Automated-detection of risky alcohol use prior to surgery using natural language processing.

BACKGROUND: Preoperative risky alcohol use is one of the most common surgical risk factors. Accurate and early identification of risky alcohol use could enhance surgical safety. Artificial Intelligence-based approaches, such as natural language processing (NLP), provide an innovative method to identify alcohol-related risks from patients' electronic health records (EHR) before surgery. METHODS: Clinical notes (n = 53,629) from pre-operative patients in a tertiary care facility were analyzed for evidence of risky alcohol use and alcohol use disorder. One hundred of these records were reviewed by experts and labeled for comparison. A rule-based NLP model was built, and we assessed the clinical notes for the entire population. Additionally, we assessed each record for the presence or absence of alcohol-related International Classification of Diseases (ICD) diagnosis codes as an additional comparator. RESULTS: NLP correctly identified 87% of the human-labeled patients classified with risky alcohol use. In contrast, diagnosis codes alone correctly identified only 29% of these patients. In terms of specificity, NLP correctly identified 84% of the non-risky cohort, while diagnosis codes correctly identified 90% of this cohort. In the analysis of the full dataset, the NLP-based approach identified three times more patients with risky alcohol use than ICD codes. CONCLUSIONS: NLP, an artificial intelligence-based approach, efficiently and accurately identifies alcohol-related risk in patients' EHRs. This approach could supplement other alcohol screening tools to identify patients in need of intervention, treatment, and/or postoperative withdrawal prophylaxis. Alcohol-related ICD diagnosis had limited utility relative to NLP, which extracts richer information within clinical notes to classify patients.

Predicting persistent opioid use after surgery using electronic health record and patient-reported data.

BACKGROUND: More than 100 million surgeries take place annually in the United States, and more than 90% of surgical patients receive an opioid prescription. A sizable minority of these patients will go on to use opioids long-term, contributing to the national opioid epidemic. METHODS: The objective of this study was to develop and validate a model to predict persistent opioid use after surgery. Participants included surgical patients (≥18 years old) enrolled in a cohort study at an academic medical center between 2015 and 2018. Persistent opioid use was defined as filling opioid prescriptions in postdischarge days 4 to 90 and 91 to 180. Predictors included electronic health record data, state prescription drug monitoring data, and patient-reported measures. Three models were developed: a full, a restricted, and a minimal model using a derivation and validation cohort. RESULTS: Of 24,040 patients, 4,879 (20%) experienced persistent opioid use. In the validation cohort, the full, restricted, and minimal model had C-statistics of 0.87 (95% CI 0.86-0.88), 0.86 (0.85-0.88), and 0.85 (0.84-0.87), respectively. All models performed better among patients with preoperative opioid use compared to opioid-naive patients (P < .001). The models slightly overpredicted risk in the validation cohort. The net benefit of using the restricted model to refer patients for preoperative counseling was 0.072 to 0.092, which is superior to evaluating no patients (net benefit of 0) or all patients (net benefit of -0.22 to -0.63). CONCLUSION: This study developed and validated a prediction model for persistent opioid use using accessible data resources. The models achieved strong performance, outperforming prior published models.

Patient perspectives on emergency department self-referral after bariatric surgery.

BACKGROUND: Reducing avoidable emergency department (ED) visits is an increasingly important target of quality improvement and cost containment efforts in bariatric surgery. Administrative and clinical registry data provide an incomplete picture of the factors contributing to postoperative ED utilization. Patient-centered interviews can help identify intervention opportunities. OBJECTIVES: We sought to understand the circumstances surrounding patient self-referral to the ED after elective, primary bariatric surgery. SETTING: A quality improvement collaborative in Michigan. METHODS: A prospective review of clinically abstracted data and patient interviews was completed across 40 hospitals participating in a statewide quality improvement collaborative. Trained nurses collected data on the circumstances surrounding patients' 30-day postoperative ED visits using a previously validated interview tool. Over a year, 201 of 633 total ED visits met the inclusion criteria, with 78% of those patients being interviewed. RESULTS: The most common reported chief complaints were abdominal pain and nausea/vomiting. Patients reported high compliance with provider-driven perioperative measures to reduce ED visits. One third of patients stated urgency as the reason for not contacting their surgeon prior to their visit. A majority of patients believed their ED visit was both necessary and unavoidable. CONCLUSIONS: Most patients experienced non-life-threatening symptoms but believed their concerns required immediate medical attention in an ED. Patients did not seek lower acuity alternatives despite the increasing availability of these lower cost options. Urgent care centers are one practical alternative for patients who need expeditious professional evaluation. Focused, patient-centered education and promotion of appropriate lower acuity options may decrease nonurgent ED utilization among postoperative bariatric patients.

Effect of Surgeon Age on Bariatric Surgery Outcomes.

OBJECTIVE: This study sought to explore the relationship of bariatric surgeon age and patient outcomes. BACKGROUND: Regulators, policy makers, and patient advocacy groups have recently been pushing to establish clear guidelines for physician retirement in the United States. Although it is often assumed that increasing physician age leads to worse patient outcomes, the relationship is lacking robust evidence, and is still unclear. METHODS: We conducted a study analyzing all bariatric surgeons in Michigan who participated in a statewide collaborative quality improvement program (n = 71) who performed primary laparoscopic Roux-en-Y Gastric Bypass, or sleeve gastrectomy operations, and data on their patients (n = 60430) over the past 10 years. Our primary outcomes were 30-day postoperative complications. Odds ratios for overall complications and serious complications were calculated for each age group, and surgery type. RESULTS: Late career surgeons had more bariatric surgery experience and had a higher average annual case volume than early career surgeons. Considering all cases in the past 10 years, older surgeons performed more Roux-en-Y Gastric Bypass (40%) and less sleeve gastrectomy (38.8%) than younger surgeons (34.7% and 51.5%). When adjusting for patient and surgeon characteristics, there were no statistically significant differences in overall or serious complication rates for either procedure among surgeon age groups. CONCLUSIONS: When evaluating bariatric surgeons in the State of Michigan, we found no statistically significant association between surgeon age and patient outcomes. Our findings do not provide evidence for age-specific retirement cut-offs, but support the development of guidelines which are holistic, and focus on evaluating and improving physician outcomes at all career levels.

Site-specific Approach to Reducing Emergency Department Visits Following Surgery.

OBJECTIVE: The aim of this study was to explore the efficacy of current bariatric perioperative measures at reducing emergency department (ED) visits following bariatric surgery in the state of Michigan. SUMMARY OF BACKGROUND DATA: Many ED visits following bariatric surgery do not result in readmission and may be preventable. Little research exists evaluating the efficacy of perioperative measures aimed at reducing ED visits in this population. Therefore, understanding the driving factors behind these preventable ED visits may be a fruitful approach to prevention. Furthermore, evaluating the efficacy of current perioperative measures may shed light on how to achieve meaningful reductions in ED visits. METHODS: We studied 48,035 eligible bariatric surgery patients across 37 Michigan Bariatric Surgical Collaborative (MBSC) sites between January 2012 and October 2015. Hospitals were ranked according to their risk- and reliability-adjusted ED visit rates. For hospitals in each ED visit rate tercile, several patient, surgery, and hospital summary characteristics were compared. We then studied whether a hospital's compliance with specific perioperative measures was significantly associated with reduced ED visit rates. RESULTS: Only 3 of the 30 surgery, hospital, and patient summary characteristics studied were significant predictors of a hospital's ED visit rate: rate of sleeve gastrectomies, rate of readmissions, and rate of venous thromboembolism complications (P = 0.04, P = 0.0065, and P = 0.0047, respectively). Also, a hospital's compliance with the perioperative measures evaluated was not a significant predictor of ED visit rates (P = 0.12). CONCLUSIONS: Current practices aimed at reducing ED visits appear to be ineffective. Due to heterogeneity in patient populations and local infrastructure, a more tailored approach to ED visit reduction may be more successful.

The putative aminophospholipid translocases, DNF1 and DNF2, are not required for 7-nitrobenz-2-oxa-1,3-diazol-4-yl-phosphatidylserine flip across the plasma membrane of Saccharomyces cerevisiae.

The regulation of phosphatidylserine (PS) distribution across the plasma membrane of eukaryotic cells has been implicated in numerous cell functions (e.g. apoptosis and coagulation). In a recent study, fluorescent phospholipids labeled in the acyl chain with 7-nitrobenz-2-oxa-1, 3-diazol-4-yl (NBD) were used to identify two members of the P4 subfamily of P-type ATPases, Dnf1p and Dnf2p, that are necessary for the inward-directed transport of phospholipids across the plasma membrane (flip) of yeast ( Pomorski, T., Lombardi, R., Riezman, H., Devaux, P. F., Van Meer, G., and Holthuis, J. C. (2003) Mol. Biol. Cell 14, 1240-1254 ). Herein, we present evidence that the flip of NBD-labeled PS (NBD-PS) across the plasma membrane does not require the expression of Dnf1p or Dnf2p. In strains in which DNF1 and DNF2 are both deleted, the flip of NBD-PS is increased approximately 2-fold over that of the isogenic parent strain, whereas the flip of NBD-labeled phosphatidylcholine and NBD-labeled phosphatidylethanolamine are reduced to approximately 20 and approximately 50%, respectively. The mechanism responsible for NBD-PS flip is similar to that for NBD-labeled phosphatidylcholine and NBD-labeled phosphatidylethanolamine in its dependence on cellular ATP and the plasma membrane proton electrochemical gradient, as well as its regulation by the transcription factors Pdr1p and Pdr3p. Based on the observation that deletion or inactivation of all four members of the DRS2/DNF essential subfamily of P-type ATPases does not affect NBD-PS flip, we conclude that the activity reflected by NBD-PS internalization is not the essential function of the DRS2/DNF subfamily of P-type ATPases.

The proton electrochemical gradient across the plasma membrane of yeast is necessary for phospholipid flip.

Recently, two members of the P4 family of P-type ATPases, Dnf1p and Dnf2p, were shown to be necessary for the internalization (flip) of fluorescent, 7-nitrobenz-2-oxa-1,3-diazol-4-yl(NBD)-labeled phospholipids across the plasma membrane of Saccharomyces cerevisiae. In the current study, we have demonstrated that ATP hydrolysis is not sufficient for phospholipid flip in the absence of the proton electrochemical gradient across the plasma membrane. This requirement was demonstrated by two independent means. First, collapse of the plasma membrane proton electrochemical gradient by the protonophore, carbonyl cyanide m-chlorophenylhydrazone (CCCP) almost completely blocked NBD-phospholipid flip while only moderately reducing the cytosolic ATP concentration. Second, strains with point mutations in PMA1, which encodes the plasma membrane proton pump that generates the proton electrochemical gradient, are defective in NBD-PC flip, whereas their cytosolic ATP content is actually increased. These results establish that the proton electrochemical gradient is required for NBD-phospholipid flip across the plasma membrane of yeast and raise the question whether it contributes an additional required driving force or whether it functions as a regulatory signal.