Chris Wirtalla

Transfer status: A risk factor for mortality in patients with necrotizing fasciitis

BACKGROUNDnNecrotizing fasciitis (NF) is a rapidly progressive disease that requires urgent surgical debridement for survival. Interhospital transfer (IT) may be associated with delay to operation, which could increase mortality. We hypothesized that mortality would be higher in patients undergoing surgical debridement for necrotizing fasciitis after IT compared to Emergency Department (ED) admission.nnnMETHODSnWe performed a retrospective cohort analysis from 2000-2006 using the Nationwide Inpatient Sample. Inclusion criteria were age >18 years, primary diagnosis of NF, and surgical therapy within 72 hours of admission. Logistic regression was used to assess the relationship between admission source, patient and hospital variables, and mortality.nnnRESULTSnWe identified 9,958 cases over the study period. Patients in the ED group were more likely to be nonwhite and of lower income when compared with patients in the IT group. Unadjusted mortality was higher in the IT group than ED group (15.5% vs 8.7%, P < .001). After adjusting for potential confounders, odds of mortality were still greater in the IT (OR 2.04, CI 95% 1.60-2.59, P < .001).nnnCONCLUSIONnInterhospital transfer is associated with increased risk of in-hospital mortality after surgical therapy for NF, a finding which persists after controlling for patient and hospital level variables.

Prevention of surgical resident attrition by a novel selection strategy.

Objective(s):We modified the resident selection strategy in an attempt to reduce resident attrition (RA). Summary Background Data:Despite implementation of the Accreditation Council for Graduate Medical Education work rules, lifestyle and generational priorities have fostered a persistent and relatively high attrition rate for surgical trainees. Methods:An independent external review of residents who left the training program and a detailed analysis of the resident selection strategy were performed by an organizational management expert. Modifications implemented in 2005 (the intervention) included standardization of the screening and interview format. Applicants were required to submit a 500 words essay related to stress management, organizational skills, future aspirations, and prioritization abilities. Their responses formed the basis of an extended, personalized, and structured interview script. Candidate characteristics and RA were compared for the 5 years before and after the intervention, using Fisher exact test or &khgr;2. Results:Age, sex, birthplace, medical school ranking, step 1 score, and American Board of Surgey In-Training Examination performance were not significantly different between the selection strategy groups. Risk factors for RA included ABSITE performance and gender. Resident performance and subsequent RA were significantly affected by the resident selection strategy. Conclusions:RA was dramatically reduced following the intervention. A custom designed process to identify candidates most likely to succeed substantially improved resident retention in a demanding academic training program.

Teaching status: the impact on emergency and elective surgical care in the US.

Objective(s): To examine the relation between hospital teaching status and surgical outcomes for both emergency and elective general surgery cases using a national database. Background: Teaching hospitals (TH) have been shown to have better outcomes for complex elective surgical cases when compared with nonteaching hospitals (NTH). Less is known about the effect of teaching status on outcomes for more common procedures, especially where emergency surgical cases are concerned. Worse outcomes seen in this cohort are often attributed to patient disease, but systems level variables such as TH status may also play a role. Methods: We performed a nationally representative retrospective cohort study of surgical admissions during 2000 to 2006 using the Nationwide Inpatient Sample. Patients were included if they were more than 18 years of age and had a general surgical procedure performed on the day of admission. We examined unadjusted and adjusted in-hospital mortality (IHM) and postoperative complications (POC) for elective and emergency patients. Results: We identified 1,052,809 admissions. Patients treated at THs were more likely to be nonwhite and at extremes of income when compared with those treated at NTH. Adjusted outcomes revealed an increased risk of IHM at TH (overall OR = 1.20; 95% CI 1.03–1.40, P = 0.017) for emergency admissions with no difference in IHM seen after elective procedures. Postoperative infections were more likely to occur at TH than NTH after elective procedures (OR = 1.14; 95% CI 1.06–1.17, P < 0.007). Postoperative fistula was more likely to occur at TH than NTH after elective surgery (OR = 1.56; 95% CI 1.32–1.85, P < 0.005) whereas postoperative ileus was less likely to occur at TH than NTH (OR = 0.82; 95% CI 0.74–0.91, P = 0.002). Conclusions: Teaching status is associated with increased risk of IHM after emergency cases. POC profiles also differ by TH status. Investigation of the way in which systems-level variables that differ between TH and NTH contribute to postoperative outcomes may identify novel targets for performance improvement.

The Malnourished Patient With Obesity: A Unique Paradox in Bariatric Surgery

BACKGROUNDnHypoalbuminemia is a known risk factor for poor outcomes following surgery. Obesity can be associated with modest to severe malnutrition. We evaluated the impact of hypoalbuminemia on surgical outcomes in patients with obesity undergoing elective bariatric surgical procedures.nnnMATERIALS AND METHODSnThe 2015 metabolic and bariatric surgery accreditation and quality improvement program database was queried. Patients ≥ 18 y with body mass index ≥35 undergoing bariatric surgery were included. Revision procedures were excluded. Patients were classified by albumin level (albumin ≥3.5xa0g/dL [normal], 3.49-3.0xa0g/dL [mild], 2.99-2.5xa0g/dL [moderate], and <2.5xa0g/dL [severe]). Independent logistic regression models were developed to estimate the adjusted odds of (1) death or serious morbidity (DSM); (2) mild to moderate complications; (3) severe complications; and (4) 30-d readmissions by albumin level. In addition, effect modification by >10% weight loss was examined.nnnRESULTSnA total of 106,577 patients were included in the study. Over 6% of patients had hypoalbuminemia. Fifty-five percent of complications were severe as categorized by the Clavien-Dindo classification. Patients with mild hypoalbuminemia had 20% increased odds of DSM (95% confidence interval: 1.1-1.4). There was increasing likelihood of DSM with severe hypoalbuminemia. Patients with mild hypoalbuminemia had 20% increased odds of 30-d readmission (confidence interval: 1.1-1.3). A >10% weight loss modified the effect of moderate to severe hypoalbuminemia on DSM.nnnCONCLUSIONSnMore than 6% of patients with obesity undergoing bariatric surgery are malnourished. Hypoalbuminemia is an important and modifiable risk factor for postoperative adverse outcomes following bariatric surgery. Preoperative weight loss >10% combined with moderate to severe hypoalbuminemia is synergistic for high rates of DSM and should be addressed before proceeding with bariatric surgery.

A preoperative prediction model for risk of multiple admissions after colon cancer surgery

BACKGROUNDnA subset of patients who undergo colon cancer surgery may be at a high risk of multiple subsequent admissions. We developed a simplified model to predict the preoperative risk of multiple postoperative admissions (MuAdm) among patients undergoing colon resection to aid in preoperative planning.nnnMETHODSnPatients aged ≥18xa0y with colon cancer who underwent elective surgical resection identified in discharge claims from California and New York (2008-2011) were included. The primary outcome, MuAdm, was defined as 2 or more admissions in the year following resection. Logistic regression models were developed to identify factors predictive of MuAdm. A weighted point system was developed using beta-coefficients (Pxa0<xa00.05). A random sample of 75% of the data was used for model development, which was validated in the remaining 25% sample.nnnRESULTSnA total of 14,780 patients underwent colon resection for cancer. Almost 30% had an admission in the year after index surgery and 9.8% had MuAdm. The significant predictors of MuAdm were higher Elixhauser comorbidity index score, metastatic disease, payer system, and the number of admissions in the year before surgery. Scores ranged from 0 to 8. Scores ≤1 had a 7% risk of MuAdm, and scores ≥6 had a >30% risk of MuAdm.nnnCONCLUSIONSnIn the year following discharge after resection of colon cancer, nearly 10% of patients are admitted 2 or more times. A simple, preoperative clinical model can prospectively predict the likelihood of multiple admissions in patients anticipating resection. This model can be used for preoperative planning and setting postoperative expectations more accurately.

The Association of Body Mass Index with Postoperative Outcomes After Elective Paraesophageal Hernia Repair

PurposeTo evaluate the association between body mass index (BMI) and postoperative outcomes in elective paraesophageal hernia (PEH) repairs.MethodsA retrospective review of patients who underwent elective PEH repair in the ACS NSQIP database (2005–2015) was performed. Patients were stratified into BMI groups (<u200918.5, 18.5–24.9, 25.0–29.9, 30.0–34.9, 35–39.9, and ≥u200940.0xa0kg/m2) according to the World Health Organization classification criteria. A multivariable logistic regression model was developed to characterize the association between BMI class and outcomes, including readmission, reoperation, postoperative complications, and mortality.ResultsThe median (IQR) age of the 9641 patients who met inclusion criteria was 64 (55–72) and 72.7% were women. Across each BMI class, age, race, gender, type of procedure, frailty index, smoking, and ASA class varied (pu2009<u20090.05). Underweight patients (BMI <u200918.5xa0kg/m2) had an increased risk of mortality (ORu2009=u20096.35, pu2009<u20090.05). Patients with a BMI 35–39.9xa0kg/m2 (ORu2009=u20090.65, pu2009<u20090.05) and ≥u200940xa0kg/m2 (ORu2009=u20090.36, pu2009<u20090.001) were associated with a decreased risk for readmissions.ConclusionUnderweight patients have an increased risk for postoperative mortality after elective PEH repair. Higher BMI was associated with a diminished risk for readmission, but not for mortality, reoperations, or overall complications.

An Instrumental Variables Design for the Effect of Emergency General Surgery

Abstract Confounding by indication is a critical challenge in evaluating the effectiveness of surgical interventions using observational data. The threat from confounding is compounded when using medical claims data due to the inability to measure risk severity. If there are unobserved differences in risk severity across patients, treatment effect estimates based on methods such a multivariate regression may be biased in an unknown direction. A research design based on instrumental variables offers one possibility for reducing bias from unobserved confounding compared to risk adjustment with observed confounders. This study investigates whether a physician’s preference for operative care is a valid instrumental variable for studying the effect of emergency surgery. We review the plausibility of the necessary causal assumptions in an investigation of the effect of emergency general surgery (EGS) on inpatient mortality among adults using medical claims data from Florida, Pennsylvania, and New York in 2012–2013. In a departure from the extant literature, we use the framework of stochastic monotonicity which is more plausible in the context of a preference-based instrument. We compare estimates from an instrumental variables design to estimates from a design based on matching that assumes all confounders are observed. Estimates from matching show lower mortality rates for patients that undergo EGS compared to estimates based in the instrumental variables framework. Results vary substantially by condition type. We also present sensitivity analyses as well as bounds for the population level average treatment effect. We conclude with a discussion of the interpretation of estimates from both approaches.