Clinical Outcomes and Cost Differences Between Patients Undergoing Primary Anterior Cervical Discectomy and Fusion Procedures with Private or Medicare Insurance: A Propensity Score-Matched Study.

OBJECTIVE: We assessed whether the insurance type reflects a patient's quality of care after an anterior discectomy and fusion (ACDF) procedure by comparing differences in the postoperative complications, readmission rates, reoperation rates, lengths of hospital stay, and cost of treatment between patients with Medicare versus private insurance. METHODS: Propensity score matching was used to match patient cohorts insured by Medicare and private insurance in the MarketScan Commercial Claims and Encounters Database (2007-2016). Age, sex, year of operation, geographic region, comorbidities, and operative factors were used to match cohorts of patients who had undergone an ACDF procedure. RESULTS: A total of 110,911 patients met the inclusion criteria. Of these patients, 97,543 patients (87.9%) were privately insured and 13,368 patients (12.1%) were insured by Medicare. The propensity score matching algorithm matched 7026 privately insured patients to 7026 Medicare patients. After matching, no significant differences were found in the 90-day postoperative complication rates, lengths of stay, or reoperation rates between the Medicare and privately insured cohorts. The Medicare group had had lower postoperative readmission rates for all time points: 30 days (1.8% vs. 4.6%; P < 0.001), 60 days (2.5% vs. 6.3%; P < 0.001), and 90 days (4.2% vs. 7.7%; P < 0.001). The median payment to physicians was significantly lower for the Medicare group ($3885 vs. $5601; P < 0.001). CONCLUSIONS: In the present study, propensity score matched patients covered by Medicare and private insurance who had undergone an ACDF procedure had had similar treatment outcomes.

Comparison of Deep Learning and Classical Machine Learning Algorithms to Predict Postoperative Outcomes for Anterior Cervical Discectomy and Fusion Procedures With State-of-the-art Performance.

STUDY DESIGN: Retrospective cohort. OBJECTIVE: Due to anterior cervical discectomy and fusion (ACDF) popularity, it is important to predict postoperative complications, unfavorable 90-day readmissions, and two-year reoperations to improve surgical decision-making, prognostication, and planning. SUMMARY OF BACKGROUND DATA: Machine learning has been applied to predict postoperative complications for ACDF; however, studies were limited by sample size and model type. These studies achieved ≤0.70 area under the curve (AUC). Further approaches, not limited to ACDF, focused on specific complication types and resulted in AUC between 0.70 and 0.76. MATERIALS AND METHODS: The IBM MarketScan Commercial Claims and Encounters Database and Medicare Supplement were queried from 2007 to 2016 to identify adult patients who underwent an ACDF procedure (N=176,816). Traditional machine learning algorithms, logistic regression, and support vector machines, were compared with deep neural networks to predict: 90-day postoperative complications, 90-day readmission, and two-year reoperation. We further generated random deep learning model architectures and trained them on the 90-day complication task to approximate an upper bound. Last, using deep learning, we investigated the importance of each input variable for the prediction of 90-day postoperative complications in ACDF. RESULTS: For the prediction of 90-day complication, 90-day readmission, and two-year reoperation, the deep neural network-based models achieved AUC of 0.832, 0.713, and 0.671. Logistic regression achieved AUCs of 0.820, 0.712, and 0.671. Support vector machine approaches were significantly lower. The upper bound of deep learning performance was approximated as 0.832. Myelopathy, age, human immunodeficiency virus, previous myocardial infarctions, obesity, and documentary weakness were found to be the strongest variable to predict 90-day postoperative complications. CONCLUSIONS: The deep neural network may be used to predict complications for clinical applications after multicenter validation. The results suggest limited added knowledge exists in interactions between the input variables used for this task. Future work should identify novel variables to increase predictive power.

Racial and socioeconomic correlates of treatment and survival among patients with meningioma: a population-based study.

BACKGROUND: Though meningioma is the most common primary brain tumor, there is a paucity of epidemiologic studies investigating disparities in treatment and patient outcomes. Therefore, we sought to explore how sociodemographic factors are associated with rates of gross total resection (GTR) and radiotherapy as well as survival. METHODS: The National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) database was queried to identify adult patients with meningioma diagnosed between 2005 and 2015. Socioeconomic status (SES) was determined using a validated composite index in which patients were stratified into tertiles and quintiles. Multivariable logistic regression and Cox proportional hazards analyses were used to identify predictors of treatment and survival, respectively. RESULTS: 71,098 patients met our inclusion criteria. Low SES quintile was associated with reduced odds of receiving GTR (OR 0.76, 95% CI 0.69-0.83, p < 0.0001) and radiotherapy (OR 0.83, 95% CI 0.76-0.91, p < 0.0001) as well as worse survival (HR 1.48, 95% CI 1.41-1.56) as compared to the highest SES quintile. Black patients had reduced odds of GTR (OR 0.74, 95% CI 0.67-0.71, p < 0.0001) and worse survival (HR 1.23, 95% CI 1.18-1.29, p < 0.0001) as compared to white patients. CONCLUSIONS: This national study of patients with meningioma found socioeconomic status and race to be independent inverse correlates of likelihood of GTR, radiotherapy, and survival. Limited access to care may underlie these disparities in part, and future studies are warranted to identify specific causes for these findings.

Risks, costs, and outcomes of cerebrospinal fluid leaks after pediatric skull fractures: a MarketScan analysis between 2007 and 2015.

OBJECTIVE: Skull fractures are common after blunt pediatric head trauma. CSF leaks are a rare but serious complication of skull fractures; however, little evidence exists on the risk of developing a CSF leak following skull fracture in the pediatric population. In this epidemiological study, the authors investigated the risk factors of CSF leaks and their impact on pediatric skull fracture outcomes. METHODS: The authors queried the MarketScan database (2007-2015), identifying pediatric patients (age < 18 years) with a diagnosis of skull fracture and CSF leak. Skull fractures were disaggregated by location (base, vault, facial) and severity (open, closed, multiple, concomitant cerebral or vascular injury). Descriptive statistics and hypothesis testing were used to compare baseline characteristics, complications, quality metrics, and costs. RESULTS: The authors identified 13,861 pediatric patients admitted with a skull fracture, of whom 1.46% (n = 202) developed a CSF leak. Among patients with a skull fracture and a CSF leak, 118 (58.4%) presented with otorrhea and 84 (41.6%) presented with rhinorrhea. Patients who developed CSF leaks were older (10.4 years vs 8.7 years, p < 0.0001) and more commonly had skull base (n = 183) and multiple (n = 22) skull fractures (p < 0.05). These patients also more frequently underwent a neurosurgical intervention (24.8% vs 9.6%, p < 0.0001). Compared with the non-CSF leak population, patients with a CSF leak had longer average hospitalizations (9.6 days vs 3.7 days, p < 0.0001) and higher rates of neurological deficits (5.0% vs 0.7%, p < 0.0001; OR 7.0; 95% CI 3.6-13.6), meningitis (5.5% vs 0.3%, p < 0.0001; OR 22.4; 95% CI 11.2-44.9), nonroutine discharge (6.9% vs 2.5%, p < 0.0001; OR 2.9; 95% CI 1.7-5.0), and readmission (24.7% vs 8.5%, p < 0.0001; OR 3.4; 95% CI 2.5-4.7). Total costs at 90 days for patients with a CSF leak averaged $81,206, compared with $32,831 for patients without a CSF leak (p < 0.0001). CONCLUSIONS: The authors found that CSF leaks occurred in 1.46% of pediatric patients with skull fractures and that skull fractures were associated with significantly increased rates of neurosurgical intervention and risks of meningitis, hospital readmission, and neurological deficits at 90 days. Pediatric patients with skull fractures also experienced longer average hospitalizations and greater healthcare costs at presentation and at 90 days.