Machine Learning Differentiates Extracorporeal Membrane Oxygenation Mortality Risk Profiles Among Trauma Patients.

BACKGROUND: Extracorporeal membrane oxygenation (ECMO) is resource intensive with high mortality. Identifying trauma patients most likely to derive a survival benefit remains elusive despite current ECMO guidelines. Our objective was to identify unique patient risk profiles using the largest database of trauma patients available. METHODS: ECMO patients ≥16 years were identified using Trauma Quality Improvement Program data (2010-2019). Machine learning K-median clustering (ML) utilized 101 variables including injury severity, demographics, comorbidities, and hospital stay information to generate unique patient risk profiles. Mortality and patient and center characteristics were evaluated across profiles. RESULTS: A total of 1037 patients were included with 33% overall mortality, mean age 32 years, and median ISS = 26. The ML identified 3 unique patient risk profile groups. Although mortality rates were equivalent across the 3 groups, groups were distinguished by (Group 1) young (median 25 years), severely injured (ISS = 34) patients with thoracic and head injuries (99%) via blunt mechanism (93%), and a high prevalence of ARDS (77%); (Group 2) relatively young (median 30 years) and moderately injured (ISS = 22) patients with exposure-related injuries (11%); and (Group 3) older (median 46 years) patients with a high proportion of comorbidities (69%) and extremity injuries (100%). There were no differences based on center ECMO volume, teaching status, or ACS-Level across all 3 groups. CONCLUSION: Machine learning compliments traditional analyses by identifying unique mortality risk profiles for trauma patients receiving ECMO. These details can further inform treatment guidelines, clinical decision making, and institutional criteria for ECMO usage.

Two-center analysis of cannabis on venous thromboembolism risk after traumatic injury: A matched analysis.

BACKGROUND: Conflicting evidence exists evaluating associations between cannabis (THC) and post-traumatic DVT. METHODS: Retrospective analysis (2014-2023) of patients ≥15yrs from two Level I trauma centers with robust VTE surveillance and prophylaxis protocols. Multivariable hierarchical regression assessed the association between THC and DVT risk. THC â€‹+ â€‹patients were direct matched to other drug use categories on VTE risk markers and hospital length of stay. RESULTS: Of 7365 patients, 3719 were drug-, 575 were THC â€‹+ â€‹only, 2583 were other drug+, and 488 were TCH+/other drug+. DVT rates by exposure group did not differ. TCH â€‹+ â€‹only patients had higher GCS scores, shorter hospital length of stay, and the lowest pelvic fracture and mortality rates. A total of 458 drug-, 453 other drug+, and 232 THC+/other drug â€‹+ â€‹patients were matched to 458, 453, and 232 THC â€‹+ â€‹only patients. There were no differences in DVT event rates in any paired sub-cohort set. Additionally, iteratively adjusted paired models did not show an association between THC and DVT. CONCLUSIONS: THC does not appear to be associated with increased DVT risk in patients with strict trauma chemoprophylaxis. Toxicology testing is useful for identifying substance abuse intervention opportunities, but not for DVT risk stratification in THC â€‹+ â€‹patients.

Lessons for the next pandemic: analysis of the timing and outcomes including post-discharge decannulation rates for tracheostomy in severe COVID-19 respiratory failure.

PURPOSE: COVID-19 patients with respiratory failure frequently require prolonged ventilatory support that would typically warrant early tracheostomy. There has been significant debate on timing, outcomes, and safety of these procedures. The purpose of this study was to determine the epidemiological, hospital, and post-discharge outcomes of this cohort, based on early (ET) versus late (LT) tracheostomy. METHODS: Retrospective review (March 2020-January 2021) in a 5-hospital system of ventilated patients who underwent tracheostomy. Demographics, hospital/ICU length of stay (LOS), procedural characteristics, APACHE II scores at ICU admission, stabilization markers, and discharge outcomes were analyzed. Long-term decannulation rates were obtained from long-term acute care facility (LTAC) data. RESULTS: A total of 97 patients underwent tracheostomy (mean 61 years, 62% male, 64% Hispanic). Despite ET being frequently performed during active COVID infection (85% vs. 64%), there were no differences in complication types or rates versus LT. APACHE II scores at ICU admission were comparable for both groups; however, > 50% of LT patients met PEEP stability at tracheostomy. ET was associated with significantly shorter ICU and hospital LOS, ventilator days, and higher decannulation rates. Of the cohort discharged to an LTAC, 59% were ultimately decannulated, 36% were discharged home, and 41% were discharged to a skilled nursing facility. CONCLUSIONS: We report the first comprehensive analysis of ET and LT that includes LTAC outcomes and stabilization markers in relation to the tracheostomy. ET was associated with improved clinical outcomes and a short LOS, specifically on days of pre-tracheostomy ventilation and in-hospital decannulation rates.

To the point: Utility of laparoscopy for operative management of stabbing abdominal trauma.

BACKGROUND: Algorithms for managing penetrating abdominal trauma are conflicting or vague regarding the role of laparoscopy. We hypothesized that laparoscopy is underutilized among hemodynamically stable patients with abdominal stab wounds. METHODS: Trauma Quality Improvement Program data (2016-2019) were used to identify stable (SBP ≥110 and GCS ≥13) patients ≥16yrs with stab wounds and an abdominal procedure within 24hr of admission. Patients with a non-abdominal AIS ≥3 or missing outcome information were excluded. Patients were analyzed based on index procedure approach: open, therapeutic laparoscopy (LAP), or LAP-conversion to open (LCO). Center, clinical characteristics and outcomes were compared according to surgical approach and abdominal AIS using non-parametric analysis. RESULTS: 5984 patients met inclusion criteria with 7 â€‹% and 8 â€‹% receiving therapeutic LAP and LCO, respectively. The conversion rate for patients initially treated with LAP was 54 â€‹%. Compared to conversion or open, therapeutic LAP patients had better outcomes including shorter ICU and hospital stays and less infection complications, but were younger and less injured. Assessing by abdominal AIS eliminated ISS differences, meanwhile LAP patients still had shorter hospital stays. At time of admission, 45 â€‹% of open patients met criteria for initial LAP opportunity as indicated by comparable clinical presentation as therapeutic laparoscopy patients. CONCLUSIONS: In hemodynamically stable patients, laparoscopy remains infrequently utilized despite its increasing inclusion in current guidelines. Additional opportunity exists for therapeutic laparoscopy in trauma, which appears to be a viable alternative to open surgery for select injuries from abdominal stab wounds. LEVEL OF EVIDENCE: Prognostic and Epidemiological; Level IV.

Modified Brain Injury Guidelines for preinjury anticoagulation in traumatic brain injury: An opportunity to reduce health care resource utilization.

INTRODUCTION: The Brain Injury Guidelines (BIG) stratify patients by traumatic brain injury (TBI) severity to provide management recommendations to reduce health care resource burden but mandates that patients on anticoagulation (AC) are allocated to the most severe tertile (BIG 3). We sought to analyze TBI patients on AC therapy using a modified BIG model to determine if this population can offer further opportunity for safe reductions in health care resource utilization. METHODS: Patients 55 years or older on AC with traumatic intracranial hemorrhage (ICH) from two centers were retrospectively stratified into BIG 1 to 3 risk groups using modified BIG criteria excluding AC as a criterion. Intracranial hemorrhage progression, neurosurgical intervention (NSI), death, and worsened discharge status were compared. RESULTS: A total of 221 patients were included, with 23%, 29%, and 48% classified as BIG 1, BIG 2, and BIG 3, respectively. The BIG 3 cohort had a higher rate of AC reversal agents administered (66%) compared with the BIG 1 (40%) and BIG 2 (54%) cohorts ( p < 0.01), as well as ICH progression discovered on repeat head computed tomography (56% vs. 38% vs. 26%, respectively; p < 0.001). No patients in the BIG 1 and 2 cohorts required NSI. No patients in BIG 1 and 3% of patients in BIG 2 died secondary to the ICH. In the BIG 3 cohort, 16% of patients required NSI and 26% died. Brain Injury Guidelines 3 patients had 15 times the odds of mortality compared with BIG 1 patients ( p < 0.01). CONCLUSION: The AC population had higher rates of ICH progression than the BIG literature, but this did not lead to more NSI or mortality in the lower tertiles of our modified BIG protocol. If the modified BIG used the original tertile management on our population, then NS consultation may have been reduced by up to 52%. These modified criteria may be a safe opportunity for further health care resource and cost savings in the TBI population. LEVEL OF EVIDENCE: Prognostic and Epidemiological; Level IV.

A Comparative Study of Laparoscopic versus Open Management of Index Small Bowel Obstruction in Children.

INTRODUCTION: There is limited literature on the optimal approach to treat adhesive small bowel obstruction (ASBO) in children. We sought to compare rates and outcomes of laparoscopic (LAP) and open (OPEN) surgery for pediatric ASBO. METHODS: A California statewide database was used to identify children (<18 years old) with an index ASBO from 2007 to 2020. The primary outcome was the type of operative management: LAP or OPEN. Secondary outcomes were hospital characteristics, patient demographics, and postoperative complications. We excluded patients treated non-operatively. RESULTS: Our study group had 545 patients. 381 (70%) underwent OPEN and 164 (30%) LAP during the index admission. Over the study period, there was increasing use of laparoscopic surgery, with higher use in older children (p < 0.001). LAP was associated with fewer overall complications (65.2% vs. 81.6%, p < 0.001), with a decreasing trend in complications over time (p < 0.001). The LAP group had significantly lower rates of bowel resection (4.9% vs. 17.1%, p < 0.001), length of stay (LOS) (17 vs. 23 days, p < 0.001), and TPN use (12.2% vs. 29.1%, p < 0.001). Mortality rates were equivalent. Although the LAP group had lower readmission rates (22.6% vs. 37.3%, p < 0.001), the length of time between discharge and readmission was similar (171 vs. 165 days, p = 0.190). DISCUSSION: The use of laparoscopic surgery for index ASBO increased over the study period. However, it was less commonly utilized in younger children. LAP had fewer overall complications as well as shorter LOS, decreased TPN use, and fewer readmissions. The benefits and risks of each approach must be weighed. LEVEL OF EVIDENCE: III.

Trends in Management of Index Adhesive Small Bowel Obstruction in Children.

INTRODUCTION: To examine practice patterns and surgical outcomes of nonoperative versus operative management (OPM) of children presenting with an index adhesive small bowel obstruction (ASBO). METHODS: A California statewide health discharge database was used to identify children (<18 y old) with an index ASBO from 2007 to 2020. The primary study outcome was evaluating initial management patterns (nonoperative versus OPM and early [≤3 d] versus late surgery [>3 d]) of ASBO. Secondary outcomes were hospital characteristics, patient demographics, and postoperative complications. RESULTS: Of the 2297 patients identified, 1948 (85%) underwent OPM for ASBO during the index admission. Of these, 14.7% underwent early surgery within 3 d. Teaching hospitals had higher operative intervention than nonteaching centers (87.1% versus 83.7%, P = 0.034). OPM was the highest in 0-5-year-olds compared to other ages (89% versus 82%, P < 0.001). In comparison to early surgery, late surgery was associated with longer length of stay (early 7[interquartile range 5-10], late 9[interquartile range 6-17], P < 0.001), increased infectious complications (16.4% versus 9.8%, P = 0.004), and greater use of total parenteral nutrition (28.0% versus 14.3%, P = 0.001); there was no difference in bowel resection (21% versus 18%, P = 0.102) or mortality (P = 0.423). CONCLUSIONS: Our pediatric study demonstrated a high rate of OPM for index ASBO, especially in newborns and toddlers. Although operative intervention, especially late surgery, was associated with increased length of stay, increased infectious complications, and increased total parenteral nutrition use, the rates of bowel resection and mortality did not differ by management strategy. These trends need to be further evaluated to optimize outcomes.

Firearm Assaults in Communities: The Impact of COVID-19 Lockdown.

BACKGROUND: Firearm violence has increased nationwide, with recent surges linked to the COVID-19 pandemic. We measured traumatic assault trends at our urban Level I trauma center and assessed rates of firearm violence over time and pre/post local COVID-19 lockdown based on levels of socioeconomic disadvantage. METHODS: We conducted a retrospective review (2016-2022) of assault patients 16 years and older. Demographics and hospital outcomes were assessed by assault mechanism (firearm, knife, blunt). Patient address was correlated to Area Deprivation Index (ADI), a measure of socioeconomic disadvantage. COVID-19 lockdown onset was defined as initial date of lockdown (3/19/2020). Trend and time-series analyses compared all assault mechanisms and firearm-specific assaults pre/post-lockdown. Poisson regression assessed firearm assault risk. RESULTS: Of the 1583 total assaults, firearm patients (n = 335) were younger (median 29 years), had longer hospital stays (median 2 days), and greater mortality (12%) than other mechanisms. The 2 years post-lockdown had significantly more firearm assaults (27% vs 15% pre-lockdown, P < .001) and time-series analysis found this abrupt and significant increase in firearm assaults occurred at lockdown onset (P = .01). Also post-lockdown, the rate of firearm assaults increased by 10% for every unit increase in socioeconomic deprivation (P < .01). There was no change in assault type by race/ethnicity. DISCUSSION: Firearm assaults increased dramatically immediately post-COVID lockdown at our center and have maintained higher rates through 2022. Greater ADI was associated with increasing firearm assaults and has magnified post-lockdown, demonstrating lower socioeconomic groups are disproportionately and increasingly affected by firearm violence.

Financial vulnerability of trauma centers: A national analysis.

OBJECTIVE: Trauma centers function as an essential safeguard in the United States health care system. However, there has been minimal study of their financial health or vulnerability. We sought to perform a nationwide analysis of trauma centers using detailed financial data and a recently developed Financial Vulnerability Score (FVS) metric. METHODS: The RAND Hospital Financial Database was used to evaluate all American College of Surgeons-verified trauma centers nationwide. The composite FVS was calculated for each center using six metrics. Financial Vulnerability Score tertiles were used to classify centers as high, medium, or low vulnerability, and hospital characteristics were analyzed and compared. Hospitals were also compared by US Census region and teaching versus nonteaching hospitals. RESULTS: A total of 311 American College of Surgeons-verified trauma centers were included in the analysis, with 100 (32%) Level I, 140 (45%) Level II, and 71 (23%) Level III. The largest share of the high FVS tier was consisted of Level III centers (62%), with the majority of Level I (40%) and Level II (42%) in the middle and low FVS tier, respectively. The most vulnerable centers had fewer beds, negative operating margins, and significantly less cash on hand. Lower FVS centers had greater asset/liability ratios, lower outpatient shares, and three times less uncompensated care. Nonteaching centers were statistically significantly more likely to have high vulnerability compared with teaching centers (46% vs. 29%). Statewide analysis showed high discrepancy among individual states. CONCLUSION: With approximately 25% of Levels I and II trauma centers at high risk for financial vulnerability, disparities in characteristics, including payer mix and outpatient status, should be targeted to reduce vulnerabilities and bolster the health care safety net. LEVEL OF EVIDENCE: Prognostic and Epidemiological; Level IV.

Pediatric penetrating thoracic trauma: Examining the impact of trauma center designation and penetrating trauma volume on outcomes.

INTRODUCTION: We analyzed the impact of treating center designation and case volume of penetrating trauma on outcomes after pediatric penetrating thoracic injuries (PTI). METHODS: PTI patients <18 years were identified from the National Trauma Data Bank (2013-2016). Centers were categorized by type (Pediatric or Adult) and designation status (Level I, Level II, and other). Performance was calculated as the difference between observed and expected mortality and standardized using the total penetrating trauma volume per center. Expected mortality was calculated using the Trauma Mortality Prediction Model. Pearson correlation and linear mixed-effects models evaluated the association between variables and performance. RESULTS: We identified 4,134 PTI patients treated at 596 trauma centers: 879 (21%) at Adult Level I, 608 (15%) at Adult Level II, 531 (13%) at Pediatric Level I, 320 (8%) at Pediatric Level II, and 1,796 (43%) at other centers. Primary injury mechanisms were firearm-related (58%) and cut/piercing (42%). Overall mortality was 16% and median predicted mortality was 3.6% (IQR: 1.5% - 11.2%). Among patients with thoracic firearm-related injuries, centers with lower penetrating case volume and total trauma care demonstrated significantly worse outcomes. Multivariable analysis revealed Adult Level I centers had superior outcomes compared with all other non-Level I centers. There was no difference in mortality between Pediatric and Adult Level I centers. DISCUSSION: Adult Level I trauma center designation and annual case volume of penetrating thoracic trauma are associated with improved mortality after pediatric firearm-related thoracic injuries. Further study is needed to identify factors in higher volume centers that improve outcomes. LEVEL OF EVIDENCE: Level III.

Validation of a miniaturized handheld arterial pressure monitor for guiding full and partial REBOA use during resuscitation.

PURPOSE: Resuscitative endovascular balloon occlusion of the aorta (REBOA) is a well-validated method for the control of noncompressible truncal hemorrhage. In lower resource or battlefield settings, the need for arterial line setup and monitoring is problematic and potentially prohibitive. We sought to evaluate the accuracy and precision of a miniaturized portable device (Centurion COMPASS®) versus standard arterial pressure monitoring using standard ER-REBOA and partial REBOA (pREBOA) as a high-fidelity and space-/time-conserving alternative. METHODS: A total of 40 swine underwent a four-phase validation/precision study (each phase using five ER-REBOAs and five pREBOAs). Phases I/II evaluated accuracy with full and pREBOA in uninjured animals. Phases III/IV duplicated the previous phases but in a severe hemorrhagic shock model. Carotid and femoral pressures were monitored with both intra-arterial pressure systems and the COMPASS® device. The vascular flow was measured by aortic flow probes. Correlation and Bland-Altman analysis were performed. RESULTS: There was a strong correlation in accuracy testing of proximal and distal COMPASS® devices compared to standard intra-arterial pressure monitoring (r = 0.94, 0.8; p < 0.005) as well as during precision testing (r = 0.98, 0.89 p < 0.005) in the uninjured phases. Similar accuracy and reliability were demonstrated in hemorrhagic shock, with a strong correlation for the proximal and distal COMPASS® devices (r = 0.98, 0.97; p < 0.005), as well as during precision testing (r = 0.99, 0.95; p < 0.005) in both full and pREBOA scenarios. Bland-Altman analysis showed extremely low bias between the COMPASS® and arterial line for both proximal (bias = 1.9) and distal (bias = 0.8) pressure measurements. CONCLUSION: The COMPASS® provides accurate and precise pressure measurements during standard and partial REBOA in both uninjured and shock conditions. This device may help extend and enhance capability in any low-resource/battlefield settings, or even eliminate the need for standard intra-arterial invasive pressure monitoring and external setup.

Financial Vulnerability of American College of Surgeons-Verified Trauma Centers: A Statewide Analysis.

BACKGROUND: Although trauma centers represent an integral part of healthcare in the US, characterization of their financial vulnerability has not been reported. We sought to characterize the financial health and vulnerability among California trauma centers and identify factors associated with high and low vulnerability. STUDY DESIGN: The RAND Hospital Data financial dataset was used to evaluate all American College of Surgeons (ACS)-verified trauma centers in California. Financial vulnerability of each center was calculated using 6 metrics to calculate a composite Financial Vulnerability Score (FVS). Tertiles of the FVS were generated to classify trauma centers as high, medium, or low financial vulnerability. Hospital characteristics were also analyzed and compared. RESULTS: Forty-seven ACS trauma centers were identified. Nine were Level I, 27 were Level II, and 8 were Level III. Level I centers encompassed the greatest proportion of the high FVS tier (44%), whereas Level II and III centers were the most likely to be in the middle and lower tiers, respectively (44%; 63%). Lower FVS centers had greater asset:liability ratios, operating margins, and days cash on hand compared with the 2 higher tiers, whereas high FVS centers showed a greater proportion of uncompensated care, outpatient share rates, outpatient surgeries, and longer days in net accounts. Lower FVS centers were more likely to be teaching hospitals and members of a larger corporate entity. CONCLUSION: Many ACS trauma centers are at moderate/high risk for financial vulnerability and disparate impacts of stressor events, and the FVS may represent a novel metric that could be used at the local or statewide level.

Why Trauma Care is Not Expensive: Detailed Analysis of Actual Costs and Cost Differences of Patient Cohorts.

BACKGROUND: Trauma patients are resource intensive, requiring a variety of medical and procedural interventions during hospitalization. These expenses often label trauma care as "high cost" based on gross hospital charges. We hypothesized that a financial metric built on actual costs and clinically relevant trauma patient cohorts would demonstrate a lower true cost of trauma care than the standardly reported gross hospital charges. METHODS: We examined all trauma patients (≥16 yr) treated in 2017 from a single institution and matched them to the institution's detailed financial accounting data. The organization's Financial Operations Division is uniquely able to allocate total operating costs across patient encounters to include medications, procedures, and salaries/fees from medical professionals and administrators. Patient subgroups were identified by Trauma Quality Improvement Program (TQIP) criteria for cost comparisons. RESULTS: Overall median cost per patient was $6,544 [IQR $4,975-14,532] for 2,548 patients. The median cost per patient increased with Injury Severity Score (ISS) ranging from $5,457(ISS ≤ 7) to $34,898(ISS ≥ 21), each accompanied by an average 548% increase in gross charges. Costs also varied widely from $13,498 [IQR $8,247-26,254] to $45,759 [IQR $22,186-113,993] across TQIP patient cohorts. Of the total cost, 91% was attributed to personnel alone. DISCUSSION: Measuring the true cost of trauma care is feasible. As hypothesized, the true cost of trauma care is lower than charges. True cost increased with injury severity with variable cost across subgroups. Non-physician staff and administration are the largest component of the cost of trauma care.

Reading the signs in penetrating cervical vascular injuries: Analysis of hard/soft signs and initial management from a nationwide vascular trauma database.

BACKGROUND: Algorithms for management of penetrating cervical vascular injuries (PCVIs) commonly call for immediate surgery with "hard signs" and imaging before intervention with "soft signs." We sought to analyze the association between initial examination and subsequent evaluation and management approaches. METHODS: Analysis of PCVIs from the American Association for the Surgery of Trauma Prospective Observational Vascular Injury Treatment vascular injury registry from 25 US trauma centers was performed. Patients were categorized by initial examination findings of hard signs or soft signs, and subsequent imaging and surgical exploration/repair rates were compared. RESULTS: Of 232 PCVI patients, 110 (47%) had hard signs (hemorrhage, expanding hematoma, or ischemia) and 122 (53%) had soft signs. With hard signs, 61 (56%) had immediate operative exploration and 44% underwent computed tomography (CT) imaging. After CT, 20 (18%) required open surgical repair, and 7% had endovascular intervention. Of note, 21 (19%) required no operative intervention. A total of 122 patients (53%) had soft signs on initial examination; 37 (30%) had immediate surgery, and 85 (70%) underwent CT imaging. After CT, 9% had endovascular repair, 7% had open surgery, and 65 (53%) were observed. No difference in mortality was observed for hard signs patients undergoing operative management versus observation alone (23% vs. 17%, p = 0.6). Those with hemorrhage as the primary hard signs most often required surgery (76%), but no interventions were required in 19% of hemorrhage, 20% of ischemia, and 24% of expanding hematoma. CONCLUSION: Although hard signs in PCVIs are associated with the need for operative intervention, initial CT imaging can facilitate endovascular options or nonoperative management in a significant subgroup. Hard signs should not be considered an absolute indication for immediate surgical exploration. LEVEL OF EVIDENCE: Prognostic/Epidemiological; Level IV.

The where, when, and why of surgical rib fixation: Utilization patterns, outcomes, and readmissions.

INTRODUCTION: There has been increasing use of surgical stabilization of rib fractures (SSRF), but most literature demonstrate outcomes of single centers during the index hospitalization. We sought to analyze statewide patterns and longer-term outcomes after SSRF. METHODS: Adult patients with >1 rib fracture in the 2016-2018 California Office of Statewide Health Planning Database were identified. SSRF and non-operatively managed (NO) patients were matched on clinical and demographic variables. Patterns and outcomes of SSRF were assessed with multivariate modeling. RESULTS: 599 SSRF patients were matched to 1191 NO patients. Readmission and readmission complication rates were similar between the groups. In a competing risks regression, admission to a high-volume SSRF center (SHR 4.6, CI95 4.0-5.4, p = 0.01) was the primary predictor of SSRF. 30-day mortality adjusted risk was lower for the SSRF vs. NO group (HR 0.47, CI 0.25-0.88, p = 0.02). DISCUSSION: Statewide utilization of SSRF varied widely and appears to be driven by center or surgeon characteristics rather than clinical factors. Efforts to expand access to SSRF based on clinical factors may be warranted.

Analysis of Readmission and Complications After Repair of Orbital Fractures Using California's Office of Statewide Health Planning and Development Database.

BACKGROUND: Orbital fractures represent one of the most common trauma-related facial fractures and may present with a variety of concomitant injuries. Many factors including age, associated diagnoses, and fracture complications are important in determining surgical candidacy. We used a statewide database to determine the prevalence of orbital fractures, rates of surgical repair during initial admission, and early patient outcomes. METHODS: A longitudinal analysis of patients with orbital fracture was performed using California's Office of Statewide Health Planning and Development patient discharge database for admissions occurring between January 2015 and December 2018.Patients were identified using International Classification of Diseases, Tenth Revision codes. The primary risk factor was surgical management of orbital fractures. The primary outcomes of interest were readmission requiring surgical management and complications of the fracture. Survival models were developed to evaluate the risk of a repair at readmission adjusting for relevant covariates. RESULTS: Of the 67,408 facial fractures included in our study, 8.7% (n = 5872) were diagnosed with orbital fractures. Among this population, 18.4% (n = 1082) underwent surgical repair during their initial admission. Patients were primarily male (71.1%; n = 4,173) and presented in a nonurgent fashion (93.7%; n = 5501). Less than half (42.8%) of patients with an urgent presentation and 16.8% of patients with a nonurgent presentation underwent repair. Centers of Medicaid & Medicare Services guidelines dictated presentation classification. Repair was associated with a significantly higher survival outcome. Orbital fractures were more frequently repaired in the setting of concomitant zygomatic, nasal, and LeFort I-III fractures. Increased risk in complications was observed in all concomitant fracture groups, and there existed a decreased risk of postsurgical complications in these same cohorts. CONCLUSIONS: Although most orbital fractures were managed nonoperatively, our analysis found that rates of repair for orbital floor, maxillary, and zygomatic fractures were greater than for other facial fractures. Concomitant fractures were associated with an increased hazard ratio for complications. Although low in prevalence overall, the most often observed postoperative complications in this population were diplopia, glaucoma, and blindness/low vision.

Using the Disruptive Score to Identify Publications That Changed Plastic Surgery Practice.

PURPOSE: The impact of academic publications is often characterized by the total number of future citations. However, this metric does not adequately characterize the true impact in terms of changing practices or paradigms. A new metric called the "disruption score" (DS) has been developed and validated in nonsurgical publications. This study aims to use the DS to identify the most disruptive publications in plastic surgery.The DS, a ratio of 2 numbers, varies between -1 and +1. Scores closer to -1 are developing papers that summarize the known literature while papers closer to +1 are disruptive-they result in a paradigm shift in the field of study. METHODS: A search was performed for all articles from 1954 to 2014 in the following journals: Plastic and Reconstructive Surgery; Aesthetic Surgery Journal; Journal of Plastic, Reconstructive, and Aesthetic Surgery; Annals of Plastic Surgery; Aesthetic Plastic Surgery; Clinics in Plastic Surgery; and Plastic Surgery. The disruptive score was calculated for each article.The top 100 papers ranked by DS were examined and any editorials/viewpoints, publications with less than 26 citations, or less than 3 references were excluded because of their subjective nature and smaller academic contribution. The remaining 64 publications were analyzed for topic, study type, and citation count. RESULTS: A total of 32,622 articles were found with a DS range from 0.385 to 0.923. The mean score of the top 64 articles was 0.539 with an average citation count of 195 and 9 references. Plastic and Reconstructive Surgery had the most disruptive papers with 50. There were no randomized controlled trials with a majority of the studies being technical descriptions or case series. CONCLUSIONS: There are many ways to measure academic success, but there are fewer ways to measure the impact of academic contributions. The DS is a novel measurement that can demonstrate when an article results in a paradigm shift as opposed to just total citation count. When applied to the plastic surgery literature, the DS demonstrates that technical innovation and creativity are the most academically impactful. Future evaluations of academic success should include the DS to measure the quality of academic contributions.

Surgical management and outcomes of adhesive small bowel obstruction: teaching versus non-teaching hospitals.

BACKGROUND: The relationship between surgical management of adhesive small bowel obstruction (ASBO) and hospital teaching status is not well known. We sought to elucidate the association between hospital teaching status and clinical metrics for ASBO. METHODS: Using the 2007-2017 California Office of Statewide Health Planning and Development database, we identified adult ASBO patients hospitalized for surgical intervention. Hospital teaching status was categorized as major teaching (MajT), minor teaching (MinT), and non-teaching (NT). Cox proportional hazards modeling was used to evaluate risk of death and other adverse outcomes. RESULTS: Of 25,047 admissions, 15.4% were at MajT, 32.0% at MinT, and 52.6% at NT; 2.9% died. Patients at MajT had longer overall hospital stays (HLOS) than those at MinT or NT (median days 9 vs. 8 vs. 8; p = 0.005), longer post-ASBO procedure HLOS (median days 7 vs. 6 vs. 6; p = 0.0001) and higher rates of small bowel resection (27.1% vs. 21.7% vs. 21.7%; p < 0.0001). Mean time to first surgery at MajT was 3.3 days compared with 2.6 days (p = 0.004) at MinT and NT. Compared with patients at NT, those at MajT were significantly less likely to die (HR 0.62, p < 0.0001), develop pneumonia (HR 0.57, p = 0.001), or experience adverse discharge disposition (HR 0.79, p < 0.0001). CONCLUSION: Mortality and morbidity of ASBO surgery were reduced at MajT; however, time to surgery, HLOS, and rate of small bowel resection were greater. These findings may guide improvements in the management of ASBO patients.

Fragmentation of care in the blunt abdominal trauma patient: Capturing our true outcomes and impact on care.

BACKGROUND: Trauma care is associated with unplanned readmissions, which may occur at facilities other than the index treatment facility. This "fragmentation of care" may be associated with adverse outcomes. We evaluated a statewide database that includes readmissions to analyze the incidence and impact of FC. METHODS: The California Office of Statewide Health Planning and Development patient discharge data set was evaluated for calendar years 2016 to 2018. Patients 15 years or older diagnosed with blunt abdominal solid organ injury during the index admission were identified. Readmissions were evaluated postdischarge at 1, 3, and 6 months. Patients readmitted within 6 months to a facility other than the index admission facility (fragmented care [FC]) were compared with those readmitted to their index admission facility (non-FC). Logistic regression modeling was used to evaluate risk of FC. RESULTS: Of the total 1,580 patients, there were 752 FC (47.6%) and 828 (52.4%) non-FC. Readmissions representing FC at months 1, 3, and 6 were 40.3%, 49.3%, and 53.4%, respectively. At index admission, the groups were demographically and clinically similar, with similar rates of abdominal operations and complications. Non-FC patients had a higher rate of abdominal reoperation at readmission (5.8% non-FC vs. 2.9% FC, p = 0.006). In an adjusted model, multiple readmissions (odds ratio [OR] 1.11, p = 0.014), readmission >30 days after index facility discharge (OR, 1.98; p < 0.001), and discharge to a nonmedical facility (OR, 2.46; p < 0.0001) were associated with increased odds of FC. Operative intervention at index admission was associated with lower odds of FC (OR, 0.77; p = 0.039). However, FC was not independently associated with demographic or insurance characteristics. CONCLUSION: The rate of FC among patients with blunt abdominal injury is high. The risk of FC is mitigated when patients are managed operatively during the index admission. Trauma systems should implement measures to ensure that these patients are followed postdischarge. LEVEL OF EVIDENCE: Prognostic and epidemiological, level III; Care management, level IV.

Isolated Intracranial Hemorrhage in Elderly Patients With Pre-Injury Anticoagulation: Is Full Trauma Team Activation Necessary?

BACKGROUND: Traumatic intracranial hemorrhage (ICH) is a highly morbid injury, particularly among elderly patients on preinjury anticoagulants (AC). Many trauma centers initiate full trauma team activation (FTTA) for these high-risk patients. We sought to determine if FTTA was superior compared with those who were evaluated as a trauma consultation (CON). METHODS: Patients aged ≥55 on preinjury AC who presented from January 2015 to December 2019 with blunt isolated head injury (non-head AIS ≤2) and confirmed ICH were identified. CON patients and FTTA patients were matched by age and head AIS. Cox proportional hazard model was used to assess patient and injury characteristics with mortality and survivor discharge disposition. REASULTS: There were 45 CON patients and 45 FTTA patients. Mean age was 80 years in both groups. Fall was the most common mechanism (98% CON vs. 92% FTTA). Glasgow Coma Score (GCS) was lower in FTTA (14 vs. 15, p<0.01). CON had a significantly longer time from arrival to CT scan (1.3 vs. 0.4 hrs, p<0.01). Hospital days were similar (CON: 3.9 vs. FTTA: 3.7 days). However, CON had increased ventilator use (p=0.03). Lower admission GCS was the only factor associated with increased risk of death. Among survivors, only head AIS increased the risk of discharge to a level of care higher than that of preinjury (p=0.01). CONCLUSION: There was no difference in mortality or adverse discharge disposition between FTTA and CON, although FTTA was associated with a more rapid evaluation and diagnosis. Any alteration in GCS was strongly associated with mortality and should prompt evaluation by FTTA.