Medicare Standard Analytical Files Research Articles

Cost savings for pVAD compared to ECMO in the management of acute myocardial infarction complicated by cardiogenic shock: An episode-of-care analysis.

Acute myocardial infarction complicated by cardiogenic shock (AMICS) occurs in up to 10% of acute myocardial infarction admissions and is associated with high mortality, frequent adverse outcomes, prolonged hospitalizations, extensive medical resource utilization, and major cost. Using hospital cost data for Medicare Fee-for-Service (FFS) patients with AMICS, we sought to evaluate in hospital and 45-day outcomes and cost, comparing patients treated with percutaneous ventricular assist device (pVAD) versus extracorporeal membrane oxygenation (ECMO). The goal of this study was to better understand clinical and economic outcomes of AMICS to help clinicians and hospitals optimize outcomes most economically for AMICS patients. A retrospective claims analysis identified patients from the full census Medicare Standard Analytic Files compiled by the Center for Medicare and Medicaid (CMS) including: Inpatient, Outpatient, Skilled Nursing Facility and Home Health files for Medicare FFS beneficiaries. Study costs were defined as the total costs incurred by providers for treating a population with AMICS. Medicare FFS beneficiaries who experienced an inpatient admission during the index period (January 1, 2015 to March 31, 2017) with a diagnosis of AMICS were eligible for study inclusion and were identified by the presence of appropriate International Classification of Diseases, Ninth and Tenth Versions (ICD-9 and ICD-10) diagnosis and procedure codes. To create a matched sample and control for any baseline differences, a 1:1 Propensity Score Matching (PSM) was performed based on criteria such as age, gender, race, geographic distribution, and 11 high-cost comorbidities (e.g., congestive HF, coronary artery disease, diabetes, etc.). Index length of stay (LOS), index costs, discharge disposition (including mortality), post-index utilization, and episode-of-care (EOC) costs were reviewed. EOC was defined as index admission for all patients plus a 45-day post index period for patients who survived the index admission. Each cohort included 338 patients. Index in-hospital mortality rates were 53% for pVAD versus 64% for ECMO (178 vs. 217; p=.0023), and total EOC in-hospital mortality rates were 66% for pVAD versus 74% for ECMO (222 vs. 250; p=.0160). Index LOS for pVAD was 27% lower versus ECMO (12.12 vs. 16.59; p=.0006). The index LOS for patients discharged alive was 25% lower for pVAD versus ECMO (17.73 vs. 23.62; p=.0016). For patients that experienced in-hospital mortality during their index stay, pVAD had a 44% lower LOS compared to ECMO (7.08 vs. 12.66; p < .0001). Following index hospitalization, the average cost savings with additional inpatient care was 31% lower for pVAD patients ($62,188 vs. $90,087; p=NS). During the EOC, pVAD patients incurred 32% lower costs compared to ECMO patients ($117,849 vs. $172,420; <.0001). This study of Medicare FFS patients demonstrates that hospitals utilizing pVAD for appropriately selected AMICS patients have reduced mortality rates and reduced index LOS with lower index facility costs and lower post index 45-day costs. The study results offer hospitals and clinicians an opportunity to improve clinical outcomes and reduce total EOC costs in treating patients with AMI complicated by CS.

Increased Risk of Humeral Fracture With Open Versus Arthroscopic Tenodesis of the Long Head of the Biceps Brachii

PurposeTo determine the risk of postoperative humeral fracture following tenotomy, open tenodesis and arthroscopic tenodesis of the long head of the biceps brachii.MethodsA retrospective review of deidentified patient data from the Medicare Standard Analytic File using the PearlDiver software was conducted to identify procedures performed between 2005 and 2014. Groups were matched by age, gender, region, and medical comorbidities.ResultsWe evaluated 157,163 patients who had undergone arthroscopic or open tenodesis or tenotomy of the long head of the biceps brachii over a 10-year period (2005-2014), and we identified 2,196 postoperative humeral fractures (1.4%). Matched subgroup analysis consisting of 44,292 patients demonstrated a statistically significant increase in humeral fracture risk in open (280; 1.26%) compared to arthroscopic tenodesis (232; 1.04%) with a P value of 0.03 and an odds ratio of 1.21. The majority of fractures were sustained by patients 65-74 years of age.ConclusionIn this study, an increased risk of postoperative humeral fracture was associated with open tenodesis of the LHB.Level of EvidenceIII, Retrospective Comparative Trial.

Hospital network participation and outcomes following elective posterior lumbar fusions – are mergers effective?

Background contextDue to financial pressures associated with healthcare reforms, an increasing number of hospitals are now merging (or consolidating) into networks (or systems). However, it remains unclear how these mergers or network participations affect quality of care and/or costs. PurposeThe current study aims to evaluate the impact of hospital network participation on 90-day complications, charges, and costs following elective posterior lumbar fusions (PLFs). Study designRetrospective review of a 100% national sample of Medicare claims from 2010 to 2014 (SAF100). Study sampleAll Medicare-eligible patients undergoing elective 1-to-3 level PLFs for degenerative lumbar pathology from 2010 to 2014. Outcome measuresNinety-day complications, charges, and costs. MethodsThe 2010 to 2014 100% Medicare Standard Analytical Files (SAF100) was used to identify patients undergoing elective 1- to 3-level PLFs for degenerative lumbar pathology. The Dartmouth Atlas for Healthcare hospital-level data, which uses a combination of American Hospital Association and additional source data, was used to identify hospitals that were part of a network (or system) between 2010 and 2014. The study sample was divided into 2 cohorts (network hospitals and non-network hospitals) for analyses. Multivariate logistic regression models were used to compare differences in 90-day complications between network and non-network hospitals, while controlling for baseline demographics (age, gender, region, year of surgery, median household income, co-morbidity burden) and hospital-level characteristics (case volume, teaching status, urban/rural location, and hospital ownership). Generalized linear regression modeling was used to assess for differences in 90-day charges and costs. ResultsA total of 145,141 patients undergoing surgery in 2,186 hospitals were included in the study, out of which 107,919 (74.4%) underwent surgery in a network hospital (N=1,526). Network hospitals were more prevalent in the South or West regions of the United States. Patients in network hospitals had a median household income less than the 5th quintile. Network hospitals were also more likely to have a higher annual case volume of elective 1- to 3-level PLFs, greater number of beds, be located in an urban location, and have a voluntary/nonprofit or proprietary/profit ownership model. Multivariate analyses showed that even though patients undergoing surgery at network hospitals (vs non-network hospitals) had a slightly increased odds of 90-day cardiac complications (7.9% vs 7.4%, odds ratio [OR] 1.07 [95% confidence interval {CI} 1.02–1.12]; p=.010), thromboembolic complications (2.4% vs 2.2%, OR 1.12 [95% CI 1.01–1.20]; p=.025) and emergency department visits (16.4% vs 16.0%, OR 1.06 [95% CI 1.02–1.09]; p=.002), the differences would not be considered clinically significant. Despite a slight decrease in risk-adjusted 90-day reimbursements (−$272), the risk-adjusted 90-day charges were actually significantly higher (+$9,959; p<.001) at network hospitals. ConclusionsEven though hospitals that are part of a network do not appear to have significantly different complication rates following elective PLFs, they do have significantly higher risk-adjusted charges as compared to non-network hospitals. Further research is required to understand market-level changes induced by hospital mergers into networks.

PMD7 ONE-YEAR COMPLICATIONS AND REIMBURSEMENT AFTER INTRAMEDULLARY NAIL FIXATION FOR DISTAL FEMUR FRACTURES AMONG THE MEDICARE POPULATION

Fractures of the distal femur, and their treatment with an intramedullary nail, are becoming increasingly more common. This study evaluated the one-year incidence of complications and their associated costs (reimbursement) for patients treated with an intramedullary nail(IMN) for fixation of distal femur fracture(). This retrospective observational cohort analysis used patient-level data from the Medicare Standard Analytic File(2015Q4 to 2018Q1). Patients were 65 years of age or older with an insurance claim for DFF and an IMN procedure during the episode of care for which Medicare was the primary payor. One-year post-fixation incidence for infection, deep infection, delayed healing, and nonunion were quantified. One-year post index reimbursement related to these complications was calculated when any of these complications were listed as primary on any insurance claim. Subgroup analysis was performed for patients with evidence of a prior total knee arthroplasty(TKA) on the same side as the DFF. A total of 3,070 patients had an IMN procedure for a DFF diagnosis and were included in the analysis; among these patients 8.0%(n=247 ) had a prior TKA. The mean (standard deviation [SD]) age for the cohort was 80(8.8) years. Eighty-three percent of the cohort was female, and most patients had a closed fracture(96.4%). The one-year incidence of complications was 3.3% for delayed healing, 3.0% for nonunion, 1.5% for deep infection, and 1.4% for infection. The respective mean(SD) one-year reimbursement for these complications was as follows: delayed healing $1,794($3,454), nonunion $7,027($11,915), deep infection $6,496($9,793), and infection $18,949($21,858). With the exception of nonunion, all other evaluated complications were similar or lower 1-year reimbursement rates among those with evidence of a prior TKA. Complications following DFF are costly and in the case of nonunion having a prior TKA increases the financial burden. Improved surgical methods and technologies could reduce these complications thereby reducing healthcare system costs.

Does Increasing Patient Complexity Have an Effect on Medical Outcomes and Lengths-of-Stay after Total Knee Arthroplasty?

A greater number of medically complex patients with multiple comorbidities are now more readily considered for total knee arthroplasty (TKA). Therefore, the purpose of this study was to determine whether comorbidity burden, measured with the Elixhauser Comorbidity Index (ECI), correlated with 90-day medical complications and longer in-hospital lengths-of-stay (LOS) in TKA patients. The PearlDiver supercomputer was queried for all primary TKA patients in the Medicare Standard Analytic Files from 2005 to 2014 using International Classification of Disease, 9th edition codes. Patients were included based on ECI scores, ranging from 1 to 5. ECI 1 patients served as the control cohort, while ECI 2, 3, 4, and 5 patients were considered study cohorts. Each study cohort was matched based on age and gender to the control cohort, resulting in a total of 715,398 patients included for analysis (ECI 1, n = 144,072; ECI 2, n = 144,072; ECI 3, n = 144,072; ECI 4, n = 144,072; ECI 5, n = 139,110). Logistic regression analyses were performed to compare 90-day medical complications and Welch's t-tests were performed to compare LOS between the cohorts. Patients with higher ECI scores were more likely to develop medical complications and have longer LOS compared with matched patients in the control cohort. Compared with matched ECI 1 patients, patients with ECI scores of 2 (odds ratio [OR]: 1.19, 95% confidence interval [CI]: 1.14-1.24), 3 (OR: 1.27, 95% CI: 1.21-1.32), 4 (OR: 1.32, 95% CI: 1.27-1.38), and 5 (OR: 1.33, 95% CI: 1.27-1.39) were significantly more likely to develop 90-day medical complications. Additionally, the mean LOS of patients in the ECI 2 (2.59 ± 1.49 vs. 2.73 ± 1.52 days), ECI 3 (2.59 ± 1.49 vs. 2.88 ± 1.51 days; p < 0.001), ECI 4 (2.59 ± 1.49 vs. 3.01 ± 1.56 days; p < 0.001), and ECI 5 (2.61 ± 1.49 vs. 3.14 ± 1.61 days; p < 0.001) groups were significantly longer than the mean LOS in the control ECI 1 group. In an increasingly complex patient population, associations between comorbidities and outcomes after TKA procedures can guide providers on how to modify their pre- and postoperative care. These results demonstrate that higher ECI scores are associated with a greater likelihood of 90-day medical complications and longer in-hospital LOS.

Racial/Ethnic Disparities in Hospice Utilization Among Medicare Beneficiaries Dying from Pancreatic Cancer

BackgroundWe sought to define the incidence and characterize the timing of hospice utilization among racial/ethnic minority patients following pancreatectomy for pancreatic cancer. MethodsThe Medicare Standard Analytic Files from 2013 to 2017 were used to identify patients with pancreatic cancer who underwent a pancreatectomy. Logistic regression was utilized to identify the association between race and patterns of hospice utilization among deceased individuals. ResultsAmong the 14,495 individuals (median age 73; 52.3% female; 6.8% racial/ethnic minority) who underwent a pancreatectomy for pancreatic cancer, 47% (n = 6859) died by the end of the follow-period. Among deceased individuals, three-fourths of patients (n = 4978, 72.6%) used hospice leading up to the time of death. Racial/ethnic minority patients were less likely, however, to have used hospice services compared with white patients (racial/ethnic minorities n = 301, 67% vs. whites: n = 4677, 73%; p = 0.024). On multivariable analysis, after controlling for clinical factors, racial/ethnic minority patients remained 22% less likely than whites to initiate hospice services prior to death (OR 0.78, 95% CI 0.63–0.96). Despite overall lower use of hospice, racial/ethnic minority patients had comparable odds of late hospice utilization (i.e., within 3 days of death) versus white patients (OR 1.5, 95% CI 0.73–1.50). DiscussionWhile most patients undergoing pancreatectomy for pancreatic cancer utilized hospice services prior to death, racial/ethnic minorities were less likely to use hospice services than whites.

An Analysis of Current Treatment Trends in Platelet-Rich Plasma Therapy in the Medicare Database.

Background:The use of platelet-rich plasma (PRP) in the Medicare population is not well described.Purpose:To investigate the national use of PRP among Medicare beneficiaries, including the incidence and conditions for which it was used in both operative and nonoperative settings, and determine charges to Medicare.Study Design:Descriptive epidemiology study.Methods:The Medicare Standard Analytical Files within the PearlDiver database were queried for PRP injections by use of Current Procedural Terminology (CPT) code 0232T from 2010 to 2014. A search of every associated International Classification of Diseases, 9th Revision, code and CPT code on the day of the injection was performed, and codes were broadly categorized as shoulder, knee, elbow, hip, and foot/ankle. These categories were then subdivided into 2 groups based on whether the injection was performed at the time of surgery or for a nonoperative condition. The patient data were analyzed by demographics and geographic region. In further analysis, the charges sent to Medicare for PRP injections were stratified by year and musculoskeletal site.Results:A total of 3654 PRP injections were coded for and administered during the study period; 57% of recipients were men and 33% were 65 to 69 years of age. We found that 42% of all PRP injections were administered in the southern geographic region. PRP injections were most commonly associated with shoulder diagnoses, followed closely by the foot and ankle and by the knee. The majority of injections given for shoulder conditions were performed at the time of surgery, whereas the majority of knee conditions treated with PRP were associated with nonoperative treatments. Annual charges to Medicare for PRP injections increased 400%, from $500,000 in 2010 to more than $2 million in 2014.Conclusion:The use and breadth of PRP therapy have increased substantially in Medicare beneficiaries. Further research is required to obtain a consensus on treatment recommendations for PRP use in this population in addition to strategies to obtain insurance reimbursement.

Refining Risk-Adjustment of 90-Day Costs Following Surgical Fixation of Ankle Fractures: An Analysis of Medicare Beneficiaries

As the current healthcare model transitions from fee-for-service to value-based payments, identifying cost-drivers of 90-day payments following surgical procedures will be a key factor in risk-adjusting prospective bundled payments and ensuring success of these alternative payment models. The 5% Medicare Standard Analytical Files data set for 2005–2014 was used to identify patients undergoing open reduction and internal fixation (ORIF) for isolated unimalleolar, bimalleolar, and trimalleolar ankle fractures. All acute care and post–acute care payments starting from day 0 of surgery to day 90 postoperatively were used to calculate 90-day costs. Patients with missing data were excluded. Multivariate linear regression modeling was used to derive marginal cost impact of patient-level (age, sex, and comorbidities), procedure-level (fracture type, morphology, location of surgery, concurrent ankle arthroscopy, and syndesmotic fixation), and state-level factors on 90-day costs after surgery. A total of 6499 patients were included in the study. The risk-adjusted 90-day cost for a female patient, aged 65 to 69 years, undergoing outpatient ORIF for a closed unimalleolar ankle fracture in Michigan was $6949 ± $1060. Individuals aged <65 or ≥70 years had significantly higher costs. Procedure-level factors associated with significant marginal cost increases were inpatient surgery (+$5577), trimalleolar fracture (+$1082), and syndesmotic fixation (+$2822). The top 5 comorbidities with the largest marginal cost increases were chronic kidney disease (+$8897), malnutrition (+$7908), obesity (+$5362), cerebrovascular disease/stroke (+$4159), and anemia (+$3087). Higher costs were seen in Nevada (+$6371), Massachusetts (+$4497), Oklahoma (+$4002), New Jersey (+$3802), and Maryland (+$3043) compared with Michigan. With the use of a national administrative claims database, the study identifies numerous patient-level, procedure-level, and state-level factors that significantly contribute to the cost variation seen in 90-day payments after ORIF for ankle fracture. Risk adjustment of 90-day costs will become a necessity as bundled-payment models begin to take over the current fee-for-service model in patients with fractures.

Risk Factors for Infection Following Total Wrist Arthroplasty and Arthrodesis: An Analysis of 6641 Patients.

Background: Infection following wrist arthroplasty (WA) or wrist fusion (WF) is an uncommon but difficult complication often resulting in explantation and prolonged courses of antibiotics. The purposes of this study are to: (1) characterize the demographic trends of individuals undergoing WA and WF; (2) determine the incidence of postoperative infection; and (3) identify risk factors for postoperative infection. Methods: The PearlDiver database was used to query 100% Medicare Standard Analytic files from 2005 to 2014. Patients undergoing WA or radiocarpal WF were identified using Current Procedural Terminology (CPT) codes. Diagnosis for infection within 1 year of operative intervention was assessed by International Classification of Diseases, Ninth Revision codes or CPT codes related to infection. Multivariable logistic regression analyses were performed to evaluate the risk factors for postoperative infection. Results: Of the 6641 patients included, 1137 (17.1%) underwent arthroplasty and 5504 (82.9%) underwent arthrodesis. Within 1 year of the index procedure, 3.5% had a diagnosis of, or procedure for, postoperative infection (WA: n = 40 of 1137; WF: n = 192 of 5504). Risk factors for infection following WA include age >85, tobacco use, depression, diabetes mellitus, and chronic kidney disease. Risk factors following radiocarpal WF include male sex, age >85, body mass index <19 kg/m2, depression, diabetes mellitus, and chronic kidney disease. Posttraumatic origin of wrist arthritis was a risk factor for infection following both WA and WF. Conclusions: Infection following WA and WF is relatively uncommon in a nationally representative Medicare database cohort. Risk factors common to both WA and WF include age >85, depression, diabetes mellitus, chronic kidney disease, and posttraumatic arthritis.

Comparison Study of Patient Demographics, Risk Factors, and Lengths of Stay for Postoperative Ileus After Primary Total Knee Arthroplasty

BackgroundPostoperative ileus is a potential complication after orthopedic surgery, which has not been well studied after total knee arthroplasty (TKA). The aims of this study were to analyze rates of postoperative ileus; patient demographic profiles; in-hospital lengths of stay (LOS); and patient-related risk factors for postoperative ileus after primary TKA. MethodsA query was performed from January 1, 2005 to March 31, 2014 using the Medicare Standard Analytical Files. Patients who underwent primary TKA and developed postoperative ileus within 3 days after their index procedure were identified. Patients who did not develop ileus represented controls. Primary outcomes analyzed and compared included patient demographics, risk factors, and in-hospital LOS. A P value less than .05 was considered statistically significant. ResultsIleus patients were older, more likely to be male, and had higher Elixhauser-Comorbidity Index scores (8 vs 6; P < .0001) compared with controls. Male patients (odds ratio [OR], 2.12; P < .0001), patients with preoperative electrolyte/fluid imbalance (OR, 3.40; P < .001), patients older than 70 years (OR, 1.62-2.33; P < .015), and body mass indices greater than 30 kg/m2 (OR, 1.79-2.00; P < .001) were at the greatest risk of developing ileus. In addition, ileus patients had significantly longer in-hospital LOS (5.42 vs 3.22 days; P < .001). ConclusionThe study demonstrated differences in patient demographics, patient-related risk factors, and an increased in-hospital LOS for ileus patients after primary TKA. The study is important as it can allow orthopedists to properly identify and optimize patients with certain risk factors to potentially mitigate this adverse event from occurring.

CMS Hospital Compare System of Star Ratings and Surgical Outcomes Among Patients Undergoing Surgery for Cancer: Do the Ratings Matter?

The Centers for Medicare and Medicaid Services (CMS) Hospital Compare star rating system has been proposed as a means to assess hospital quality performance. The current study aimed to investigate outcomes and payments among patients undergoing surgery for colorectal, lung, esophageal, pancreatic, and liver cancer across hospital star rating groups. The Medicare Standard Analytic Files (SAF) from 2013 to 2015 were used to derive the analytic cohort. The association of star ratings to perioperative outcomes and expenditures was examined. Among 119,854 patients, the majority underwent surgery at a 3-star (n = 34,901, 29.1%) or 4-star (n = 30,492, 25.4%) hospital. Only 12.2% (n = 14,732) were treated at a 5-star hospital. Across all procedures examined, patients who underwent surgery at a 1-star hospital had greater odds of death within 90days than patients who had surgery at a 5-star hospital (colorectal, 1.41 [95% confidence interval {CI}, 1.25-1.60]; lung, 1.97 [95% CI 1.56-2.48]; esophagectomy, 1.83 [95% CI 0.81-4.16]; pancreatectomy, 1.70 [95% CI 1.20-2.41]; hepatectomy, 1.63 [95% CI 0.96-2.77]). A similar trend was noted for failure to rescue (FTR), with the greatest odds of FTR associated with 1-star hospitals. The median expenditure associated with an abdominal operation was $1661 more at a 1-star hospital than at a 5-star hospital (1-star: $17,399 vs 5-star: $15,738). A similar trend was noted for thoracic operations. The risk of FTR, 90-day mortality, and increased hospital expenditure were allhigher at a 1-star hospital. Further research is needed to investigate barriers to care at 5-star-rated hospitals and to target specific interventions to improve outcomes at 1-star hospitals.

Racial/ethnic disparities in hospice utilization among Medicare beneficiaries dying from pancreatic cancer.

41 Background: While racial/ethnic treatment disparities among patients with pancreatic cancer have been documented, variation in other aspects of care including hospice utilization have not. As such, we sought to define the incidence, as well as characterize the timing, of hospice utilization among racial/ethnic minority patients following pancreatectomy for pancreatic cancer. Methods: The Medicare Standard Analytic Files were used to identify white and racial/ethnic minorities (African Americans and Hispanics) who underwent pancreatectomy for pancreatic cancer. Trends and timing related to overall hospice utilization among racial/ethnic minority and white patients were analyzed. Results: Among the 6,530 individuals (median age: 73, IQR 69-78; 51.5% female; 6.6% % racial/ethnic minority) who underwent a pancreatectomy for pancreatic cancer, 64.6% (n = 4221) had died by the end of the follow-period. Among deceased individuals, three-fourths of patients (n = 3149, 74.6%) had used hospice leading up to the time of death. Among individuals who died, age and comorbidity burden were similar among racial/ethnic minority and white patients; racial/ethnic minority patients were less likely, however, to have used hospice services compared with white patients (racial/ethnic minorities: n = 188, 68.9%, whites: n = 2,961, 75.0%; p = 0.024). On multivariable analysis, after controlling for clinical factors, racial/ethnic minority patients remained 27% less likely than whites to initiate hospice services prior to death (OR 0.73, 95%CI 0.56-0.95, p = 0.021). Despite overall lower use of hospice, racial/ethnic minority patients had comparable odds of late hospice utilization (i.e. within 3 days of death) versus white patients (OR 0.75, 95% CI 0.49-1.14, p = 0.18). Conclusions: While most patients undergoing pancreatectomy for pancreatic cancer utilized hospice services prior to death, ethnic/racial minority were less likely to use hospice services than whites. Further research should seek to better understand possible barriers to hospice initiation among racial/ethnic minority patients with cancer.

Evaluation of costs and outcomes of physician-owned hospitals across common surgical procedures

IntroductionThe Affordable Care Act introduced restrictions on the creation of new physician-owned hospitals (POH). We sought to define whether POH status was associated with differences in care. MethodsPatients undergoing one of ten surgical procedures were identified using Medicare Standard Analytic Files. Patient and hospital-level characteristics and outcomes between POH and non-POH were compared. ResultsAmong 1,255,442 patients identified, 14,560 (1.2%) were treated at POH. A majority of POHs were in urban areas (n = 30, 90.9%) and none were in low socioeconomic status areas. Patients at POH were slightly younger (POH:72, IQR:68–77 vs. non-POH:73, IQR:69–79) and healthier (CCI; POH:2; IQR: 1–3 vs. non-POH: 3; IQR: 1–4). Patients at non-POH had higher odds of postoperative complications (OR:1.67, 95%CI:1.55–1.80) and slightly higher medical expenditures (POH:$11,347, IQR:$11,139-$11,936 vs. non-POH:$13,389, IQR:$11,381–$19,592). ConclusionsPOH were more likely to be located in socioeconomic advantaged areas, treat healthier patients and have lower associated expenditures.

INCORPORATING ROUTINE PASSIVE LEG RAISE ASSESSMENTS IMPROVES SEP-1 PERFORMANCE AND PATIENT CARE

SESSION TITLE: Wednesday Abstract Posters SESSION TYPE: Original Investigation Posters PRESENTED ON: 10/23/2019 09:45 AM - 10:45 AM PURPOSE: Reporting on hospital compliance to SEP-1 measures became publicly available at medicare.gov starting in July 2018. Performing a dynamic assessment of fluid responsiveness through passive leg raise (PLR) has been incorporated as a component of SEP-1 compliance. Passive Leg Raise (PLR) is an established method to predict fluid responsiveness or cardiac response to preload expansion. As new technologies allow continuous non-invasive bedside monitoring of cardiac output, the PLR has been used to guide fluid administration in sepsis improving outcomes. The Starling SV monitor (Cheetah Medical) measures a change in stroke volume (increased ≥ 10%) at the bedside to determine fluid responsiveness. We hypothesize there may be a relationship between Starling SV use, hospital SEP-1 compliance, and hospital length of stay. METHODS: The data set was comprised of 120 hospitals routinely using the Starling SV prior to 2017. Hospitals were grouped into 5 quintiles based on their ratio of Starling SV use / sepsis cases. SEP-1 compliance data was obtained from the Medicare dataset. Data from each quintile was compared against the average SEP-1 compliance rate for all hospitals within each quintile. A one-way ANOVA was performed to test the relationship between SEP-1 compliance and Starling SV utilization. The national average length of hospital stay (ALOS) was calculated by averaging the LOS from all hospitals that had a DRG 870, 871, 872 claim within the commercially available Definitive Database, analyzing Annual Medicare Data from the Medicare Standard Analytical Files (n=4163), and was compared to the available data on ALOS for hospitals using the Starling SV in the treatment of sepsis in >60% of patients (n=9) RESULTS: Hospital with the lowest Starling SV utilization (Group A, <20% use) also exhibited the lowest average compliance with SEP-1 (44%). Hospitals with the highest Starling SV utilization (Group E, >80% use) exhibited the highest average compliance with Sep-1 measures (67%) (p<0.0001). Across all quintiles, a significant difference was found between SEP-1 hospital compliance across the different Starling SV utilization groups (p=0.0003). Further, hospitals that routinely used the Starling SV in the treatment of sepsis exhibited a decreased ALOS (4.72 days, 2.1-4.8 days) compared to the national LOS (6.89 days, 1.7-23.9 days). CONCLUSIONS: Hospitals with a high usage of the Starling SV in the treatment of sepsis exhibited significantly increased compliance with SEP-1. Increased use of the PLR to check fluid responsiveness may be an important step on the pathway to increased SEP-1 performance and improved patient care. CLINICAL IMPLICATIONS: Performing a dynamic assessment of fluid responsiveness prior to administering treatment fluid may result in improved patient care. DISCLOSURES: Employee relationship with Cheetah Medical Please note: >$100000 Added 03/12/2019 by Shane Cooke, source=Web Response, value=Salary Employee relationship with Cheetah Medical Please note: >$100000 Added 03/12/2019 by Douglas Hansell, source=Web Response, value=Salary Employee relationship with Cheetah Medical, Inc. Please note: >$100000 Added 03/12/2019 by Chris Hutchison, source=Web Response, value=Salary Board relationship with Cheetah Medical, Inc. Please note: >$100000 Added 03/12/2019 by Chris Hutchison, source=Web Response, value=Salary Employee relationship with Cheetah Medical Please note: >$100000 Added 03/12/2019 by Robert Jaquith, source=Web Response, value=Salary Employee relationship with Cheetah Medical Please note: >$100000 Added 03/06/2019 by Jennifer Sahatjian, source=Web Response, value=Salary

Refining Risk-Adjustment of 90-day Costs Following Surgical Fixation of Ankle Fractures: Implications for a Bundled Payment Model

Category:TraumaIntroduction/Purpose:Current literature revolving around understanding cost-determinants of 90-day payments following ankle fractures is limited to single-institution studies. As the current healthcare model transitions from fee-for-service to value- based payments, risk-adjustment of 90-day payments will be a key factor driving success of these bundled payment models. The current study, utilizes a national Medicare database, to understand patient-level, procedure-level and state-level variation in 90-day payments following open reduction internal fixation (ORIF) of isolated ankle fractures.Methods:The 2005-2014 5% Medicare SAF (Standard Analytical Files) database was queries using Current Procedural Terminology codes to identify patients undergoing open reduction internal fixation (ORIF) for uni-malleolar (27766, 27769, 27792), bi-malleolar (27814) and tri-malleolar (27822, 27823) ankle fractures. Patients with polytrauma or those undergoing a concurrent surgical fixation of the upper extremity, hip, femur, knee or tibia were removed from the study to capture a relevant cohort of isolated ankle fracture patients. All payments starting from day 0 of surgery up to day 90 post-operatively were used to calculate 90-day costs. Patients with missing data were excluded. Multi-variate linear regression modeling was used to derive marginal cost-impact of patient-level (age, gender, co-morbidities), procedure-level (fracture type, morphology, location of surgery, concurrent ankle arthroscopy and syndesmotic fixation) and state-level factors on 90-day costs following surgery.Results:Following application of inclusion/exclusion criteria, a total of 6,499 patients were included in the study. The risk- adjusted 90-day price of a non-geriatric (age<65) female patient undergoing outpatient ORIF for a closed uni-malleolar ankle fracture was $8,915 ± $1,054. Individuals aged 65-69 vs. <65 had significantly lower costs (-$1,967). Procedure-level factors associated with significant marginal cost-increases were inpatient surgery (+$5,577), tri-malleolar fracture (+$1,082) and syndesmotic fixation (+$2,822). The top 5 co-morbidities with largest marginal cost-increases were chronic kidney disease (+$8,897), malnutrition (+$7,908), obesity (+$5,362), cerebrovascular disease/stroke (+$4,159) and anemia (+$3,087). Significant state-level variation in 90-day costs was seen with Nevada (+$6,371), Massachusetts (+$4,497), Oklahoma (+$4,002), New Jersey (+$3,802) and Maryland (+$3,043) having the highest marginal cost-increase and Idaho (-$6,025) having the lowest.Conclusion:Using a national administrative claims database, the study identifies numerous patient-level, procedure-level and state-level factors that significantly contribute to the cost-variation seen in 90-day payments following ORIF for ankle fracture. Risk adjustment of 90-day costs will become a necessity as bundled-payment models begin to take over the current fee-for-service model in fracture patients.

Delayed Healing in Metatarsal Fractures: Role of Low-Intensity Pulsed Ultrasound Treatment

The most common fracture in primary care is metatarsal fracture, but it is controversial whether to treat this fracture conservatively or surgically. We performed a cohort study to contrast metatarsal fractures that heal normally with fractures that show delayed healing. We analyzed 5% Medicare Standard Analytic Files, selecting all metatarsal fractures in 2011 to 2013, excluding patients with multiple fractures. Delayed healing was defined as treatment >14 days postfracture with either low-intensity pulsed ultrasound or surgery. Treatment for delayed healing was identified using the Current Procedural Terminology and International Classification of Diseases, Revision 9, Clinical Modification codes. Among 9482 metatarsal fractures, 256 (2.7%) showed delayed healing. Patients with delayed healing were younger (p < .0001); more likely to receive specialist referral (p < .001); more likely to have obesity (p = .005), psychosis (p = .003), chronic lung disease (p = .012), or iron deficiency anemia (p = .016); and more likely to receive surgery before ultrasound (p < .0001). Patients more likely to be treated with surgery than ultrasound included younger patients (p < .0001), obese patients (p = .02), and patients who first see a specialist (p < .05).

Cost Analysis of Medicare Patients with Varying Complexities Who Underwent Total Knee Arthroplasty

The effort to reduce overall healthcare costs may affect more complex patients, as their pre- and postoperative care can be substantially involved. Therefore, the purpose of this study was to use a large nationwide insurance database to compare (1) costs, (2) reimbursements, and (3) net losses of 90-day episodes of care (EOC) for total knee arthroplasty (TKA) patients according to Elixhauser's Comorbidity Index (ECI) scores. All TKAs performed between 2005 and 2014 in the Medicare Standard Analytic Files were extracted from the database and stratified based on ECI scores, ranging from 1 to 5. ECI 1 patients served as the control cohort, while ECI 2, 3, 4, and 5 patients were considered study cohorts. Each study cohort and control cohort were matched based on age and sex, resulting in a total of 715,398 patients included for analysis. Total EOC costs, reimbursements, and total net losses (defined as total EOC costs minus total EOC reimbursements) were compared between the cohorts. Overall, total EOC costs increased with ECI. For example, compared with the matched ECI 1 cohorts, the total EOC costs for ECI 5 patients ($56,589.19 vs. $51,747.54) were significantly greater (p < 0.01). Although reimbursements increased with increasing ECI, so did net losses. The net losses for ECI 5 patients were greater than that for ECI 1 patients ($42,309.39 vs. $40,007.82). The bundled payments for care improvement (BPCI) and comprehensive care for joint replacement (CJR) are alternative payment models that might de-incentivize treatment of more complex patients. Our study found that despite increasing reimbursements, overall costs, and therefore net losses, were greater for more complex patients with higher ECI scores.

93. Certificate-of-Need state laws and elective posterior lumbar fusions: a Medicare trends, costs and outcomes analysis

BACKGROUND CONTEXT Certificate-of-Need (CON) state laws have been a major mandate of health policies aimed at forming centralized health planning as a key-core of the US health care system. Though the federal CON mandate was repealed in 1986, several states still continue to use it as a means of limiting health care resource utilization while attempting to provide quality care in high-volume facilities. PURPOSE The current study aimed to utilized a national Medicare claims dataset to assess whether elective posterior lumbar fusions (PLFs) performed in CON vs No CON states had lower utilization rates, lower costs, and better quality of care. STUDY DESIGN/SETTING Retrospective review of administrative claims from the 100% Medicare Standard Analytical Files (SAF100). PATIENT SAMPLE The 2005-2014 100% Medicare Standard Analytical File (SAF100) was queried to identify patients undergoing elective 1-to-3 level PLF. OUTCOME MEASURES Trends in utilization during the study period (2005-2014), average 90-day costs/reimbursements, 90-day complications and 90-day readmissions. METHODS Patients were categorized into ‘CON’ and ‘No CON’ based on existing state laws where they received the procedure. For each group (CON vs No CON), differences in per-capita utilization, 90-day reimbursements and proportion of high-volume facilities (>150 procedures/year) vs low-volume facilities (≤50 procedures/year) have been reported. Multivariate analyses were used analyze 90-day complication/readmission rates between the two groups. RESULTS A total of 188,687 patients underwent an elective 1- to 3-level PLF in a CON state and 167,642 patients received the procedure in a No CON state from 2005 to 2014. The average per capita utilization of PLFs was lower in CON states as compared to No CON states (14.5 vs 15.4 per 10,000 population; p CONCLUSIONS While the presence of CON laws was associated with lower utilization of elective 1-to-3 level PLFs, and a greater presence of high-volume facilities, the mandate's effect on enhancing quality-of-care, by reduction of 90-day readmissions and 90-day complications is minimally significant. Future research, involving other surgical subspecialties, is needed before health policy makers may consider the removal or alteration of these laws. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs.

128. Posterior lumbar fusions (PLFs) at physician-owned hospitals: reconsidering the restrictions of the Affordable Care Act (ACA)

BACKGROUND CONTEXT Due to perceptions and concerns about unnecessary high resource utilization/costs and low quality of care provided in physician owned hospitals, the Affordable Care Act (ACA) of 2010 imposed restrictions on existing physician-owned hospitals from expanding. PURPOSE The current study aimed to assess differences in 90-day costs and outcomes between patients undergoing elective 1-to-3 level posterior lumbar fusions (PLFs) at physician-owned hospitals vs non-physician–owned hospitals. STUDY DESIGN/SETTING Retrospective review of 2007-2014 100% Medicare Standard Analytical Files (SAF100) database. PATIENT SAMPLE The 2007-2014 Medicare 100% Standard Analytical Files (SAF100) database was queried using International Classification of Diseases, 9th Edition (ICD-9) procedure code for patients undergoing elective 1- to 3-level posterior lumbar fusions (81.07, 81.08 and 81.62). The Medicare Hospital Compare database was used to identify provider codes for physician-owned hospitals and were cross-referenced to identify records of patients receiving elective posterior lumbar fusions at these hospitals from the SAF100 database. OUTCOME MEASURES Ninety-day complications, readmissions, charges and costs. METHODS Multivariate logistic and linear regression analyses were used to assess significant differences in 90-day complications, readmissions, charges and costs between the two groups. RESULTS A total of 6,679 (2.9%) patients received an elective posterior lumbar fusion at a physician-owned hospital (N=39; 2.2%) whereas 225,090 (97.1%) received surgery at non-physician–owned hospital (N=1,774; 97.8%). After controlling for age, gender, region, hospital factors (socioeconomic status area, urban vs rural location and volume) and ECI, undergoing surgery at physician-owned hospital was associated with lower odds of thromboembolic complications (OR 0.66 [95% CI 0.53-0.82]; p CONCLUSIONS Our results suggest that patients undergoing elective 1-to-3 level PLFs at physician-owned hospitals have fewer charges and costs over the 90-day episode of care, while having lower odds of experiencing urinary tract infections, thromboembolic complications and renal complications within 90-days of the surgery. The findings call into the need for revaluation/reconsideration of the ACAs restriction on the expansion of these physician-owned hospitals. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs.

80. Bariatric surgery prior to elective anterior cervical discectomy and fusion in obese patients is associated with a reduced risk of 90-day postoperative complications and readmissions

BACKGROUND CONTEXT Obese patients, particularly those with a body mass index (BMI) ≥35 kg/m2, pose a major challenge for a spine surgeon. Though obesity has previously been shown to be linked with adverse outcomes following elective spine surgical procedures, the impact of prior bariatric surgery on altering postoperative outcomes following elective anterior cervical discectomy and fusions (ACDFs) has not been explored. PURPOSE To evaluate whether bariatric surgery prior to ACDF in obese Medicare patients reduces the risk of experiencing adverse postoperative outcomes. STUDY DESIGN/SETTING Retrospective review of 100% Medicare Claims Database. PATIENT SAMPLE The PearlDiver program was used to query the 2007-2013 100% Medicare Standard Analytical Files (SAF100) for patients undergoing an elective ACDF using International Classification of Diseases 9th Edition (ICD-9) procedure codes 81.02. Records were filtered to include only those patients who underwent a 1- to 2-level surgery using ICD-9 procedure code 81.62. Only those patients receiving an ACDF due to degenerative spine pathologies were included in the study. Finally, patients who did not have active enrollment up to 2 years prior and 1 year after the surgery were excluded. A total of 121,382 patients undergoing an elective 1- to 2-level ACDF were included in the final cohort. OUTCOME MEASURES ICD-9 diagnosis codes were used to identify patients having at least moderate/Class II obesity (V85.35-V85.45, V85.4 and 278.01) within the last 2 years prior to an ACDF. Based on previously published literature, ICD-9 procedure codes for high gastric bypass (44.31), laparoscopic gastroenterostomy (44.38), other gastroenterostomy without gastrectomy (44.39), laparoscopic gastroplasty (44.68), laparoscopic gastric restrictive procedure (44.95-44.98), insertion of gastric bubble (44.93), other repair of stomach (44.69), open and other partial gastrectomy (43.89), isolation of intestinal segment (45.50-45.51), intestinal anastomosis (45.90, 45.91), partial gastrectomy with anastomosis to esophagus (43.5), duodenum (43.6) and jejunum (43.7) and other operations on stomach, unspecified (44.99) being coded on the same day as an obesity code were used to categorize patients receiving a bariatric surgical procedure.Only those patients receiving a bariatric surgery within 2 years prior to an ACDF were included. The 2-year mark was chosen as an arbitrary value. The study cohort was divided into two groups: (1) obese ACDF patients (BMI ≥35) receiving a bariatric surgery procedure within two years prior to an ACDF and (2) obese ACDF patients (BMI ≥35) without a known history of a bariatric surgical procedure within the last 2 years. Primary outcomes included 90-day complications (wound, pulmonary, cardiac, sepsis, renal, deep venous thrombosis, pain and dysphagia), 90-day readmissions, 90-day and 1-year revision cervical fusions. METHODS Multivariate analyses were used to assess the impact of prior bariatric surgery on 90-day outcomes following ACDF, after adjusting for age, gender, region and Elixhauser Co-morbidity Index. RESULTS A total of 6,128 patients with a BMI >35kg/m2 underwent ACDF, of which 411 ACDF patients underwent a bariatric surgical procedure within the 2 years prior to an ACDF. Multivariate analysis adjusting for age, gender, region and Elixhauser Co-morbidity Index (ECI), demonstrated that a history of prior bariatric surgery was associated with a significantly reduced rate of 90-day pulmonary complications (6.6% vs 12.7%, OR 0.53 [95% CI 0.34-0.78]; p=0.002), cardiac complications (15.3% vs 24.2%, OR 0.69 [95% CI 0.51-0.92]; p=0.012), sepsis (9.5% vs 15.1%, OR 0.69 [95% CI 0.48-0.96]; p=0.035), renal complications (2.9% vs 7.1%, OR 0.54 [95% CI 0.28-0.94]; p=0.044) and 90-day readmissions (3.9% vs 8.0%, OR 0.53 [95% CI 0.30-0.85]; p=0.015). Prior bariatric surgery had no statistically significant impact on the rates of wound complications (p=0.325), deep venous thrombosis (p=0.176), pain complications (p=0.606), dysphagia (p=0.171), 90-day revision (p=0.875) and 1-year revisions (p=0.313). CONCLUSIONS Surgery-induced weight loss prior to an ACDF in obese patients is associated with reduced 90-day complication and readmission rates. Orthopedic and bariatric surgeons should counsel obese patients on the benefits of bariatric surgery following ACDFs. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs.