York State Statewide Planning Research Articles

Does structural compromise of the aorta in patients with aortic pathologies predict increased spinal and vascular complications and reoperations in patients undergoing anterior approach to the spine?

Objectives: Anterior spinal fusion (ASF) presents unique challenges, including the proximity of critical anatomic structures. Previous reports have detailed vascular injuries during exposure/approach; however, it is not well-documented whether structural aortic pathology, such as aneurysm, dissection, atherosclerosis, aortitis, or aortic tumors, impacts postoperative outcomes following anterior approach to the spine for spinal fusion. Materials and Methods: Using the New York State Statewide Planning and Research Cooperative System, thoracolumbar ASF patients with a history of aneurysm, dissection, atherosclerosis, aortitis, or aortic tumors (APath) were identified and matched to patients with no aortic pathologies (No-APath). The two cohorts were compared at 90-day and 2-year follow-up for complications, readmissions, and revisions. Multivariate binary stepwise logistic regression identified independent predictors of these outcomes. Results: Ninety-nine and 64 patients were included at 90-day and 2-year follow-ups, respectively. APath and No-APath had comparable demographics. Through 90-day follow-up, both cohorts had similar vascular complications and overall complications. No-APath patients had higher surgical complications (11.1% vs. 0%, P = 0.021). At 2-year follow-up, cohorts had comparable vascular complications, overall complications, and all other outcomes (P > 0.05). Neither group was observed to require any revisions through 2-year follow-up. The presence of a structural aortic pathology was not associated with increased odds of incurring adverse outcomes at 90-day and 2-year follow-ups. Conclusions: Structural aortic compromise in the setting of thoracolumbar ASF did not predispose patients to adverse vascular, spinal-related, or medical/surgical complications, nor to anterior spinal or vascular revision or repair through 2-year follow-up. These results highlight the need for well-powered studies to further delineate the risk of anterior spinal surgery in this cohort of patients.

Predicting hospital length of stay using machine learning on a large open health dataset

BackgroundGovernments worldwide are facing growing pressure to increase transparency, as citizens demand greater insight into decision-making processes and public spending. An example is the release of open healthcare data to researchers, as healthcare is one of the top economic sectors. Significant information systems development and computational experimentation are required to extract meaning and value from these datasets. We use a large open health dataset provided by the New York State Statewide Planning and Research Cooperative System (SPARCS) containing 2.3 million de-identified patient records. One of the fields in these records is a patient’s length of stay (LoS) in a hospital, which is crucial in estimating healthcare costs and planning hospital capacity for future needs. Hence it would be very beneficial for hospitals to be able to predict the LoS early. The area of machine learning offers a potential solution, which is the focus of the current paper.MethodsWe investigated multiple machine learning techniques including feature engineering, regression, and classification trees to predict the length of stay (LoS) of all the hospital procedures currently available in the dataset. Whereas many researchers focus on LoS prediction for a specific disease, a unique feature of our model is its ability to simultaneously handle 285 diagnosis codes from the Clinical Classification System (CCS). We focused on the interpretability and explainability of input features and the resulting models. We developed separate models for newborns and non-newborns.ResultsThe study yields promising results, demonstrating the effectiveness of machine learning in predicting LoS. The best R2 scores achieved are noteworthy: 0.82 for newborns using linear regression and 0.43 for non-newborns using catboost regression. Focusing on cardiovascular disease refines the predictive capability, achieving an improved R2 score of 0.62. The models not only demonstrate high performance but also provide understandable insights. For instance, birth-weight is employed for predicting LoS in newborns, while diagnostic-related group classification proves valuable for non-newborns.ConclusionOur study showcases the practical utility of machine learning models in predicting LoS during patient admittance. The emphasis on interpretability ensures that the models can be easily comprehended and replicated by other researchers. Healthcare stakeholders, including providers, administrators, and patients, stand to benefit significantly. The findings offer valuable insights for cost estimation and capacity planning, contributing to the overall enhancement of healthcare management and delivery.

Predictive interpretable analytics models for forecasting healthcare costs using open healthcare data

Healthcare expenditure, a considerable proportion of national budgets, has risen rapidly. Consequently, considerable research is devoted to controlling healthcare costs. Many efforts are underway to improve medical price transparency. Price transparency will help patients become better informed, allowing them to shop for care they can afford, eventually leading to efficiency in healthcare markets. This first requires medical pricing data to be made available publicly. Since the raw pricing data can be large and cover multiple conditions, it is necessary to provide an engine to process the data to facilitate its usage and understanding. We recommend creating computational models that predict healthcare costs for various patient conditions and demographics. Patients and providers can interrogate the underlying data to understand the variation of healthcare costs concerning medical conditions and demographic variables of interest, including age. We demonstrate our approach by creating predictive models using recent machine learning techniques. We analyzed anonymous patient data from the New York State Statewide Planning and Research Cooperative System, consisting of 2.34 million records from 2019. We built models to predict costs from over two dozen patient variables, including diagnosis codes, severity of illness, age, and other demographic variables. We investigated three models: regression, decision trees, and random forests. These models are explainable. We analyzed features to determine those that were predictive of total costs. We determined that the diagnosis code, severity of illness, and length of stay were good predictors of total costs, whereas race and gender are not useful in predicting total costs. We obtained the best performance using a catboost regressor, which yielded an R2 score of 0.85, better than the values reported in the literature.

An Analysis of New York Data: Fluctuations in Hospital Capacity Are Driven by Variability in Elective Admissions and Discharge Activity.

Background Hospital overcrowding compromises patient safety. The contribution of variability in admissions and discharges to overall hospital capacity needs to be quantified. This study describes the statewide day-to-day fluctuation in the volume of hospitalized patients, the variability and pattern of hospital admissions and discharges throughout the week, and the contribution of Emergency Department (ED) vs. elective (non-ED) admissions and discharges to the overall variability in the system across the week. Methodology This is a retrospective analysis of the New York State Statewide Planning and Research Cooperative System database, in which all New York healthcare facilities submit patient-level data monthly. The study period was from January 01 to December 31, 2015. Outcomes included total volumes of admissions and discharges and length of stay sorted by patient origin (ED vs. non-ED admits (elective)) and service type (medicine vs. surgery) by day of the week. Results We studied 1,692,090 hospital admissions. Admissions were highest on Mondays and Tuesdays and steadily decreased throughout the week. There was little variability in the ED admissions throughout the week. Surgical elective admissions had significant variability throughout the week, with higher admissions at the beginning of the week. There was a significant difference (p < 0.01) between admissions on weekdays vs. weekends. Discharges increased from Monday to Friday, with a dramatic drop on the weekends, for both ED and elective pathways. Systemwide, on Monday, hospitals were 21% above the mean volume, and on Fridays, hospitals were 32% below the mean volume. Conclusions Overall hospital capacity shows dramatic variability throughout the week, driven primarily by elective admissions and discharges from any source throughout the week. Because elective admissions are schedulable, hospitals can reduce variability by smoothing scheduling. Increased weekend discharges will also improve capacity.

Comparable performance of machine learning algorithms in predicting readmission and complications following total joint arthroplasty with external validation

BackgroundThe purpose of the study was to use Machine Learning (ML) to construct a risk calculator for patients who undergo Total Joint Arthroplasty (TJA) on the basis of New York State Statewide Planning and Research Cooperative System (SPARCS) data and externally validate the calculator on a single TJA center.MethodsSeven ML algorithms, i.e., logistic regression, adaptive boosting, gradient boosting (Xg Boost), random forest (RF) classifier, support vector machine, and single and a five-layered neural network were trained on the derivation cohort. Models were trained on 68% of data, validated on 15%, tested on 15%, and externally validated on 2% of the data from a single arthroplasty center.ResultsValidation of the models showed that the RF classifier performed best in terms of 30-d mortality AUROC (Area Under the Receiver Operating Characteristic) 0.78, 30-d readmission (AUROC 0.61) and 90-d composite complications (AUROC 0.73) amongst the test set. Additionally, Xg Boost was found to be the best predicting model for 90-d readmission and 90-d composite complications (AUC 0.73). External validation demonstrated that models achieved similar AUROCs to the test set although variation occurred in top model performance for 90-d composite complications and readmissions between our test and external validation set.ConclusionThis was the first study to investigate the use of ML to create a predictive risk calculator from state-wide data and then externally validate it with data from a single arthroplasty center. Discrimination between best performing ML models and between the test set and the external validation set are comparable.Level of EvidenceIII.

Comparing hospital costs and length of stay for cancer patients in New York State Comprehensive Cancer Centers versus nondesignated academic centers and community hospitals.

To explore differences in costs and lengths of stay for cancer patients admitted to National Cancer Institute-designated Comprehensive Cancer Centers, nondesignated academic medical centers, and community hospitals in New York State. We use patient-level data from the New York State Statewide Planning and Research Cooperative System Hospital Inpatient Discharges dataset for the years 2017-2019. We employ ordinary least squares and Poisson regressions to compare hospital costs and length of stay for cancer patients, controlling for hospital type, patient demographics, and patient health. Our key outcomes are differences in costs and lengths of stay. We use data on patient demographics, total treatment costs, and lengths of stay for patients discharged from New York hospitals with cancer-related diagnoses between 2017 and 2019. We determine that inpatient costs were 27% higher (95% CI 0.252, 0.285), but length of stay was 12% shorter (95% CI -0.131, -0.100), in Comprehensive Cancer Centers relative to community hospitals. The results imply that, in New York State, Comprehensive Cancer Centers are a magnet for more complex oncology cases and administer more expensive treatments. That expertise, however, seems to be responsible for more efficient care delivery and thorough discharge planning, allowing for shorter average lengths of stay.

A call to "own the bone": osteoporosis is a predictor for adverse two-year outcomes following total hip and knee arthroplasty.

While bone health is instrumental in orthopedic surgery, few studies have described the long-term outcomes of osteoporosis (OP) in patients undergoing total hip (THA) or knee (TKA) arthroplasties. Using the New York State statewide planning and research cooperative system database, all patients who underwent primary TKA or THA for osteoarthritis from 2009 to 2011 with minimum 2-year follow-up were identified. They were divided based on their OP status (OP and non-OP) and 1:1 propensity score matched for age, sex, race, and Charlson/Deyo index. Cohorts were compared for demographics, hospital-related parameters, and 2-year postoperative complications and reoperations. Multivariate binary logistic regression was utilized to identify significant independent associations with 2-year medical and surgical complications and revisions. A total of 11,288 TKA and 8248 THA patients were identified. OP and non-OP TKA patients incurred comparable overall hospital charges for their surgical visit and hospital length of stay (LOS) (both, p ≥ 0.125). Though OP and non-OP THA patients incurred similar mean hospital charges for their surgical visit, they experienced longer hospital LOS (4.3 vs. 4.1days, p = 0.035). For both TKA and THA, OP patients had higher rates of overall and individual medical and surgical complications (all, p < 0.05). OP was independently associated with the 2-year occurrence of any overall, surgical, and medical complications, and any revision in TKA and THA patients (all, OR ≥ 1.42, p < 0.001). Our study found OP was associated with a greater risk of 2-year adverse outcomes following TKA or THA, including medical, surgical, and overall complications as well as revision operations compared to non-OP patients.

The Impact of Osteoporosis on 2-Year Outcomes in Patients Undergoing Long Cervical Fusion

Osteoporosis affects nearly 200 million individuals worldwide. Given this notable disease burden, there have been increased efforts to investigate complications in patients with osteoporosis undergoing cervical fusion (CF). However, there are limited data regarding long-term outcomes in osteoporotic patients in the setting of ≥4-level cervical fusion. The New York State Statewide Planning and Research Cooperative System database was used to identify patients who underwent posterior or combined anterior-posterior ≥4-level CF for cervical radiculopathy or myelopathy from 2009 to 2011, with a minimum follow-up surveillance of 2 years. The following were compared between patients with and without osteoporosis: demographics, hospital-related parameters, medical/surgical complications, readmissions, and revisions. Binary multivariate stepwise logistic regression was used to identify independent predictors of outcomes. A total of 2,604 patients were included (osteoporosis: n = 136 (5.2%); nonosteoporosis: n = 2,468). Patients with osteoporosis were older (66.9 ± 11.2 vs. 60.0 ± 11.4 years, P < 0.001), more often female (75.7% vs. 36.2%, P < 0.001), and White (80.0% vs. 65.3%, P = 0.007). Both cohorts had comparable comorbidity burdens (Charlson/Deyo: 1.1 ± 1.2 vs. 1.0 ± 1.3, P = 0.262), total hospital charges ($100,953 ± 94,933 vs. $91,618 ± 78,327, P = 0.181), and length of stay (9.7 ± 10.4 vs. 8.4 ± 9.6 days, P = 0.109). Patients with osteoporosis incurred higher rates of overall medical complication rates (41.9% vs. 29.4%, P = 0.002) and individual surgical complications, such as nonunion (2.9% vs. 0.7%, P = 0.006). Osteoporosis was associated with medical complications (OR = 1.57, P = 0.021), surgical complications (OR = 1.52, P = 0.030), and readmissions (OR = 1.86, P = 0.003) at 2 years. Among patients who underwent multilevel cervical fusion, those with osteoporosis had higher risk of adverse postoperative outcomes at two years. These data indicate that preoperative screening and management of osteoporosis may be important for optimizing long-term outcomes in patients who require multilevel CF. The data used in this study are available for public use at https://www.health.ny.gov/statistics/sparcs/.

113. Impact of iron deficiency anemia on postoperative outcomes of thoracolumbar spinal fusion (≥2-level) on patients with adult spinal deformity with minimum two-year follow-up surveillance

BACKGROUND CONTEXT Iron deficiency anemia (IDA) is a common hematological disorder and cause of low hemoglobin. Preoperative anemia has been shown to increase the risk of adverse outcomes after posterior cervical fusion and other spinal surgery. The need for a transfusion during lumbar fusion has been shown to increase the length of stay. The impact of IDA on outcomes after spinal fusion for ASD is unknown. PURPOSE This study sought to compare outcomes and complication rates between ASD patients with and without IDA undergoing 2+ level spinal fusion. STUDY DESIGN/SETTING Retrospective review. PATIENT SAMPLE The New York State Statewide Planning and Research Cooperative System (NYS SPARCS) was queried from 2009-2011 to identify all patients undergoing 2+ level spinal fusion for ASD with two-year follow-up. OUTCOME MEASURES Two-year rates of medical and surgical complications as well as readmissions, reoperations and mortality. METHODS NY SPARCS was queried from 2009-11 to identify all patients undergoing 2-level thoracolumbar spinal fusion (primary and revision) for ASD with 2-year follow-up. Patients were then stratified by the presence (IDA) or absence (no-IDA) of IDA. IDA and no-IDA were 1:1 propensity score-matched based on age, sex and obesity. Univariate analysis compared demographics, hospital parameters, and rates of adverse outcomes. Multivariate binary logistic regression was used to identify independent predictors of adverse outcomes. RESULTS A total of 524 patients (262 IDA and 262 no-IDA) were identified. There was no difference in age (56.8 vs 56.8 years), sex (79.8% vs 79.8% female), or obesity (13.0% vs 13.0%), all p=1.0. IDA patients had higher hospital cost ($95, 673 vs $59,537) and LOS (6.0 vs 4.0 days), all p < 0.001. IDA patients experienced higher rates of overall surgical complications (50.4% vs 23.7%, p < 0.001), wound complications (3.4% vs 0.4%, p=0.011), and blood transfusion (10.3% vs 6.5%, p <0.001). There was no difference in the rate of overall medical complications. IDA and no-IDA had comparable rates of readmission (8.0% vs 13.0%, p=0.064), though IDA had lower rates of reoperation (7.6% vs 13.0%, p=0.044). There was no mortality in either cohort. IDA independently was associated with wound complications (OR=10.6, p=0.028), blood transfusion (OR=3.9, p <0.001), and surgical complications (OR=3.5, p <0.001). CONCLUSIONS Baseline IDA was predictive of increased wound complications, postoperative blood transfusion and overall surgical complications following ASD surgery. This data can be used to better optimize and risk-stratify patients prior to surgery. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs. Iron deficiency anemia (IDA) is a common hematological disorder and cause of low hemoglobin. Preoperative anemia has been shown to increase the risk of adverse outcomes after posterior cervical fusion and other spinal surgery. The need for a transfusion during lumbar fusion has been shown to increase the length of stay. The impact of IDA on outcomes after spinal fusion for ASD is unknown. This study sought to compare outcomes and complication rates between ASD patients with and without IDA undergoing 2+ level spinal fusion. Retrospective review. The New York State Statewide Planning and Research Cooperative System (NYS SPARCS) was queried from 2009-2011 to identify all patients undergoing 2+ level spinal fusion for ASD with two-year follow-up. Two-year rates of medical and surgical complications as well as readmissions, reoperations and mortality. NY SPARCS was queried from 2009-11 to identify all patients undergoing 2-level thoracolumbar spinal fusion (primary and revision) for ASD with 2-year follow-up. Patients were then stratified by the presence (IDA) or absence (no-IDA) of IDA. IDA and no-IDA were 1:1 propensity score-matched based on age, sex and obesity. Univariate analysis compared demographics, hospital parameters, and rates of adverse outcomes. Multivariate binary logistic regression was used to identify independent predictors of adverse outcomes. A total of 524 patients (262 IDA and 262 no-IDA) were identified. There was no difference in age (56.8 vs 56.8 years), sex (79.8% vs 79.8% female), or obesity (13.0% vs 13.0%), all p=1.0. IDA patients had higher hospital cost ($95, 673 vs $59,537) and LOS (6.0 vs 4.0 days), all p < 0.001. IDA patients experienced higher rates of overall surgical complications (50.4% vs 23.7%, p < 0.001), wound complications (3.4% vs 0.4%, p=0.011), and blood transfusion (10.3% vs 6.5%, p <0.001). There was no difference in the rate of overall medical complications. IDA and no-IDA had comparable rates of readmission (8.0% vs 13.0%, p=0.064), though IDA had lower rates of reoperation (7.6% vs 13.0%, p=0.044). There was no mortality in either cohort. IDA independently was associated with wound complications (OR=10.6, p=0.028), blood transfusion (OR=3.9, p <0.001), and surgical complications (OR=3.5, p <0.001). Baseline IDA was predictive of increased wound complications, postoperative blood transfusion and overall surgical complications following ASD surgery. This data can be used to better optimize and risk-stratify patients prior to surgery.

P125. Evaluating the impact of multiple sclerosis on 2-year postoperative outcomes following long fusion for adult spinal deformity: a propensity score matched analysis

<h3>BACKGROUND CONTEXT</h3> The impact of neuromuscular disorders such as multiple sclerosis (MS) on outcomes following long segment (4+ level) fusion is underreported. <h3>PURPOSE</h3> To identify the impact of MS on two-year (2Y) postoperative complications and revisions following 4 level fusion for ASD. <h3>STUDY DESIGN/SETTING</h3> Retrospective cohort. <h3>PATIENT SAMPLE</h3> Patients undergoing 4+ level fusion for ASD were identified from the New York State Statewide Planning and Research Cooperative System. <h3>OUTCOME MEASURES</h3> Two-year postoperative complications and reoperations. <h3>METHODS</h3> Patients undergoing 4-level fusion for ASD were identified from the New York Statewide Planning and Research Cooperative System. Patients with a baseline diagnosis of MS were also identified. Patients with infectious/traumatic/neoplastic indications were excluded. Subjects were 1:1 propensity score-matched (MS to no-MS) by age, sex, and race and compared for rates of 2-year postoperative complications and reoperations. Logistic regression models were utilized to determine risk factors for adverse outcomes at 2 years. <h3>RESULTS</h3> A total of 86 patients were included (n=43 per group). Age (50.1 vs 50.1 years, p=0.225), sex and race were comparable between groups. MS pts incurred higher charges for their surgical visit ($125,906 vs $84,006, p=0.007) with similar LOS (8.1 vs 5.3 days, p>0.05). MS patients experienced comparable rates of overall medical complications (30.1% vs 25.6%) and surgical complications (34.9% vs 30.2%) all p>0.05. MS pts had similar rates of 2-year revisions (16.3% vs 9.3%, p=0.333). MS was not associated with medical, surgical or overall complications or revisions at minimum 2-year follow-up. <h3>CONCLUSIONS</h3> Patients with MS experienced similar postoperative course compared to those without MS following 4-level fusion for ASD. This data supports the findings of multiple previously published case series' that long segment fusions for ASD can be performed relatively safely in patients with MS. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.

The Impact of Osteoporosis on Adverse Outcomes After Short Fusion for Degenerative Lumbar Disease.

Osteoporosis affects nearly 200 million individuals worldwide. There are little available data regarding outcomes in patients with osteoporosis who undergo short-segment lumbar fusion for degenerative disk disease (DDD). We sought to identify a relationship between osteoporosis and risk of adverse outcomes in patients with DDD undergoing short-segment lumbar fusion. Using the New York State Statewide Planning and Research Cooperative System, all patients with DDD who underwent 2- to 3-level lumbar fusion from 2009 through 2011 were identified. Patients with bone mineralization disorders and other systemic and endocrine disorders and surgical indications of trauma, systemic disease(s), and infection were excluded. Patients were stratified by the presence or absence of osteoporosis and compared for demographics, hospital-related parameters, and 2-year complications and revision surgeries. Multivariate binary logistic regression models were used to identify notable predictors of complications. A total of 29,028 patients (osteoporosis = 1,353 [4.7%], nonosteoporosis = 27,675 [95.3%]) were included. Patients with osteoporosis were older (66.9 vs 52.6 years), more often female (85.1% vs 48.4%), and White (82.8% vs 73.5%) (all P < 0.001). The Charlson/Deyo comorbidity index did not significantly differ between groups. Hospital lengths of stay and total charges were higher for patients with osteoporosis (4.9 vs 4.1 days; $74,484 vs $73,724; both P < 0.001). Medical complication rates were higher in patients with osteoporosis, including acute renal failure and deep-vein thrombosis (both P < 0.01). This cohort also had higher rates of implant-related (3.4% vs 1.9%) and wound (9.8% vs 5.9%) complications (both P < 0.01). Preoperative osteoporosis was strongly associated with 2-year medical and surgical complications (odds ratios, 1.6 and 1.7) as well as greater odds of revision surgeries (odds ratio, 1.3) (all P < 0.001). Patients with osteoporosis undergoing 2- to 3-level lumbar fusion for DDD were at higher risk of 2-year medical and surgical complications, especially implant-related and wound complications. These findings highlight the importance of rigorous preoperative metabolic workup and patients' optimization before spinal surgery.

151. Association Between Outpatient Antibiotic Prescribing, Antimicrobial Resistance, and Initial Presentation to Inpatient Setting for Urinary Tract Infections Among Older Adults in New York State

BackgroundAntibiotic prescribing (AP) and resistance (AR) may influence severity of illness in urinary tract infection (UTI). Limited data exist assessing the relationship between county-level AP and AR on initial presentation to hospital for UTI. This study evaluated the association between county-level AP and AR on UTI severity of illness among hospitalized patients in New York State.MethodsRetrospective, cross-sectional analysis, combining data from New York State Statewide Planning and Research Cooperative System (SPARCS) and previously published data on countywide antimicrobial resistance and antimicrobial prescribing. Inclusion criteria: female patients admitted to a New York inpatient setting in 2017, UTI (CCS 159), Medicare insurance. Exclusion criteria: missing countywide prescribing or resistance. All-patient refined (APR) clinical severity ≥ 3 was the primary outcome. Counties were classified as prescribing above or below the median prescribing proportion, and above or below the median prevalence of E. coli resistance for TMP-SMX and NTF. Countywide prescribing practices, antimicrobial resistance, patient factors, and location factors were evaluated for association with APR clinical severity ≥ 3 using chi-squared and logistic regression.Results8,024 patients met study criteria. Baseline characteristics are presented in Table 1. 3,597 (44.8%) had an APR severity of ≥ 3. Factors associated with APR severity ≥ 3 include age group (P < 0.001), ethnicity (P = 0.013), hospital county (P < 0.001), first line prescribing ≥ 45.4% (P = 0.049), E. coli TMP-SMX resistance ≥ 29.0% (P < 0.001) via chi-squared test. In the logistic regression analysis counties with higher first line prescribing was associated with decreased odds for severe infection (aOR: 0.83 [0.72 – 0.97]). Additional factors associated with severe infection are presented in Table 2. ConclusionPrescribing patterns may have a significant influence on initial presentation to the hospital for urinary tract infections. Outpatient antimicrobial stewardship should endeavor to promote guideline driven prescribing. Further research is needed to corroborate the findings from this cross-sectional study.Disclosures All Authors: No reported disclosures

Multimorbidity and Inpatient Utilization Among Older Adults with Opioid Use Disorder in New York City.

Nationally, there is a sharp increase in older adults with opioid use disorder (OUD). However, we know little of the acute healthcare utilization patterns and medical comorbidities among this population. This study describes the prevalence of chronic conditions, patterns of inpatient utilization, and correlates of high inpatient utilization among older adults with OUD in New York City (NYC). Retrospective longitudinal cohort study. Patients aged ≥55 with OUD hospitalized in NYC in 2012 identified using data from New York State's Statewide Planning and Research Cooperative System (SPARCS). The prevalence of comorbid substance use diagnoses, chronic medical disease, and mental illness was measured using admission diagnoses from the index hospitalization. We calculated the ICD-Coded Multimorbidity-Weighted Index (MWI-ICD) for each patient to measure multimorbidity. We followed the cohort through September 30, 2015 and the outcome was the number of rehospitalizations for inpatient services in NYC. We compared patient-level factors between patients with the highest use of inpatient services (≥7 rehospitalizations) during the study period to low utilizers. We used multiple logistic regression to examine possible correlates of high inpatient utilization. Of 3669 adults aged ≥55 with OUD with a hospitalization in 2012, 76.4% (n=2803) had asubsequent hospitalization and accounted for a total of 22,801 rehospitalizations during the study period. A total of 24.7% of the cohort (n=906) were considered high utilizers and had a higher prevalence of alcohol and cocaine-related diagnoses, congestive heart failure, diabetes, schizophrenia, and chronic obstructive pulmonary disease. Multivariable predictors of high utilization included being a Medicaid beneficiary (adjusted odds ratio [aOR]=1.70, 95% confidence interval [CI]=1.37-2.11), alcohol-related diagnoses (aOR=1.43, 95% CI: 1.21-1.69), and increasing comorbidity measured by MWI-ICD (highest MWI-ICD quartile: aOR=1.98, 95% CI=1.59-2.48). Among older adults with OUD admitted to the hospital, multimorbidity is strongly associated with high inpatient utilization.

Evaluation of Data Sources for Carbon Monoxide Poisoning Surveillance in New York City.

Carbon monoxide (CO) exposure can be life-threatening. Suspected and confirmed cases of CO poisoning warranting health care in New York City (NYC) are reportable to the NYC Poison Control Center (PCC). We evaluated 4 hospital-based sources of CO surveillance data to identify ways to improve data capture and reporting. Suspected and confirmed CO poisoning records from October 2015 through December 2016 were collected from the NYC emergency department (ED) syndromic surveillance system, New York State Statewide Planning and Research Cooperative System (SPARCS) ED billing data, NYC PCC calls made from hospitals, and the Electronic Clinical Laboratory Reporting System (ECLRS). Syndromic and SPARCS records were person- and visit-matched. SPARCS and ECLRS records were also matched to PCC records on combinations of name, demographic characteristics, and visit information. Hospitals in NYC. Individuals who visited NYC hospitals for CO-related health effects. We assessed the validity of syndromic data, with SPARCS records as the gold standard. We matched SPARCS and ECLRS records to PCC records to analyze reporting rates by case characteristics. The sensitivity of syndromic surveillance was 60% (225 true-positives detected among 372 visit-matched SPARCS cases), and positive predictive value was 46%. Syndromic records often missed CO flags because of a nonspecific or absent International Classification of Diseases code in the diagnosis field. Only 15% of 428 SPARCS records (total includes 56 records not visit-matched to syndromic) and 16% of 199 ECLRS records were reported to PCC, with male sex and younger age associated with higher reporting. Mandatory reporting makes PCC useful for tracking CO poisoning in NYC, but incomplete reporting and challenges in distinguishing between confirmed and suspected cases limit its utility. Simultaneous tracking of the systems we evaluated can best reveal surveillance patterns.

How to choose your surgeon for colorectal cancer: The influence of fellowship training on outcomes.

e18603 Background: Rectal cancer has a well-established volume-outcome relationship. However, the impact of fellowship training on outcomes following colorectal cancer resection is unknown. Methods: Patients who underwent stage I-III colorectal adenocarcinoma resection (2004-2014) were identified in the New York State Cancer Registry and Statewide Planning &amp; Research Cooperative System. Hierarchical analyses assessed the association between surgeon American Board of Colorectal Surgery (ABCRS) certification and postoperative outcomes and survival. Derived from volume tertiles, surgeons who performed ≥15 annual colon cancer resections or ≥10 annual rectal cancer resections were characterized as high-volume colon (HVCSs) or rectal (HVRSs) surgeons, respectively. Results: 47,491 colorectal cancer resections (right colon = 23,824, 50.2%; left colon = 18,154, 38.2%; rectal cancer = 5,513, 11.6%) met inclusion criteria. ABCRS-certified HVCSs (N = 92) comprised only 5.5% of the surgeon cohort but performed 20.6% of the colon cancer resections. Similarly, ABCRS-certified HVRSs (N = 31) comprised only 3.8% of the surgeon cohort but performed 16.5% of the rectal cancer resections. Despite no significant difference in complications, ABCRS-certified HVCS/HVRSs overall had better postoperative outcomes and were associated with improved survival following colorectal cancer resection. Conclusions: These results suggest that for the best outcomes and long-term survival following colon and rectal cancer resection, referral to high-volume, ABCRS-certified surgeons may be the best choice.[Table: see text]

Robotic-assisted simple prostatectomy versus open simple prostatectomy: a New York statewide analysis of early adoption and outcomes between 2009 and 2017.

The factors driving early adoption of robotic-assisted simple prostatectomy (RASP) for large gland BPH have not yet been identified. This study aims to determine the patient, provider, and facility level differences and predictors in undergoing RASP versus OSP. This population-based cohort study used data from the all-payer New York State Statewide Planning and Research Cooperative System (SPARCS) database. Patient, provider, and facility characteristics for each cohort were analyzed, and a multivariate analysis was conducted to identify predictive factors associated with undergoing RASP versus OSP. From 2009 to 2017, 1881 OSP and 216 RASP cases were identified. RASP utilization increased from 2.6% of all cases in 2009 to 16.8% in 2017. Patient demographics were similar between both cohorts. Median length of stay was shorter for RASP patients (3 vs. 4days, p < 0.001), and OSP was associated with a long length of stay (> 7days) (p < 0.001). There were no significant differences in 30- and 90-day readmission rates or 1-year mortality. More OSP patients were discharged to continued care facilities than RASP patients (p = 0.049), and more RASP patients were discharged to home compared to OSP patients (p = 0.035). Positive predictors for undergoing RASP included teaching hospital status, medium and high hospital bed volume, high hospital operative volume, high surgeon volume, and surgeons that graduated within 15years of surgery. As RASP shows favorable perioperative outcomes, the diffusion of robotic technology and newer graduates entering the workforce may augment the upward trend of RASP utilization.

P137. Osteoporosis is a predictor of two-year adverse outcomes following short fusion for degenerative lumbar disease

BACKGROUND CONTEXT Spinal fusion is increasingly considered for management of patients with osteoporosis (OP) and degenerative disc disease (DDD). Little data is available regarding long-term outcomes in OP patients undergoing short-segment, lumbar fusion for DDD. PURPOSE Analyze the impact of OP on long-term outcomes in patients who have had 2-3-level lumbar fusions for degenerative disc disease. STUDY DESIGN/SETTING Retrospective review of the New York State Statewide Planning and Research Cooperative System (NYS SPARCS) database. PATIENT SAMPLE Patients with DDD undergoing 2-3-level lumbar fusion. OUTCOME MEASURES Frequency of 2-year medical and surgical complications and reoperations. METHODS Utilizing SPARCS, all patients from 2009-2011 with ICD-9 codes for DDD who underwent 2-3-level lumbar fusion were identified. Patients with bone mineralization disorders, systemic and endocrine disorders affecting bone quality were excluded. Surgical indications of trauma, systemic disease(s), and infection were excluded. Patients with and without OP were compared for demographics, hospital-related parameters, and 2-year complications and reoperations. Logistic regression models were utilized to identify predictors of complications. RESULTS A total of 29,028 patients (OP=1,353 (4.7%), No-OP=27,675) were included. OP patients were older (66.9 vs 52.6 years) as well as more often female (85.1% vs 48.4%) and white (82.8% vs 73.5%), all p CONCLUSIONS Patients with OP undergoing 2-3-level lumbar fusion for DDD were at higher risk of 2-year medical and surgical complications, and 3.4% of OP patients experienced implant-related complication. These findings highlight the importance of rigorous preoperative metabolic workup prior to spinal surgery. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs.

122. Variation in 90-day neurological complications across increasing fusion levels for posterior cervical fusion: a five-year analysis

BACKGROUND CONTEXT While neurological complications are known to occur following cervical fusions, few studies with sufficient power have quantified the difference in neurological outcomes as posterior cervical fusion (PCF) surgical invasiveness increases. PURPOSE Compare longer vs shorter PCF complication rates. STUDY DESIGN/SETTING Retrospective review of the New York State Statewide Planning and Research Cooperative System (SPARCS) database. PATIENT SAMPLE All patients undergoing ≥2 level PCF. OUTCOME MEASURES Ninety-day neurological or medical complications status post short-cervical (SC), long-cervical (LC), or long cervicothoracic (LCT). METHODS From the NYS Statewide Planning and Research Cooperative System (SPARCS) database, retrospective review of all patients who underwent ≥2 level PCF from 2009-13 with ≤ 90-day follow-up were identified and stratified by levels fused: 2-3 (SC), 4-8 (LC), and ≥ 9 (LCT). Data on demographics, hospital-related parameters, and 90-day neurological and other complications, readmissions, and revisions were collected. Regression analysis identified independent predictors of neurologic and overall complications. RESULTS A total of 6,981 patients were included (SC, n=2,964, LC, n=3,899, LCT, n=118). LC patients were older than SC and LCT (60.8 vs 58.2 and 56.1), while LCT patients were more often female (59.3% vs LC 42.1% and SC 44%) and had higher total charges ($187,996 vs LC $99,020 and SC $82,239) and LOS (12.4 vs 6.1 and 6.9 days), all p CONCLUSIONS Compared to 2-3 levels, longer PCF had higher C5-C7 nerve palsy rates (3.3%) and 70% increased odds of sustaining ≥1 neurological complication. Longer PCF was also associated with increased rates of medical and total complications. This data can improve the ability to counsel patients regarding the risks and expectations of potential adverse outcomes preceding cervical fusion via posterior or combined anterior-posterior approach. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs.

P34. Evaluating the impact of multiple sclerosis on two-year postoperative outcomes following ACDF for cervical degenerative pathology: a propensity score-matched analysis

BACKGROUND CONTEXT The impact of neuromuscular disorders such as multiple sclerosis (MS) on outcomes following anterior cervical discectomy and fusion (ACDF) is underreported. PURPOSE Identify the impact of MS on two-year (2Y) postoperative complications and revisions following 2-3-level ACDF for cervical radiculopathy (CR) or myelopathy (CM). STUDY DESIGN/SETTING Retrospective review of the New York State Statewide Planning and Research Cooperative System (SPARCS) database. PATIENT SAMPLE Patients undergoing 2-3 level ACDF for CR/CM. OUTCOME MEASURES Two-year postoperative complications and reoperations. METHODS Patients undergoing 2-3 level ACDF for CR/CM were identified from SPARCS. Patients with a baseline diagnosis of MS were also identified. Patients with infectious/traumatic/neoplastic were excluded. The remaining patients were 1:1 propensity score-matched (MS to no-MS) for age, sex, race, Deyo index and compared for rates of 2Y postoperative complications and reoperations. Logistic regression models were utilized to determine significant predictors of these outcomes. RESULTS A total of 302 total patients were identified (MS, n=151; non-MS, n=151). MS patients were more frequently female (72.8% vs 27.2%, p 0.05. MS patients had similar rates of 2Y revisions (6.6% vs 5.3%, p=0.627). MS was not associated with medical, surgical, or overall complications or revisions at minimum 2Y follow-up. CONCLUSIONS Despite differences in demographics and higher rates of postoperative DVT, MS patients experienced comparable postoperative course with respect to other individual and overall complications and revisions following 2-3-level CF for degenerative spinal pathologies. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs.

Abstract 142: Hepatitis B And C Is Associated With Greater Postoperative Complications Following Mastectomy And Breast Reconstruction

Purpose: Prior studies have analyzed the effects of Hepatitis B and C (HB and HC) on outcomes following hepatic surgeries. Fewer have investigated this association in plastic and reconstructive surgeries. We present a large-scale database analysis of outcomes following mastectomy and breast reconstruction in patients with HB/HC compared to patients without infectious hepatitis. Methods: Data from the New York State Statewide Planning and Research Cooperative System (NYS SPARCS) from 2008-2013 was queried. Patients with confirmed HB and/or HC who underwent mastectomy and/or breast reconstruction procedures (identified using International Classification of Disease [ICD-9] codes) were compared to propensity-score matched patients without HB or HC who underwent these procedures to assess differences in post-surgical outcomes. Univariate analysis assessed differences in demographic information, baseline health characteristics, and perioperative factors as possible risk factors. Post-surgical outcomes were assessed by collecting infection rate, re-operation rate, postoperative complications, and length of hospital stay. Multivariate analysis revealed the effects of the possible risk factors on postoperative outcomes. Results: 36,072 patients were identified. The majority (35,898) had no history of HB/HC infection, and 174 patients had documented history of HB/HC. There was no difference in age between the cohorts (56.3 vs. 56.6 years for non-HB/HC and HB/HC patients, respectively). The majority of patients were female (98.6% of non-HB/HC, 95.9% of HB/HC patients, p=0.005). There were significantly more Black/African American, Native American/Alaska Native, and Asian patients in the HB/HC cohort (p<0.001). Insurance types also differed with a larger proportion of HB/HC patients utilizing Medicaid (37.9% vs. 12.0%, p<0.001). Hepatitis patients had more frequent comorbid HIV/AIDS, deficiency anemias, chronic pulmonary disease, coagulopathies, depression, diabetes without complications, drug abuse, hypertension, fluid/electrolyte disorders, psychoses, and renal failure (p<0.04). These were used as covariates in the multivariate analysis. HB and HC patients had a higher incidence of postoperative hemorrhage (8.0% vs. 3.1%, p<0.01), urinary system complications (1.1% vs. 0.3%, p=0.030), death during hospitalization (0.6% vs. 0.1%, p=0.003), and a lengthier overall hospital stay (2.39 vs. 2.84 days, p=0.03). Other complication rates were comparable, including shock, infection, and cardiac, respiratory, and digestive complications. HB and HC patients were more likely to have any postoperative complication (10.3% vs. 5.4%, p=0.004), and to have two or more postoperative complications (1.7% vs. 0.4%, p=0.008). Multivariate analysis revealed that HB/HC patients had greater odds of postoperative hemorrhage (OR=2.45, 95% CI 1.39-4.32, p=0.002) and greater odds of having at least one postoperative complication (OR=1.72, 95% CI 1.04-2.87, p=0.03). Conclusion: Patients with comorbid HB and/or HC had greater postoperative complications after undergoing mastectomy and breast reconstruction procedures compared to similar patients without HB and/or HC. They had greater odds of postoperative hemorrhage and of developing postoperative complications overall. They were more likely to be ethnic minorities and to be supported by Medicaid. Our results reveal that hepatitis infection may be an understudied contributor to historically demonstrated racial and insurance-based differences in operative complication rates.