Penn Evidence-Based Literature Review (PEBLR)

Summarized highlights from contemporary literature in surgical and allied disciplines for general surgery residents.

August 2016

  1. Evaluation of ProPublica Surgeon Scorecard "Adjusted Complication Rate" Measure Specifications
  2. Ban et al,, Annals of Surgery 2016 (PDF)
    Contributor: Becky Hoffman
    Discipline: Quality Improvement, Health Services Research, Surgical Education

Synopsis: The ProPublica Surgeon Scorecard ( is the first nationwide, multi-specialty public reporting of individual surgeon outcomes. They report mortality and complications for 8 elective procedures (most notably for us, lap chole), but the methodology for reporting complication rate has been criticized (eg, it only looks at inpatient stay and most lap chole's are outpatient unless complicated, age >65, etc.). This study assessed (1) the proportion of cases excluded by Propublica's methods, (2) proportion of complications occurring as an inpatient and (3) validity of ProPublica's by comparing it to other, previously validated, outcome measures. The authors used 2 years of NSQIP data; the ProPublica "rules" were applied to define the study population and the complication rate and then compared to the cases which were excluded. The authors found that ProPublica's analysis excludes 82% of cases (96% of lap choles), misses 84% of postop complications and does not correlate with validated outcome measures. This "surgeon scorecard" is highly questionable and draws from a very narrow and biased surgical population. The study has the following limitations: using NSQIP, the authors were unable to look at outcomes on an individual surgeon level, which is what ProPublic does. (NSQIP data is available at the hospital level). However, surgeon-level data not necessary to assess the validity of outcome metrics.

  1. Randomized Trial of Thymectomy in Myasthenia Gravis
  2. Wolfe et al., New England Journal of Medicine 2016 (PDF)
    Contributor: Grace Lee
    Discipline: Cardiothoracic/Vascular; Non-Surgical Disciplines

Synopsis: Thymectomy is the mainstay of treatment for myasthenia gravis (MG) when thymoma is present; however, retrospective analyses for thymectomy with nonthymomatous MG demonstrate variable response. The alternative to thymectomy involves long-term prednisone and other immunosuppressive agents. The MGTX study group performed an international, randomized, single-blinded trial comparing Trans-sternal Thymectomy + Prednisone vs. Prednisone Alone for patients with nonthymomatous MG. Results demonstrated that thymectomy patients had significantly improved clinical outcomes (9% vs 37% hospitalized for exacerbation) and lower average prednisone dose requirements (44 mg vs. 60 mg) over a 3-year period. The major limitations of this study included the method for prednisone dose measurement, which relied on pill count through blister packs rather than actual intake, and a higher than expected prednisone dose in both groups compared to routine practice due to the stipulation that prednisone was prescribed to maintain minimal disease manifestation.

  1. Prevention of Atrial Fibrillation in High-Risk Patients Undergoing Lung Cancer Surgery: The PRESAGE Trial
  2. Cardinale et al., Annals of Surgery 2016 (PDF)
    Contributor: Ann Gaffey
    Discipline: Cardiothoracic/Vascular

Synopsis: Postoperative Afib is a well-recognized complication after lung cancer surgery, with an incidence of ~30% (2-4% s/p wedge resection, 10-15% s/p lobectomy, 20-30% s/p pneumonectomy). The exact mechanism for postop Afib is not known but thought to be related to surgical stress, such as sympathetic nerve activity, right ventricular overload, hypoxia, and systemic inflammatory response. Prophylaxis for post-op Afib has been studied extensively in cardiac surgery, but its use in lung cancer surgery is more limited and/or varied in clinical practice. Published guidelines indicate inadequate data to recommend routine Afib prophylaxis, and recommend evaluation on a case-by-case basis. Peri-op NT-proBNP is known to be a strong independent predictor for postop Afib. The authors enrolled 1116 patients with NT-proBNP levels measured preoperatively; 320 pts had elevated levels. Patients were randomly assigned to three groups: 108 metoprolol (12.5mg BID starting up to 50 qd), 102 losartan (12.5 mg BID up to 50 mg qd), and 110 to the control group. Therapy was started within 12 hours of surgery. Incidence of Afib was 20% overall with 6% in metoprolol, 12% in losartan, and 40% in the control. The risk reduction was 32% in the metoprolol cohort and 28% in the losartan cohort. The major limitations of this study are single center design and lack of blinding during treatment randomization.

  1. Interval Appendectomy: Finding the Breaking Point for Cost-Effectiveness
  2. Senekjian et al., Journal of the American College of Surgeons 2016 (PDF)
    Contributor: Charles Vasquez
    Discipline: General Surgery/MIS/Colorectal

Synopsis: Patient with acute complicated appendicitis with phlegmon/abscess are initially managed non-operatively with antibiotics with or without image-guided drainage. After resolution of the acute episode, most patients are recommended to undergo interval appendectomy (IA), with the rationale being that this avoids risk of recurrent appendicitis and diagnoses rarer causes of appendicitis, such as malignancy or IBD. However, the benefits of IA must be weighed with the risks, potential complications, and costs related to appendectomy. This study developed a model to assess the cost-effectiveness of IA across different age groups (ages 18, 35 and 50). Results showed that the cost of IA at age 35 is $8989.16 with 9.1 QALYs gained compared to “no IA” which costs $6614.61 with 9.09 QALYs gained. Furthermore, IA was cost-effective for patients <34 years, but not for older subgroups. The major limitation of this study was that this model is primarily based on age and does not take into account patient baseline health characteristics or family history, which may shift the decision to manage a patient either more or less aggressively.