Evaluating Global Surgery Partnerships From Low and Middle Income Country Perspectives.

INTRODUCTION: Little is known about perceptions of low-income and middle-income country (LMIC) partners regarding global surgery collaborations with high-income countries (HICs). METHODS: A survey was distributed to surgeons from LMICs to assess the nature and perception of collaborations, funding, benefits, communication, and the effects of COVID-19 on partnerships. RESULTS: We received 19 responses from LMIC representatives in 12 countries on three continents. The majority (83%) had participated in collaborations within the past 5 y with 39% of collaborations were facilitated virtually. Clinical and educational partnerships (39% each) were ranked most important by respondents. Sustainability of the partnership was most successfully achieved in domains of education/training (78%) and research (61%). The majority (77%) of respondents reported expressing their needs before HIC team arrival. However, 54% of respondents were the ones to initiate the conversation and only 47% said HIC partners understood the overall environment well at arrival to LMIC. Almost all participants (95%) felt a formal process of collaboration and a structured partnership would benefit all parties in assessing needs. During the COVID-19 pandemic, 87% of participants reported continued collaborations; however, 44% of partners felt that relationships were weaker, 31% felt relationships were stronger, and 25% felt they were unchanged. CONCLUSIONS: Our study provides a snapshot of LMIC surgeons' perspectives on collaboration in global surgery. Independent of location, LMIC partners cite inadequate structure for long-term collaborations. We propose a formal pathway and initiation process to assess resources and needs at the outset of a partnership.

Simulation of undiagnosed patients with novel genetic conditions.

Rare Mendelian disorders pose a major diagnostic challenge and collectively affect 300-400 million patients worldwide. Many automated tools aim to uncover causal genes in patients with suspected genetic disorders, but evaluation of these tools is limited due to the lack of comprehensive benchmark datasets that include previously unpublished conditions. Here, we present a computational pipeline that simulates realistic clinical datasets to address this deficit. Our framework jointly simulates complex phenotypes and challenging candidate genes and produces patients with novel genetic conditions. We demonstrate the similarity of our simulated patients to real patients from the Undiagnosed Diseases Network and evaluate common gene prioritization methods on the simulated cohort. These prioritization methods recover known gene-disease associations but perform poorly on diagnosing patients with novel genetic disorders. Our publicly-available dataset and codebase can be utilized by medical genetics researchers to evaluate, compare, and improve tools that aid in the diagnostic process.

Subdural hemorrhage, macrocephaly, rash, and developmental delay in an infant: A pathogenic variant in NLRP3 causes CINCA/NOMID.

Subdural hemorrhages (SDHs) in children are most often observed in abusive head trauma (AHT), a distinct form of traumatic brain injury, but they may occur in other conditions as well, typically with clear signs and symptoms of an alternative diagnosis. We present a case of an infant whose SDH initially raised the question of AHT, but multidisciplinary evaluation identified multiple abnormalities, including rash, macrocephaly, growth failure, and elevated inflammatory markers, which were all atypical for trauma. These, along with significant cerebral atrophy, ventriculomegaly, and an absence of other injuries, raised concerns for a genetic disorder, prompting genetic consultation. Clinical trio exome sequencing identified a de novo likely pathogenic variant in NLRP3, which is associated with chronic infantile neurological, cutaneous, and articular (CINCA) syndrome, also known as neonatal-onset multisystem inflammatory disease (NOMID). He was successfully treated with interleukin-1 blockade, highlighting the importance of prompt treatment in CINCA/NOMID patients. This case also illustrates how atraumatic cases of SDH can be readily distinguished from AHT with multidisciplinary collaboration and careful consideration of the clinical history and exam findings.

Machine Learning and Statistics in Clinical Research Articles-Moving Past the False Dichotomy.

This Viewpoint describes the false dichotomy between statistics and machine learning and suggests considerations in building and evaluating clinical prediction models.

The Diagnostic and Triage Accuracy of the GPT-3 Artificial Intelligence Model.

Importance: Artificial intelligence (AI) applications in health care have been effective in many areas of medicine, but they are often trained for a single task using labeled data, making deployment and generalizability challenging. Whether a general-purpose AI language model can perform diagnosis and triage is unknown. Objective: Compare the general-purpose Generative Pre-trained Transformer 3 (GPT-3) AI model's diagnostic and triage performance to attending physicians and lay adults who use the Internet. Design: We compared the accuracy of GPT-3's diagnostic and triage ability for 48 validated case vignettes of both common (e.g., viral illness) and severe (e.g., heart attack) conditions to lay people and practicing physicians. Finally, we examined how well calibrated GPT-3's confidence was for diagnosis and triage. Setting and Participants: The GPT-3 model, a nationally representative sample of lay people, and practicing physicians. Exposure: Validated case vignettes (<60 words; <6th grade reading level). Main Outcomes and Measures: Correct diagnosis, correct triage. Results: Among all cases, GPT-3 replied with the correct diagnosis in its top 3 for 88% (95% CI, 75% to 94%) of cases, compared to 54% (95% CI, 53% to 55%) for lay individuals (p<0.001) and 96% (95% CI, 94% to 97%) for physicians (p=0.0354). GPT-3 triaged (71% correct; 95% CI, 57% to 82%) similarly to lay individuals (74%; 95% CI, 73% to 75%; p=0.73); both were significantly worse than physicians (91%; 95% CI, 89% to 93%; p<0.001). As measured by the Brier score, GPT-3 confidence in its top prediction was reasonably well-calibrated for diagnosis (Brier score = 0.18) and triage (Brier score = 0.22). Conclusions and Relevance: A general-purpose AI language model without any content-specific training could perform diagnosis at levels close to, but below physicians and better than lay individuals. The model was performed less well on triage, where its performance was closer to that of lay individuals.

Geographic Proximity of Family Members and Healthcare Utilization After Complex Surgical Procedures.

OBJECTIVE: We sought to determine the relationship between a patient's proximal familial social support, defined as the geographic proximity of family members, and healthcare utilization after complex cardiovascular and oncologic procedures. BACKGROUND: Social support mechanisms are increasingly identified as modifiable risk factors for healthcare utilization. METHODS: We performed a retrospective cohort study of 60,895 patients undergoing complex cardiovascular procedures or oncologic procedures. We defined healthcare utilization outcomes as 30-day all-cause readmission unplanned readmission, nonindex hospital readmission, index hospital length of stay, and home discharge disposition. For each patient, we aggregated the number of first-degree relatives (FDR) living within 30 miles of the patient's home address at the time of the surgical procedure into the following categories: 0 to 1, 2 to 3, 4 to 5, 6+ FDRs. We developed hierarchical multivariable regression models to determine the relationship between the number of FDR living within 30 miles of the patient and the healthcare utilization outcomes. RESULTS: Compared with patients with 0 to 1 FDRs, patients with 6+ FDRs living in close proximity had significantly lower rates of all-cause readmission (12.1% vs 13.5%, P <0.001), unplanned readmission (10.9% vs 12.0%, P =0.001), nonindex readmission (2.6% vs 3.2%, P =0.003); higher rates of home discharge (88.0% vs 85.3%, P <0.001); and shorter length of stay (7.3 vs 7.5 days, P =0.02). After multivariable adjustment, a larger number of FDRs living within 30 miles of the patient was significantly associated with a lower likelihood of all-cause readmission ( P <0.001 for trend), 30-day unplanned readmission ( P <0.001), nonindex readmission ( P <0.001); higher likelihood of home discharge ( P <0.001); and shorter index length of stay ( P <0.001). CONCLUSIONS: The geographic proximity of family members is significantly associated with decreased healthcare utilization after complex cardiovascular and oncologic surgical procedures.

Hospital Costs Following Surgical Complications: A Value-driven Outcomes Analysis of Cost Savings Due to Complication Prevention.

OBJECTIVE: Surgical complications have substantial impact on healthcare costs. We propose an analysis of the financial impact of postoperative complications. SUMMARY OF BACKGROUND DATA: Both complications and preoperative patient risk have been shown to increase costs following surgery. The extent of cost increase due to specific complications has not been well described. METHODS: A single institution's American College of Surgeons National Surgical Quality Improvement Program data was queried from 2012 to 2018 and merged with institutional cost data for each encounter. A mixed effects multivariable generalized linear model was used to estimate the mean relative increase in hospital cost due to each complication, adjusting for patient and procedure-level fixed effects clustered by procedure. Potential savings were calculated based on projected decreases in complication rates and theoretical hospital volume. RESULTS: There were 11,897 patients linked between the 2 databases. The rate of any American College of Surgeons National Surgical Quality Improvement Program complication was 11.7%. The occurrence of any complication resulted in a 1.5-fold mean increase in direct hospital cost [95% confidence interval (CI) 1.49-1.58]. The top 6 most costly complications were postoperative septic shock (4.0-fold, 95% CI 3.58-4.43) renal insufficiency/failure (3.3-fold, 95% CI 2.91-3.65), any respiratory complication (3.1-fold, 95% CI 2.94-3.36), cardiac arrest (3.0-fold, 95% CI 2.64-3.46), myocardial infarction (2.9-fold, 95% CI 2.43-3.42) and mortality within 30 days (2.2-fold, 95% CI 2.01-2.48). Length of stay (6.5 versus 3.2 days, P < 0.01), readmission rate (29.1% vs 3.1%, P < 0.01), and discharge destination outside of home (20.5% vs 2.7%, P < 0.01) were significantly higher in the population who experienced complications. CONCLUSIONS: Decreasing complication rates through preoperative optimization will improve patient outcomes and lead to substantial cost savings.

Comparison of Patient Outcomes Among Integrated Residency Versus Fellowship-trained Vascular Surgeons.

OBJECTIVE: This study aims to evaluate whether graduates of integrated vascular surgery residency (IVSR) programs achieve similar surgical outcomes in clinical practice as compared to graduates of vascular surgery fellowships (VSF). SUMMARY OF BACKGROUND DATA: Early sub-specialization through IVSR programs decreases the total years of surgical training. However, it is unclear whether IVSR graduates achieve comparable outcomes to fellowship-trained surgeons once in clinical practice. METHODS: We identified all vascular surgeons who finished IVSR and VSF programs between 2013-2017 using American Board of Surgery data, which was linked to the Vascular Quality Initiative registry (2013-2019) to evaluate provider-specific clinical outcomes following carotid, lower extremity, and aortic aneurysm repair procedures. The association between training models and the composite outcome of 1-year mortality, major adverse cardiac events and/or other major complications were analyzed using mixed-effects logistic regression models. RESULTS: A total of 338 surgeons (31% IVSR, 69% VSF) submitted cases into the Vascular Quality Initiative registry, including 8155 carotid, 21,428 lower extremity, and 5800 aortic aneurysm repair procedures. Composite 1-year outcome rates were comparable between IVSR and VSF-trained surgeons following carotid endarterectomy (8%-IVSR vs 7%-VSF), lower extremity revascularization (19%-IVSR vs 16%-VSF), and aortic aneurysm repair (13%-IVSR vs 13%-VSF) procedures. These findings among IVSR-trained surgeons persisted following risk adjustment for severity of patient disease and indications for undertaking carotid [aOR: 1.04 (0.84-1.28)], lower extremity [aOR: 1.03 (0.84-1.26)], and aortic [aOR: 0.96 (0.76-1.21)] procedures when compared to VSF-trained surgeons. CONCLUSIONS: Despite fewer total years of training, graduates of IVSR programs achieve equivalent surgical outcomes as fellowship-trained vascular surgeons once in practice. These results suggest that concerns about differential competence among integrated residency graduates are not warranted.

Precision Medicine: Using Artificial Intelligence to Improve Diagnostics and Healthcare.

The continued generation of large amounts of data within healthcare-from imaging to electronic medical health records to genomics and multi-omics -necessitates tools and methods to parse and interpret these data to improve healthcare outcomes. Artificial intelligence, and in particular deep learning, has enabled researchers to gain new insights from large scale and multimodal data. At the 2022 Pacific Symposium on Biocomputing (PSB) session entitled "Precision Medicine: Using Artificial Intelligence to Improve Diagnostics and Healthcare", we showcase the latest research, influenced and inspired by the idea of using technology to build a more fair, tailored, and cost-effective healthcare system after the COVID-19 pandemic.

Value Analysis of Methods of Inguinal Hernia Repair.

OBJECTIVE: Value is defined as health outcomes important to patients relative to cost of achieving those outcomes: Value = Quality/Cost. For inguinal hernia repair, Level 1 evidence shows no differences in long-term functional status or recurrence rates when comparing surgical approaches. Differences in value reside within differences in cost. The aim of this study is to compare the value of different surgical approaches to inguinal hernia repair: Open (Open-IH), Laparoscopic (Lap-IH), and Robotic (R-TAPP). METHODS: Variable and fixed hospital costs were compared among consecutive Open-IH, Lap-IH, and R-TAPP repairs (100 each) performed in a university hospital. Variable costs (VC) including direct materials, labor, and variable overhead ($/min operating room [OR] time) were evaluated using Value Driven Outcomes, an internal activity-based costing methodology. Variable and fixed costs were allocated using full absorption costing to evaluate the impact of surgical approach on value. As cost data is proprietary, differences in cost were normalized to Open-IH cost. RESULTS: Compared to Open-IH, VC for Lap-IH were 1.02X higher (including a 0.81X reduction in cost for operating room [OR] time). For R-TAPP, VC were 2.11X higher (including 1.36X increased costs for OR time). With allocation of fixed cost, a Lap-IH was 1.03X more costly, whereas R-TAPP was 3.18X more costly than Open-IH. Using equivalent recurrence as the quality metric in the value equation, Lap-IH decreases value by 3% and R-TAPP by 69% compared to Open-IH. CONCLUSIONS: Use of higher cost technology to repair inguinal hernias reduces value. Incremental health benefits must be realized to justify increased costs. We expect payors and patients will incorporate value into payment decisions.

Natural language processing for the surveillance of postoperative venous thromboembolism.

BACKGROUND: The objective of this study was to develop a portal natural language processing approach to aid in the identification of postoperative venous thromboembolism events from free-text clinical notes. METHODS: We abstracted clinical notes from 25,494 operative events from 2 independent health care systems. A venous thromboembolism detected as part of the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) was used as the reference standard. A natural language processing engine, easy clinical information extractor-pulmonary embolism/deep vein thrombosis (EasyCIE-PEDVT), was trained to detect pulmonary embolism and deep vein thrombosis from clinical notes. International Classification of Diseases (ICD) discharge diagnosis codes for venous thromboembolism were used as baseline comparators. The classification performance of EasyCIE-PEDVT was compared with International Classification of Diseases codes using sensitivity, specificity, area under the receiver operating characteristic curve, using an internal and external validation cohort. RESULTS: To detect pulmonary embolism, EasyCIE-PEDVT had a sensitivity of 0.714 and 0.815 in internal and external validation, respectively. To detect deep vein thrombosis, EasyCIE-PEDVT had a sensitivity of 0.846 and 0.849 in internal and external validation, respectively. EasyCIE-PEDVT had significantly higher discrimination for deep vein thrombosis compared with International Classification of Diseases codes in internal validation (area under the receiver operating characteristic curve: 0.920 vs 0.761; P < .001) and external validation (area under the receiver operating characteristic curve: 0.921 vs 0.794; P < .001). There was no significant difference in the discrimination for pulmonary embolism between EasyCIE-PEDVT and ICD codes. CONCLUSION: Accurate surveillance of postoperative venous thromboembolism may be achieved using natural language processing on clinical notes in 2 independent health care systems. These findings suggest natural language processing may augment manual chart abstraction for large registries such as NSQIP.

Cross-modal representation alignment of molecular structure and perturbation-induced transcriptional profiles.

Modeling the relationship between chemical structure and molecular activity is a key goal in drug development. Many benchmark tasks have been proposed for molecular property prediction, but these tasks are generally aimed at specific, isolated biomedical properties. In this work, we propose a new cross-modal small molecule retrieval task, designed to force a model to learn to associate the structure of a small molecule with the transcriptional change it induces. We develop this task formally as multi-view alignment problem, and present a coordinated deep learning approach that jointly optimizes representations of both chemical structure and perturbational gene expression profiles. We benchmark our results against oracle models and principled baselines, and find that cell line variability markedly influences performance in this domain. Our work establishes the feasibility of this new task, elucidates the limitations of current data and systems, and may serve to catalyze future research in small molecule representation learning.

Portable Automated Surveillance of Surgical Site Infections Using Natural Language Processing: Development and Validation.

OBJECTIVES: We present the development and validation of a portable NLP approach for automated surveillance of SSIs. SUMMARY OF BACKGROUND DATA: The surveillance of SSIs is labor-intensive limiting the generalizability and scalability of surgical quality surveillance programs. METHODS: We abstracted patient clinical text notes after surgical procedures from 2 independent healthcare systems using different electronic healthcare records. An SSI detected as part of the American College of Surgeons' National Surgical Quality Improvement Program was used as the reference standard. We developed a rules-based NLP system (Easy Clinical Information Extractor [CIE]-SSI) for operative event-level detection of SSIs using an training cohort (4574 operative events) from 1 healthcare system and then conducted internal validation on a blind cohort from the same healthcare system (1850 operative events) and external validation on a blind cohort from the second healthcare system (15,360 operative events). EasyCIE-SSI performance was measured using sensitivity, specificity, and area under the receiver-operating-curve (AUC). RESULTS: The prevalence of SSI was 4% and 5% in the internal and external validation corpora. In internal validation, EasyCIE-SSI had a sensitivity, specificity, AUC of 94%, 88%, 0.912 for the detection of SSI, respectively. In external validation, EasyCIE-SSI had sensitivity, specificity, AUC of 79%, 92%, 0.852 for the detection of SSI, respectively. The sensitivity of EasyCIE-SSI decreased in clean, skin/subcutaneous, and outpatient procedures in the external validation compared to internal validation. CONCLUSION: Automated surveillance of SSIs can be achieved using NLP of clinical notes with high sensitivity and specificity.

Measuring the migration of surgical specialists.

BACKGROUND: The lack of access to essential surgical care in low-income countries is aggravated by emigration of locally-trained surgical specialists to more affluent regions. Yet, the global diaspora of surgeons, obstetricians, and anesthesiologists from low-income and middle-income countries has never been fully described and compared with those who have remained in their country of origin. It is also unclear whether the surgical workforce is more affected by international migration than other medical specialists. In this study, we aimed to quantify the proportion of surgical specialists originating from low-income and middle-income countries that currently work in high-income countries. METHODS: We retrieved surgical workforce data from 48 high-income countries and 102 low-income and middle-income countries using the database of the World Health Organization Global Surgical Workforce. We then compared this domestic workforce with more granular data on the country of initial medical qualification of all surgeons, anesthesiologists, and obstetricians made available for 14 selected high-income countries to calculate the proportion of surgical specialists working abroad. RESULTS: We identified 1,118,804 specialist surgeons, anesthesiologists, or obstetricians from 102 low-income and middle-income countries, of whom 33,021 (3.0%) worked in the 14 included high-income countries. The proportion of surgical specialists abroad was greatest for the African and South East Asian regions (12.8% and 12.1%). The proportion of specialists abroad was not greater for surgeons, anesthesiologists, or obstetricians than for physicians and other medical specialists (P = .465). Overall, the countries with the lowest remaining density of surgical specialists were also the countries from which the largest proportion of graduates were now working in high-income countries (P = .011). CONCLUSION: A substantial proportion of all surgeons, anesthesiologists, and obstetricians from low-income and middle-income countries currently work in high-income countries. In addition to decreasing migration from areas of surgical need, innovative strategies to retain and strengthen the surgical workforce could involve engaging this large international pool of surgical specialists and instructors.

Home Disposal Kits for Leftover Opioid Medications After Surgery: Do They Work?

BACKGROUND: Postoperative overprescribing is common, and many patients will have excess medications. An effective method to encourage disposal is lacking. We hypothesized that a convenient home disposal kit will result in more appropriate disposal of excess opioids. MATERIALS AND METHODS: We conducted a single-center prospective observational pilot study to evaluate the effectiveness of a postoperative opioid disposal kit. Patients in the intervention group received an opioid disposal kit and educational handout before discharge from the hospital. At the first follow-up visit, patients completed a survey in which they reported the remaining amount of pain medications from their original prescription and their plan for the excess medication. Patients were asked about risk factors for chronic opioid use. We used multivariable Poisson regression to identify independent factors associated with an increased likelihood of appropriate opioid disposal. RESULTS: The survey was offered to 904 patients with a response rate of 91.7%. After excluding those with missing data, 571 patients were included in the study. Overall, 83 (14.5%) patients never filled an opioid prescription, and 286 (60.0%) patients had tablets remaining at the time of the follow-up visit. Among those with tablets remaining, 52 received a home disposal kit, whereas 234 patients with tablets remaining did not. Patients who received the kit were more likely to dispose of opioid medications (54.9% versus 34.8%, relative risk = 1.8, 95% CI 1.3-2.5). No confounders were identified during multivariable analysis that increased a patient's likelihood of disposing excess medications. CONCLUSIONS: The provision of a convenient home disposal kit postoperatively increased patient-reported opioid disposal.

Academic Advancement in Global Surgery: Appointment, Promotion, and Tenure: Recommendations From the American Surgical Association Working Group on Global Surgery.

: There is growing interest in global surgery among US academic surgical departments. As academic global surgery is a relatively new field, departments may have minimal experience in evaluation of faculty contributions and how they integrate into the existing academic paradigm for promotion and tenure. The American Surgical Association Working Group on Global Surgery has developed recommendations for promotion and tenure in global surgery, highlighting criteria that: (1) would be similar to usual promotion and tenure criteria (eg, publications); (2) would likely be undervalued in current criteria (eg, training, administrative roles, or other activities that are conducted at low- and middle-income partner institutions and promote the partnerships upon which other global surgery activities depend); and (3) should not be considered (eg, mission trips or other clinical work, if not otherwise linked to funding, training, research, or building partnerships).

Examining the Use of Real-World Evidence in the Regulatory Process.

The 21st Century Cures Act passed by the United States Congress mandates the US Food and Drug Administration to develop guidance to evaluate the use of real-world evidence (RWE) to support the regulatory process. RWE has generated important medical discoveries, especially in areas where traditional clinical trials would be unethical or infeasible. However, RWE suffers from several issues that hinder its ability to provide proof of treatment efficacy at a level comparable to randomized controlled trials. In this review article, we summarized the advantages and limitations of RWE, identified the key opportunities for RWE, and pointed the way forward to maximize the potential of RWE for regulatory purposes.

Patient Satisfaction With Decision Making Does Not Correlate With Patient Centeredness of Surgeons.

BACKGROUND: High-quality decision making is important in patient-centered care. Although patient involvement in decision making varies widely, most patients desire to share in decision making. The Press-Ganey Patient Satisfaction survey includes questions that measure patients' perceptions of their providers' efforts to involve them in decision making (PGDM). We hypothesized that higher PGDM scores would correlate with higher scores on a validated measure of patient centeredness. MATERIALS AND METHODS: Surgical providers at a university hospital who routinely receive Press-Ganey scores received a survey that included the Patient-Practitioner Orientation Scale (PPOS), a validated tool that measures the provider's orientation toward patient centeredness on a continuous six-point scale: score ≥5 = high, 4.57-5 = moderate, and <4.57 = low and includes nine-item "caring" or "sharing" subscales. We compared PPOS scores to PGDM scores, averaged from April 2015 to January 2016. RESULTS: Eighty-six of 112 (75%) of surgical providers responded to the survey. Fifty-two (46%) had PGDM scores available and 26% achieved a perfect score on the PGDM. The overall PPOS scores were low, with a mean of 4.2 (SD = 0.5). The PPOS was not correlated with the PGDM, correlation coefficient (rs) = -0.07 (CI: -0.34-0.21, P = 0.63). Similarly, the two subscales of the PPOS did not correlate with the PGDM with rs = -0.15 (CI: -0.41-0.13, P = 0.29) for "caring" and rs = -0.04 (CI: -0.31-0.23, P = 0.76) for "sharing". CONCLUSIONS: Although surgical providers scored low in patient centeredness using the PPOS, over one-quarter (26%) of them rank in the top 1% on the PGDM. No correlation was found between providers' patient centeredness and their patients' perceptions of efforts to include them in decision making.