Healthcare Cost And Utilization Project Research Articles

Measuring case severity: a novel tool for benchmarking and clinical documentation improvement

BackgroundSeverity of illness (SOI) is an All Patients Refined Diagnosis Related Groups (APR DRG) modifier based on comorbidity capture. Tracking SOI helps hospitals improve performance and resource distribution. Furthermore, benchmarking SOI plays a key role in Quality Improvement (QI) efforts such as Clinical Documentation Improvement (CDI) programs. The current SOI system highly relies on the 3 M APR DRG grouper that is updated annually, making it difficult to track severity longitudinally and benchmark against hospitals with different patient populations. Here, we describe an alternative SOI scoring system that is grouper-independent and that can be tracked longitudinally.MethodsAdmission data for 2019–2020 U.S. News and World Report Honor Roll facilities were downloaded from the Vizient Clinical Database and split into training and testing datasets. Elixhauser comorbidities, body systems developed from the Healthcare Cost and Utilization Project (HCUP), and ICD-10-CM complication and comorbidity (CC/MCC) indicators were selected as the predictors for orthogonal polynomial regression models to predict patients’ admission and discharge SOI. Receiver operating characteristic (ROC) and Precision-Recall (PR) analysis, and prediction accuracy were used to evaluate model performance.ResultsIn the training dataset, the full model including both Elixhauser comorbidities and body system CC/MCC indicators had the highest ROC AUC, PR AUC and predication accuracy for both admission (ROC AUC: 92.9%; PR AUC: 91.0%; prediction accuracy: 85.4%) and discharge SOI (ROC AUC: 93.6%; PR AUC: 92.8%; prediction accuracy: 86.2%). The model including only body system CC/MCC indicators had similar performance for admission (ROC AUC: 92.4%; PR AUC: 90.4%; prediction accuracy: 84.8%) and discharge SOI (ROC AUC: 93.1%; PR AUC: 92.2%; prediction accuracy: 85.6%) as the full model. The model including only Elixhauser comorbidities exhibited the lowest performance. Similarly, in the validation dataset, the prediction accuracy was 86.2% for the full model, 85.6% for the body system model, and 79.3% for the comorbidity model. With fewer variables and less model complexity, the body system model was more efficient and was determined to be the optimal model. The probabilities generated from this model, named J_Score and J_Score_POA, successfully measured SOI and had practical applications in assessment of CDI performance.ConclusionsThe J_Scores generated from the body system model have significant value in evaluating admission and discharge severity of illness. We believe that this new scoring system will provide a useful tool for healthcare institutions to benchmark patients’ illness severity and augment Quality Improvement (QI) efforts.

P4. INCREASED COST OF UNPLANNED HOSPITALIZATIONS FOR RACIAL MINORITIES AFTER COMMON PLASTIC SURGERY PROCEDURES

PURPOSE: The association of systems-level inequities with surgical outcomes resulting in racial disparities has been identified in the literature. Nonetheless, the effect for patients in the postoperative period is not well understood. We aimed to identify the effect of race on the rate and cost of unplanned hospitalizations after reconstructive plastic surgery. METHODS: An analysis of the Healthcare Cost and Utilization Project (HCUP) from 2006 to 2015 was used to identify adult patients undergoing autologous and implant-based breast reconstruction and abdominoplasty. Multivariable analyses were performed to determine the association of race with readmission rate and costs. RESULTS: Five percent of the 17,739 patients included in our sample experienced a readmission. There was no significant difference in the readmission rate for Black and Hispanic patients compared to White patients (Black patients OR:0.87, 95% CI:0.70-1.09, Hispanic patients OR:1.10, 95% CI:0.86-1.40). However, Black (OR:1.28, 95% CI:1.05-1.56) and Hispanic patients (OR:1.27, 95% CI:1.03-1.58) had increased odds of incurring greater cost compared to White patients, despite controlling for income and comorbidities. CONCLUSION: Racial minorities bear a substantially higher cost for readmission for breast and abdominoplasty procedures, resulting in discrepancies in access and care provision. Systems-level issues may influence access to care which can lead to delayed diagnoses. Future research is needed to examine the role of different payment models on inequities and identify solutions to foster more equitable care.

What Factors Correlate With Length of Stay and Readmission After Limb Lengthening Procedures? A Large-database Study.

Indications and techniques for limb lengthening procedures have evolved over the past two decades. Although there are several case series reporting on the complications and efficacy of these techniques, limited data are available on length of stay and hospital readmission rates after these procedures. (1) What is the median length of stay after lower limb lengthening procedures, and is variability in patient demographics, preoperative diagnosis, and surgical technique associated with length of stay? (2) What is the 1-year readmission rate after lower limb lengthening procedures? (3) Is variability in patient demographics, preoperative diagnosis, and surgical technique associated with varying rates of hospital readmission? Patients who underwent femoral or tibial lengthening from 2005 to 2015 in seven states were identified using the Healthcare Cost and Utilization Project (HCUP) State Inpatient Databases. These databases include a large, diverse group of patients across a wide range of hospitals and socioeconomic backgrounds with inclusion of patients regardless of payer. Between 2005 and 2015, there were 3979 inpatient admissions that were identified as involving femoral and/or tibial lengthening procedures based on ICD-9 procedure codes; of those, 2% (97 of 3979) of the inpatient admissions were excluded from analysis because they had ICD-9 procedure codes for primary or revision hip or knee arthroplasty, and 10% (394 of 3979) of the inpatient admissions were excluded because they involved repeated admissions of patients with previous hospitalization data within the database. This yielded 3488 patients for analysis. The median (interquartile range) age of patients was 18 years (12 to 41), and 42% (1481 of 3488) of patients were women. A total of 49% (1705 of 3469) of patients were children (younger than 18 years), 19% (675 of 3469) were young adults (18 to 34 years), 24% (817 of 3469) were adults (35 to 59 years), and 8% (272 of 3469) were seniors (60 years and older). Length of stay and rates of readmission at 1 year after the lengthening procedure were calculated. Univariate analysis was performed to examine associations between age, race, payment method, underlying diagnosis, bone lengthened, and lengthening technique with length of stay and readmission rate. Factors found to be significantly associated with the outcome variables (p < 0.05) were further examined with a multivariate analyses. Included patients had a median (IQR) length of hospital stay of 3 days (2 to 4). Given the poor explanatory power of the multivariate model for length of stay (R 2 = 0.03), no meaningful correlations could be drawn between age, race, underlying diagnosis, lengthening technique, and length of stay. The overall 1-year readmission rate was 35% (1237 of 3488). There were higher readmission rates among adult patients compared with pediatric patients (odds ratio 1.78 [95% confidence interval 1.46 to 2.18]; p < 0.001), patients with government insurance compared with commercial insurance (OR 1.28 [95% CI 1.05 to 1.54]; p = 0.01), and patients undergoing lengthening via external fixation (OR 1.61 [95% CI 1.29 to 2.02]; p < 0.001) or hybrid fixation (OR 1.81 [95% CI 1.38 to 2.37]; p < 0.001) compared with lengthening with internal fixation only. When counseling patients who may be candidates for limb lengthening, providers should inform individual patients and their caretakers on the anticipated length of hospital stay and likelihood of hospital readmission based on our findings. Adult patients, those with government insurance, and patients undergoing hybrid or external fixator limb lengthening procedures should be advised that they are at greater risk for hospital readmission. The relationship of specific patient-related factors (such as severity of deformity or associated comorbidities) and treatment-related variables (such as amount of lengthening, compliance with physical therapy, or surgeon's experience) with clinical outcomes after lower limb lengthening and the burden of care associated with hospital readmission needs further study. Level III, therapeutic study.

Trend Analysis of Substance Use Disorder During Pregnancy.

Objectives: This study aims to analyze the trends in substance use among pregnant women in the United States.Methodology: In this retrospective study, we utilized the National Inpatient Sample (NIS) dataset sponsored by the Agency for Healthcare Research and Quality (AHRQ) under the Healthcare Cost and Utilization Project (HCUP). Major Diagnostic Category (MDC) 14 (Pregnancy, Childbirth, and the Puerperium) and International Classification of Disease (ICD 10) codes were used to identify pregnancy-related diagnoses and presentations with any of the substance use disorder (SUD) indicators that met the inclusion criteria among the birthing population in the NIS dataset (2016-2018). We analyzed the demographic and regional characteristics between 2015 and 2018.Results: Among the population, a total of 23,475 (2.7%) had a primary or secondary diagnosis of SUD, and 851,428 (97.3%) did not. In the study group of 332,275 (2.8%) that met the inclusion criteria, 12,750 (0.1%) use alcohol, 108,960 (0.9%) had opioid use disorder (OUD), 171,490 (1.4%) use cannabis, 6,375 (0.1%) use sedatives, 28,075 (0.2%) use cocaine, 48,765 (0.4%) use other stimulants, 1,155 (0%) use hallucinogens, 115 (0%) use inhalants, and 23,950 (0.2%) had other psychoactive diagnosis. Further analysis comparing the risk of severity and mortality at presentation, procedure type, delivery method, and cost of care shows statistically significant differences (p < 0.005) between the study and control groups.Conclusion: The current trends necessitate a further assessment and implementation of comprehensive community-based treatment programs tailored to the most frequent regional SUD presentations, which could aid in mitigating drug use during pregnancy.

Issues With Big Data: Variability in Reported Demographics and Complications Associated With Posterior Spinal Fusion in Pediatric Patients.

Clinical and administrative registries provide large volumes of data that can be used for clinical research. However, there are several limitations relating to the quality, consistency, and generalizability of big data. In this study, we aim to compare reported demographics and certain outcomes in patients undergoing posterior spinal fusion (PSF) for adolescent idiopathic scoliosis (AIS), neuromuscular scoliosis (NS), and Scheuermann kyphosis (SK) between 3 commonly utilized databases in pediatric orthopaedic research. We used International Classification of Diseases, Ninth Revision (ICD-9), International Classification of Diseases, 10th Revision (ICD-10), and Current Procedural Terminology (CPT) codes to identify patients in the National Surgical Quality Improvement Program (NSQIP), Healthcare Cost and Utilization Project (HCUP), and Pediatric Health Information System (PHIS) between the ages of 10 to 18 that underwent PSF for AIS, SK, and NS from 2012 to 2015. We compared various demographic factors, such as sex, race/ethnicity, age, and rates of postsurgical infection and 30-day readmissions. Data was analyzed with descriptive and univariate statistics. We identified 9891 patients that underwent PSF in NSQIP, 10,771 patients in PHIS, and 4335 patients in HCUP over the study period. There were significant differences in patient demographics, readmission rates, and infection rates between all patients that underwent PSF across the databases (P<0.01), as well as specifically in patients with AIS (P<0.01). HCUP had the highest proportion of Hispanic patients that underwent PSF (13.5%), as well as patients who had AIS (13.3%) or NS (17.9%). The PHIS database had the highest proportion of patients undergoing PSF for SK. Among patients with NS, there were significant differences in race across the databases (P<0.01), but no significant differences in sex, ethnicity, or readmission (P>0.05). In addition, there were significant differences in race (P=0.04) and readmission (P=0.01) across databases for patients with SK, but no differences in sex or ethnicity (P>0.05). NSQIP reported the highest rate of 30-day readmissions for patients undergoing PSF (17.9%) compared with other databases (HCUP 4.1%, PHIS 12.1%). There are significant differences in patient demographics, sample sizes, and rates of complications for pediatric patients undergoing PSF across 3 commonly utilized US administrative databases. Given the variability in reported outcomes and demographics, generalizability is difficult to extrapolate from these large data sources. In addition, certain databases should be selected to appropriately power studies focusing on particular patient populations or outcomes.

Sociodemographic characteristics of patients and their use of post-bariatric contouring surgery in the US

BackgroundExpansion of healthcare insurance coverage to bariatric surgeries has led to an increased demand from patients for post-bariatric contouring surgeries. This study examined the relationship between the use of contouring procedures on post-bariatric surgery patients, clinical need and sociodemographic factors.MethodsData were extracted from the Healthcare Cost and Utilization Project (HCUP) Nationwide Ambulatory Surgery Sample (NASS) regarding hospital-owned ambulatory surgical centres in the US. Episodes without missing values relating to patients, 18 years old or above were the primary unit of analysis. Episodes were excluded if the field regarding expected payer was reported as “no charge.” The primary outcome was the likelihood of panniculectomy, abdominoplasty, and mastopexy among post-bariatric surgery patients; and the degree to which uptake of these types of contouring surgery were associated with age, gender, median households’ income, expected payer, patient’s location, and comorbidity.ResultsA total of 66,368 weighted episodes of care received by patients who had had bariatric surgery were extracted (54,684 female [82.4%]; mean age, 51.3 [SD, 12.1]; 6219 episodes had contouring surgeries [9.37%]). Panniculectomy was the most common post-bariatric contouring procedure (3.68%). Uptake of post-bariatric contouring procedures was associated with age, sex, payment type, area-based measures of median household income, and patient location. Compared to Medicare insured patients, the odds of receiving contouring surgery among self-payers were 1.82 (95% CI, 1.47 to 2.26) for panniculectomy, 14.79 (95% CI, 12.19 to 17.93) for abdominoplasty and 47.97 (95% CI, 32.76 to 70.24) for mastopexy. Rank order of comorbidity profiles also differed between insured and self-paying recipients of contouring surgery.ConclusionsInsurance status of bariatric surgery patients and their sex were strongly associated with receipt of a range of contouring procedures. Self-payments were associated with a doubling of the odds of having panniculectomy and an increase in the odds to approximately 14 times for abdominoplasty and 48 times for mastopexy. Thus, access to contouring surgery by post-bariatric patients may be disproportionately dependent on personal preference supported by ability to pay rather than clinical need. Further research is needed to examine the impact of contouring or delayed/denied contouring on health status.

497 Developing Predictive Models to Anticipate Shunt Complications in 33,248 Pediatric Patients with Shunted Hydrocephalus Utilizing Machine Learning

INTRODUCTION: Hydrocephalus is a common pediatric neurosurgical pathology and is typically treated with a ventricular shunt. Shunt complications are common in pediatric hydrocephalus, and ∼30% of patients experience shunt failure during the first year after shunt placement. METHODS: We queried the Healthcare Cost and Utilization Project (HCUP) National Readmissions Database (NRD) from 2016 to 2017 using ICD-10 codes and found 33,248 pediatric patients who underwent shunt placement. Elixhauser comorbidity index (ECI) scores, Johns Hopkins Adjusted Clinical Groups (JHACG) frailty-defining criteria, and Major Diagnostic Category (MDC) classifications were obtained. To determine predictors of shunt complication, the database was divided into a training dataset (n=19,948), a validation dataset (n=6,650), and a testing dataset (n=6,650). Multivariable analysis was performed using the training dataset to identify significant independent predictors of shunt complications in the other datasets. Statistically significant variables were used to develop logistic regression models, and post-hoc receiver operating characteristic (ROC) curves were created. RESULTS: Average age of the 33,248 pediatric patients was 6.9±5.7 years. Variables found to be positively correlated with shunt complications included: number of diagnoses during primary admission (OR: 1.05, 95% CI: 1.04-1.07) and neurological MDCs (OR: 3.83, 95% CI: 3.33-4.42). Variables found to be negatively correlated with shunt complications included: ECI (OR: 0.82, 95%CI: 0.77-0.86), Female sex (OR: 0.87, 95%CI: 0.76-0.99), and elective admissions (OR: 0.62, 95%CI: 0.53-0.72). The ROC curve for the regression model utilizing all significant predictors of readmission in multivariate analysis demonstrated an area under the curve (AUC) of 0.733. CONCLUSION: Expeditious and safe treatment of pediatric hydrocephalus is of paramount importance. Our machine learning algorithm established numerous variables predictive of shunt complications in the pediatric population. Our AUC is suggestive of excellent predictive value. These risk factors may be of clinical importance when evaluating pediatric patients with hydrocephalus.

Abstract EP03: 30-day Readmission Rates After Targeted Temperature Management For Cardiac Arrest

Background: In its 2015 guidelines, the American Heart Association instituted a Class 1 recommendation for use of Targeted Temperature Management (TTM) in patients with cardiac arrest (CA) . Prior observational studies have suggested benefits using TTM in these patients, while many randomized control trials have failed to replicate these results. Regardless, there have been studies with contradictory conclusions addressing readmission rates in patients receiving TTM. Thus, we sought to ascertain this relationship further by identifying 30-day readmission rates in Out-of-Hospital Cardiac Arrest (OHCA) patients receiving TTM. Methods: We utilized the Healthcare Cost and Utilization Project (HCUP) National Readmission Database (NRD) for data collection and International Classification of Diseases, Ninth and Tenth Revision, Clinical Modification codes to identify 353,380 cardiac arrest cases and clinical variables. The primary outcome includes all-cause readmission rates within 30 days from cardiac arrest index admission discharge. We then conducted a logistic regression analysis after accounting for variables to assess odds of 30-day readmission in these patients. Results: TTM following CA was associated with a decreased likelihood of readmission at 30 days(OR: 0.822, p = 0.006, 95% CI [0.716-0.945]). Comorbidities associated with an increased risk of readmission after controlling for the other variables include congestive heart failure(OR: 1.259, p = 0.027, 95% CI [1.207-1.313]) and renal failure (OR: 1.363, p = 0.031, 95% CI [1.304- 1.425]). Of these patients receiving TTM compared to those who did not receive TTM, there was a greater proportion of cardiogenic shock (23.7% vs. 12.3%, p &lt; 0.001), acute heart failure (20.1% vs. 17.3%, p &lt; 0.001), coronary artery disease (53.0% vs. 46.2%, p &lt; 0.001), ischemic cardiomyopathy (61.7% vs. 53.1%, p &lt; 0.001), ventricular tachycardia (9.6% vs. 8.6%, p = 0.045), and ventricular fibrillation (56.6% vs. 24.1%, p&lt;0.001). Conclusion: TTM following CA is associated with decreased odds of readmission at 30 days. Prior long-term comorbidities represented most of the covariates for readmissions in this patient population.

Abstract WP115: Venous Thromboembolism Readmission In Acute Ischemic Stroke

Introduction: Venous thromboembolism (VTE) is a frequent and potentially life-threatening complication of acute ischemic stroke (AIS). We performed a nationwide analysis to evaluate rate and risk factors for VTE readmission in patients with AIS. Methods: Using the Healthcare Cost and Utilization Project (HCUP) Nationwide Readmission Database, we included adult patients with a principal discharge diagnosis of AIS from 2016 to 2018. AIS, pulmonary embolism, deep vein thrombosis and other diagnosis were identified based on standard ICD-10 CM codes. Patients who had VTE diagnosis during the index admission were excluded. We determined 90-day VTE readmission rates and trends in patients with a principal diagnosis of AIS stratified by 30-day epochs. We then constructed a stepwise binary logistic regression model to determine odds ratios (OR) of demographic and clinical factors associated with VTE readmission rates. Results: Of the 1,023,478 patients with AIS, 8378 (0.82%) had VTE during readmission and 2906 (0.28%) had VTE as principal diagnosis for readmission within 90 days of discharge. Among them, more than half (4557, 54.39% and 1581, 54.40%, respectively) of patients were readmitted within 30 days of discharge. The rate of VTE readmission decreased further away from index event (P &lt; 0.001). In the Cox regression model, obesity (OR 1.48, 95% CI 1.24-1.76, P &lt; 0.001), plegia of at least one limb (OR 1.24, 95% CI 1.08-1.43, P = 0.003), longer hospital length of stay (OR 1.02, 95% CI 1.01-1.02, P &lt; 0.001), higher NIHSS (OR 1.04, 95% CI 1.03-1.05, P &lt; 0.001) were associated with VTE readmission. Conversely, VTE readmission rates were lower in patients with a history of atrial fibrillation/flutter (OR 0.72, 95% CI 0.62-0.84, P &lt; 0.001). Conclusion: Patients with obesity, paralysis, higher NIHSS score, or prolonged hospital length of stay are at higher risk for VTE readmission. AF strokes are less likely to have DVT/PE, perhaps due to anticoagulant use in such patients. Studies are needed to determine whether early mobilization and mechanical and/or chemical prophylaxis reduces VTE risk in high-risk patients.

Abstract TP215: Principal Hospital Admission Of Hypertension Foreshadows 3-month Ischemic And Hemorrhagic Stroke

Introduction: Hypertension is a well-recognized risk factor for acute ischemic stroke (AIS) and intraparenchymal hemorrhagic (IPH). We performed a nationwide analysis to evaluate the rate of AIS and IPH readmission comparing patients with hypertension as a principal versus non-principal diagnosis. Methods: Using the Healthcare Cost and Utilization Project (HCUP) Nationwide Readmission Database, we included patients with index admission diagnosis (principal or non-principal) of hypertension from 2016 to 2018. Hypertension, AIS, IPH, and other diagnoses were identified based on standard ICD-10 CM codes. Patients with AIS or IPH diagnoses during the index admission were excluded. We determined post-discharge 90-day admission trends by Kaplan-Meier estimates in patients with principal versus non-principal diagnosis of hypertension. We then constructed Cox regression models to determine associations between principal diagnosis of hypertension and AIS or IPH admission rates, with adjustment to age, gender, insurance type, zip code income quartile, index admission disposition, and Elixhauser comorbidities. Results: Of the 36,173,682 patients identified, 586,613 (1.62%) had a principal diagnosis of hypertension during the index admission. Patients with a principal diagnosis of hypertension were more likely to have an AIS or IPH during 90-day follow up versus non-principal diagnosis of hypertension (Figure). In the Cox regression model after the aforementioned adjustments, a principal diagnosis of hypertension was significantly associated with AIS (HR 2.37, 95% CI 2.23-2.53, P &lt; 0.001) and IPH admission (HR 2.25, 95% CI 1.97-2.56, P &lt; 0.001). Conclusion: Patients with hypertensive conditions warranting hospitalization as principal diagnosis are at increased risk of AIS and IPH in the first 90-days after discharge. Post-discharge planning should include intensive ambulatory blood pressure monitoring and control by outpatient providers.

Abstract WMP85: Postpartum Risk Of Cerebral Venous Thrombosis

Background: The postpartum state is associated with a heightened risk of thrombosis. The duration of heightened risk for postpartum cerebral venous thrombosis (CVT) is uncertain. Methods: Using claims data from the Healthcare Cost and Utilization Project (HCUP) from all emergency departments and acute care hospitalizations in Florida from 2005-2015 and New York from 2006-2015, we identified women aged ≥18 years old who were hospitalized for labor and delivery. CVT was ascertained using previously validated International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM), diagnosis codes. For women with multiple labor-related hospitalizations during a single 40-week period, we excluded cases of false labor by identifying delivery as the latest hospitalization during that time. Patients with claims for CVT before their first recorded delivery, and women with a second delivery during the follow-up period were also excluded. We compared the likelihood of a first-ever recorded CVT during postpartum days 0-41 compared with the same period 1 year later. We repeated this crossover-cohort analysis for consecutive 6-weeks periods after delivery, as compared with the corresponding 6-week period 1 year later. We used McNemar's to calculate odds ratios for each 6-week interval. Results: Among the 1,406,447 women with a first recorded delivery, the risk of CVT was markedly higher within the 6 weeks after delivery than in the same period 1 year later (22 versus 3 CVTs per million deliveries). This corresponded to an absolute risk difference of 19 events per million (95% confidence interval [CI], 11-27) and an odds ratio of 10.0 (95% CI, 3.1-51.2). There was no significant increase in the risk of CVT during the period of 7 to 12 weeks after delivery as compared with the same period 1 year later with an absolute risk difference of 6.0 events (95% CI, 0-11) per million deliveries and an odds ratio of 3.0 (95% CI, 0.9-12.8). Conclusion: There appears to be a heightened risk of CVT for 6 weeks after delivery.

Association of Infantile Hemangiomas and Retinopathy of Prematurity: Analysis of the Multicenter KID

Introduction: Retinopathy of prematurity (ROP) and infantile hemangiomas (IHs) both have similar proposed pathophysiological mechanisms. IH is more common in preterm than term infants. Hypoxia-induced mediators like vascular endothelial growth factor have been found elevated in children with hemangiomas. The aim of our study was to determine if there is an association between ROP and IH in preterm infants and to investigate racial/ethnic and gender differences of ROP and IHs in this cohort. Methods: We accessed the national multicenter Kids’ Inpatient Database (KID) Healthcare Cost and Utilization Project (HCUP) including admissions at age ≤28 days. Eligible infants were identified by using ICD-9 codes of ROP and IH in infants with gestational age (GA) ≤32 weeks and/or birth weight ≤1,500 g during the years 2003, 2006, 2009, and 2012. A weight-based analysis was performed using SAS Enterprise Guide 7.1 for complex sample design. Results: In the cohort of 1,068,502 eligible infants, the prevalence of IH was 4.7 per 1,000 preterm admissions (<32 weeks). ROP prevalence was 16% for GA ≤26 weeks, 12.5% for GA 27–30 weeks, and 2.7% for GA 31–32 weeks. IH was significantly higher in infants with ROP; this relationship was consistent among all stages of ROP. Regression analysis showed that females are at increased risk of IH with ROP compared to males (adjusted odds ratio [aOR]: 2.00 [1.85–2.56]). White non-Hispanic premature infants had an increased risk of IH with concomitant ROP compared to both African American (aOR: 3.9 [2.63–4.76]) and Hispanic (aOR: 1.2 [1.14–1.38]) infants. However, African American infants had an increased risk of ROP compared to white non-Hispanic infants (aOR: 1.16 [1.07–1.14]). These genders and racial/ethnic disparities were consistent among GA categories. Conclusions: To our knowledge, this is the largest cohort based on a national multicenter database comparing an association between ROP and IH. A strong association between ROP and IH may suggest similar risk factors and/or pathophysiology. A further role of genetic factors could explain racial/ethnic differences in both conditions despite similar pathogenesis. These findings may open up new bases of research for management and prevention strategies.

Predictors and Cost Comparison of Subsequent Urinary Stone Care at Index Versus Non-Index Hospitals

ObjectiveTo examine the effects of care fragmentation, or the engagement of different health care systems along the continuum of care, on patients with urinary stone disease. MethodsAll-payer data from the 2016 Healthcare Cost and Utilization Project (HCUP) State Databases from Florida (FL) and New York (NY) were used to identify a cohort of adult patients with an emergency department visit for a diagnosis of urolithiasis, who subsequently re-presented to an index or non-index hospital for renal colic and/or urological intervention. Patient demographics, regional data, and procedural information were collected and 30-day episode-based costs were calculated. Multivariable logistic and gamma generalized linear regression were utilized to identify predictors of receiving subsequent care at an index hospital and associated costs, respectively. ResultsOf the 33,863 patients who experienced a subsequent encounter related to nephrolithiasis, 9593 (28.3%) received care at a non-index hospital. Receiving subsequent care at the index hospital was associated with fewer acute care encounters prior to surgery (2.5 vs 2.7; P <.001) and less days to surgery (29 vs 42; P < .001). Total episode-based costs were higher in the non-index setting, with a mean difference of $783 (Non-index: $13,672, 95% CI $13,292-$14,053; Index: $12,889, 95% CI $12,677 - $13,102; P < .001). ConclusionRe-presentation to a unique healthcare facility following an initial diagnosis of urolithiasis is associated with a greater number of episode-related health encounters, longer time to definitive surgery, and increased costs.

Self-inflicted craniomaxillofacial gunshot wounds by handguns are more deadly than those by shotguns: an NIS study.

The purpose of the present study is to determine whether handguns discharged at the craniomaxillofacial region are associated with greater fatality in suicide attempts relative to shotguns. The following retrospective cohort study was completed using the Nationwide Inpatient Sample (NIS), a Healthcare Cost and Utilization Project (HCUP) database. The primary predictor variable was the type of firearm (handguns vs. shotguns). The primary outcome variable was prognosis (fatality vs. no fatality). A binary logistic regression was performed to determine the risk of mortality of all predictor variables. Our final sample consisted of 223 patients. Handguns (85.2%) were the most popular firearm. Relative to black patients, white patients were over three times more likely to die (p < 0.05). Compared to patients who were transferred in, patients not transferred in were eight times more likely to die (p = 0.000). Relative to patients who were admitted electively, patients who were not admitted electively were 16.7 times more likely to die (p < 0.01). Finally, relative to patients who used shotguns, patients who used handguns were 3.4 times more likely to die (p = 0.002). Self-inflicted gunshot wounds to the maxillofacial region by handguns were more lethal than shotguns.

CROHN’S DISEASE AND ACUTE PANCREATITIS; ASSOCIATION, OUTCOMES, AND PREDICTORS OF 90-DAY READMISSION WITH ACUTE PANCREATITIS. INSIGHTS FROM THE NATIONWIDE READMISSION DATABASE

Abstract INTRODUCTION It’s a controversy if pancreatitis can be considered one of the extraintestinal manifestations of Crohn’s disease (CD). (Barthet et al AJG) CD patients have a fourfold increase in the risk of acute pancreatitis (AP) compared to the general population. (Rasmussen et al Scan J Gastroenterol) The annual incidence of AP in the general population ranges from 13 to 45/100,000 persons. (Dhiraj et al Gastroenterology) METHODS We collected data from the publicly available Healthcare Cost and Utilization Project (HCUP) Nationwide Readmission Databases (NRD) 2017-2018. We identified individuals discharged with a diagnosis of CD or one of its complications. The case selection process is summarized in Figure 1. We studied the burden of AP in patients with Crohn’s disease, hospitalization outcomes, and predictors of hospital readmissions with AP. Median and IQR were used to describe Continuous variables, and proportions were used with categorical variables. Comparison between groups was performed by Mann Whitney test for continuous variables and Chi-Square test for Categorical variables. We performed a binary logistic regression analysis for predictors of 90-day readmission with AP. Statistical analyses were performed using SPSS Version 25 (IBM Corporation, Armonk, NY, USA). RESULTS We analyzed 147,501 index hospital discharges with Crohn’s disease or one of its complications, 0.9% had chronic pancreatitis, and 1.6% had a diagnosis of AP. Around 14% of AP was biliary, 13% was Alcohol-induced, 7% was drug-induced, while 66% was classified as idiopathic, other, or unspecified AP. Table 1 shows a Comparison between CD patients with AP and CD patients without AP. Having AP on index hospitalization wasn’t related to CD complication status as those without AP had a higher prevalence of complications (P values &amp;lt;0.001). After excluding those who died in index hospitalization, those with Chronic Pancreatitis, and those discharged in Oct-Dec, 0.3% of patients had 90-day readmission with AP. Characteristics of CD patients with 90-day readmission with AP are illustrated in Table 2. A binary logistic regression for predictors of 90-day readmission with AP is illustrated in Table 3. The chi-square test results indicate that our regression model is a significant improvement in fit relative to an intercept-only model (χ2= 498.489, p&amp;lt;.001). CONCLUSION AP is associated with higher morbidity and mortality in patients with CD. Age, number of comorbidities, and having AP in index hospitalization are significant predictors of 90-day readmission with AP. Further biochemical studies would be needed to investigate the association between AP and CD. The study was limited by the lack of a specific ICD-10 code for autoimmune pancreatitis and the inability to account for common CD drugs associated with Pancreatitis.

INFLAMMATORY BOWEL DISEASE IN APPALACHIA: AN INVESTIGATION INTO DISPARATE OUTCOMES WITHIN THE STATE OF KENTUCKY

Abstract INTRODUCTION Inflammatory Bowel Disease (IBD) affects an estimated 3.1 million US adults, with an increasing prevalence in recent years1. Of these, 1.1 million live in the CDC classified “Southern” region, which includes the state of Kentucky. The eastern portion of the state is part of the Appalachian region, a low-resourced area with a dense, vulnerable population. The majority of Kentucky’s Appalachian counties are classified as distressed (ranked bottom 10% in the US based on economic development), and suffer a higher prevalence of chronic disease2. To our knowledge, there is no published data quantifying the healthcare burden of IBD or evaluating disease outcomes in this vulnerable community. METHODS We utilized the hospital inpatient discharge and outpatient services databases available through the healthcare cost and utilization project (HCUP) to assess outcomes among patients with a diagnosis of either Crohn’s disease (CD) or Ulcerative Colitis (UC) in the state of Kentucky. Outcomes of interest were aggregated at the county level and included number of hospital admissions, number of total outpatient visits and visits to an emergency department, inpatient and outpatient surgeries. Data was reported as number of incident visits per 100,000 population/year, collected between the years 2009-2019. RESULTS We found that CD patients residing in Appalachian counties were more likely to be admitted to the hospital (42.72 vs. 29.69, p-value= 0.0001), require inpatient surgery (6.91 vs 5.35, p-value= 0.006), and be admitted for abdominal pain as a complication of their disease (3.82 vs 2.28, p-value = 0.0031) than non-Appalachians. Appalachian CD patients also had a higher number of ER visits (84.06 vs. 52.64, p-value = 0.0102), outpatient surgery visits (161.24 vs 112.35, p-value = 0.0024), and clinic visits for abdominal pain (39.8 vs 24.34, p-value=0.0002) than non-Appalachians. UC patients in Appalachian counties were more likely to require inpatient surgery (1.36 vs 0.74, p-value 0.0007) than non-Appalachians. There was no significant difference in rates of fistula, perforation, or malabsorption in either CD or UC based on Appalachian county data. CONCLUSION These data support the notion of disparate outcomes in IBD based on Appalachian county residence in the state of Kentucky. This was most striking when evaluating Crohn’s disease in Appalachians, as evidenced by hospital admissions, surgeries, and ER visits. Appalachian inpatients with UC had a significantly higher rate of surgical resection. These findings highlight a need for aggressive investigation into root causes of these disparate outcomes, with particular attention to identifying and subsequently alleviating barriers to IBD care. Whether these data infer treatment of Appalachian IBD patients should be managed more aggressively earlier in the disease course awaits further investigation.

BURDEN OF NON-ALCOHOLIC FATTY LIVER DISEASE AMONG PATIENTS WITH CROHN’S DISEASE: ANALYSIS OF THE NATIONWIDE READMISSION DATABASE

Abstract INTRODUCTION Non-alcoholic fatty liver disease (NAFLD) is a spectrum that ranges from hepatic steatosis to Non-alcoholic steatohepatitis (NASH). NASH can lead to liver fibrosis, cirrhosis, and hepatocellular carcinoma. (Chao CY et al. World J Gastroenterol). Both NAFLD and Crohn’s disease are associated with increased gut permeability. (Miele L et al Hepatology, Pearson AD et al. Br Med J). Current evidence has shown fatty liver disease to be the most common explanation for abnormal liver tests in patients with IBD. (Gisbert JP et al. Inflamm Bowel Dis). A practical approach may be pursued in identifying patients with IBD at risk for having NAFLD and subsequently identifying patients at an increased risk of morbidity and mortality. METHODS We collected data from the publicly available Healthcare Cost and Utilization Project (HCUP) Nationwide Readmission Databases (NRD) 2016, 2017, and 2018. We identified and analyzed the burden of NAFLD amongst individuals discharged with a diagnosis of CD or one of its complications (Figure 1). We studied the characteristics, co-morbidities, and hospitalization outcomes of CD patients with NAFLD. A subgroup analysis was performed for those with NASH. Patients who died in index hospitalization or were discharged in Dec. were excluded from the 30-day readmission analysis. Median and IQR were used to describe Continuous variables, and proportions were used with categorical variables. Comparison between groups was performed by Mann Whitney test for continuous variables and Chi-Square test for Categorical variables. We performed a binary logistic regression analysis for predictors of NAFLD among patients with CD. Statistical analyses were performed using SPSS Version 25 (IBM Corporation, Armonk, NY, USA). RESULTS We analyzed 215,049 index hospital discharges with Crohn’s disease or one of its complications. 2.4% had NAFLD. Hypertension, diabetes, Dyslipidemia, Obesity, and metabolic syndrome are the five most recognized risk factors for NAFLD in the general population. 35.9% of CD patients with NAFLD had none of those risk factors (Table 1). 0.42 % of CD patients had a diagnosis of NASH. Subgroup analysis for CD patients with NASH showed increased mortality, increased length of stay, and a higher rate of 30-day readmission among CD patients with NASH (Table 2). Predictors of NAFLD in patients with CD are illustrated in Table 3. The chi-square test results indicate that our regression model is a significant improvement in fit relative to an intercept-only model (χ2 = 1593.171, p&amp;lt;.001). CONCLUSION The presence of NAFLD in patients with CD is associated with metabolic syndrome components and CD-specific risk factors. The development of NASH is associated with increased morbidity and mortality in patients with CD. Further studies are needed to investigate the pathogenesis of NAFLD in CD patients.

CROHN'S DISEASE AND ACUTE PANCREATITIS; ASSOCIATION, OUTCOMES, AND PREDICTORS OF 90-DAY READMISSION WITH ACUTE PANCREATITIS. INSIGHTS FROM THE NATIONWIDE READMISSION DATABASE

It’s a controversy if pancreatitis can be considered one of the extraintestinal manifestations of Crohn’s disease (CD). (Barthet et al AJG) CD patients have a fourfold increase in the risk of acute pancreatitis (AP) compared to the general population. (Rasmussen et al Scan J Gastroenterol) The annual incidence of AP in the general population ranges from 13 to 45/100,000 persons. (Dhiraj et al Gastroenterology) We collected data from the publicly available Healthcare Cost and Utilization Project (HCUP) Nationwide Readmission Databases (NRD) 2017-2018. We identified individuals discharged with a diagnosis of CD or one of its complications. The case selection process is summarized in Figure 1. We studied the burden of AP in patients with Crohn’s disease, hospitalization outcomes, and predictors of hospital readmissions with AP. Median and IQR were used to describe Continuous variables, and proportions were used with categorical variables. Comparison between groups was performed by Mann Whitney test for continuous variables and Chi-Square test for Categorical variables. We performed a binary logistic regression analysis for predictors of 90-day readmission with AP. Statistical analyses were performed using SPSS Version 25 (IBM Corporation, Armonk, NY, USA). We analyzed 147,501 index hospital discharges with Crohn’s disease or one of its complications, 0.9% had chronic pancreatitis, and 1.6% had a diagnosis of AP. Around 14% of AP was biliary, 13% was Alcohol-induced, 7% was drug-induced, while 66% was classified as idiopathic, other, or unspecified AP. Table 1 shows a Comparison between CD patients with AP and CD patients without AP. Having AP on index hospitalization wasn’t related to CD complication status as those without AP had a higher prevalence of complications (P values <0.001). After excluding those who died in index hospitalization, those with Chronic Pancreatitis, and those discharged in Oct-Dec, 0.3% of patients had 90-day readmission with AP. Characteristics of CD patients with 90-day readmission with AP are illustrated in Table 2. A binary logistic regression for predictors of 90-day readmission with AP is illustrated in Table 3. The chi-square test results indicate that our regression model is a significant improvement in fit relative to an intercept-only model (χ2 = 498.489, p<.001). AP is associated with higher morbidity and mortality in patients with CD. Age, number of comorbidities, and having AP in index hospitalization are significant predictors of 90-day readmission with AP. Further biochemical studies would be needed to investigate the association between AP and CD. The study was limited by the lack of a specific ICD-10 code for autoimmune pancreatitis and the inability to account for common CD drugs associated with Pancreatitis.Figure 2Table 1, 2 and 3View Large Image Figure ViewerDownload Hi-res image Download (PPT)

Utilizing Data Mining Algorithms and Geospatial Analysis to Study Cataract Surgeries in Florida

Purpose:To determine statewide cataract surgery rates with cataract extraction with intraocular lens implantation (CEIOL) in Florida from 2005 to 2014 among Caucasians, African–Americans, Hispanics, and Asian/Pacific Islanders.Methods:This is a retrospective database study analyzing ambulatory surgical data in Florida from 2005 to 2014. Using the Agency for Healthcare Research and Quality's Healthcare Cost and Utilization Project (HCUP) and State Ambulatory Surgery and Services Databases (SASD), the authors utilized data mining algorithms to analyze and graphically represent disparities in the delivery of cataract surgery, changes in surgery volume, and demographic characteristics in patients 65 years and older in all Florida counties from 2005 to 2014.Results:Cataract surgeries performed in patients ≥65 years of age represented 1,892,132 (14.90%) of the 12,695,932 total ambulatory surgical procedures from 2005 to 2014 in the HCUP-SASD Florida database. More surgeries were performed in females versus males, P < 0.001. Caucasians, African–Americans, and Hispanics represented 82.23%, 4.95%, and 10.69% of the utilization rate of all CEIOLs, respectively. From 2005 to 2014, the average surgery volume increased by an average rate of change of 1.29%. Cataract surgery penetration in the general population observed a steady decrease from 18.82% in 2005 to 16.66% in 2014.Conclusions:Cataract surgery in Florida exhibited an unequal distribution with respect to gender and race, and select counties exhibited marked changes in surgical volume over the past 11 years. This study establishes a method for data mining and geospatial analysis to study surgical and epidemiological trends and identify disparities in delivery of healthcare.

Outcomes of endoscopic retrograde cholangio-pancreatography in patients with liver transplant.

Biliary complications are the leading causes of morbidity and mortality after liver transplant (LT). However, national data on endoscopic retrograde cholangiopancreatography (ERCP) usage and outcomes in LT patients are lacking. Our study aims to identify the trends, outcomes, and predictors of ERCP and related complications in this patient subgroup. We derived our study cohort from the Nationwide Inpatient Sample (NIS) of the Healthcare Cost and Utilization Project (HCUP) between 2007 and 2017. LT patients were identified using ICD-9/10CM diagnosis codes and patients who underwent ERCP were identified by ICD-9/10-CM procedure codes. We utilized the Cochrane-Armitage trend test and multivariate logistic regression to analyze temporal trends, outcomes, and predictors. A total of 372,814 hospitalizations occurred in LT patients between 2007 and 2017. ERCP was performed in 2.05% (n = 7632) of all hospitalizations. There was a rise in ERCP procedures from 1.96% (n = 477) in 2007 to 2.05% (n = 845) in 2017. Among LT patients who underwent ERCP, the in-hospital mortality rate was 1% (n = 73) and 8% (n = 607) were discharged to facilities. Mean length of hospital stay was 7 ±0.3 days. Septicemia was the most common periprocedural complication (18.3%, n = 1399) followed by post-ERCP pancreatitis (8.8%, n = 674). There has been an increase in ERCP procedures over the past decade among LT patients. Our study highlights the periprocedural complications and outcomes of ERCP in LT patients from a nationally representative dataset.