under the curve. Results: 890 patients were identified who met inclusion criteria. The most commonly performed procedures were distal pancreatectomy (559, 63%) and pancreaticoduodenectomy (246, 27%). The majority of patients were less than 65 years of age and ASA Class III. The most common comorbidities were hypertension (396, 45%) and diabetes (186, 21%). Complications among these patients based on ACS NSQIP definitions included any (225, 25%), serious (215, 24%), readmission (153, 17%), surgical site infection (153, 17%), return to the OR (41, 5%), urinary tract infection (39, 5%), venous thromboembolism (28, 3%), discharge to a facility (25, 3%), pneumonia (25, 3%), cardiac (9, 1%), and death (12, 1%). The calculator provided reasonable estimates of risk for discharge to nursing facility (AUC = 0.713) and performed poorly (AUC < 0.7) for all other complications (Fig. 1). Conclusion:The ACS NSQIP Risk Calculator estimates similar proportion of risk to actual events in patients with PNET but has low specificity for identifying the correct patients. The tool may require modification for some patient populations. Most importantly, surgeons must be cautious when estimating risk and use of the calculator alone may be inappropriate.

Diaphragmatic pacing, also known as electrophrenic respiration or phrenic pacing has been used for quite some time in patients with respiratory failure due to diaphragmatic paralysis. We present a case of patient with respiratory insufficiency due to unilateral phrenic nerve injury as a result of radiation therapy for breast cancer. The patient had a diaphragmatic pacer inserted, with significant recovery on her symptoms. Three years after this procedure, the patient recovered her phrenic nerve function, and this was confirmed by nerve conduction study.