American Black Authorship Has Decreased Across All Clinical Specialties Despite an Increasing Number of Black Physicians Between 1990 and 2020 in the USA

PURPOSE

Many diversity, equity, and inclusion initiatives assume that attainment of a racially diverse healthcare workforce will translate to increased diversity elsewhere in the healthcare system (e.g., leadership roles or academic authorship). We sought to investigate these trends over time by examining the evolution of physician demographics in the USA, in concert with demographic changes in US authorship in US medical journals from 1990 to 2020 across 25 specialties.

METHODS

We reviewed all articles indexed in PubMed, with a primary author affiliation located in the USA and limited to journals based in the USA, relative to the proportion of medical professionals in the CMS National Provider Registry. We employed a previously peer-reviewed/validated algorithm called "averaging-of-proportions" that probabilistically predicts racial identity from surname using the US Census to assess the relationship between diversity among medical professionals and diversity in medical journal authorship.

RESULTS

Data reveals a sharp disconnect between the demographic breakdown of physicians and authors. Despite an increase in the number of Black physicians (from 8.5% in 2005 to 9.1% in 2020), there has been a decrease in Black early-career authorship from 7.2% in 1990 to 5.8% in 2020. The percentage of Black early-career authors across all specialties in 2020 is lower than the average per specialty in 1990. Similar trends were noted for Black senior authorship, decreasing from 7.6% in 1990 to 6.2% in 2020, as well as a plateau in Hispanic authorship over the same time interval despite an increasing number of Hispanic physicians.

CONCLUSION

Modest advances in physician diversity have not translated to increased diversity in academic authorship. Increasing diversity requires initiatives focused beyond recruitment of underrepresented minorities to medical schools or residencies.

Insurance Not Socioeconomic Status is Associated With Access to Postacute Care After Injury: A Multicenter Cohort Study

INTRODUCTION

Access to postacute care services in rehabilitation or skilled nursing facilities is essential to return trauma patients to their preinjury functional level but is often hindered by systemic barriers. We sought to study the association between the type of insurance, socioeconomic status (SES) measures, and postacute care utilization after injury.

METHODS

Adult trauma patients with an Injury Severity Score (ISS) ≥9 admitted to one of three Level I trauma centers were contacted 6-12 mo after injury to gather long-term functional and patient-centered outcome measures. In addition to SES inquiry specifically focused on education and income levels, patients were asked to subjectively categorize their perceived SES (p-SES) as high, mid-high, mid-low, or low. Insurance and income data were retrieved from trauma registries. Multivariable regression models were built to determine the association between type of insurance, SES, and discharge disposition after adjusting for patient and injury characteristics and hospitalization events.

RESULTS

A total of 1373 patients were included, of which 44% were discharged to postacute care facilities. The median age (IQR) was 65 (46, 76) years, 56% of patients were male, 11% were on Medicaid, 68% had attained education higher than high school, 27% had low income, and 29% reported a low/mid-low p-SES. Medicaid patients were less likely to be discharged to postacute care compared to privately insured (OR [95% CI]: 0.41 [0.29-0.58]) and Medicare patients (OR [95% CI]: 0.29 [0.16-0.50]). The latter relationship was true across p-SES categories. P-SES, income and educational level were not associated with discharge destination.

CONCLUSIONS

Insurance status, specifically having Medicaid, can pose a barrier to access to postacute care services in the trauma patient population across patients of all SES. Initiatives and policies that aim at reducing these access disparities are warranted.

Use of artificial intelligence for nonlinear benchmarking of surgical care

BACKGROUND

Existent methodologies for benchmarking the quality of surgical care are linear and fail to capture the complex interactions of preoperative variables. We sought to leverage novel nonlinear artificial intelligence methodologies to benchmark emergency surgical care.

METHODS

Using a nonlinear but interpretable artificial intelligence methodology called optimal classification trees, first, the overall observed mortality rate at the index hospital's emergency surgery population (index cohort) was compared to the risk-adjusted expected mortality rate calculated by the optimal classification trees from the American College of Surgeons National Surgical Quality Improvement Program database (benchmark cohort). Second, the artificial intelligence optimal classification trees created different "nodes" of care representing specific patient phenotypes defined by the artificial intelligence optimal classification trees without human interference to optimize prediction. These nodes capture multiple iterative risk-adjusted comparisons, permitting the identification of specific areas of excellence and areas for improvement.

RESULTS

The index and benchmark cohorts included 1,600 and 637,086 patients, respectively. The observed and risk-adjusted expected mortality rates of the index cohort calculated by optimal classification trees were similar (8.06% [95% confidence interval: 6.8-9.5] vs 7.53%, respectively, P = .42). Two areas of excellence and 4 for improvement were identified. For example, the index cohort had lower-than-expected mortality when patients were older than 75 and in respiratory failure and septic shock preoperatively but higher-than-expected mortality when patients had respiratory failure preoperatively and were thrombocytopenic, with an international normalized ratio ≤1.7.

CONCLUSION

We used artificial intelligence methodology to benchmark the quality of emergency surgical care. Such nonlinear and interpretable methods promise a more comprehensive evaluation and a deeper dive into areas of excellence versus suboptimal care.

Artificial intelligence versus surgeon gestalt in predicting risk of emergency general surgery

BACKGROUND

Artificial intelligence (AI) risk prediction algorithms such as the smartphone-available Predictive OpTimal Trees in Emergency Surgery Risk (POTTER) for emergency general surgery (EGS) are superior to traditional risk calculators because they account for complex nonlinear interactions between variables, but how they compare to surgeons' gestalt remains unknown. Herein, we sought to: (1) compare POTTER to surgeons' surgical risk estimation and (2) assess how POTTER influences surgeons' risk estimation.

STUDY DESIGN

A total of 150 patients who underwent EGS at a large quaternary care center between May 2018 and May 2019 were prospectively followed up for 30-day postoperative outcomes (mortality, septic shock, ventilator dependence, bleeding requiring transfusion, pneumonia), and clinical cases were systematically created representing their initial presentation. POTTER's outcome predictions for each case were also recorded. Thirty acute care surgeons with diverse practice settings and levels of experience were then randomized into two groups: 15 surgeons (SURG) were asked to predict the outcomes without access to POTTER's predictions while the remaining 15 (SURG-POTTER) were asked to predict the same outcomes after interacting with POTTER. Comparing to actual patient outcomes, the area under the curve (AUC) methodology was used to assess the predictive performance of (1) POTTER versus SURG, and (2) SURG versus SURG-POTTER.

RESULTS

POTTER outperformed SURG in predicting all outcomes (mortality-AUC: 0.880 vs. 0.841; ventilator dependence-AUC: 0.928 vs. 0.833; bleeding-AUC: 0.832 vs. 0.735; pneumonia-AUC: 0.837 vs. 0.753) except septic shock (AUC: 0.816 vs. 0.820). SURG-POTTER outperformed SURG in predicting mortality (AUC: 0.870 vs. 0.841), bleeding (AUC: 0.811 vs. 0.735), pneumonia (AUC: 0.803 vs. 0.753) but not septic shock (AUC: 0.712 vs. 0.820) or ventilator dependence (AUC: 0.834 vs. 0.833).

CONCLUSION

The AI risk calculator POTTER outperformed surgeons' gestalt in predicting the postoperative mortality and outcomes of EGS patients, and when used, improved the individual surgeons' risk prediction. Artificial intelligence algorithms, such as POTTER, could prove useful as a bedside adjunct to surgeons when preoperatively counseling patients.

LEVEL OF EVIDENCE

Prognostic and Epidemiological; Level II.

Development of a Machine Learning-Based Prescriptive Tool to Address Racial Disparities in Access to Care After Penetrating Trauma

Importance

The use of artificial intelligence (AI) in clinical medicine risks perpetuating existing bias in care, such as disparities in access to postinjury rehabilitation services.

Objective

To leverage a novel, interpretable AI-based technology to uncover racial disparities in access to postinjury rehabilitation care and create an AI-based prescriptive tool to address these disparities.

Design, Setting, and Participants

This cohort study used data from the 2010-2016 American College of Surgeons Trauma Quality Improvement Program database for Black and White patients with a penetrating mechanism of injury. An interpretable AI methodology called optimal classification trees (OCTs) was applied in an 80:20 derivation/validation split to predict discharge disposition (home vs postacute care [PAC]). The interpretable nature of OCTs allowed for examination of the AI logic to identify racial disparities. A prescriptive mixed-integer optimization model using age, injury, and gender data was allowed to "fairness-flip" the recommended discharge destination for a subset of patients while minimizing the ratio of imbalance between Black and White patients. Three OCTs were developed to predict discharge disposition: the first 2 trees used unadjusted data (one without and one with the race variable), and the third tree used fairness-adjusted data.

Main Outcomes and Measures

Disparities and the discriminative performance (C statistic) were compared among fairness-adjusted and unadjusted OCTs.

Results

A total of 52 468 patients were included; the median (IQR) age was 29 (22-40) years, 46 189 patients (88.0%) were male, 31 470 (60.0%) were Black, and 20 998 (40.0%) were White. A total of 3800 Black patients (12.1%) were discharged to PAC, compared with 4504 White patients (21.5%; P < .001). Examining the AI logic uncovered significant disparities in PAC discharge destination access, with race playing the second most important role. The prescriptive fairness adjustment recommended flipping the discharge destination of 4.5% of the patients, with the performance of the adjusted model increasing from a C statistic of 0.79 to 0.87. After fairness adjustment, disparities disappeared, and a similar percentage of Black and White patients (15.8% vs 15.8%; P = .87) had a recommended discharge to PAC.

Conclusions and Relevance

In this study, we developed an accurate, machine learning-based, fairness-adjusted model that can identify barriers to discharge to postacute care. Instead of accidentally encoding bias, interpretable AI methodologies are powerful tools to diagnose and remedy system-related bias in care, such as disparities in access to postinjury rehabilitation care.

Underuse and Variability in Substance Use Screening Among Adult Trauma Patients in the U.S.: A Nationwide Retrospective Cohort Study

OBJECTIVE

To characterize the rates and variability in substance screening among adult trauma patients in the U.S.

SUMMARY BACKGROUND DATA

Emergency Department trauma visits provide a unique opportunity to identify patients with substance use disorders. Despite the existence of screening guidelines, underscreening and variability in screening practices remain.

METHODS

Retrospective cohort study including adult trauma patients (18- 64-year-old) from the ACS-TQIP 2017-18 database. Multivariable logistic regressions were performed to adjust for demographics, clinical, and facility factors, and marginal probabilities were calculated using these multivariable models. The primary outcomes were substance screening and positivity, which were defined relative to the observation-weighted grand mean (mean).

RESULTS

2,048,176 patients were contained in the TQIP dataset, 809,878 (39.5%) were screened for alcohol (20.8% positive), and 617,129 (30.1%) were screened for drugs (37.3% positive). After all exclusion criteria were applied, 765,897 patients were included in the analysis, 394,391 (52.9%) were screened for alcohol (22.1% tested positive), and 279,531 (36.5%) were screened for drugs (44.3% tested positive). Among the patients included in our study, significant variability in screening rates existed with respect to demo-graphic, trauma mechanism, injury severity, and facility factors. Furthermore, in several cases, patient subpopulations who were less likely to be screened were in fact more likely to screen positive or vice versa.

CONCLUSIONS

Effective substance-screening guidelines should be predicated on achieving universal screening. Current lapses in screening, along with the observed variability, likely affect different patient populations in disparate manners and lead to both under-detection as well as waste of valuable resources.

Trauma patients with limited English proficiency: Outcomes from two level one trauma centers

BACKGROUND

Outcomes for surgical patients with limited English proficiency (LEP) may be worse compared to patients with English proficiency. We sought to evaluate the association of LEP with outcomes for trauma patients.

METHODS

Admitted adult patients on trauma service at two Level One trauma centers from 2015 to 2019 were identified.

RESULTS

12,562 patients were included in total; 7.3% had LEP. On multivariable analyses, patients with LEP had lower odds of discharge to post-acute care versus home compared to patients with English proficiency (OR 0.69; 95% CI 0.58-0.83; p < 0.001) but had similar length of stay (Beta coefficient 1.16; 95% CI 0.00-2.32; p = 0.05), and 30-day readmission (OR 1.08; 95% CI 0.87-1.35; p = 0.46).

CONCLUSIONS

Trauma patients with LEP had comparable short-term outcomes to English proficient patients but were less likely to be discharged to post-acute care facilities. The role of structural barriers, family preferences, and other factors merit future investigation.

Contemporary Surgical Management of Pancreatic Necrosis

Importance

For decades, infected or symptomatic pancreatic necrosis was managed by open surgical necrosectomy, an approach that has now been largely supplanted by an array of techniques referred to as the step-up approach.

Observations

This review describes the evidence base behind the step-up approach, when to use the different techniques, and their technical basics. The most common treatment strategies are included: percutaneous drainage, video-assisted retroperitoneal debridement, sinus tract endoscopy, endoscopic transgastric necrosectomy, and surgical transgastric necrosectomy. Also included is the evidence base around management of common complications that can occur during step-up management, such as hemorrhage, intestinal fistula, and thrombosis, in addition to associated issues that can arise during step-up management, such as the need for cholecystectomy and disconnected pancreatic duct syndrome.

Conclusions and Relevance

The treatment strategies highlighted in this review are those most commonly used during step-up management, and this review is designed as a guide to the evidence base underlying these strategies, as surgeons tailor their therapeutic approach to individual patients.

Hospital Academic Status and the Volume-Outcome Association in Postoperative Patients Requiring Intensive Care: Results of a Nationwide Analysis of Intensive Care Units in the United States

Objective: To determine whether the outcomes of postoperative patients admitted directly to an intensive care unit (ICU) differ based on the academic status of the institution and the total operative volume of the unit. Methods: This was a retrospective analysis using the eICU Collaborative Research Database v2.0, a national database from participating ICUs in the United States. All patients admitted directly to the ICU from the operating room were included. Transfer patients and patients readmitted to the ICU were excluded. Patients were stratified based on admission to an ICU in an academic medical center (AMC) versus non-AMC, and to ICUs with different operative volume experience, after stratification in quartiles (high, medium-high, medium-low, and low volume). Primary outcomes were ICU and hospital mortality. Secondary outcomes included the need for continuous renal replacement therapy (CRRT) during ICU stay, ICU length of stay (LOS), and 30-day ventilator free days. Results: Our analysis included 22,180 unique patients; the majority of which (15,085[68%]) were admitted to ICUs in non-AMCs. Cardiac and vascular procedures were the most common types of procedures performed. Patients admitted to AMCs were more likely to be younger and less likely to be Hispanic or Asian. Multivariable logistic regression indicated no meaningful association between academic status and ICU mortality, hospital mortality, initiation of CRRT, duration of ICU LOS, or 30-day ventilator-free-days. Contrarily, medium-high operative volume units had higher ICU mortality (OR = 1.45, 95%CI = 1.10-1.91, p-value = 0.040), higher hospital mortality (OR = 1.33, 95%CI = 1.07-1.66, p-value = 0.033), longer ICU LOS (Coefficient = 0.23, 95%CI = 0.07-0.39, p-value = 0.038), and fewer 30-day ventilator-free-days (Coefficient = -0.30, 95%CI = -0.48 - -0.13, p-value = 0.015) compared to their high operative volume counterparts. Conclusions: This study found that a volume-outcome association in the management of postoperative patients requiring ICU level of care immediately after a surgical procedure may exist. The academic status of the institution did not affect the outcomes of these patients.

Preperitoneal Pelvic Packing is Associated With Increased Risk of Venous Thromboembolism

INTRODUCTION

Preperitoneal pelvic packing (PPP) is an important intervention for control of severe pelvic hemorrhage in blunt trauma patients. We hypothesized that PPP is associated with an increased incidence of deep vein thrombosis (DVT) and pulmonary embolism (PE).

METHODS

A retrospective cohort analysis of blunt trauma patients with severe pelvic fractures (AIS ≥4) using the 2015-2017 American College of Surgeons-Trauma Quality Improvement Program database was performed. Patients who underwent PPP within four hours of admission were matched to patients who did not using propensity score matching. Matching was performed based on demographics, comorbidities, injury- and resuscitation-related parameters, vital signs at presentation, and initiation and type of prophylactic anticoagulation. The rates of DVT and PE were compared between the matched groups.

RESULTS

Out of 5129 patients with severe pelvic fractures, 157 (3.1%) underwent PPP within four h of presentation and were matched with 157 who did not. No significant differences were detected between the two matched groups in any of the examined baseline variables. Similarly, mortality and end-organ failure rates were not different. However, PPP patients were significantly more likely to develop DVT (12.7% versus 5.1%, P = 0.028) and PE (5.7% versus 0.0%, P = 0.003).

CONCLUSIONS

PPP in severe pelvic fractures secondary to blunt trauma is associated with an increased risk of DVT and PE. A high index of suspicion and a low threshold for screening for these conditions should be maintained in patients who undergo PPP.

Role of Transfusion Volume and Transfusion Rate as Markers of Futility During Ultramassive Blood Transfusion in Trauma

BACKGROUND

Using a large national database, we evaluated the relationship between RBC transfusion volume, RBC transfusion rate, and in-hospital mortality to explore the presence of a futility threshold in trauma patients receiving ultramassive blood transfusion.

STUDY DESIGN

The ACS-TQIP 2013 to 2018 database was analyzed. Adult patients who received ultramassive blood transfusion (≥20 units of RBC/24 hours) were included. RBC transfusion volume and rate were captured at the only 2 time points available in TQIP (4 hours and 24 hours), or time of death, whichever came first.

RESULTS

Among 5,135 patients analyzed, in-hospital mortality rate was 62.1% (n = 3,190), and 4-hour and 24-hour mortality rates were 17.53% (n = 900) and 42.41% (n = 2,178), respectively. RBC transfusion volumes at 4 hours (area under the receiver operating characteristic curve [AUROC] 0.59 [95% CI 0.57 to 0.60]) and 24 hours (AUROC 0.59 [95% CI 0.57 to 0.60]) had low discriminatory ability for mortality and were inconclusive for futility. Mean RBC transfusion rates calculated within 4 hours (AUROC 0.65 [95% CI 0.63 to 0.66]) and 24 hours (AUROC 0.85 [95% CI 0.84 to 0.86]) had higher discriminatory ability than RBC transfusion volume. A futility threshold was not found for the mean RBC transfusion rate calculated within 4 hours. All patients with a final mean RBC transfusion rate of ≥7 U/h calculated within 24 hours of arrival experienced in-hospital death (n = 1,326); the observed maximum length of survival for these patients during the first 24 hours ranged from 24 hours for a rate of 7 U/h to 4.5 hours for rates ≥21 U/h.

CONCLUSION

RBC transfusion volume within 4 or 24 hours and mean RBC transfusion rate within 4 hours were not markers of futility. The observed maximum length of survival per mean RBC transfusion rate could inform resuscitation efforts in trauma patients receiving ongoing transfusion between 4 and 24 hours.

POTTER-ICU: An artificial intelligence smartphone-accessible tool to predict the need for intensive care after emergency surgery

BACKGROUND

Delays in admitting high-risk emergency surgery patients to the intensive care unit result in worse outcomes and increased health care costs. We aimed to use interpretable artificial intelligence technology to create a preoperative predictor for postoperative intensive care unit need in emergency surgery patients.

METHODS

A novel, interpretable artificial intelligence technology called optimal classification trees was leveraged in an 80:20 train:test split of adult emergency surgery patients in the 2007-2017 American College of Surgeons National Surgical Quality Improvement Program database. Demographics, comorbidities, and laboratory values were used to develop, train, and then validate optimal classification tree algorithms to predict the need for postoperative intensive care unit admission. The latter was defined as postoperative death or the development of 1 or more postoperative complications warranting critical care (eg, unplanned intubation, ventilator requirement ≥48 hours, cardiac arrest requiring cardiopulmonary resuscitation, and septic shock). An interactive and user-friendly application was created. C statistics were used to measure performance.

RESULTS

A total of 464,861 patients were included. The mean age was 55 years, 48% were male, and 11% developed severe postoperative complications warranting critical care. The Predictive OpTimal Trees in Emergency Surgery Risk Intensive Care Unit application was created as the user-friendly interface of the complex optimal classification tree algorithms. The number of questions (ie, tree depths) needed to predict intensive care unit admission ranged from 2 to 11. The Predictive OpTimal Trees in Emergency Surgery Risk Intensive Care Unit application had excellent discrimination for predicting the need for intensive care unit admission (C statistics: 0.89 train, 0.88 test).

CONCLUSION

We recommend the Predictive OpTimal Trees in Emergency Surgery Risk Intensive Care Unit application as an accurate, artificial intelligence-based tool for predicting severe complications warranting intensive care unit admission after emergency surgery. The Predictive OpTimal Trees in Emergency Surgery Risk Intensive Care Unit application can prove useful to triage patients to the intensive care unit and to potentially decrease failure to rescue in emergency surgery patients.

Adding to the story, did penetrating trauma really increase? changes in trauma patterns during the COVID-19 pandemic: A multi-institutional, multi-region investigation

BACKGROUND

Results from single-region studies suggest that stay at home orders (SAHOs) had unforeseen consequences on the volume and patterns of traumatic injury during the initial months of the Coronavirus disease 2019 (COVID-19). The aim of this study was to describe, using a multi-regional approach, the effects of COVID-19 SAHOs on trauma volume and patterns of traumatic injury in the US.

METHODS

A retrospective cohort study was performed at four verified Level I trauma centers spanning three geographical regions across the United States (US). The study period spanned from April 1, 2020 - July 31, 2020 including a month-matched 2019 cohort. Patients were categorized into pre-COVID-19 (PCOV19) and first COVID-19 surge (FCOV19S) cohorts. Patient demographic, injury, and outcome data were collected via Trauma Registry queries. Univariate and multivariate analyses were performed.

RESULTS

A total 5,616 patients presented to participating study centers during the PCOV19 (2,916) and FCOV19S (2,700) study periods.  Blunt injury volume decreased (p = 0.006) due to a significant reduction in the number of motor vehicle collisions (MVCs) (p = 0.003). Penetrating trauma experienced a significant increase, 8% (246/2916) in 2019 to 11% (285/2,700) in 2020 (p = 0.007), which was associated with study site (p = 0.002), not SAHOs. Finally, study site was significantly associated with changes in nearly all injury mechanisms, whereas SAHOs accounted for observed decreases in calculated weekly averages of blunt injuries (p < 0.02) and MVCs (p = 0.003).

CONCLUSION

Results of this study suggest that COVID-19 and initial SAHOs had variable consequences on patterns of traumatic injury, and that region-specific shifts in traumatic injury ensued during initial SAHOs. These results suggest that other factors, potentially socioeconomic or cultural, confound trauma volumes and types arising from SAHOs. Future analyses must consider how regional changes may be obscured with pooled cohorts, and focus on characterizing community-level changes to aid municipal preparation for future similar events.

Planning for the next Pandemic: Trauma Injuries Require Pre-COVID-19 Levels of High-Intensity Resources

As hospital systems plan for health care utilization surges and stress, understanding the necessary resources of a trauma system is essential for planning capacity. We aimed to describe trends in high-intensity resource utilization (operating room [OR] usage and intensive care unit [ICU] admissions) for trauma care during the initial months of the COVID-19 pandemic. Trauma registry data (2019 pre-COVID-19 and 2020 COVID-19) were collected retrospectively from 4 level I trauma centers. Direct emergency department (ED) disposition to the OR or ICU was used as a proxy for high-intensity resource utilization. No change in the incidence of direct ED to ICU or ED to OR utilization was observed (2019: 24%, 2020 23%; P = .62 and 2019: 11%, 2020 10%; P = .71, respectively). These results suggest the need for continued access to ICU space and OR theaters for traumatic injury during national health emergencies, even when levels of trauma appear to be decreasing.

Does An ERAS Protocol Reduce Postoperative Opiate Prescribing in Plastic Surgery?

Background

Enhanced recovery after surgery (ERAS) protocols are effective at reducing inpatient opiate use. There is a paucity of studies on the effects of an ERAS protocol on outpatient opiate prescriptions. The aim of this study was to determine whether an ERAS protocol for plastic and reconstructive surgery would reduce opiate use in the outpatient postoperative setting.

Methods

A statewide (Massachusetts, USA) controlled substance prescription monitoring database was retrospectively reviewed to assess the prescribing patterns of a single academic plastic surgeon performing common plastic surgical outpatient operations. The time period prior to implementation of the ERAS protocol was then compared with the time period following protocol implementation. An additional three months of post-implementation data were then compared with those of each of the previous time periods to investigate whether the results were sustained.

Results

A comparison of opiate prescriptions in pre-ERAS, immediate post-ERAS procedures, and follow-up ERAS implementation procedures revealed a statistically significant decrease in opiate prescriptions after ERAS protocol implementation. This decrease in the quantity of opiates prescribed was sustained over time.

Conclusions

ERAS protocols are effective at reducing outpatient opiate prescriptions after a variety of plastic surgery operations. Appropriate patient and physician education is paramount for success.

Increasing Frequency of Interpreting Services is Associated With Shorter Peri-operative Length of Stay

BACKGROUND

Patients with limited English proficiency have barriers to accessing care. Rather than a binary use or no use, this study uses granular data on frequency of interpreting services to determine if this frequency is associated with differences in peri-operative length of stay for patients with limited English proficiency.

MATERIALS AND METHODS

This is a cross sectional study on length of stay for peri-operative admissions of at least one night during 2018, for patients who used medical interpreting services in an academic medical center in Boston, Massachusetts. The participants are split into quartiles of ascending number of interpreting events per day. The exposure for the primary outcome is the frequency of interpreting events per day during peri-operative admission. The primary study outcome measurement is peri-operative length of stay in days.

RESULTS

There was a statistically significant decrease in length of stay for patients in the highest two quartiles of interpreting service frequency, compared to the lowest quartile: quartile 2 trended shorter by 1.4 d (95% CI -4.5 to 1.7, P = 0.37), quartile 3 was 4.2 d shorter (95% CI -7.6 to -0.7, P = 0.02), and quartile 4 was 4.6 d shorter (95% CI -8.1 to -1.1, P = 0.01).

CONCLUSIONS

More frequent interpreting services per day during peri-operative admission are associated with shorter length of stay in adjusted analysis. The findings merit further study in an intervention to increase use of interpreting services for surgical patients with limited English proficiency to study the impact of increased frequency of culturally competent care.

Lower Mortality with Cryoprecipitate During Massive Transfusion in Penetrating but Not Blunt Trauma

BACKGROUND

Balanced blood product transfusion improves the outcomes of trauma patients with exsanguinating hemorrhage, but it remains unclear whether administering cryoprecipitate improves mortality. We aimed to examine the impact of early cryoprecipitate transfusion on the outcomes of the trauma patients needing massive transfusion (MT).

METHODS

All MT patients 18 years or older in the 2017 Trauma Quality Improvement Program (TQIP) were retrospectively reviewed. MT was defined as the transfusion of ≥10 units of blood within 24 hours. Propensity score analysis (PSA) was used to 1:1 match then compare patients who received and those who did not receive cryoprecipitate in the first 4 hours after injury. Outcomes included in-hospital mortality, 1-day mortality, in-hospital complications and transfusion needs at 24 hours.

RESULTS

Of 1,004,440 trauma patients, 1,454 MT patients received cryoprecipitate and 2,920 did not. After PSA, 877 patients receiving cryoprecipitate were matched to 877 patients who did not. In-hospital mortality was lower among patients who received cryoprecipitate (49.4% v. 54.9%, P = 0.022), as was 1-day mortality. Sub-analyses showed that mortality was lower with cryoprecipitate in patients with penetrating (37.5% versus. 48%, adjusted P = 0.008), but not blunt trauma (58.5% versus. 59.8%, adjusted P = 1.000). In penetrating trauma, the cryoprecipitate group also had lower 1-day mortality (21.8% versus. 38.6%, P <0.001) and a higher rate of hemorrhage control surgeries performed within 24 hours (71.4% versus. 63.3%, P = 0.018).

CONCLUSIONS

Cryoprecipitate in MT is associated with improved survival in penetrating, but not blunt, trauma. Randomized trials are needed to better define the role of cryoprecipitate in MT.

Non-English Primary Language is Associated with Emergency Surgery for Diverticulitis

BACKGROUND

Language barriers can limit access to care for patients with a non-English primary language (NEPL). The objective of this study was to define the association between primary language and emergency versus elective surgery among diverticulitis patients.

MATERIALS AND METHODS

Retrospective cohort study of adult patients from the 2009-2014 New Jersey State Inpatient Database. Patients were included if they had primary language data and underwent a partial colon resection for diverticulitis. Primary language was dichotomized into NEPL versus English primary language (EPL). The primary outcome was surgical admission type - urgent/emergent (referred to as "emergency") versus elective. Descriptive and multivariable analyses were performed.

RESULTS

A total of 9,453 patients underwent surgery for diverticulitis, of which 592 (6.3%) had NEPL. Among NEPL patients, 300 (51%) had Spanish as primary language and 292 (49%) had another non-Spanish primary language. Patients with NEPL and EPL were similar in age (median age 58 versus 59 years; P = 0.54) and sex (52% versus 53% female; P = 0.45). Patients with NEPL were less likely to have commercial insurance (45% versus 59%; P <0.001). On multivariable analysis, compared to patients with EPL, NEPL was associated with increased odds of emergency surgery for diverticulitis (OR 1.35; 95% Confidence Interval 1.13-1.62; P = 0.001) CONCLUSION: Patients with NEPL have higher odds of emergency versus elective surgery for diverticulitis compared to patients with EPL. Further research is needed to examine differences in referral pathways, patient-provider communication, and health literacy that may hinder access to elective surgery in patients with diverticulitis.

Machine Learning and Artificial Intelligence for Surgical Decision Making

Background: The use of machine learning (ML) and artificial intelligence (AI) in medical research continues to grow as the amount and availability of clinical data expands. These techniques allow complex interpretation of data and capture non-linear relations not immediately apparent by classic statistical techniques. Methods: This review of the ML/AI literature provides a brief overview for practicing surgeons and clinicians of the current and future roles these methods will have within surgical infection research. Results: A conceptual overview of the techniques is provided along with concrete examples in the surgical infections literature. Further examples of ML/AI techniques in clinical decision support as well as therapy discovery with model-based deep reinforcement learning are illustrated. Conclusions: Artificial intelligence and ML are important and increasingly utilized techniques within the expanding body of surgical infection research. This article provides a minimal baseline literacy in ML/AI to be able to view such projects in an appropriately critical fashion.

Trauma outcome predictor: An artificial intelligence interactive smartphone tool to predict outcomes in trauma patients

BACKGROUND

Classic risk assessment tools often treat patients' risk factors as linear and additive. Clinical reality suggests that the presence of certain risk factors can alter the impact of other factors; in other words, risk modeling is not linear. We aimed to use artificial intelligence (AI) technology to design and validate a nonlinear risk calculator for trauma patients.

METHODS

A novel, interpretable AI technology called Optimal Classification Trees (OCTs) was used in an 80:20 derivation/validation split of the 2010 to 2016 American College of Surgeons Trauma Quality Improvement Program database. Demographics, emergency department vital signs, comorbidities, and injury characteristics (e.g., severity, mechanism) of all blunt and penetrating trauma patients 18 years or older were used to develop, train then validate OCT algorithms to predict in-hospital mortality and complications (e.g., acute kidney injury, acute respiratory distress syndrome, deep vein thrombosis, pulmonary embolism, sepsis). A smartphone application was created as the algorithm's interactive and user-friendly interface. Performance was measured using the c-statistic methodology.

RESULTS

A total of 934,053 patients were included (747,249 derivation; 186,804 validation). The median age was 51 years, 37% were women, 90.5% had blunt trauma, and the median Injury Severity Score was 11. Comprehensive OCT algorithms were developed for blunt and penetrating trauma, and the interactive smartphone application, Trauma Outcome Predictor (TOP) was created, where the answer to one question unfolds the subsequent one. Trauma Outcome Predictor accurately predicted mortality in penetrating injury (c-statistics: 0.95 derivation, 0.94 validation) and blunt injury (c-statistics: 0.89 derivation, 0.88 validation). The validation c-statistics for predicting complications ranged between 0.69 and 0.84.

CONCLUSION

We suggest TOP as an AI-based, interpretable, accurate, and nonlinear risk calculator for predicting outcome in trauma patients. Trauma Outcome Predictor can prove useful for bedside counseling of critically injured trauma patients and their families, and for benchmarking the quality of trauma care.

Multisystem outcomes and predictors of mortality in critically ill patients with COVID-19: Demographics and disease acuity matter more than comorbidities or treatment modalities

BACKGROUND

We sought to describe characteristics, multisystem outcomes, and predictors of mortality of the critically ill COVID-19 patients in the largest hospital in Massachusetts.

METHODS

This is a prospective cohort study. All patients admitted to the intensive care unit (ICU) with reverse-transcriptase-polymerase chain reaction-confirmed severe acute respiratory syndrome coronavirus 2 infection between March 14, 2020, and April 28, 2020, were included; hospital and multisystem outcomes were evaluated. Data were collected from electronic records. Acute respiratory distress syndrome (ARDS) was defined as PaO2/FiO2 ratio of ≤300 during admission and bilateral radiographic pulmonary opacities. Multivariable logistic regression analyses adjusting for available confounders were performed to identify predictors of mortality.

RESULTS

A total of 235 patients were included. The median (interquartile range [IQR]) Sequential Organ Failure Assessment score was 5 (3-8), and the median (IQR) PaO2/FiO2 was 208 (146-300) with 86.4% of patients meeting criteria for ARDS. The median (IQR) follow-up was 92 (86-99) days, and the median ICU length of stay was 16 (8-25) days; 62.1% of patients were proned, 49.8% required neuromuscular blockade, and 3.4% required extracorporeal membrane oxygenation. The most common complications were shock (88.9%), acute kidney injury (AKI) (69.8%), secondary bacterial pneumonia (70.6%), and pressure ulcers (51.1%). As of July 8, 2020, 175 patients (74.5%) were discharged alive (61.7% to skilled nursing or rehabilitation facility), 58 (24.7%) died in the hospital, and only 2 patients were still hospitalized, but out of the ICU. Age (odds ratio [OR], 1.08; 95% confidence interval [CI], 1.04-1.12), higher median Sequential Organ Failure Assessment score at ICU admission (OR, 1.24; 95% CI, 1.06-1.43), elevated creatine kinase of ≥1,000 U/L at hospital admission (OR, 6.64; 95% CI, 1.51-29.17), and severe ARDS (OR, 5.24; 95% CI, 1.18-23.29) independently predicted hospital mortality.Comorbidities, steroids, and hydroxychloroquine treatment did not predict mortality.

CONCLUSION

We present here the outcomes of critically ill patients with COVID-19. Age, acuity of disease, and severe ARDS predicted mortality rather than comorbidities.

LEVEL OF EVIDENCE

Prognostic, level III.

Differences in outcomes after emergency general surgery between Hispanic subgroups in the New Jersey State Inpatient Database (2009-2014): The Hispanic population is not monolithic

BACKGROUND

Our aim was to examine differences in clinical outcomes between Hispanic subgroups who underwent emergency general surgery (EGS).

METHODS

Retrospective cohort study of the HCUP State Inpatient Database from New Jersey (2009-2014), including Hispanic and non-Hispanic White (NHW) adult patients who underwent EGS. Multivariable analyses were performed on outcomes including 7-day readmission and length of stay (LOS).

RESULTS

125,874 patients underwent EGS operations. 22,971 were Hispanic (15,488 with subgroup defined: 7,331 - Central/South American; 4,254 - Puerto Rican; 3,170 - Mexican; 733 - Cuban). On multivariable analysis, patients in the Central/South American subgroup were more likely to be readmitted compared to the Mexican subgroup (OR 2.02; p < 0.001, respectively). Puerto Rican and Central/South American subgroups had significantly shorter LOS than Mexican patients (Puerto Rico -0.58 days; p < 0.001; Central/South American -0.30 days; p = 0.016).

CONCLUSIONS

There are significant differences in EGS outcomes between Hispanic subgroups. These differences could be missed when data are aggregated at Hispanic ethnicity.

Artificial Intelligence, Machine Learning, and Surgical Science: Reality Versus Hype

Artificial intelligence (AI) has made increasing inroads in clinical medicine. In surgery, machine learning-based algorithms are being studied for use as decision aids in risk prediction and even for intraoperative applications, including image recognition and video analysis. While AI has great promise in surgery, these algorithms come with a series of potential pitfalls that cannot be ignored as hospital systems and surgeons consider implementing these technologies. The aim of this review is to discuss the progress, promise, and pitfalls of AI in surgery.

Validation of the Artificial Intelligence-Based Predictive Optimal Trees in Emergency Surgery Risk (POTTER) Calculator in Emergency General Surgery and Emergency Laparotomy Patients

BACKGROUND

The Predictive Optimal Trees in Emergency Surgery Risk (POTTER) tool is an artificial intelligence-based calculator for the prediction of 30-day outcomes in patients undergoing emergency operations. In this study, we sought to assess the performance of POTTER in the emergency general surgery (EGS) population in particular.

METHODS

All patients who underwent EGS in the 2017 American College of Surgeons NSQIP database were included. The performance of POTTER in predicting 30-day postoperative mortality, morbidity, and 18 specific complications was assessed using the c-statistic metric. As a subgroup analysis, the performance of POTTER in predicting the outcomes of patients undergoing emergency laparotomy was assessed.

RESULTS

A total of 59,955 patients were included. Median age was 50 years and 51.3% were women. POTTER predicted mortality (c-statistic = 0.93) and morbidity (c-statistic = 0.83) extremely well. Among individual complications, POTTER had the highest performance in predicting septic shock (c-statistic = 0.93), respiratory failure requiring mechanical ventilation for 48 hours or longer (c-statistic = 0.92), and acute renal failure (c-statistic = 0.92). Among patients undergoing emergency laparotomy, the c-statistic performances of POTTER in predicting mortality and morbidity were 0.86 and 0.77, respectively.

CONCLUSIONS

POTTER is an interpretable, accurate, and user-friendly predictor of 30-day outcomes in patients undergoing EGS. POTTER could prove useful for bedside counseling of patients and their families and for benchmarking of EGS care.

Risk of gallstone-related complications in necrotizing pancreatitis patients treated with a step-up approach: The experience of two tertiary care centers

BACKGROUND

A minimally invasive step-up approach to necrotizing biliary pancreatitis often requires multiple interventions, delaying cholecystectomy. The risk of gallstone-related complications during this time interval is unknown, as is the feasibility and safety of cholecystectomy after minimally invasive step-up treatment. In this paper, we analyzed both.

METHODS

Necrotizing pancreatitis patients treated with a minimally invasive step-up approach who underwent interval cholecystectomy at 2 tertiary care centers between 2014 and 2019 were included. Gallstone-related complications prior to cholecystectomy were examined, as were surgical approaches to cholecystectomy and complications. Necrotizing pancreatitis patients treated without mechanical intervention were also examined.

RESULTS

Seven of 31 patients developed gallstone-related complications between minimally invasive step-up treatment initiation and cholecystectomy. One patient developed biliary colic. Six patients developed acute cholecystitis. Two of these patients also developed choledocholithiasis, and 1 developed cholangitis, all requiring endoscopic retrograde cholangiopancreatography. Cholecystectomy was performed laparoscopically in 27 of 31 patients. One patient required open conversion, and 3 patients underwent planned cholecystectomy during another open operation. Four patients developed postoperative complications. Two of 14 necrotizing pancreatitis patients treated without mechanical intervention developed recurrent pancreatitis while awaiting cholecystectomy.

CONCLUSION

Over 20% of necrotizing pancreatitis patients treated by a minimally invasive step-up approach developed gallstone-related complications while awaiting cholecystectomy. Laparoscopic cholecystectomy is feasible and safe in the great majority of necrotizing pancreatitis patients treated by a minimally invasive step-up approach.

Systolic Blood Pressure <110 mm Hg as a Threshold of Hypotension in Patients with Isolated Traumatic Brain Injuries

Hypotension is a known risk factor for poor neurologic outcomes after traumatic brain injury (TBI). Current guidelines suggest that higher systolic blood pressure (SBP) thresholds likely confer a mortality benefit. However, there is no consensus on the ideal perfusion pressure among different age groups (i.e., recommended SBP ≥100 mm Hg for patients age 50-69 years; ≥ 110 mm Hg for all other adults). We hypothesize that admission SBP ≥110 mm Hg will be associated with improved outcomes regardless of age group. A retrospective database review of the 2010-2016 Trauma Quality Improvement Program database was performed for adults (≥ 18 years) with isolated moderate-to-severe TBIs (head Abbreviated Injury Scale [AIS] ≥3; all other AIS <3). Sub-analyses were performed after dividing patients by SBP and age; comparison groups were matched with propensity score matching. Primary outcomes were early (6 h, 12 h, and 1 day) and overall in-hospital mortality. Overall, 154,725 patients met the inclusion criteria (mean age 62.8 ± 19.8 years, 89,431 [57.8%] males, Injury Severity Score13.9 ± 6.8). Multi-variate logistic regression showed that the risk of in-hospital mortality decreased with increasing SBP, plateauing at 110 mm Hg. Among patients of all ages, SBP ≥110 mm Hg was associated with improved mortality (SBP 110-129 vs. 90-109 mm Hg: 12 h 0.4% vs. 0.8%, p = 0.001; 1 day 0.8% vs. 1.4%, p = 0.004; overall 3.2% vs. 4.9%, p < 0.001). Among patients age 50-69 years, SBP ≥110 mm Hg was associated with improved mortality (SBP 110-119 vs. 100-109 mm Hg: 12 h 0.3% vs. 0.9%, p = 0.018; 1 day 0.5% vs. 1.5%, p = 0.007; overall 2.7% vs. 4.3%, p = 0.015). In conclusion, SBP ≥110 mm Hg is associated with lower in-hospital mortality in adult patients with isolated TBIs, including patients age 50-69 years. SBP <110 mm Hg should be used to define hypotension in adult patients of all ages.

Gastrointestinal Complications in Critically Ill Patients With and Without COVID-19

This cohort study compares the incidence of gastrointestinal complications (transaminitis, ileus, Ogilvie syndrome, mesenteric ischemia) among critically ill patients with coronavirus disease 2019 (COVID-19)– vs non–COVID-19–induced acute respiratory distress syndrome (ARDS) at a single US academic hospital between March and May 2020.

Protecting Our Own: Equity for Employees as Hospitals Battle COVID-19

As medical providers garner praise during the coronavirus disease 2019 (COVID-19) pandemic, “nonclinical” health care workers remain largely overlooked. Although these essential workers face similar, if not greater, risks of contracting severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) as others on the frontlines, many hospitals have fallen short in providing this vulnerable population with needed protections. Instead, hospitals should implement policies that guarantee all staff receive the information, equipment, and support necessary for battling the current crisis. This is critical not only for promoting the safety of these workers and their families, but also for ensuring the well-being of the community as a whole.

Leveraging interpretable machine learning algorithms to predict postoperative patient outcomes on mobile devices

Setting patient and family expectations for postoperative outcomes is an important aspect of care, a cornerstone of which is accurate, personalized, and explainable risk estimation. Modern machine learning offers a plethora of models that can effectively capture the complex, nonlinear contributions of preoperative risk factors to the surgical patient's overall risk. However, most of these models produce risk estimates that are not interpretable, which compromises trust in these systems, renders them unaccountable, and limits their widespread adoption. Recent developments in machine learning have been successful at creating risk calculators that address this gap, producing explainable risk estimates without compromising accuracy. In this work, we describe how the state of the art in postoperative risk estimation addresses the shortcomings of older methods, and how they have been applied in a clinical setting. We conclude with a discussion of the potential of machine learning models to be systematically integrated in health care more broadly and future prospects beyond passive risk prediction.

A low dose heparinized saline protocol is associated with improved duration of arterial line patency in critically ill COVID-19 patients

Purpose

Critically ill patients with Coronavirus Disease 2019 (COVID-19) have high rates of line thrombosis. Our objective was to examine the safety and efficacy of a low dose heparinized saline (LDHS) arterial line (a-line) patency protocol in this population.

Materials and Methods

In this observational cohort study, patients ≥18 years with COVID-19 admitted to an ICU at one institution from March 20–May 25, 2020 were divided into two cohorts. Pre-LDHS patients had an episode of a-line thrombosis between March 20–April 19. Post-LDHS patients had an episode of a-line thrombosis between April 20–May 25 and received an LDHS solution (10 units/h) through their a-line pressure bag.

Results

Forty-one patients (pre-LDHS) and 30 patients (post-LDHS) were identified. Baseline characteristics were similar between groups, including age (61 versus 54 years; p = 0.24), median Sequential Organ Failure Assessment score (6 versus 7; p = 0.67) and systemic anticoagulation (47% versus 32%; p = 0.32). Median duration of a-line patency was significantly longer in post-LDHS versus pre-LDHS patients (8.5 versus 2.9 days; p < 0.001). The incidence of bleeding complications was similar between cohorts (13% vs. 10%; p = 0.71).

Conclusions

A LDHS protocol was associated with a clinically significant improvement in a-line patency duration in COVID-19 patients, without increased bleeding risk.

•

Critically ill COVID-19 patients have frequent arterial line (a-line) thrombosis.

•

A low dose heparinized saline (LDHS) a-line patency protocol was examined.

•

Two a-line thrombosis cohorts were identified: “post-LDHS” and “pre-LDHS”.

•

A-line patency was longer in post- vs. pre-LDHS patients (8.5 vs. 2.9 d; p < 0.001).

•

Bleeding complications were similar between groups (13% vs. 10%, p = 0.7).

Telemedicine, COVID-19, and disparities: Policy implications

•

COVID-19 has resulted in a dramatic increase in the use of telemedicine but this could worsen disparities.

•

We review recent telemedicine policy changes and their implications regarding disparities.

•

We also discuss what systems can do to improve access to telemedicine and to best meet the needs of underserved patients.

While the rapid expansion of telemedicine in response to the COVID-19 pandemic highlights the impressive ability of health systems to adapt quickly to new complexities, it also raises important concerns about how to implement these novel modalities equitably. As the healthcare system becomes increasingly virtual, it risks widening disparities among marginalized populations who have worse health outcomes at baseline and limited access to the resources necessary for the effective use of telemedicine. In this article, we review recent policy changes and outline important recommendations that governments and health care systems can adopt to improve access to telemedicine and to tailor the use of these technologies to best meet the needs of underserved patients. We suggest that by making health equity integral to the implementation of telemedicine now, it will help to ensure that all can benefit from its use going forward and that this will be increasingly integral to care delivery.

Axon fasciculation in the developing olfactory nerve

Olfactory sensory neuron (OSN) axons exit the olfactory epithelium (OE) and extend toward the olfactory bulb (OB) where they coalesce into glomeruli. Each OSN expresses only 1 of approximately 1,200 odor receptors (ORs). OSNs expressing the same OR are distributed in restricted zones of the OE. However, within a zone, the OSNs expressing a specific OR are not contiguous - distribution appears stochastic. Upon reaching the OB the OSN axons expressing the same OR reproducibly coalesce into two to three glomeruli. While ORs appear necessary for appropriate convergence of axons, a variety of adhesion associated molecules and activity-dependent mechanisms are also implicated. Recent data suggest pre-target OSN axon sorting may influence glomerular convergence. Here, using regional and OR-specific markers, we addressed the spatio-temporal properties associated with the onset of homotypic fasciculation in embryonic mice and assessed the degree to which subpopulations of axons remain segregated as they extend toward the nascent OB. We show that immediately upon crossing the basal lamina, axons uniformly turn sharply, usually at an approximately 90° angle toward the OB. Molecularly defined subpopulations of axons show evidence of spatial segregation within the nascent nerve by embryonic day 12, within 48 hours of the first OSN axons crossing the basal lamina, but at least 72 hours before synapse formation in the developing OB. Homotypic fasciculation of OSN axons expressing the same OR appears to be a hierarchical process. While regional segregation occurs in the mesenchyme, the final convergence of OR-specific subpopulations does not occur until the axons reach the inner nerve layer of the OB.