- Email: [email protected]
Guiding the Reluctant
IAP Guidelines: Commentary Pancreatology 2003;3:139–143 DOI: 10.1159/000070082
Guiding the Reluctant A Primer on Guidelines in General and Pancreatitis in Particular
Edward L. Bradley III Clinical Sciences (Surgery), Florida State University, Tallahassee, Fla., USA
Key Words Guidelines W Evidence-based medicine W Acute pancreatitis W Standard of care
Abstract Within the past 5 years, no fewer than 6 guidelines for the management of acute pancreatitis have appeared in the literature, including the current submission from the International Association of Pancreatology. When these collected guidelines are subjected to comparison, however, marked similarities emerge betweeen the proposals, and significant differences are rare. Surprisingly, neither the passage of time nor the application of evidence-based medicine techniques to the creation of guidelines for acute pancreatitis has resulted in substantive changes to the original guidelines offered by the Atlanta Symposium. Moreover, it is important to realize that, despite claims of objectivity, opinion may enter into the process of guideline creation whenever proposals are based upon lesser levels of evidence. Until the guidelines common to these collected proposals have been supported by randomized controlled trials, it is unlikely that we need any more guidelines for acute pancreatitis. Copyright © 2003 S. Karger AG, Basel and IAP
As far as the author is aware, modern guidelines for the management of patients with acute pancreatitis initially appeared in 1994 [1]. These guidelines were the consensus
ABC
© 2003 S. Karger AG, Basel and IAP 1424–3903/03/0032–0139$19.50/0
Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com
Accessible online at: www.karger.com/pan
practice recommendations of 40 internationally recognized experts in acute pancreatitis from six medical specialties and 15 countries. They were assembled in Atlanta for the twin purposes of: (1) describing their approach to the diagnosis and management of patients with acute pancreatitis, and (2) developing a clinically based classification system for acute pancreatitis and its complications. While the clinical classification system for acute pancreatitis proposed by the attendees has been widely adopted by the international medical community [2], the practice guidelines developed at the same meeting, and later published in book form, have received considerably less attention. More recently, however, a veritable epidemic of guidelines created to assist clinicians in the diagnosis and management of patients with acute pancreatitis has spread through the literature. Within the past 5 years, no fewer than six large-scale efforts have been made to disseminate proposed guidelines for the diagnosis and treatment of acute pancreatitis and its complications [3–8]; ranging in scope from generalized descriptions [5] to encyclopedic recommendations [3, 6, 7]. The rationale for this evident outbreak of guidelines devoted to acute pancreatitis is not entirely clear. Some may have appeared in an effort to update those guidelines proposed by participants in the Atlanta Symposium. Others may have been generated by the current penchant for ‘evidence-based’ evaluation, a technique of data analysis in which the strength of clinical evidence supporting individual proposals is objectively graded on a continuum of scientific merit. Finally, at least
E.L. Bradley III, MD 1600 Baywood Way Sarasota, FL 34231 (USA) Tel. +1 941 923 1331, Fax +1 941 923 3113 E-Mail [email protected]
Table 1. Congruence of specific guidelines for acute pancreatitis Guideline
1 Stratification of severity
Atlanta (1994)
BSG (1998)
SSAT (1998)
Santorini (1999)
WCG (2002)
JSAEM (2002)
IAP (2003)
APACHE II CRP CECT
APACHE II CRP CECT
NS
APACHE II
APACHE II
APACHE II CRP
NS
2 CECT scanning
severe AP
severe AP
NS
severe AP
severe AP
severe AP
NS
3 Prophylactic antibiotics
necrotizing pancreatitis
necrotizing pancreatitis
necrotizing pancreatitis
necrotizing pancreatitis
necrotizing pancreatitis
necrotizing pancreatitis
necrotizing pancreatitis
4 Endoscopic sphincterotomy for biliary pancreatitis
cholangitis jaundice
cholangitis jaundice severe AP
cholangitis
cholangitis jaundice severe AP
cholangitis jaundice severe AP
cholangitis jaundice severe AP
cholangitis jaundice
5 Timing of cholecystectomy for biliary pancreatitis
same 2–4 weeks hospitalization
same NS hospitalization
same NS hospitalization
same hospitalization
6 Surgical therapy for sterile necrotizing pancreatitis
rarely
NS
NS
rarely
rarely
rarely
rarely
7 FNA to identify infected pancreatic necrosis
yes
NS
NS
yes
yes
yes
yes
8 Enteral nutrition
NS
NS
NS
yes
yes
yes
yes
9 Efficacy of antiproteases
NS
NS
NS
no
no
yes
NS
yes
yes
yes
yes
yes
yes
NS
10 Specialty centers for severe pancreatitis
Reference numbers: Atlanta Symposium [1]; British Society of Gastroenterology [3]; Society for Surgery of the Alimentary Tract [4]; Santorini Consensus Conference [5]; World Congress of Gastroenterology [6]; Japanese Society of Abdominal Emergency Medicine [7]; International Association of Pancreatology [8]. AP = Acute pancreatitis; NS = not stated; CECT = contract enhanced computed tomography; FNA = fine needle aspiration; CRP = C-reactive protein.
in one case, generation of guidelines for acute pancreatitis seemed to have resulted from an organizational programmatic affinity for guidelines in general. Seen from the most charitable viewpoint, the recent appearance of six additional sets of guidelines for acute pancreatitis may, in part, reflect attempts to organize and collate the burgeoning basic science and clinical data arising from truly remarkable advances in technology. However, when the collected practice guidelines are compared, it becomes readily apparent that widespread overlap exists among these seven proposals (table 1). More precisely, only a very few differences exist between them. The single substantial difference between all of these proposed guidelines concerns the efficacy of antiproteases in acute pancreatitis; the guidelines from the Japanese Society of Abdominal Emergency Medicine claim gabexate mesilate as established therapy, while the Santorini and World Congress of Gastroenterology guidelines emphatically reject antiproteases as effective. Furthermore, aside from two guidelines concerning enteral nutrition and antiproteases, it is interesting to note that there are no significant differences between the original Atlanta guidelines and the six subsequent proposals. While it may be true that
140
Pancreatology 2003;3:139–143
repetition has an intrinsic value, one cannot help but wonder about the frequency that quite similar information is being shared with the medical profession. Even more curious, however, is the observation that not all of the authors claiming ‘evidence-based’ guidelines agree as to the scientific merit of the available evidence supporting or rejecting an individual guideline! Supporters of ‘evidence-based’ analysis assume that a scientifically dispassionate review of existing data should lead different reviewers to the same conclusion; i.e. scientific proof is not reviewer-dependent. However, as shown in table 2, diverse sets of reviewers failed to agree upon the precise level of scientific proof for nine of the ten ‘evidencebased’ guidelines examined, suggesting that considerable subjectivity exists in an allegedly objective process. Truth, like beauty, may well lie in the eye of the beholder! Moreover, the enthusiasm exhibited by proponents of ‘evidence-based’ analysis, although eminently justifiable, cannot be allowed to advance the inverse proposition that a thorough evaluation of existing data cannot have been conducted in the absence of ‘evidence-based’ activity. Presumably, the principal reasons for the development of any set of practice guidelines are to identify, and then
Bradley
Table 2. Evidence-based guidelines for acute pancreatitis: group assessment of evidence level
Guideline
BSG*
Santorini
WCG
JSAEM
IAP
1 2 3 4 5 6 7 8 9 10
B B NS A B NS NS NS NS B
A A B B NS A A A A C
NS NS A B B B NS A A NS
A B B B NS B A B B B
NS NS A NS B B B B NS NS
Stratification of severity CECT scanning Prophylactic antibiotics Endoscopic sphincterotomy for biliary pancreatitis Timing of cholecystectomy for biliary pancreatitis Surgical therapy for sterile necrotizing pancreatitis FNA to identify infected pancreatic necrosis Enteral nutrition Efficacy of antiproteases Specialty centers for severe pancreatitis
Levels of scientific evidence: A = randomized controlled trials (RCTs); B = non-randomized clinical studies; C = expert opinion. CECT = Contrast-enhanced CT scanning; FNA = fine-needle aspiration; NS = not stated. * References similar to table 1.
to encourage, ‘best clinical practice’. Certainly, no one would choose to contest such worthy goals, provided that they were based upon good science, and were patient-centered. While these are admirable goals to be sure, it is perhaps only fair to question whether or not proposed practice guidelines have actually met these goals. Can ‘best practice’ be reliably identified? Will practice guidelines proposed by specific interest groups be followed by the medical profession at large? Are there any risks to the establishment of practice guidelines? Validation of the impact of guidelines on clinical practice has only just begun, making definitive answers to our posed questions problematic at best. Many of the evaluative efforts regarding guidelines have come from researchers in the United Kingdom. Following the publication of the Guidelines for the Management of Acute Pancreatitis proposed by the British Society of Gastroenterology in 1998 [3], a number of investigators have examined the impact that these guidelines have had on clinical practice. Norton et al. [9] analyzed the actual clinical practice records of 210 patients with acute pancreatitis admitted to a single district hospital in England against the BSG guidelines. They found that despite notable discrepancies between actual hospital practice and the ‘evidence-based’ BSG guidelines, mortality in their institution was almost precisely similar to the mortality obtained by strict adherence to the guidelines. This observation raises some doubt about the validity of those BSG guidelines with implications for mortality. How could such a paradox occur, since the BSG Guidelines were offered as ‘evidence-based’ guidelines?
Because the state of existing knowledge dictates whether or not ‘evidence-based’ guidelines must necessarily accept lesser levels of scientific ‘proof’, it should be quite clear that any specific guideline which is not founded upon the highest levels of proof (randomized controlled trials: RCTs) may ultimately prove to be in error. Physicians must remain wary of unproved guidelines, and should be alert to the possibility that future studies may change, or even reverse, current recommendations. As an equally important corollary to this statement, any proposed guideline that is not firmly based upon a RCT should automatically be identified as fertile ground for future investigation. Not only can reasonable doubt exist as to the validity of an individual guideline in ‘evidence-based’ proposals, but wholesale acceptance of guidelines, regardless of merit, may be an even greater issue for the medical profession. Despite national recommendations, several groups have documented the failure of UK physicians to incorporate the ‘evidence-based’ BSG Guidelines into their clinical practice. Barnard and Siriwardena [10] reported pooled data from five audits of acute pancreatitis containing 545 patients that were analyzed for compliance with the BSG Guidelines. They found that compliance with the National Guideline advocating definitive treatment of gallstone pancreatitis within 4 weeks of the index attack of acute pancreatitis was only 49% (range 16–65%). Furthermore, the National Guideline recommending admission of severe acute pancreatitis to a high intensity or intensive care unit was being followed but 52% of the time (range 20– 100%). Moreover, not only was compliance with the BSG
Guiding the Reluctant
Pancreatology 2003;3:139–143
141
recommendations subject to striking geographic variations, but disregarding the guidelines altogether was common. Others have found significant differences in the implementation of the BSG National Guidelines between types of physicians; those claiming hepatobiliary specialization accepted and employed the guidelines, while nonspecialists were notably reluctant to follow the recommendations [11]. From these reports, we might conclude that, for whatever reasons, dissemination of national guidelines for the management of acute pancreatitis did not result in widespread practice incorporation within the UK. As to whether or not guidelines carry any risk, concerns have been raised in various quarters that the establishment of ANY type of guidelines creates a legal risk for those physicians who, for one reason or another, choose not to follow the proposals. Some members of the plaintiff’s bar have attempted to equate the creation of guidelines to the legal term ‘standard-of-care’. ‘Standard-ofcare’ is a concept that is variously defined in different venues, but is most often characterized as ‘those actions that a reasonably prudent physician, practicing in similar circumstances, would take’. Perceived deviations from this concept of ‘standard-of-care’ enable accusations of medical negligence, or malpractice. While the concept of ‘standard-of-care’ seems quite clear to attorneys, it is certainly less so for physicians. Encoded within the legal definition are the tacit presumptions that all ‘prudent’ physicians are in agreement with a specific approach, and that there is little variation between patients. Due to the complexity and variation inherent in medicine, it is extremely unlikely that ANY set of guidelines will ever be written that can cover all conceivable aspects of a particular condition. Clinical judgment will continue to reign as the physicians’ most powerful tool. Accordingly, clinical judgment may quite properly dictate courses of action beyond the restrictive confines of practice guidelines. Furthermore, as we have noted, clinical guidelines that are not founded upon results from RCTs should be considered to be ‘current recommendations subject to future change’, and not ‘representations of established truths accepted by all reasonably prudent physicians’. When guidelines are temporary proposals and subject to future change, they cannot reasonably be considered to be ‘standard-of-care’, since the legal concept of a TEMPORARY ‘standard-of-care’ does not exist. Moreover, as we have seen, not even expert physicians agree upon which data should be regarded as ‘proved’, and therefore not temporary. Given the unproved potentially temporary nature of the majority of medical data, there should be little confusion as to whether or not guidelines
142
Pancreatology 2003;3:139–143
represent ‘standard-of-care’ mandates. In our desire to educate the members of our profession to ‘best current practice’, we cannot permit the Law to equate the promotion of educational guidelines within the medical profession to the legal concept of ‘standard-of-care’. Medicine entails infinite variability; the Law proscribes variation. Physicians must recognize this basic difference between the professions. Putting aside these general comments regarding guidelines, how, then, should we view the proposed ‘IAP Guidelines for the Surgical Management of Acute Pancreatitis’? In one sense, since 10 of the 11 proposals made by the authors are based upon suggestive data (level B) rather than proven studies (level A), these guidelines merely represent the opinions of the authors as to the best practices at this time. They are also subject to change in the future as definitive evidence becomes available. Do the IAP Guidelines break any new ground? Of the 11 recommendations proposed by the IAP, 10 were initially proposed by the members of the Atlanta Symposium in their prior publication on therapy [1]. The eleventh, ‘Endoscopic sphincterotomy is an alternative to cholecystectomy in those who are not fit to undergo surgery, in order to lower the risk of recurrence of gallstoneassociated acute pancreatitis’ has been previously recommended in other guidelines [6]. What can we say of the specific IAP Guidelines themselves? Although most pancreatologists would find themselves in agreement with these proposals as re-stated by the IAP, the sections dealing with sterile pancreatic necrosis seem to contain some unnecessarily vague language. With regard to Recommendation #5, the ‘selected cases’ of sterile pancreatic necrosis that should be subjected to surgical intervention, rather than persisting in conservative management, are not otherwise characterized. Nor are the ‘specific recommendations’ further identified for which abandonment of conservative management of sterile necrosis may be necessary within the first 14 days after onset. In fairness to the authors, these areas have not been well investigated, and the absence of more specific recommendations may not only reflect the inadequacy of our current knowledge, but also a natural reluctance on the part of the authors to include Expert Opinions (level C) in their overall proposals. Still, the authors have not been reluctant to include Clinical Trials evidence (evidence level B) as support for their proposals, evidence which may be only somewhat less likely to be overturned in the future. Level C information does exist to suggest that small subpopulations of patients with sterile necrosis may be
Bradley
surgical candidates; information not previously set forth in prior attempts at creation of guidelines for acute pancreatitis. Some patients recovering from necrotizing pancreatitis will develop recurrent abdominal pain and pancreatic inflammation whenever attempts are made at restoring oral alimentation. This is usually seen 4 or more weeks following recovery from the initial episode of necrotizing pancreatitis, and has been variously referred to as ‘persistent pancreatitis’ [12], ‘re-feeding pancreatitis’ [13], or ‘disconnected duct syndrome’ [14]. The underlying pathophysiology seems to be either obstruction of the pancreatic duct by the necrotic process [14] or leakage of pancreatic juice from a necrotic duct into an enclosed section of the retroperitoneal space [15]. ERCP or MRI studies will usually detect the interruption or leakage from the duct. Surgical removal of the involved portion of duct is most often curative. Even in the most favorable light, however, this indication for surgery would not rise above the level of Expert Opinion, and further study targeting this recommendation is necessary. Finally, while the overall success of conservative management for sterile necrotizing pancreatitis has been demonstrated beyond reasonable doubt, some patients will continue to deteriorate and die despite appropriate and enthusiastic resuscitative measures. Whether or not patients with sterile pancreatic necrosis and progressive organ deterioration can be salvaged by surgical debride-
ment is not known. Currently, no credible evidence exists to suggest that surgical debridement of sterile necrosis prevents, eliminates, or even ameliorates co-existing organ failure. However, false-negative fine-needle aspirations of infected pancreatic necrosis have been described, and while quite rare (!5%), can result in potentially preventable death. Nevertheless, any decision to offer surgical exploration to a progressively deteriorating patient with FNA aspirate ‘proved’ sterile pancreatic necrosis, in the hope of finding cryptic infected necrosis, should include a frank discussion of the 20:1 odds against finding infected necrosis at surgical exploration, and the real possibility that exploration could accelerate the patient’s decline. In summary, one could reasonably propose that because there have been few guidelines for acute pancreatitis over the past decade that have actually broken new ground, it may be time to call for a moratorium on new sets of guidelines that do not result in significant advances in clinical practice. Rather than re-stating or fine-tuning previous proposals, future efforts might be more profitably expended toward raising the level of proof for existing guidelines. The IAP counts among its members some of the brightest and best workers in pancreatic disease. With direction from its officers and co-operation among its members, randomized controlled trials designed to definitively address less desirable levels of evidence could be conducted. The need is both clear and urgent.
References 1 Bradley EL III (ed): Acute Pancreatitis: Diagnosis and Therapy. New York, Raven Press, 1994. 2 Bradley EL III: A clinically based classification system for acute pancreatitis: Summary of the international symposium on acute pancreatitis, Atlanta, Ga., September 11 through 13, 1992. Arch Surg 1994;128:586–590. 3 United Kingdom Guidelines for the Management of Acute Pancreatitis. Gut 1998;42(suppl 2):1S–13S. 4 SSAT Patient Care Guidelines: Treatment of acute pancreatitis. J Gastrointest Surg 1998;2: 487–488. 5 Dervenis C, Johnson CD, Bassi C, Bradley EL III, Imrie CW, McMahon MJ, Modlin I: Diagnosis, objective assessment of severity, and management of acute pancreatitis: Santorini consensus conference. Int J Pancreatol 1999; 25:195–210. 6 Toouli J, Brooke-Smith M, Bassi C, Carr-Locke D, Telford J, Freeney P, Imrie C, Tandon R: Working party report: Guidelines for the management of acute pancreatitis. J Gastroenterol Hepatol 2002;17(suppl):S15–S39.
Guiding the Reluctant
7 Mayumi T, Ura H, Arata S, Kitamura N, Kiriyama I, Shibuya K, Sekimoto M, Nago N, Hirota M, Yoshida M, Ito Y, Hirata K, Takada T: Evidence-based clinical practice guidelines for acute pancreatitis: Proposals. J Hepatobiliary Panc Surg 2002;9:413–422. 8 Uhl W, Warshaw A, Imrie C, Bassi C, McKay CJ, Lankisch PG, Carter R, DiMagno E, Banks PA, Whitcomb DC, Dervenis C, Ulrich CD, Satake K, Ghaneh P, Hartwig W, Werner J, McEntee G, Neoptolemos JP, Büchler MW: IAP guidelines for the surgical management of acute pancreatitis. Pancreatology 2002;2:565– 573. 9 Norton SA, Cheruvu CV, Collins J, Dix FP, Eyre-Brook IA: An assessment of clinical guidelines for the management of acute pancreatitis. Ann R Coll Surg Engl 2001;83:399–405. 10 Barnard J, Siriwardena AK: Variations in implementation of current national guidelines for the treatment of acute pancreatitis: Implications for acute surgical service provision. Ann R Coll Surg Eng 2002;84:79–81.
11 Aly EA, Milne R, Johnson CD: Non-compliance with national guidelines in the management of acute pancreatitis in the United Kingdom. Dig Surg 2002;19:192–198. 12 Rutledge PL, Warshaw AL: Persistent acute pancreatitis: A variant treated by pancreatoduodenectomy. Arch Surg 1988;123:597–600. 13 Bradley EL III, Balthazar EJ: Pancreatic Necrosis, chap 33; in: Hepatobiliary and Pancreatic Disease: The Team Approach to Management. New York, Little, Brown, 1995, pp 366–375. 14 Howard TJ, Rhodes GJ, Selzer DJ, Sherman S, Fogel E, Lehman GA: Roux-en-Y drainage is the best surgical option to treat patients with disconnected duct syndrome after severe acute pancreatitis. Surgery 2001;130:714–721. 15 Uomo G, Visconti M, Manes G, Calise F, Laccetti M, Rabbitti PG: Nonsurgical treatment of acute necrotizing pancreatitis. Pancreas 1996; 12:142–148.
Pancreatology 2003;3:139–143
143
Guiding the Reluctant A Primer on Guidelines in General and Pancreatitis in Particular
Edward L. Bradley III Clinical Sciences (Surgery), Florida State University, Tallahassee, Fla., USA
Key Words Guidelines W Evidence-based medicine W Acute pancreatitis W Standard of care
Abstract Within the past 5 years, no fewer than 6 guidelines for the management of acute pancreatitis have appeared in the literature, including the current submission from the International Association of Pancreatology. When these collected guidelines are subjected to comparison, however, marked similarities emerge betweeen the proposals, and significant differences are rare. Surprisingly, neither the passage of time nor the application of evidence-based medicine techniques to the creation of guidelines for acute pancreatitis has resulted in substantive changes to the original guidelines offered by the Atlanta Symposium. Moreover, it is important to realize that, despite claims of objectivity, opinion may enter into the process of guideline creation whenever proposals are based upon lesser levels of evidence. Until the guidelines common to these collected proposals have been supported by randomized controlled trials, it is unlikely that we need any more guidelines for acute pancreatitis. Copyright © 2003 S. Karger AG, Basel and IAP
As far as the author is aware, modern guidelines for the management of patients with acute pancreatitis initially appeared in 1994 [1]. These guidelines were the consensus
ABC
© 2003 S. Karger AG, Basel and IAP 1424–3903/03/0032–0139$19.50/0
Fax + 41 61 306 12 34 E-Mail [email protected] www.karger.com
Accessible online at: www.karger.com/pan
practice recommendations of 40 internationally recognized experts in acute pancreatitis from six medical specialties and 15 countries. They were assembled in Atlanta for the twin purposes of: (1) describing their approach to the diagnosis and management of patients with acute pancreatitis, and (2) developing a clinically based classification system for acute pancreatitis and its complications. While the clinical classification system for acute pancreatitis proposed by the attendees has been widely adopted by the international medical community [2], the practice guidelines developed at the same meeting, and later published in book form, have received considerably less attention. More recently, however, a veritable epidemic of guidelines created to assist clinicians in the diagnosis and management of patients with acute pancreatitis has spread through the literature. Within the past 5 years, no fewer than six large-scale efforts have been made to disseminate proposed guidelines for the diagnosis and treatment of acute pancreatitis and its complications [3–8]; ranging in scope from generalized descriptions [5] to encyclopedic recommendations [3, 6, 7]. The rationale for this evident outbreak of guidelines devoted to acute pancreatitis is not entirely clear. Some may have appeared in an effort to update those guidelines proposed by participants in the Atlanta Symposium. Others may have been generated by the current penchant for ‘evidence-based’ evaluation, a technique of data analysis in which the strength of clinical evidence supporting individual proposals is objectively graded on a continuum of scientific merit. Finally, at least
E.L. Bradley III, MD 1600 Baywood Way Sarasota, FL 34231 (USA) Tel. +1 941 923 1331, Fax +1 941 923 3113 E-Mail [email protected]
Table 1. Congruence of specific guidelines for acute pancreatitis Guideline
1 Stratification of severity
Atlanta (1994)
BSG (1998)
SSAT (1998)
Santorini (1999)
WCG (2002)
JSAEM (2002)
IAP (2003)
APACHE II CRP CECT
APACHE II CRP CECT
NS
APACHE II
APACHE II
APACHE II CRP
NS
2 CECT scanning
severe AP
severe AP
NS
severe AP
severe AP
severe AP
NS
3 Prophylactic antibiotics
necrotizing pancreatitis
necrotizing pancreatitis
necrotizing pancreatitis
necrotizing pancreatitis
necrotizing pancreatitis
necrotizing pancreatitis
necrotizing pancreatitis
4 Endoscopic sphincterotomy for biliary pancreatitis
cholangitis jaundice
cholangitis jaundice severe AP
cholangitis
cholangitis jaundice severe AP
cholangitis jaundice severe AP
cholangitis jaundice severe AP
cholangitis jaundice
5 Timing of cholecystectomy for biliary pancreatitis
same 2–4 weeks hospitalization
same NS hospitalization
same NS hospitalization
same hospitalization
6 Surgical therapy for sterile necrotizing pancreatitis
rarely
NS
NS
rarely
rarely
rarely
rarely
7 FNA to identify infected pancreatic necrosis
yes
NS
NS
yes
yes
yes
yes
8 Enteral nutrition
NS
NS
NS
yes
yes
yes
yes
9 Efficacy of antiproteases
NS
NS
NS
no
no
yes
NS
yes
yes
yes
yes
yes
yes
NS
10 Specialty centers for severe pancreatitis
Reference numbers: Atlanta Symposium [1]; British Society of Gastroenterology [3]; Society for Surgery of the Alimentary Tract [4]; Santorini Consensus Conference [5]; World Congress of Gastroenterology [6]; Japanese Society of Abdominal Emergency Medicine [7]; International Association of Pancreatology [8]. AP = Acute pancreatitis; NS = not stated; CECT = contract enhanced computed tomography; FNA = fine needle aspiration; CRP = C-reactive protein.
in one case, generation of guidelines for acute pancreatitis seemed to have resulted from an organizational programmatic affinity for guidelines in general. Seen from the most charitable viewpoint, the recent appearance of six additional sets of guidelines for acute pancreatitis may, in part, reflect attempts to organize and collate the burgeoning basic science and clinical data arising from truly remarkable advances in technology. However, when the collected practice guidelines are compared, it becomes readily apparent that widespread overlap exists among these seven proposals (table 1). More precisely, only a very few differences exist between them. The single substantial difference between all of these proposed guidelines concerns the efficacy of antiproteases in acute pancreatitis; the guidelines from the Japanese Society of Abdominal Emergency Medicine claim gabexate mesilate as established therapy, while the Santorini and World Congress of Gastroenterology guidelines emphatically reject antiproteases as effective. Furthermore, aside from two guidelines concerning enteral nutrition and antiproteases, it is interesting to note that there are no significant differences between the original Atlanta guidelines and the six subsequent proposals. While it may be true that
140
Pancreatology 2003;3:139–143
repetition has an intrinsic value, one cannot help but wonder about the frequency that quite similar information is being shared with the medical profession. Even more curious, however, is the observation that not all of the authors claiming ‘evidence-based’ guidelines agree as to the scientific merit of the available evidence supporting or rejecting an individual guideline! Supporters of ‘evidence-based’ analysis assume that a scientifically dispassionate review of existing data should lead different reviewers to the same conclusion; i.e. scientific proof is not reviewer-dependent. However, as shown in table 2, diverse sets of reviewers failed to agree upon the precise level of scientific proof for nine of the ten ‘evidencebased’ guidelines examined, suggesting that considerable subjectivity exists in an allegedly objective process. Truth, like beauty, may well lie in the eye of the beholder! Moreover, the enthusiasm exhibited by proponents of ‘evidence-based’ analysis, although eminently justifiable, cannot be allowed to advance the inverse proposition that a thorough evaluation of existing data cannot have been conducted in the absence of ‘evidence-based’ activity. Presumably, the principal reasons for the development of any set of practice guidelines are to identify, and then
Bradley
Table 2. Evidence-based guidelines for acute pancreatitis: group assessment of evidence level
Guideline
BSG*
Santorini
WCG
JSAEM
IAP
1 2 3 4 5 6 7 8 9 10
B B NS A B NS NS NS NS B
A A B B NS A A A A C
NS NS A B B B NS A A NS
A B B B NS B A B B B
NS NS A NS B B B B NS NS
Stratification of severity CECT scanning Prophylactic antibiotics Endoscopic sphincterotomy for biliary pancreatitis Timing of cholecystectomy for biliary pancreatitis Surgical therapy for sterile necrotizing pancreatitis FNA to identify infected pancreatic necrosis Enteral nutrition Efficacy of antiproteases Specialty centers for severe pancreatitis
Levels of scientific evidence: A = randomized controlled trials (RCTs); B = non-randomized clinical studies; C = expert opinion. CECT = Contrast-enhanced CT scanning; FNA = fine-needle aspiration; NS = not stated. * References similar to table 1.
to encourage, ‘best clinical practice’. Certainly, no one would choose to contest such worthy goals, provided that they were based upon good science, and were patient-centered. While these are admirable goals to be sure, it is perhaps only fair to question whether or not proposed practice guidelines have actually met these goals. Can ‘best practice’ be reliably identified? Will practice guidelines proposed by specific interest groups be followed by the medical profession at large? Are there any risks to the establishment of practice guidelines? Validation of the impact of guidelines on clinical practice has only just begun, making definitive answers to our posed questions problematic at best. Many of the evaluative efforts regarding guidelines have come from researchers in the United Kingdom. Following the publication of the Guidelines for the Management of Acute Pancreatitis proposed by the British Society of Gastroenterology in 1998 [3], a number of investigators have examined the impact that these guidelines have had on clinical practice. Norton et al. [9] analyzed the actual clinical practice records of 210 patients with acute pancreatitis admitted to a single district hospital in England against the BSG guidelines. They found that despite notable discrepancies between actual hospital practice and the ‘evidence-based’ BSG guidelines, mortality in their institution was almost precisely similar to the mortality obtained by strict adherence to the guidelines. This observation raises some doubt about the validity of those BSG guidelines with implications for mortality. How could such a paradox occur, since the BSG Guidelines were offered as ‘evidence-based’ guidelines?
Because the state of existing knowledge dictates whether or not ‘evidence-based’ guidelines must necessarily accept lesser levels of scientific ‘proof’, it should be quite clear that any specific guideline which is not founded upon the highest levels of proof (randomized controlled trials: RCTs) may ultimately prove to be in error. Physicians must remain wary of unproved guidelines, and should be alert to the possibility that future studies may change, or even reverse, current recommendations. As an equally important corollary to this statement, any proposed guideline that is not firmly based upon a RCT should automatically be identified as fertile ground for future investigation. Not only can reasonable doubt exist as to the validity of an individual guideline in ‘evidence-based’ proposals, but wholesale acceptance of guidelines, regardless of merit, may be an even greater issue for the medical profession. Despite national recommendations, several groups have documented the failure of UK physicians to incorporate the ‘evidence-based’ BSG Guidelines into their clinical practice. Barnard and Siriwardena [10] reported pooled data from five audits of acute pancreatitis containing 545 patients that were analyzed for compliance with the BSG Guidelines. They found that compliance with the National Guideline advocating definitive treatment of gallstone pancreatitis within 4 weeks of the index attack of acute pancreatitis was only 49% (range 16–65%). Furthermore, the National Guideline recommending admission of severe acute pancreatitis to a high intensity or intensive care unit was being followed but 52% of the time (range 20– 100%). Moreover, not only was compliance with the BSG
Guiding the Reluctant
Pancreatology 2003;3:139–143
141
recommendations subject to striking geographic variations, but disregarding the guidelines altogether was common. Others have found significant differences in the implementation of the BSG National Guidelines between types of physicians; those claiming hepatobiliary specialization accepted and employed the guidelines, while nonspecialists were notably reluctant to follow the recommendations [11]. From these reports, we might conclude that, for whatever reasons, dissemination of national guidelines for the management of acute pancreatitis did not result in widespread practice incorporation within the UK. As to whether or not guidelines carry any risk, concerns have been raised in various quarters that the establishment of ANY type of guidelines creates a legal risk for those physicians who, for one reason or another, choose not to follow the proposals. Some members of the plaintiff’s bar have attempted to equate the creation of guidelines to the legal term ‘standard-of-care’. ‘Standard-ofcare’ is a concept that is variously defined in different venues, but is most often characterized as ‘those actions that a reasonably prudent physician, practicing in similar circumstances, would take’. Perceived deviations from this concept of ‘standard-of-care’ enable accusations of medical negligence, or malpractice. While the concept of ‘standard-of-care’ seems quite clear to attorneys, it is certainly less so for physicians. Encoded within the legal definition are the tacit presumptions that all ‘prudent’ physicians are in agreement with a specific approach, and that there is little variation between patients. Due to the complexity and variation inherent in medicine, it is extremely unlikely that ANY set of guidelines will ever be written that can cover all conceivable aspects of a particular condition. Clinical judgment will continue to reign as the physicians’ most powerful tool. Accordingly, clinical judgment may quite properly dictate courses of action beyond the restrictive confines of practice guidelines. Furthermore, as we have noted, clinical guidelines that are not founded upon results from RCTs should be considered to be ‘current recommendations subject to future change’, and not ‘representations of established truths accepted by all reasonably prudent physicians’. When guidelines are temporary proposals and subject to future change, they cannot reasonably be considered to be ‘standard-of-care’, since the legal concept of a TEMPORARY ‘standard-of-care’ does not exist. Moreover, as we have seen, not even expert physicians agree upon which data should be regarded as ‘proved’, and therefore not temporary. Given the unproved potentially temporary nature of the majority of medical data, there should be little confusion as to whether or not guidelines
142
Pancreatology 2003;3:139–143
represent ‘standard-of-care’ mandates. In our desire to educate the members of our profession to ‘best current practice’, we cannot permit the Law to equate the promotion of educational guidelines within the medical profession to the legal concept of ‘standard-of-care’. Medicine entails infinite variability; the Law proscribes variation. Physicians must recognize this basic difference between the professions. Putting aside these general comments regarding guidelines, how, then, should we view the proposed ‘IAP Guidelines for the Surgical Management of Acute Pancreatitis’? In one sense, since 10 of the 11 proposals made by the authors are based upon suggestive data (level B) rather than proven studies (level A), these guidelines merely represent the opinions of the authors as to the best practices at this time. They are also subject to change in the future as definitive evidence becomes available. Do the IAP Guidelines break any new ground? Of the 11 recommendations proposed by the IAP, 10 were initially proposed by the members of the Atlanta Symposium in their prior publication on therapy [1]. The eleventh, ‘Endoscopic sphincterotomy is an alternative to cholecystectomy in those who are not fit to undergo surgery, in order to lower the risk of recurrence of gallstoneassociated acute pancreatitis’ has been previously recommended in other guidelines [6]. What can we say of the specific IAP Guidelines themselves? Although most pancreatologists would find themselves in agreement with these proposals as re-stated by the IAP, the sections dealing with sterile pancreatic necrosis seem to contain some unnecessarily vague language. With regard to Recommendation #5, the ‘selected cases’ of sterile pancreatic necrosis that should be subjected to surgical intervention, rather than persisting in conservative management, are not otherwise characterized. Nor are the ‘specific recommendations’ further identified for which abandonment of conservative management of sterile necrosis may be necessary within the first 14 days after onset. In fairness to the authors, these areas have not been well investigated, and the absence of more specific recommendations may not only reflect the inadequacy of our current knowledge, but also a natural reluctance on the part of the authors to include Expert Opinions (level C) in their overall proposals. Still, the authors have not been reluctant to include Clinical Trials evidence (evidence level B) as support for their proposals, evidence which may be only somewhat less likely to be overturned in the future. Level C information does exist to suggest that small subpopulations of patients with sterile necrosis may be
Bradley
surgical candidates; information not previously set forth in prior attempts at creation of guidelines for acute pancreatitis. Some patients recovering from necrotizing pancreatitis will develop recurrent abdominal pain and pancreatic inflammation whenever attempts are made at restoring oral alimentation. This is usually seen 4 or more weeks following recovery from the initial episode of necrotizing pancreatitis, and has been variously referred to as ‘persistent pancreatitis’ [12], ‘re-feeding pancreatitis’ [13], or ‘disconnected duct syndrome’ [14]. The underlying pathophysiology seems to be either obstruction of the pancreatic duct by the necrotic process [14] or leakage of pancreatic juice from a necrotic duct into an enclosed section of the retroperitoneal space [15]. ERCP or MRI studies will usually detect the interruption or leakage from the duct. Surgical removal of the involved portion of duct is most often curative. Even in the most favorable light, however, this indication for surgery would not rise above the level of Expert Opinion, and further study targeting this recommendation is necessary. Finally, while the overall success of conservative management for sterile necrotizing pancreatitis has been demonstrated beyond reasonable doubt, some patients will continue to deteriorate and die despite appropriate and enthusiastic resuscitative measures. Whether or not patients with sterile pancreatic necrosis and progressive organ deterioration can be salvaged by surgical debride-
ment is not known. Currently, no credible evidence exists to suggest that surgical debridement of sterile necrosis prevents, eliminates, or even ameliorates co-existing organ failure. However, false-negative fine-needle aspirations of infected pancreatic necrosis have been described, and while quite rare (!5%), can result in potentially preventable death. Nevertheless, any decision to offer surgical exploration to a progressively deteriorating patient with FNA aspirate ‘proved’ sterile pancreatic necrosis, in the hope of finding cryptic infected necrosis, should include a frank discussion of the 20:1 odds against finding infected necrosis at surgical exploration, and the real possibility that exploration could accelerate the patient’s decline. In summary, one could reasonably propose that because there have been few guidelines for acute pancreatitis over the past decade that have actually broken new ground, it may be time to call for a moratorium on new sets of guidelines that do not result in significant advances in clinical practice. Rather than re-stating or fine-tuning previous proposals, future efforts might be more profitably expended toward raising the level of proof for existing guidelines. The IAP counts among its members some of the brightest and best workers in pancreatic disease. With direction from its officers and co-operation among its members, randomized controlled trials designed to definitively address less desirable levels of evidence could be conducted. The need is both clear and urgent.
References 1 Bradley EL III (ed): Acute Pancreatitis: Diagnosis and Therapy. New York, Raven Press, 1994. 2 Bradley EL III: A clinically based classification system for acute pancreatitis: Summary of the international symposium on acute pancreatitis, Atlanta, Ga., September 11 through 13, 1992. Arch Surg 1994;128:586–590. 3 United Kingdom Guidelines for the Management of Acute Pancreatitis. Gut 1998;42(suppl 2):1S–13S. 4 SSAT Patient Care Guidelines: Treatment of acute pancreatitis. J Gastrointest Surg 1998;2: 487–488. 5 Dervenis C, Johnson CD, Bassi C, Bradley EL III, Imrie CW, McMahon MJ, Modlin I: Diagnosis, objective assessment of severity, and management of acute pancreatitis: Santorini consensus conference. Int J Pancreatol 1999; 25:195–210. 6 Toouli J, Brooke-Smith M, Bassi C, Carr-Locke D, Telford J, Freeney P, Imrie C, Tandon R: Working party report: Guidelines for the management of acute pancreatitis. J Gastroenterol Hepatol 2002;17(suppl):S15–S39.
Guiding the Reluctant
7 Mayumi T, Ura H, Arata S, Kitamura N, Kiriyama I, Shibuya K, Sekimoto M, Nago N, Hirota M, Yoshida M, Ito Y, Hirata K, Takada T: Evidence-based clinical practice guidelines for acute pancreatitis: Proposals. J Hepatobiliary Panc Surg 2002;9:413–422. 8 Uhl W, Warshaw A, Imrie C, Bassi C, McKay CJ, Lankisch PG, Carter R, DiMagno E, Banks PA, Whitcomb DC, Dervenis C, Ulrich CD, Satake K, Ghaneh P, Hartwig W, Werner J, McEntee G, Neoptolemos JP, Büchler MW: IAP guidelines for the surgical management of acute pancreatitis. Pancreatology 2002;2:565– 573. 9 Norton SA, Cheruvu CV, Collins J, Dix FP, Eyre-Brook IA: An assessment of clinical guidelines for the management of acute pancreatitis. Ann R Coll Surg Engl 2001;83:399–405. 10 Barnard J, Siriwardena AK: Variations in implementation of current national guidelines for the treatment of acute pancreatitis: Implications for acute surgical service provision. Ann R Coll Surg Eng 2002;84:79–81.
11 Aly EA, Milne R, Johnson CD: Non-compliance with national guidelines in the management of acute pancreatitis in the United Kingdom. Dig Surg 2002;19:192–198. 12 Rutledge PL, Warshaw AL: Persistent acute pancreatitis: A variant treated by pancreatoduodenectomy. Arch Surg 1988;123:597–600. 13 Bradley EL III, Balthazar EJ: Pancreatic Necrosis, chap 33; in: Hepatobiliary and Pancreatic Disease: The Team Approach to Management. New York, Little, Brown, 1995, pp 366–375. 14 Howard TJ, Rhodes GJ, Selzer DJ, Sherman S, Fogel E, Lehman GA: Roux-en-Y drainage is the best surgical option to treat patients with disconnected duct syndrome after severe acute pancreatitis. Surgery 2001;130:714–721. 15 Uomo G, Visconti M, Manes G, Calise F, Laccetti M, Rabbitti PG: Nonsurgical treatment of acute necrotizing pancreatitis. Pancreas 1996; 12:142–148.
Pancreatology 2003;3:139–143
143