Skip to main content
SpringerLink
Log in

Risk Stratification for Distal Pancreatectomy Utilizing ACS-NSQIP: Preoperative Factors Predict Morbidity and Mortality

  • 2010 SSAT Plenary Presentation
  • Published:
Journal of Gastrointestinal Surgery Aims and scope

Abstract

Background

Evaluation of risk factors for adverse outcomes following distal pancreatectomy (DP) has been limited to data collected from retrospective, primarily single-institution studies. Using a large, multi-institutional prospectively collected dataset, we sought to define the incidence of complications after DP, identify the preoperative and operative risk factors for the development of complications, and develop a risk score that can be utilized preoperatively.

Methods

The American College of Surgeons National Surgical Quality Improvement Program participant use file was utilized to identify patients who underwent DP from 2005 to 2008 by Current Procedural Terminology codes. Multivariate logistic regression analysis was performed to identify variables associated with 30-day morbidity and mortality. A scoring system was developed to allow for preoperative risk stratification.

Results

In 2,322 patients who underwent DP, overall 30-day complication and mortality were 28.1% and 1.2%, respectively. Serious complication occurred in 22.2%, and the most common complications included sepsis (8.7%), surgical site infection (5.9%), and pneumonia (4.7%). On multivariate analysis, preoperative variables associated with morbidity included male gender, high BMI, smoking, steroid use, neurologic disease, preoperative SIRS/sepsis, hypoalbuminemia, elevated creatinine, and abnormal platelet count. Preoperative variables associated with 30-day mortality included esophageal varices, neurologic disease, dependent functional status, recent weight loss, elevated alkaline phosphatase, and elevated blood urea nitrogen. Operative variables associated with both morbidity and mortality included high intraoperative transfusion requirement (≥3 U) and prolonged operation time (>360 min). Weighted risk scores were created based on the preoperatively determined factors that predicted both morbidity (p < 0.001) and mortality (p < 0.001) after DP.

Discussion

The rate of serious complication after DP is 22%. The DP-specific preoperative risk scoring system described in this paper may be utilized for patient counseling and informed consent discussions, identifying high-risk patients who would benefit from disease optimization, and risk adjustment when comparing outcomes between institutions.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2

Similar content being viewed by others

References

  1. Balcom JHt, Rattner DW, Warshaw AL, Chang Y, Fernandez-del Castillo C: Ten-year experience with 733 pancreatic resections: changing indications, older patients, and decreasing length of hospitalization. Arch Surg 2001, 136:391-398.

    Article  PubMed  Google Scholar 

  2. Fernandez-del Castillo C, Rattner DW, Warshaw AL: Standards for pancreatic resection in the 1990 s. Arch Surg 1995, 130:295-299; discussion 299-300.

    CAS  PubMed  Google Scholar 

  3. Le Borgne J, de Calan L, Partensky C: Cystadenomas and cystadenocarcinomas of the pancreas: a multiinstitutional retrospective study of 398 cases. French Surgical Association. Ann Surg 1999, 230:152-161.

    Article  PubMed  Google Scholar 

  4. Lillemoe KD, Kaushal S, Cameron JL, Sohn TA, Pitt HA, Yeo CJ: Distal pancreatectomy: indications and outcomes in 235 patients. Ann Surg 1999, 229:693-698; discussion 698-700.

    Article  CAS  PubMed  Google Scholar 

  5. Weber SM, Cho CS, Merchant N, Pinchot S, Rettammel R, Nakeeb A, Bentrem D, Parikh A, Mazo AE, Martin RC, 3 rd, et al: Laparoscopic left pancreatectomy: complication risk score correlates with morbidity and risk for pancreatic fistula. Ann Surg Oncol 2009, 16:2825-2833.

    Article  PubMed  Google Scholar 

  6. Goh BK, Tan YM, Chung YF, Cheow PC, Ong HS, Chan WH, Chow PK, Soo KC, Wong WK, Ooi LL: Critical appraisal of 232 consecutive distal pancreatectomies with emphasis on risk factors, outcome, and management of the postoperative pancreatic fistula: a 21-year experience at a single institution. Arch Surg 2008, 143:956-965.

    Article  PubMed  Google Scholar 

  7. Kleeff J, Diener MK, Z'Graggen K, Hinz U, Wagner M, Bachmann J, Zehetner J, Muller MW, Friess H, Buchler MW: Distal pancreatectomy: risk factors for surgical failure in 302 consecutive cases. Ann Surg 2007, 245:573-582.

    Article  PubMed  Google Scholar 

  8. Rodriguez JR, Germes SS, Pandharipande PV, Gazelle GS, Thayer SP, Warshaw AL, Fernandez-del Castillo C: Implications and cost of pancreatic leak following distal pancreatic resection. Arch Surg 2006, 141:361-365; discussion 366.

    Article  PubMed  Google Scholar 

  9. Bilimoria MM, Cormier JN, Mun Y, Lee JE, Evans DB, Pisters PW: Pancreatic leak after left pancreatectomy is reduced following main pancreatic duct ligation. Br J Surg 2003, 90:190-196.

    Article  CAS  PubMed  Google Scholar 

  10. Fahy BN, Frey CF, Ho HS, Beckett L, Bold RJ: Morbidity, mortality, and technical factors of distal pancreatectomy. Am J Surg 2002, 183:237-241.

    Article  PubMed  Google Scholar 

  11. Pannegeon V, Pessaux P, Sauvanet A, Vullierme MP, Kianmanesh R, Belghiti J: Pancreatic fistula after distal pancreatectomy: predictive risk factors and value of conservative treatment. Arch Surg 2006, 141:1071-1076; discussion 1076.

    Article  PubMed  Google Scholar 

  12. Sledzianowski JF, Duffas JP, Muscari F, Suc B, Fourtanier F: Risk factors for mortality and intra-abdominal morbidity after distal pancreatectomy. Surgery 2005, 137:180-185.

    Article  CAS  PubMed  Google Scholar 

  13. Yoshioka R, Saiura A, Koga R, Seki M, Kishi Y, Morimura R, Yamamoto J, Yamaguchi T: Risk factors for clinical pancreatic fistula after distal pancreatectomy: analysis of consecutive 100 patients. World J Surg 2010;34:121–125.

    Article  PubMed  Google Scholar 

  14. Sierzega M, Niekowal B, Kulig J, Popiela T: Nutritional status affects the rate of pancreatic fistula after distal pancreatectomy: a multivariate analysis of 132 patients. J Am Coll Surg 2007, 205:52-59.

    Article  PubMed  Google Scholar 

  15. Anonymous: User Guide for the 2008 Participant Use Data File. American College of Surgeons National Surgical Quality Improvement Program 2009.

  16. Hamilton BH, Ko CY, Richards K, Hall BL: Missing data in the American College of Surgeons National Surgical Quality Improvement Program are not missing at random: implications and potential impact on quality assessments. J Am Coll Surg 2010; 210:125–139.e122.

    Article  PubMed  Google Scholar 

  17. Schafer J: NORM: multiple imputation of incomplete multivariate data under a normal model (version 2.03).

  18. Harrell FE, Jr., Lee KL, Mark DB: Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996, 15:361-387.

    Article  PubMed  Google Scholar 

  19. Tekkis PP, McCulloch P, Steger AC, Benjamin IS, Poloniecki JD: Mortality control charts for comparing performance of surgical units: validation study using hospital mortality data. Bmj 2003, 326:786-788.

    Article  PubMed  Google Scholar 

  20. Simons JP, Ng SC, Hill JS, Shah SA, Bodnari A, Zhou Z, Tseng JF: In-hospital mortality for liver resection for metastases: a simple risk score. J Surg Res 2009, 156:21-25.

    Article  PubMed  Google Scholar 

  21. Simons JP, Hill JS, Ng SC, Shah SA, Zhou Z, Whalen GF, Tseng JF: Perioperative mortality for management of hepatic neoplasm: a simple risk score. Ann Surg 2009, 250:929-934.

    Article  PubMed  Google Scholar 

  22. Hill JS, Zhou Z, Simons JP, Ng SC, McDade TP, Whalen GF, Tseng JF: A simple risk score to predict in-hospital mortality after pancreatic resection for cancer. Ann Surg Oncol 2010; 17:1802–1807.

    Article  PubMed  Google Scholar 

  23. Finan KR, Cannon EE, Kim EJ, Wesley MM, Arnoletti PJ, Heslin MJ, Christein JD: Laparoscopic and open distal pancreatectomy: a comparison of outcomes. Am Surg 2009, 75:671-679; discussion 679-680.

    PubMed  Google Scholar 

  24. Bassi C, Dervenis C, Butturini G, Fingerhut A, Yeo C, Izbicki J, Neoptolemos J, Sarr M, Traverso W, Buchler M: Postoperative pancreatic fistula: an international study group (ISGPF) definition. Surgery 2005, 138:8-13.

    Article  PubMed  Google Scholar 

  25. Gibbs J, Cull W, Henderson W, Daley J, Hur K, Khuri SF: Preoperative serum albumin level as a predictor of operative mortality and morbidity: results from the National VA Surgical Risk Study. Arch Surg 1999, 134:36-42.

    Article  CAS  PubMed  Google Scholar 

  26. Daley J, Khuri SF, Henderson W, Hur K, Gibbs JO, Barbour G, Demakis J, Irvin G, 3 rd, Stremple JF, Grover F, et al: Risk adjustment of the postoperative morbidity rate for the comparative assessment of the quality of surgical care: results of the National Veterans Affairs Surgical Risk Study. J Am Coll Surg 1997, 185:328-340.

    CAS  Google Scholar 

  27. Khuri SF, Henderson WG, Daley J, Jonasson O, Jones RS, Campbell DA, Jr., Fink AS, Mentzer RM, Jr., Neumayer L, Hammermeister K, et al: Successful implementation of the Department of Veterans Affairs' National Surgical Quality Improvement Program in the private sector: the Patient Safety in Surgery study. Ann Surg 2008, 248:329-336.

    Article  PubMed  Google Scholar 

  28. Kooby DA, Gillespie T, Bentrem D, Nakeeb A, Schmidt MC, Merchant NB, Parikh AA, Martin RC, 2nd, Scoggins CR, Ahmad S, et al: Left-sided pancreatectomy: a multicenter comparison of laparoscopic and open approaches. Ann Surg 2008, 248:438-446.

    PubMed  Google Scholar 

Download references

Disclaimer

The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in it represent the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or for the conclusions derived by the authors.

Conflicts of interest

The authors declare no conflicts of interest.

Author information

Authors and Affiliations

  1. Department of Surgery, University of Wisconsin, Madison, WI, USA

    Kaitlyn Jane Kelly, David Yu Greenblatt, Yin Wan, Robert J. Rettammel, Emily Winslow & Clifford S. Cho

  2. Clinical Science Center-H4/752, 600 Highland Avenue, Madison, WI, 53792-3236, USA

    Sharon M. Weber

Authors
  1. Kaitlyn Jane Kelly
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. David Yu Greenblatt
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Yin Wan
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Robert J. Rettammel
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Emily Winslow
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Clifford S. Cho
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Sharon M. Weber
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Sharon M. Weber.

Additional information

Discussant

Dr. David B. Adams (Charleston, SC): For pancreatic surgeons, experience trumps evidence. And the practicing surgeon would prefer to jaw about pancreatic fistula prevention than to discuss a NSQIP analysis. So there are gains and loss dimensions to NSQIP reviews, and we need to remind ourselves, as you have reminded us, of the NSQIP weaknesses.

NSQIP 30-day mortality rates underestimate mortality rates for complicated GI procedures such as pancreatectomy. NSQIP does not capture readmission data. NSQIP is based on a limited sample that diminishes the opportunity to identify infrequent but serious complications, such as class C pancreatic fistula. And that’s what we all are going to carp about in this analysis. What about the pancreatic fistula or pancreatic duct occlusion failure, the rate-limiting complication of distal pancreatectomy? How can you assess risks for distal pancreatectomy and not know the pancreatic fistula rate? And that’s just a rhetorical question.

So here is my real, one and only question. If you were to advise the College on how to improve NSQIP to make it a better tool to assess risk and improve outcomes in distal pancreatectomy, what data would you add and what data would you subtract from the current model?

Dr. Kelly, I salute you and your mentors on your premium work and your poised presentation today.

Closing Discussant

Dr. Kaitlyn Jane Kelly: You bring up an excellent point about the lack of some of the pancreas-specific or pancreatectomy-specific postoperative data that we are currently not able to capture with NSQIP, such as, most importantly, pancreatic fistula.

We do think that in this analysis, clinically significant fistulas, defined as grade B or C, are most likely picked up in patients with organ space infection or sepsis, which are outcomes collected by NSQIP.

To improve the database, I would recommend adding more variables for postoperative factors such as the incidence of pancreatic fistula, as well as postpancreatectomy hemorrhage and delayed gastric emptying, particularly for pancreaticoduodenectomy procedures.

I think it would also be useful to reduce some of the other variables currently collected in NSQIP, and to do this selectively. Some of the laboratory valuables like albumin, platelet count, and BUN have repeatedly been shown to be predictors of complications after various general surgery procedures. Those variables should clearly be kept. But it would certainly be helpful in terms of cost and enabling more hospitals to participate in NSQIP, if we could reduce some of the variables currently in the dataset.

Discussant

Dr. Henry Pitt (Indianapolis, IN): Let me help you answer Dr. Adams’ question. In fact, the College is doing just what he and you have suggested, meaning that the number of variables that don’t really play into all these logistic regressions is being reduced. This last year, the variables that we have been talking about, which are pancreas surgery-specific, have been built into ACS-NSQIP and will be rolled out in January 2011. Therefore, the key will be for all of us to switch from the current “classic” ACS-NSQIP to the new “high-risk” module, which will include pancreatectomy and hepatectomy. In working with the statisticians at the College, and with Karl Bilimoria, we also analyzed risk factors for pancreatic surgery. However, we were advised to not examine just Whipple or just distal but also the spectrum of pancreatic surgery. Having procedures that had even higher and lower mortality, and increasing the numbers, actually adds to the validity of these risk models. In fact, we probably don’t even have enough numbers with pancreatectomy, and need to lump hepatectomy and complex biliary to create an HPB Risk Calculator. When we complete this task, we will all have even a better mousetrap than any of us have developed. The ACS-NSQIP HPB Risk Calculator will be on their Web site and will provide the overall morbidity, the serious morbidity, and the mortality. Eventually, the risk of fistula will be available on these patients, and also we will have hospital-specific and surgeon-specific data.

Discussant

Dr. Lygia Stewart (San Francisco, CA): I take it this was an elective distal pancreatectomy database; is that correct? Can you explain to me the preoperative sepsis piece? Because it would seem to me that no pancreatic surgeon would take somebody for an elective pancreatectomy who had preoperative sepsis. Now, that would make sense if it was necrotizing pancreatitis. So can you explain that? And that was pretty important to your morbidity calculations, so it didn’t make a lot of sense to me.

Closing Discussant

Dr. Kaitlyn Jane Kelly: That is a good point. And we initially considered excluding patients that fell into that group of being defined as having preoperative sepsis. But when we looked back, there were a total of 31 patients in the sample who qualified as having preoperative SIRS or sepsis. These variables were defined very specifically-SIRS as having two or more of the following: a temperature above >38° or <36°, a respiratory rate >20, heart rate >90, PCO2 of <32. Those factors plus a known source of infection is defined as sepsis.

We thought it was interesting that such a large number of patients did fit this definition and still underwent elective DP. Given these definitions that are based on the very specific vital signs or laboratory parameters, that it’s feasible in clinical practice, a patient could fit the definition but really not look ill overall, but this could be going unrecognized.

So we thought it was important to point out that we should pay attention to these things. Patients that do fit this definition are obviously at increased risk.

Discussant

Dr. John Chabot (New York, NY): Help some of us who are a little less sophisticated in these analyses with the C index concept. You told us that at 0.5, the predictability of this is random; and at 1.0, it’s perfect. What does 0.64 mean? How useful is this to predict outcome for a specific patient with a C index of 0.64?

Closing Discussant

Dr. Kaitlyn Jane Kelly: It is, as you pointed out, a range. A C index of 0.64 is quite good and is comparable to many other predictive nomograms and models that have been used and published recently.

Just to mention, the NSQIP predictive scoring system for general non-cardiac surgery, called the probability of morbidity score, had a C index of 0.62 when we tested it in our validation sample.

Discussant

Dr. Shimul A. Shah (Worcester, MA): If I may make two editorial comments about NSQIP. We have to be careful about creating risk scores with every database that exists for certain complex procedures because we are limited by the variables that are in each database. So for instance, in NSQIP, we don’t have spleen-preserving versus distal pancreatectomy with splenectomy, which, as we all know, would increase the morbidity or the complication rate, or the size of a cyst in the distal pancreas, or whether it’s in the body or in the tail.

Secondly, NSQIP is 200 centers which voluntarily decide to join the database. It would be nice if we can—one other caveat for NSQIP that I would ask is that we could get 1,000 hospitals to join it. And in that way, we would have a more well-rounded distribution of hospitals that are involved in the risk assessment course that we make up.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Kelly, K.J., Greenblatt, D.Y., Wan, Y. et al. Risk Stratification for Distal Pancreatectomy Utilizing ACS-NSQIP: Preoperative Factors Predict Morbidity and Mortality. J Gastrointest Surg 15, 250–261 (2011). https://doi.org/10.1007/s11605-010-1390-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11605-010-1390-9

Keywords

Search

Navigation