Machine learning prediction of hospital discharge disposition for inpatients with acute ischemic stroke following mechanical thrombectomy in the United States.

BACKGROUND AND PURPOSE: Predicting patient recovery and discharge disposition following mechanical thrombectomy remains a challenge in patients with ischemic stroke. Machine learning offers a promising prognostication approach assisting in personalized post-thrombectomy care plans and resource allocation. As a large national database, National Inpatient Sample (NIS), contain valuable insights amenable to data-mining. The study aimed to develop and evaluate ML models predicting hospital discharge disposition with a focus on demographic, socioeconomic and hospital characteristics. MATERIALS AND METHODS: The NIS dataset (2006-2019) was used, including 4956 patients diagnosed with ischemic stroke who underwent thrombectomy. Demographics, hospital characteristics, and Elixhauser comorbidity indices were recorded. Feature extraction, processing, and selection were performed using Python, with Maximum Relevance - Minimum Redundancy (MRMR) applied for dimensionality reduction. ML models were developed and benchmarked prior to interpretation of the best model using Shapley Additive exPlanations (SHAP). RESULTS: The multilayer perceptron model outperformed others and achieved an AUROC of 0.81, accuracy of 77 %, F1-score of 0.48, precision of 0.64, and recall of 0.54. SHAP analysis identified the most important features for predicting discharge disposition as dysphagia and dysarthria, NIHSS, age, primary payer (Medicare), cerebral edema, fluid and electrolyte disorders, complicated hypertension, primary payer (private insurance), intracranial hemorrhage, and thrombectomy alone. CONCLUSION: Machine learning modeling of NIS database shows potential in predicting hospital discharge disposition for inpatients with acute ischemic stroke following mechanical thrombectomy in the NIS database. Insights gained from SHAP interpretation can inform targeted interventions and care plans, ultimately enhancing patient outcomes and resource allocation.

Management of severe traumatic brain injury in regions with limited resources.

IMPORTANCE: Severe traumatic brain injury (sTBI) is a critical health problem in regions of limited resources (RLRs). Younger populations are among the most impacted. The objective of this review is to analyze recent consensus-based algorithms, protocols and guidelines proposed for the care of patients with TBI in RLRs. OBSERVATIONS: The principal mechanisms for sTBI in RLRs are road traffic injuries (RTIs) and violence. Limitations of care include suboptimal or non-existent pre-hospital care, overburdened emergency services, lack of trained human resources, and surgical and intensive care. Low-cost neuromonitoring systems are currently in testing, and formal neurotrauma registries are forming to evaluate both long-term outcomes and best practices at every level of care from hospital transport to the emergency department (ED), to the operating room and intensive care unit (ICU). CONCLUSIONS AND RELEVANCE: The burden of sTBI is highest in RLRs. As working-age adults are the predominantly affected age-group, an increase in disability-adjusted life years (DALYs) generates a loss of economic growth in regions where economic growth is needed most. Four multi-institutional collaborations between high-income countries (HICs) and LMICs have developed evidence and consensus-based documents focused on capacity building for sTBI care as a means of addressing this substantial burden of disease.

Temporal Trends, Predictors, and Outcomes of Acute Ischemic Stroke in Acute Myocardial Infarction in the United States.

Background There are limited contemporary data prevalence and outcomes of acute ischemic stroke (AIS) complicating acute myocardial infarction (AMI). Methods and Results Adult (>18 years) AMI admissions using the National Inpatient Sample database (2000-2017) were evaluated for in-hospital AIS. Outcomes of interest included in-hospital mortality, hospitalization costs, length of stay, discharge disposition, and use of tracheostomy and percutaneous endoscopic gastrostomy. The discharge destination was used to classify survivors into good and poor outcomes. Of a total 11 622 528 AMI admissions, 183 896 (1.6%) had concomitant AIS. As compared with 2000, in 2017, AIS rates increased slightly among ST-segment-elevation AMI (adjusted odds ratio, 1.10 [95% CI, 1.04-1.15]) and decreased in non-ST-segment-elevation AMI (adjusted odds ratio, 0.47 [95% CI, 0.46-0.49]) admissions (P<0.001). Compared with those without, the AIS cohort was on average older, female, of non-White race, with greater comorbidities, and higher rates of arrhythmias. The AMI-AIS admissions received less frequent coronary angiography (46.9% versus 63.8%) and percutaneous coronary intervention (22.7% versus 41.8%) (P<0.001). The AIS cohort had higher in-hospital mortality (16.4% versus 6.0%; adjusted odds ratio, 1.75 [95% CI, 1.72-1.78]; P<0.001), longer hospital length of stay, higher hospitalization costs, greater use of tracheostomy and percutaneous endoscopic gastrostomy, and less frequent discharges to home (all P<0.001). Among AMI-AIS survivors (N=153 318), 57.3% had a poor functional outcome at discharge with relatively stable temporal trends. Conclusions AIS is associated with significantly higher in-hospital mortality and poor functional outcomes in AMI admissions.

Racial and economic disparities in the access to treatment of unruptured intracranial aneurysms are persistent problems.

BACKGROUND AND PURPOSE: Previous studies have documented disparate access to cerebrovascular neurosurgery for patients of different racial and socioeconomic backgrounds. We further investigated the effect of race and insurance status on access to treatment of unruptured intracranial aneurysms (UIAs) and compared it with data on patients with aneurysmal subarachnoid hemorrhage (aSAH). METHODS: Through the use of a national database, admissions for clipping or coiling of an UIA and for aSAH were identified. Demographic characteristics of patients were characterized according to age, sex, race/ethnicity, and insurance status, and comparisons between patients admitted for treatment of an UIA versus aSAH were performed. RESULTS: There were 10 545 admissions for clipping or coiling of an UIA and 33 166 admissions for aSAH between October 2014 and July 2018. White/non-Hispanic patients made up a greater proportion of patients presenting for treatment of an UIA than those presenting with aSAH (64.3% vs 48.2%; P<0.001), whereas black/Hispanic patients presented more frequently with aSAH than for treatment of an UIA (29.3% vs 26.1%; P=0.006). On multivariate linear regression analysis, the proportion of patients admitted for management of an UIA relative to those admitted for aSAH increased with the proportion of patients who were women (P<0.001) and decreased with the proportion of patients with a black/Hispanic background (P=0.010) and those insured with Medicaid or without insurance (P=0.003). CONCLUSION: For patients with UIAs, racial, ethnic, and socioeconomic backgrounds appear to continue to influence access to treatment.

The impact of drug shortages on patients with cardiovascular disease: causes, consequences, and a call to action.

Shortages of cardiovascular drugs have become increasingly common, representing an ongoing public health crisis. Given few therapeutic alternatives to many of the drugs in short supply, these shortages also pose a major challenge for cardiovascular care professionals. Although changes in the regulatory environment have led to some improvements in recent years, problems involving manufacturing processes remain the most common underlying cause. Because of the complex nature of drug shortages, sustainable solutions to prevent and mitigate them will require collaboration between regulatory agencies, drug manufacturers, and other key stakeholder groups. In this report, we describe the scope of the cardiovascular drug shortage crisis in the United States, including its underlying causes and the efforts currently being made to address it. Furthermore, we provide specific recommendations for how cardiovascular care professionals can be involved in efforts to limit the impact of drug shortages on patient care as well as policy changes aimed at preventing and mitigating them.

Disparities in imaging utilization for acute ischemic stroke based on patient insurance status.

OBJECTIVE: Previous studies have shown socioeconomic disparities in imaging utilization for both acute and chronic diseases. We studied a nationwide database to determine whether insurance-based disparities exist in the utilization of imaging for acute ischemic stroke. MATERIALS AND METHODS: Inpatients with a primary diagnosis of acute ischemic stroke from November 2005 through December 2011 were identified from the Perspective database. Patients were stratified into four groups according to insurance status as follows: uninsured, Medicaid, Medicare, and private insurance. Utilization rates of head CT, perfusion CT, head MRI, noninvasive head angiography (including head CT angiography [CTA] and head MR angiography [MRA]), noninvasive neck angiography (including neck CTA and neck MRA), carotid ultrasound, and echocardiography were compared using a chi-square test. A multivariable logistic regression model adjusting for potential confounding variables was fit to determine the association between insurance status and imaging utilization. RESULTS: A total of 210,212 patients were included in this study: 10,396 patients (5.0%) were uninsured, 14,243 patients (6.8%) had Medicaid, 153,209 patients (72.9%) had Medicare, and 32,364 patients (15.4%) had private insurance. Even after we had controlled for confounding variables, significant disparities existed in imaging utilization. Compared with patients with private insurance, uninsured patients had significantly lower odds of noninvasive head angiography (odds ratio [OR] = 0.78, 95% CI = 0.74-0.81, p < 0.0001), neck angiography (OR = 0.79, 95% CI = 0.76-0.83, p < 0.0001), and head MRI (OR = 0.77, 95% CI = 0.74-0.81, p < 0.0001). The same was true for Medicaid and Medicare patients. CONCLUSION: Disparities exist in the utilization of noninvasive head and neck imaging, MRI, and echocardiography for patients with acute ischemic stroke based on patient insurance status. More research is needed to understand these disparities.

Socioeconomic disparities in the utilization of mechanical thrombectomy for acute ischemic stroke.

BACKGROUND: Previous studies have demonstrated that socioeconomic disparities in access to treatment of cerebrovascular diseases exist. We studied the Nationwide Inpatient Sample (NIS) to determine if disparities exist in utilization of mechanical thrombectomy for acute ischemic stroke. METHODS: Using the NIS for the years 2006-2010, we selected all discharges with a primary diagnosis of acute ischemic stroke. Patients who received mechanical thrombectomy for stroke were identified by using the International Classification of Diseases, Ninth Revision, procedure code 39.74. We examined the utilization rates of mechanical thrombectomy by race/ethnicity (white, black, Hispanic, and Asian/Pacific Islander), income quartile (first, second to third, and fourth), and insurance status (Medicare, Medicaid, self-pay, and private). We also studied thrombectomy utilization rates at hospitals that performed thrombectomy. RESULTS: From 2006 to 2010, 2,087,017 patients were hospitalized with a primary diagnosis of acute ischemic stroke; 8946 patients (.4%) received mechanical thrombectomy. Compared with white patients, black patients had significantly lower rates of overall mechanical thrombectomy utilization (odds ratio [OR] = .59, 95% confidence interval [CI] = .55-.64, P < .0001) and at centers that offered mechanical thrombectomy (OR = .44, 95% CI = .41-.47, P < .0001). Compared with patients in the highest income quartile, patients in the lowest income quartile had significantly lower rates of mechanical thrombectomy utilization both overall (OR = .66, 95% CI = .62-.70, P < .0001) and at centers that offered mechanical thrombectomy (OR = .80, 95% CI = .75-.84, P < .0001). Compared with patients with private insurance, self-pay patients had significantly lower mechanical thrombectomy utilization both overall (OR = .71, 95% CI = .64-.78, P < .0001) and at centers that offered mechanical thrombectomy (OR = .81, 95% CI = .74-.90, P < .0001). CONCLUSIONS: Significant socioeconomic disparities exist in the utilization of mechanical thrombectomy in the United States.

Racial and ethnic disparities in the treatment of unruptured intracranial aneurysms: a study of the Nationwide Inpatient Sample 2001-2009.

BACKGROUND AND PURPOSE: Minorities in the United States have less access to healthcare system resources, especially preventative treatments. We sought to determine whether racial and sex disparities existed in the treatment of unruptured intracranial aneurysms. METHODS: Using the Nationwide Inpatient Sample, hospitalizations for clipping and coiling of intracranial aneurysms from 2001 to 2009 were identified by cross-matching International Classification of Diseases, 9th Revision codes for diagnosis of unruptured aneurysm and subarachnoid hemorrhage (SAH) with procedure codes for clipping or coiling of cerebral aneurysms. Demographic information analyzed included age (<50, 50-64, 65-79, and ≥80 years), race (white, black, Hispanic, Asian/Pacific Islander), sex, income quartile, primary payer (Medicare, Medicaid, private insurance, self-pay, no charge, other), and Charlson comorbidity index. RESULTS: When compared with patients treated for SAH, those treated for unruptured intracranial aneurysm were significantly more likely to be women (75.0% versus 69.0%; P<0.0001). In all, 9.7% of patients receiving treatment for SAH were self-payers versus 3.0% of patients being treated for unruptured intracranial aneurysm (P<0.0001). In all, 62.2% of patients receiving treatment for SAH were white compared with 76.4% of patients being treated for unruptured intracranial aneurysm (P<0.0001). There was a higher proportion of black, Hispanic, and Asian patients in the SAH treatment group when compared with the unruptured aneurysm treatment group (P<0.0001 for all groups). CONCLUSIONS: When compared with patients undergoing treatment for SAH, patients undergoing surgical and endovascular treatment for unruptured intracranial aneurysm are generally from higher socioeconomic strata and are more likely to be insured, women, and white. Future studies are needed to determine the underlying causes and solutions for this disparity.

Timing of clinical grade assessment and poor outcome in patients with aneurysmal subarachnoid hemorrhage.

OBJECT: Timing of clinical grading has not been fully studied in patients with aneurysmal subarachnoid hemorrhage (SAH). The primary objective of this study was to identify at which time point clinical assessment using the World Federation of Neurosurgical Societies (WFNS) grading scale and the Glasgow Coma Scale (GCS) is most predictive of poor functional outcome. METHODS: This study is a retrospective cohort study on the association between poor outcome and clinical grading determined at presentation, nadir, and postresuscitation. Poor functional outcome was defined as a Glasgow Outcome Scale score of 1-3 at 6 months after SAH. RESULTS: The authors identified 186 consecutive patients admitted to a teaching hospital between January 2002 and June 2008. The patients' mean age (±SD) was 56.9±13.7 years, and 63% were women. Twenty-four percent had poor functional outcome (the mortality rate was 17%). On univariable logistic regression analyses, GCS score determined at presentation (OR 0.80, p<0.0001), nadir (OR 0.73, p<0.0001), and postresuscitation (OR 0.53, p<0.0001); modified Fisher scale (OR 2.21, p=0.0013); WFNS grade assessed at presentation (OR 1.92, p<0.0001), nadir (OR 3.51, <0.0001), and postresuscitation (OR 3.91, p<0.0001); intracerebral hematoma on initial CT (OR 4.55, p<0.0002); acute hydrocephalus (OR 2.29, p=0.0375); and cerebral infarction (OR 4.84, p<0.0001) were associated with poor outcome. On multivariable logistic regression analysis, only cerebral infarction (OR 5.80, p=0.0013) and WFNS grade postresuscitation (OR 3.43, p<0.0001) were associated with poor outcome. Receiver operating characteristic/area under the curve (AUC) analysis demonstrated that WFNS grade determined postresuscitation had a stronger association with poor outcome (AUC 0.90) than WFNS grade assessed upon admission or at nadir. CONCLUSIONS: Timing of WFNS grade assessment affects its prognostic value. Outcome after aneurysmal SAH is best predicted by assessing WFNS grade after neurological resuscitation.

Hospitalization costs for acute ischemic stroke patients treated with intravenous thrombolysis in the United States are substantially higher than medicare payments.

BACKGROUND AND PURPOSE: It is important to know the costs for hospitalization for ischemic stroke patients treated with intravenous thrombolysis so that comparisons can be made with payments to hospitals. METHODS: Using the National Inpatient Sample, we evaluated hospitalization costs for patients treated with intravenous thrombolysis for acute ischemic stroke in the United States from 2001 to 2008. Cost data were correlated with demographics and clinical outcome. RESULTS: Intravenous thrombolysis for acute ischemic stroke was received by 63 472 patients; 24 094 patients were younger than age 65 years and 40 780 patients were 65 years or older. Median hospital costs in 2008 dollars were $14 102 (interquartile range, $9987-$20 819) for patients with good outcome, $18 856 (interquartile range, $13 145-$30 423) for patients with severe disability, and $19 129 (interquartile range, $11 966-$30 781) for patients with in-hospital mortality. Average 2008 Medicare payments were $10 098 for intravenous thrombolysis without complication and $13 835 for intravenous thrombolysis with major complication. CONCLUSIONS: Hospitalization costs for patients treated with intravenous thrombolysis are substantially higher than Medicare payments.

Hospitalization costs for patients with acute ischemic stroke treated with endovascular embolectomy in the United States.

BACKGROUND AND PURPOSE: It is important to know the costs of hospitalization for patients undergoing endovascular embolectomy so that comparisons can be made with payments to hospitals. METHODS: Using the National Inpatient Sample, we evaluated the costs of hospitalization for patients treated with endovascular embolectomy in the United States from 2006 to 2008. The primary end point examined in this study was total hospital costs, and these were correlated with clinical outcome. RESULTS: A total of 3864 patients received endovascular embolectomy. A total of 1649 patients were <65 years old and 2205 patients were ≥65 years old. Median hospital costs in 2008 dollars were $36,999 (interquartile range, $26,662-$56,405) for patients with good outcome, $50,628 (interquartile range, $33,135-$76,063) for patients with severe disability, and $35,109 (interquartile range, $25,053-$62,621) for patients with mortality. CONCLUSIONS: Hospitalization costs for patients treated with endovascular embolectomy are rather high, probably due to the serious nature of their illness. Medicare payments have not been adequate reimbursement for these hospitalizations.

Endovascular neurosurgery in the United States: a survey of 59 vascular neurosurgeons with endovascular training.

OBJECTIVE: We sought to assess general aspects of endovascular neurosurgery training, practice, and management for the treatment of intracranial aneurysms, arteriovenous malformations, acute ischemic stroke, and extracranial or intracranial occlusive disease within neurosurgery training and practice in the United States. METHODS: A questionnaire was sent electronically to 80 U.S. neurosurgeons with endovascular training. RESULTS: Fifty-nine surveys (74%) were returned. Survey responses illustrated different practice patterns and varying management of cerebrovascular disorders by neurosurgeons with endovascular training. CONCLUSION: Our findings provide a snapshot of current neurosurgical endovascular practices in the United States.

Electrolyte orders in the neuroscience intensive care unit: worth the value or waste?

BACKGROUND: To assess the value of the practice of obtaining frequent electrolyte measurements in patients with extended stay in a neuroscience intensive care unit (NICU). METHODS: We identified consecutive patients 18 years or older, admitted to the NICU between January 1 and July 31, 2009 with length of stay ≥ 5 days. We collected potassium, sodium, magnesium, ionized calcium, phosphorus laboratory measurements and hemoglobin levels, and recorded electrolyte replacement orders and red blood cell transfusions. Average laboratory costs were estimated. RESULTS: 93 patients were included in the study (54 men, mean age 54 years, range 18-85 years). Mean length of stay was 10.4 days (range 5-36 days). Sodium and potassium were the electrolytes most frequently measured (averages of 14.1 and 13.1 per patient, respectively). More than 75% of the results were within normal range for all electrolytes measured and critical values were extremely uncommon. The number of phlebotomies for electrolyte measurements was strongly associated with the degree of hemoglobin drop (P < 0.0001). Electrolyte panels were ordered much more often than individual electrolytes with average cost exceeding $2200 per patient. Replacing half of these electrolyte panels with single sodium or potassium orders would have resulted in savings greater than $100,000 in our population. CONCLUSIONS: Electrolytes measurements are very frequent in the NICU, but results are most often normal and only exceptionally critical. The phlebotomies required for these tests significantly worsen hemoglobin levels. A more conservative use of electrolyte measurements can result in reduction of blood loss and substantial cost savings.