Todd D. Beyer

Predictors of Hospital Readmission after Bariatric Surgery

BACKGROUND Identification of factors that might predict readmission after bariatric surgery could help surgeons target high-risk patients. The purpose of this study was to identify comorbidities, surgical variables, and postoperative complications associated with readmission. STUDY DESIGN Patients with bariatric surgery as their primary procedure were identified from the 2012 American College of Surgeons (ACS) NSQIP database. Patient variables, operative times, and major postoperative complications were analyzed for predictors of readmission. The ACS NSQIP estimated probability of morbidity (MORBPROB) was also considered. Chi-square tests and Poisson regression were used for statistical analysis to identify significant predictors. RESULTS There were 18,186 patients who met inclusion criteria. There were 1,819 who had a laparoscopic gastric band, 9,613 who had laparoscopic Roux-en-Y gastric bypass (RYGB), 6,439 who had gastroplasties (vertical banded gastroplasty and sleeve), and 315 who had open RYGB. Age, sex, BMI, American Society of Anesthesiologists (ASA) class, diabetes, hypertension, steroid use, type of procedure, and operative time all were significantly associated with readmission within 30 days of operation. All major postoperative complications were significant predictors of readmission. Patients expected to be at high risk based on the ACS NSQIP MORBPROB had a significantly higher rate of readmissions. The overall readmission rate for patients undergoing bariatric surgery was 5%. The readmission rate among patients with any major complication was 31%. CONCLUSIONS Bariatric surgery is a low-risk procedure. Complexity of operation, ASA class, prolonged operative time, and major postoperative complications are important determinants of high risk for readmission. The ACS NSQIP MORBPROB may be a useful tool to identify and target patients at risk for readmission.

A prospective evaluation of the effect of sample collection site on intraoperative parathormone monitoring during parathyroidectomy.

BACKGROUND Sample collection site may affect the dynamics of intraoperative parathyroid hormone monitoring (IPM) and influence surgical decisions. METHODS We prospectively studied 45 patients undergoing parathyroidectomy for primary hyperparathyroidism. The IPM cure criterion was a decrease in peripheral vein (PV) parathyroid hormone (PTH) of >50% at 10 minutes after gland excision. PTH samples were collected simultaneously from PV and central vein (CV) and compared for PTH decay, the incidence of >50% PTH decay, and the incidence of normal PTH values after gland excision. RESULTS Mean PTH levels were significantly higher from the CV before and after gland excision. Mean PTH decay 10 minutes after gland excision was 89% PV versus 88% CV, resulting in mean PTH levels of 27 +/- 23 and 39 +/- 35 pg/mL, respectively (P < .0001). At 5 minutes, >50% decay in PTH was present in 98% PV versus 88% CV samples. By 10 minutes, the incidence of >50% PTH decay was equivalent (98%). This yielded normal range PTH levels from the PV versus CV in 90% versus 76% of patients at 5 minutes, 96% versus 89% at 10 minutes, and 95% versus 81% at 20 minutes. Of 45 patients, 44 (98%) are normocalcemic at a mean follow-up of 6.3 months. IPM predicted the single operative failure. CONCLUSIONS CV sampling produces significantly higher PTH levels. Surgeons sampling from a PV may observe a >50% decrease in PTH and normal range PTH values starting 5 minutes after gland excision. Surgeons who sample from the CV and require normalization of PTH levels may have to wait longer and/or continue potentially unnecessary neck exploration.

A NSQIP Risk Assessment for Thyroid Surgery Based on Comorbidities

BACKGROUND Thyroid surgery is associated with low mortality and morbidity and often is performed in an ambulatory setting. The majority of patients undergoing thyroidectomy have an uncomplicated outcome, but common comorbidities may increase mortality and morbidity. Due to low complication rates, studies using single surgeon or single institutional data to identify risk factors for adverse outcomes may be limited by inadequate patient volume. STUDY DESIGN This retrospective cohort study used data from the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP). The study group included all thyroidectomy patients over a 6-year period (2005 to 2010). Common patient comorbidities were identified and analyzed using logistic regression. Risk of adverse outcomes was calculated for single and multiple comorbidities. Statistical significance was set at p < 0.05. RESULTS The study group included 38,577 consecutive patients. Thirty-day mortality and postoperative morbidity were 0.06% and 1.49%, respectively. The risk factors independently associated with morbidity included hypertension, diabetes, advanced age greater than 70 years, COPD, dialysis, malignant thyroid disease, and surgical approach (total thyroidectomy). Substernal thyroidectomy, hypertension, diabetes, age greater than 70 years, COPD, and dialysis were significant predictors (unadjusted) of mortality. Multiple comorbidities resulted in significant cumulative risk. The presence of 3 or more comorbidities was associated with a postoperative morbidity of 5.1% (p < 0.001) and mortality as high as 12.5%. CONCLUSIONS Thyroid surgery is generally safe. Common comorbidities significantly increase the risk of adverse outcomes and death. Clinically applicable risk calculation based on overall health may improve patient selection, surgical management, and informed consent.

Does homocysteine contribute to bone disease in hyperparathyroidism

BACKGROUND Osteoporosis is a complication of hyperparathyroidism (HPT). Hyperhomocysteinemia (HHCy) is an independent risk factor for osteoporotic fractures. We hypothesize that HHCy correlates with bone disease in HPT. METHODS A prospectively collected database of 250 patients treated for HPT was reviewed. Patients were categorized into 3 groups: group I, normal renal function; group 2, mild renal insufficiency; and group 3, secondary HPT with end-stage renal disease on dialysis. Serum homocysteine levels, markers of bone metabolism, and bone density studies were examined. RESULTS The prevalence of HHCy in group 1 (208 patients) was 5%, in group 2 (23 patients), 82%, and in group 3 (19 patients), 78%. Mean (+/-SD) preoperative homocysteinemia (HCy) levels in groups 1, 2, and 3 were 9.3 +/- 4.0, 20 +/- 10.2, and 20.6 +/- 12.3 micromol/L, respectively. Elevated serum markers of bone metabolism increased significantly with decreasing renal function. CONCLUSIONS Prevalence of HHCy is low in HPT patients with normal renal function. It is significantly greater in those with dialysis-independent and -dependent renal insufficiency. HHCy correlates with other serum markers of bone metabolism in HPT and may be useful for monitoring progression or improvement.

Sharps and Needlestick Injuries Among Medical Students, Surgical Residents, Faculty, and Operating Room Staff at a Single Academic Institution

BACKGROUND The hospital is a place of high risk for sharps and needlestick injuries (SNI) and such injuries are historically underreported. METHODS This institutional review board approved study compares the incidence of SNI among all surgical personnel at a single academic institution via an anonymous electronic survey distributed to medical students, surgical residents, general surgery attendings, surgical technicians, and operating room nurses. RESULTS The overall survey response rate was 37% (195/528). Among all respondents, 55% (107/195) had a history of a SNI in the workplace. The overall report rate following an initial SNI was 64%. Surgical staff reported SNIs more frequently, with an incidence rate ratio (IRR) of 1.33 (p = 0.085) when compared with attendings. When compared with surgical attendings, medical students (IRR of 2.86, p = 0.008) and residents (IRR of 2.21, p = 0.04) were more likely to cite fear as a reason for not reporting SNIs. Approximately 65% of respondents did not report their exposure either because of the time consuming process or the patient involved was perceived to be low-risk or both. CONCLUSIONS The 2 most common reasons for not reporting SNIs at our institution are because of the inability to complete the time consuming reporting process and fear of embarrassment or punitive response because of admitting an injury. Further research is necessary to mitigate these factors.