our knowledge about PM&R residency applicants and suggest surrogate markers for a successful match.

Oophorectomy in NMDA receptor encephalitis and negative pelvic imaging.

Ovarian teratomas are found in one-third of females presenting with anti-N-methyl-D-aspartate receptor (NMDAR) encephalitis. When a teratoma is detected on imaging, its removal is first-line therapy. Even with multiple imaging modalities, occasionally, the teratoma is found only on subsequent imaging, long after initial presentation. Very rarely, patients have undergone oophorectomy despite negative imaging, with pathology demonstrating teratoma, and resulting clinical improvement. We present a patient in whom removal of a teratoma, not visible on conventional imaging, resulted in marked clinical improvement. Such cases present a major clinical challenge, needing to consider the risks of oophorectomy, including sterilisation and early menopause, versus the possibility of death in the absence of response to first-line (eg, corticosteroids, plasma exchange, intravenous immunoglobulin), second-line (eg, rituximab) and third-line (eg, bortezomib) immunosuppression. This decision is made more difficult as patients are usually females of childbearing age who at the time lack capacity to make medical decisions. This case also highlights the lack of consensus and guidelines for imaging modalities used to detect teratoma and when to pursue oophorectomy.

Prone cardiopulmonary resuscitation: A scoping and expanded grey literature review for the COVID-19 pandemic.

AIM: To identify and summarize the available science on prone resuscitation. To determine the value of undertaking a systematic review on this topic; and to identify knowledge gaps to aid future research, education and guidelines. METHODS: This review was guided by specific methodological framework and reporting items (PRISMA-ScR). We included studies, cases and grey literature regarding prone position and CPR/cardiac arrest. The databases searched were MEDLINE, Embase, CINAHL, Cochrane CENTRAL, Cochrane Database of Systematic Reviews, Scopus and Google Scholar. Expanded grey literature searching included internet search engine, targeted websites and social media. RESULTS: Of 453 identified studies, 24 (5%) studies met our inclusion criteria. There were four prone resuscitation-relevant studies examining: blood and tidal volumes generated by prone compressions; prone compression quality metrics on a manikin; and chest computed tomography scans for compression landmarking. Twenty case reports/series described the resuscitation of 25 prone patients. Prone compression quality was assessed by invasive blood pressure monitoring, exhaled carbon dioxide and pulse palpation. Recommended compression location was zero-to-two vertebral segments below the scapulae. Twenty of 25 cases (80%) survived prone resuscitation, although few cases reported long term outcome (neurological status at hospital discharge). Seven cases described full neurological recovery. CONCLUSION: This scoping review did not identify sufficient evidence to justify a systematic review or modified resuscitation guidelines. It remains reasonable to initiate resuscitation in the prone position if turning the patient supine would lead to delays or risk to providers or patients. Prone resuscitation quality can be judged using end-tidal CO2, and arterial pressure tracing, with patients turned supine if insufficient.

Risk and recovery among high school athletes who sustained at least one sports-related concussion.

AIM: Evidence of factors explaining sports-related concussion (SRC) risk and recovery among high school athletes remains inconclusive. MATERIALS & METHODS: Prospective study of a real-world sample of high school athletes (n = 77) who sustained ≤1 SRC. Among those with multiple SRCs, recovery time between events was investigated. To investigate concussion risk, baseline characteristics of athletes with a single versus multiple SRC(s) were compared. RESULTS: Recovery time did not differ across events. There were no differences between those with a single versus multiple SRCs. CONCLUSION: Recovery time between initial and subsequent concussive events did not differ, suggesting that prior concussion may not prolong recovery. Baseline characteristics did not explain heightened concussion risk. Investigation of these relationships using more representative samples is needed.

Can osmotic demyelination syndrome be a complication of liver failure?

This case demonstrates that osmotic demyelination syndrome (ODS) can occur in absence of hyponatremia in patients with fulminant liver failure and markedly high bilirubin levels. Extremely high bilirubin levels, such as >900 µmol/L in the case presented here, may lead to blood brain barrier dysfunction by disrupting blood vessel endothelial cell function as well as increase the release of inflammatory cytokines. As demonstrated in the case here, even small fluctuations in electrolytes may make the brain increasingly more vulnerable to ODS. Clinicians should keep ODS high on their differential even in eunatremic patients with liver failure who have decreased levels of consciousness or coma.

Sepsis in Traumatic Brain Injury: Epidemiology and Outcomes.

BACKGROUND: Traumatic brain injury (TBI) is a leading cause of death and disability. Risk factors for in-hospital mortality include older age, co-morbidity, and TBI severity. Few studies have investigated the role of sepsis in individuals with TBI. METHODS: We studied adult patients with TBI admitted to intensive care over a 5-year period. Patient characteristics were identified by linking clinical and administrative databases. Charts of individuals with TBI and sepsis were manually reviewed. Predictors of ICU and hospital mortality were identified using logistic regression modeling. RESULTS: Four hundred eighty-six individuals with TBI were admitted to intensive care. Sixteen (3.3%) developed sepsis. Pneumonia was the most common source (94%). Staphylococcus aureus was the most common pathogen (75%). ICU lengths of stay (LOS) (12.2 days [interquartile range (IQR) 4.4-23.5] versus 3.7 days [IQR 1.7-8.2]; p < 0.001) and hospital LOS (28.0 days [IQR 11.8-41.4] versus 15.3 days [IQR 5.0-30.9]; p = 0.017) were longer in patients with TBI and sepsis. Sepsis was not associated with ICU (adjusted odds ratio [aOR] 0.51; 95%CI 0.12-2.27; p = 0.38) or hospital (aOR 0.78; 95% CI 0.21-2.96; p = 0.78) mortality, though age (aOR 1.02; 95% CI 1.00-1.04; p = 0.014 for hospital mortality), severe TBI (aOR 3.71; 95% CI 1.52-9.08; p = 0.004 for ICU mortality and 4.10; 95% CI 1.95-8.65; p < 0.001 for hospital mortality), and APACHE II score (aOR 1.19; 95% CI 1.11-1.28; p < 0.001 for ICU mortality and 1.22; 95% CI 1.14-1.31; p < 0.001 for hospital mortality) were. CONCLUSION: Sepsis in patients with TBI was not associated with mortality; however, sepsis was associated with increased health care utilization (ICU and hospital LOS).

Prophylactic treatment of osteoporosis after SCI: promising research, but not yet indicated.

In persons with spinal cord injury (SCI), osteoporosis and associated fragility fractures are a prevalent phenomenon with clinically meaningful morbidity and mortality. Prevention of osteoporosis utilizing both physical modalities and pharmacological therapies is an area of high-clinical importance. In our perspective, the current body of research cannot provide clear guidance on prophylactic interventions to prevent osteoporosis specifically to stratify SCI subjects to their risk for fragility fractures. Without this critical research, clinicians cannot weigh the risk versus benefits of interventions, such as bisphosphonates, which is not a benign treatment. Other treatments such as physical modalities provide little risk and have other therapeutic benefit. This perspective is an argument that the current research does not indicate prophylactic pharmacological intervention to prevent osteoporosis in the SCI population.

Proposed diagnostic and treatment paradigm for high-grade neurological complications of immune checkpoint inhibitors.

Immune checkpoint inhibitors such as antibodies to cytotoxic lymphocyte-associated protein 4 (ipilimumab) and programmed cell-death 1 (pembrolizumab, nivolumab) molecules have been used in non-small cell lung cancer, metastatic melanoma, and renal-cell carcinoma, among others. With these agents, immune-related adverse events (irAEs) can occur, including those affecting the neurological axis. In this review, high-grade neurological irAEs associated with immune checkpoint inhibitors including cases of Guillain-Barré syndrome (GBS) and myasthenia gravis (MG) are analyzed. Based on current literature and experience at our institution with 4 cases of high-grade neurological irAEs associated with immune checkpoint inhibitors (2 cases of GBS, 1 case of meningo-radiculitis, and 1 case of myelitis), we propose an algorithm for the investigation and treatment of high-grade neurological irAEs. Our algorithm incorporates both peripheral nervous system (meningo-radiculitis, GBS, MG) and central nervous system presentations (myelitis, encephalopathy). It is anticipated that our algorithm will be useful both to oncologists and neurologists who are likely to encounter neurological irAEs more frequently in the future as immune checkpoint inhibitors become more widely used.