How can I be sure that the patient has acute pancreatitis?

Clinicians often have difficulty managing patients with acute pancreatitis because the disease is complicated by an obscure pathogenesis, numerous causes, few effective remedies, and an often unpredictable outcome. The diagnosis of acute pancreatitis is typically established by two of the following three features: (1) appropriate clinical symptoms, such as epigastric pain, nausea, and vomiting; (2) an elevation of the amylase and/or lipase at more than 3 times the upper limit of normal; and/or (3) imaging confirmation of the diagnosis, computed tomography (CT), or magnetic resonance imaging (MRI). However, the diagnosis can often be difficult when these criteria are not met clearly (see following discussion).

A tabular or chart listing of features and signs and symptoms

What are the problems in establishing the diagnosis of acute pancreatitis?

The symptoms of acute pancreatitis may vary; however, it is almost uniformly described as a burning, stabbing pain, often radiating to the back, that may or may not be accompanied by nausea and/or vomiting. The pain of acute pancreatitis is not diagnostic and resembles the pain felt in many other diseases affecting the upper abdomen, including dissecting aneurysms, cholecystitis, perforated viscous such as peptic ulcer disease, and severe enteritis caused by food poisoning. All of these disorders can raise the amylase and/or lipase, although rarely above 3 to 5 times the upper limit of normal.

Neither amylase nor lipase are specific to the pancreas, and, for this reason, elevations of these enzymes in the absence of symptoms and/or imaging are not diagnostic of pancreatitis. Amylase and Lipase are found in numerous tissues, including the salivary glands. It has been shown that salivary amylase and lipase can rise in the blood when inflammation of the bowel occurs due to increased absorption from the digestive tract. Any inflammatory process in the abdomen (e.g., peptic ulcer disease, infectious enteritis, cholecystitis, diverticulitis) can affect the pancreas. This inflammatory effect can result in a mild two- to threefold elevation of the amylase and lipase as the inflammatory event in the abdomen leads to some pancreatic release.

Continue Reading

Also, the amylase and lipase are cleared through the kidneys; thus, renal insufficiency often results in elevations of amylase and lipase in the absence of acute pancreatitis. In addition, some asymptomatic patients have large molecular forms of amylase (macroamylasemia) and, less common, lipase, resulting in higher levels in the blood (sometimes >3 times normal) due to decreased renal clearance.

A lipase level is not necessarily required to establish the diagnosis of acute pancreatitis when the amylase is more than 3 times the upper limit of normal in the appropriate clinical setting. However, certain conditions, such as alcoholic pancreatitis and hypertriglyceridemia, may limit the “rise” of serum amylase. In these patients, a serum lipase offers a higher predictive value to the diagnosis.

How can I confirm the diagnosis?

Transabdominal ultrasound, although exquisitely sensitive for the presence of gallstones, has little role in the evaluation of the pancreas in acute pancreatitis. Most often, the ultrasound demonstrates air from the stomach, colon, and duodenum blocking the pancreas from examination. With its capacity to shield patients from radiation exposure, MRI has been shown to be equal to CT in the evaluation of acute pancreatitis. MRI has the advantage through MRCP (magnetic resonance cholangiopancreatography) to visualize the pancreatic-biliary ducts. At present, MRCP is reserved for patients who later in the course of acute pancreatitis undergo a more thorough evaluation for retained common bile duct stones and/or pancreatic tumors. Moreover, due to limitations in MRI-interpretation expertise in the setting of acute pancreatitis, CT remains the gold standard for imaging.

Due to the location of the body and tail of the pancreas extending to the splenorenal ligament and the large amounts of peritoneal fat in this region, radiologists often comment that they see “focal pancreatitis” in the pancreatic tail or body. This observation is inaccurate because acute pancreatitis is a diffuse inflammatory disease. The radiologist is typically commenting on the finding of peritoneal fat stranding, suggestive of pancreatitis, most commonly seen off the pancreatic body and tail. The head of the pancreas typically sits with little fat around it and, thus, when inflamed in the setting of acute pancreatitis, may appear normal merely due to the lack of peripancreatic fat in this area.

Intravenous contrast is needed to determine the presence of necrosis. A dynamic spiral CT with a flash phase of intravenous contrast best identifies necrotic pancreatic tissue because the lack of perfusion results in decreased attenuation of these areas. Although intravenous contrast is not needed to establish the diagnosis of acute pancreatitis, it is required in determining the presence of necrosis. Although acute pancreatitis is a diffusely inflammatory process, the complications of acute pancreatitis, such as pancreatic necrosis and/or pseudocysts, can be localized to the head, body, and/or tail of the pancreas.

If neither the serum amylase nor lipase is conclusive or if the clinical setting is unclear, a noncontrast CT should be performed. CT remains the gold standard in imaging to define the presence of acute pancreatitis. As with all inflammatory events in the abdomen, a noncontrast CT can easily identify any area of inflammation through the identification of fluid and/or mesenteric stranding adjacent to the inflamed organ.

What other diseases, conditions, or complications should I look for in patients with acute pancreatitis?

What complications can occur early in the management of patients with acute pancreatitis?

The problem of distinguishing severe and mild disease in patients: The management of acute pancreatitis is complicated by the inability to distinguish patients with mild disease from patients with severe disease during the early stages. Most patients who have complicated courses and who even eventually die from acute pancreatitis initially present with what appears to be mild disease, characterized by the absence of organ failure and/or pancreatic necrosis. It is imperative that clinicians do not label a patient with mild disease within the first 48 hours of admission. This is a common problem leading to substantial morbidity and mortality as patients’ are often left unmonitored after being falsely labeled as having mild disease.

Historically, prospective scoring systems that use clinical criteria have been developed to determine severity in patients with acute pancreatitis. These systems include the Ranson criteria, Imrie/Glasgow criteria, and APACHE score. Unfortunately, these systems are cumbersome because they require multiple measurements. More important, the systems are not accurate until 48 hours after presentation.

Severity is better defined by the Atlanta Symposium. The Atlanta Symposium utilizes the outcome of disease as the determining factor for severity, pancreatic necrosis, and/or organ failure, such as cardiovascular, pulmonary, and renal insufficiency and/or gastrointestinal bleeding. The Atlanta Symposium does not attempt to predict severity but rather defines severity as the presence of the factors known to be associated with mortality.

The Atlanta Symposium has its limitations. Over the last decade, it has become clear that organ failure is a more important determinant of severity than pancreatic necrosis. In the setting of multisystem organ failure, the mortality rate of patients with acute pancreatitis is almost 50%. Conversely, regardless of the presence of necrosis, in the absence of organ failure, mortality is 0%.

Many patients with pancreatic necrosis do well and do not develop significant organ failure. There is no clear relationship between the presence of organ failure and pancreatic necrosis. Furthermore, the extent of pancreatic necrosis and infection of the necrosis do not appear to correlate with the risk of developing organ failure. Some investigators have proposed that patients with pancreatic necrosis, in the absence of organ failure, who merely have prolonged hospital courses be labeled as having “moderately severe acute pancreatitis.” There also is a difference between patients who develop transient organ failure and those with persistent organ failure.

It is the author’s understanding that the Atlanta symposium is being revised to reflect these findings.

Predicting severe disease: It is imperative to differentiate the concept of “predicting” severe disease from “establishing” the presence of severe disease. Severe disease is not difficult to establish: the presence of organ failure, pancreatic necrosis, and/or death.

The more difficult issue is: Can clinicians accurately “predict” whether a patient with mild disease at admission will progress to severe disease? The answer is no! To date, there have been no studies establishing the superiority of one scoring system, imaging, or laboratory parameter with the ability to predict the future development of severe disease (i.e., organ failure, pancreatic necrosis, and/or death) in patients with acute pancreatitis early in the course of the disease. In general, clinicians should focus on a combination of patient characteristics that increase the likelihood of a patient’s developing severe disease. These characteristics include physical findings, laboratory parameters, scoring systems, and imaging.

A variety of studies have shown that patients at risk of developing severe disease are

– of older age (>55 years)

– obese (BMI >30)

– often present with their first attack of acute pancreatitis and have many comorbid conditions

– Many are found to have pleural effusion and/or infiltrates on initial chest radiography.

The presence of systemic inflammatory response syndrome (SIRS) also has been shown to signify a risk for severe disease. SIRS is defined by the presence of two or more of the following criteria:

  • – pulse: >90 beats/min
  • – rectal temperature: <36º C or >38º C
  • – white blood count: <4000 or >12,000 per mm3
  • – respirations: >20/min or PCO2 <32 mm Hg

Many single laboratory tests have been studied as markers of severity, with little success. The height of elevation of the serum amylase and lipase do not correlate with severity.

It has been shown that hematocrit (HCT), blood urea nitrogen (BUN), and creatinine may serve as early predictors of severity in patients with acute pancreatitis. Unlike other markers of severity studied, such as C-reactive protein, the HCT, BUN, and creatinine are not surrogate markers of inflammation but reflect complications in pancreatic perfusion (discussed as follows).

A hematocrit of more than 44% on admission, or that fails to decrease over the first 24 hours, is associated with pancreatic necrosis. Similarly, an elevated or a rise in the BUN within the first 24 hours after admission is associated with increased mortality in patients with acute pancreatitis. Similar to BUN and HCT, an elevated creatinine, as a marker of decreased intravascular volume, and decreased renal perfusion has been shown to be associated with severe disease (pancreatic necrosis).

On occasion, imaging may be effective in identifying patients with severe disease early in the course of acute pancreatitis. Contrast-enhanced CT and MRI have been shown to be sensitive for the identification of pancreatic necrosis.

The use of early imaging in the determination of severity is limited by several important factors:

  • Only a quarter of patients with acute pancreatitis develop necrosis;
  • Pancreatic necrosis may not develop until after 24 to 48 hours; and
  • The presence of pancreatic necrosis and the amount of pancreatic necrosis do not correlate with the development of organ failure.

Patients with one or more of these characteristics may require treatment in a highly supervised area, such as a step-down or an intensive care unit (ICU). At present, regardless of risk factors, predictors, and scores on the variety of systems, all patients should be monitored closely for the development of organ failure. This monitoring does not require an ICU, unless there are obvious signs or organ failure. However, depending on the institution, an ICU or step-down unit may be necessary to provide the monitoring that patients need during the first 48 hours until it becomes clear that they will have uncomplicated, mild outcomes. If organ failure (e.g., renal, cardiovascular, or pulmonary insufficiency) develops, aggressive treatment in an ICU has been associated with a survival benefit if the organ failure is reversed.

Increased morbidity and mortality is associated with persistent organ failure lasting beyond the first 48 hours. Transient organ failure is of less importance in predicting morbidity and mortality.

What is the right therapy for the patient with disease X?


What is the most effective initial therapy?

Preventing severe disease with vigorous intravenous hydration

The most important advancement in the management of patients with acute pancreatitis is the recognition of the importance of early aggressive hydration. More than a decade ago, it was shown that an elevated hematocrit is associated with increased morbidity and pancreatic necrosis. It is now understood that the relationship of hematocrit to severity is related to hemoconcentration. As the inflammatory process progresses early in the course of the disease, there is an extravasation of protein-rich intravascular fluid into the peritoneal cavity, resulting in hemoconcentration.

The intravascular hypovolemia that accompanies acute pancreatitis subsequently leads to a decrease in pancreatic blood flow. Pancreatic ischemia leads to the activation of inflammatory mediators. The decreased blood flow also leads to stasis and thrombi, in turn, leading to subsequent necrosis, which then exacerbates the inflammatory process. The decreased perfusion pressure into the pancreas leads to microcirculatory changes, leading to widespread pancreatic necrosis in some patients. A vicious cycle develops in which pancreatic inflammation leads to more extravasation of protein-rich intravascular fluid into the peritoneum, leading to more necrosis. Currently, the only way to halt this cycle is to provide vigorous intravenous hydration for intravascular volume repletion, hemodilution, increased pancreatic perfusion, and decreased pancreatic necrosis. Unfortunately, despite multiple guidelines explaining the importance of early aggressive intravenous hydration, clinicians too often give suboptimal intravenous hydration to patients with acute pancreatitis.

How much fluid is needed early in the treatment of acute pancreatitis? Acute pancreatitis is a hypercatabolic disease leading to increased respiratory losses in addition to the intravascular losses. If one considers that liters of fluid typically enter the peritoneum from the intravascular space, one gets a better appreciation of the amount of intravenous fluid that is needed for hydration.

One of the markers of severity previously defined by Ranson and colleagues (i.e., Ranson criteria) is related to intravascular losses. Ranson and colleagues found that a sequestration of more than 6 liters of fluids during the first 48 hours was an independent predictor of severity. If this amount is added to the minimal intravenous fluid requirements of a 70-kg person during the first 48 hours, intravenous hydration should be at least 250 cc to 500 cc per hour, depending on BMI (

The rate of hydration appears to be most important during the first 6 to 12 hours. Unfortunately, this period may be when the patient with acute pancreatitis is in the emergency room; in the process of being evaluated by emergency, medical, and surgical staff; and of being transferred. This poses difficulty in providing early aggressive hydration because the point of care may not be clear. Close monitoring of the patient is needed to titrate hydration to the hematocrit and BUN, which should both decrease over the first 6 to12 hours.

Listing of usual initial therapeutic options, including guidelines for use, along with expected result of therapy.

When should I consider early endoscopic retrograde cholangiopancreatography in patients with acute pancreatitis?

Role of urgent ERCP in gallstone pancreatitis: The pathogenesis of gallstone pancreatitis depends on the presence of a common bile duct stone. Although the most common cause of acute pancreatitis, the vast majority of these stones pass easily and quickly from the common bile duct (CBD) to the duodenum to be excreted in the stool. In some patients, gallstones can persist in the CBD and may lead to severe acute pancreatitis, complicated by biliary sepsis. Defining the presence of a persistent CBD stone as the cause of severe, complicated acute pancreatitis can be problematic.

Although considered the gold standard for cholelithiasis, abdominal ultrasonography in the setting of acute pancreatitis is not sensitive for the detection of choledocholithiasis, CBD stones. Gallstones may be present in the CBD and missed due to air in the duodenum. Even in the absence of biliary ductal dilatation on abdominal ultrasound, CBD stones can exist in a patient with gallstone pancreatitis complicated by biliary sepsis.

Laboratory testing may assist in the early identification of CBD stones. Although elevated transaminases have a poor sensitivity for determining gallstone pancreatitis, a high specificity can be reached with laboratory testing. A higher than 3-fold elevation of AST or ALT in the presence of acute pancreatitis has a positive predictive value of 95% in diagnosing gallstones as the etiology of pancreatitis.

On multivariate analysis, serum total bilirubin on hospital day 2 was the best predictor of a persistent CBD stone. A serum total bilirubin level higher than 1.35 mg/dL has a sensitivity of more than 90%. Unfortunately, the specificity for CBD stones is only 63%. Other investigators have found that a rising bilirubin or rising transaminases within 24 to 48 hours of admission in patients with acute pancreatitis predicted a persistent CBD stone.

Regardless of findings on laboratory testing and ultrasonography, ERCP (endoscopic retrograde cholangiopancreatography) remains the gold standard in determining whether gallstones are retained in the CBD. ERCP is safe in a patient with biliary sepsis, especially if the CBD is dilated. However, in the patient with a normal CBD diameter and relatively normal LFTs, the risk of post-ERCP pancreatitis is higher. In these patients, if needed, endoscopic ultrasound and MRI provide excellent visualization of the CBD and can be used to determine the presence of CBD stones with less risk.

The role of early ERCP in the setting of acute pancreatitis continues to evolve. Since the days of Opie, Osler, and Halstead, clinicians have pondered the following question: Will removing a stone impacted at the ampulla affect the course of acute pancreatitis? As most of the stones pass, when should an endoscopist perform ERCP to remove a suspected stone? Early ERCP, within 24 to 72 hours, is effective in patients with severe acute pancreatitis when there is evidence of biliary obstruction, cholangitis, and an elevated bilirubin. There is no evidence that urgent ERCP alters the course in patients with severe acute pancreatitis in the absence of concomitant biliary obstruction by a stone.

A listing of a subset of second-line therapies, including guidelines for choosing and using these salvage therapies


Listing of these, including any guidelines for monitoring side effects.


How should I monitor the patient with acute pancreatitis?

What are the late complications of acute pancreatitis?

Pseudocysts: Although during the early course of acute pancreatitis, liters of fluid typically leave the intravascular compartment, forming fluid collections in the abdomen, most of the fluid is eventually reabsorbed. In almost a third of patients, the fluid will persist and eventually develop a fibrous wall. Early in the course of acute pancreatitis, within the first 2 to 3 weeks, calling a fluid collection a pseudocyst is inappropriate. In a patient, with no medical history of chronic pancreatitis, who presents with acute pancreatitis and is found on admission to have a cystic lesion that appears to be a pseudocyst, the lesion may very likely be a cystic neoplasm rather than a pseudocyst.

Anatomically, another important differentiation that will affect the clinical care of patients with acute pancreatitis is whether the lesion is a pseudocyst or a walled-off pancreatic necrosis. A walled-off fluid collection appears similar to a walled-off pancreatic necrosis on CT. Although cystic fluid in the pancreas has an attenuation on CT that is much lower than that of normal pancreatic tissue, pancreatic necrosis appears similar.

Pancreatic necrosis, especially walled-off pancreatic necrosis (WOPN), cannot be differentiated from a pseudocyst by CT. This is important to keep in mind because the treatment of a symptomatic or infected pseudocyst (abscess) and pancreatic necrosis are quite different. The differentiation between pancreatic necrosis and pseudocysts can be determined only by the location of the cyst in relation to the pancreas. Whereas pseudocysts are located outside the parenchyma of the pancreas or adjacent to it, pancreatic necrosis always involves the pancreas.

MRI and endoscopic ultrasound (EUS) can determine the presence of tissue in WOPN, but these imaging modalities are typically not used.

In the past, the dogma was that only pseudocysts larger than 6 cm, that appear enlarged on serial imaging, or that become symptomatic warrant drainage. Asymptomatic pseudocysts that develop after an attack of acute pancreatitis, regardless of size, can be managed conservatively, with no intervention.

Pseudocysts can become infected and, when this occurs, they are best described as abscesses (and, abscesses require drainage). Pseudocysts can become painful, causing early satiety and weight loss when their size affects the stomach and bowel. When confronted with a patient with a symptomatic pseudocyst, regardless of infection, drainage is recommended.

Drainage can be performed via endoscopic, radiologic, or surgical techniques, depending on the location of the cyst and the expertise available. No randomized prospective trials have compared these methods. The myriad of pseudocyst size, location, and anatomy, and the unavailability of local expertise, make prospective randomized trials difficult.

Pancreatic necrosis: Most patients with acute pancreatitis have interstitial pancreatitis, characterized by normal pancreatic perfusion but accompanied by edema of the pancreas, peripancreatic stranding, and fluid collections (Figure 1). The gross destruction of the pancreatic gland, pancreatic necrosis is seen in 20% of patients with acute pancreatitis. In the absence of autopsy or laparotomy, pancreatic necrosis is defined as more than 30% of nonenhancement on contrast-enhanced CT (
Figure 2). Pancreatic necrosis is an early complication of acute pancreatitis, usually recognized within 4 days of the onset of symptoms. Although pancreatic necrosis can be identified on a CT obtained at admission, necrosis may develop over the next 48 to 72 hours (or later).

Figure 1
Interstitial pancreatitis.

Figure 2
Pancreatic necrosis.

Pancreatic necrosis typically is considered sterile early in the course of disease but can become infected later, and is clinically apparent after 7 to14 days. Both infected pancreatic necrosis and sterile pancreatic necrosis can lead to organ failure, cardiopulmonary insufficiency, renal failure, and gastrointestinal bleeding. Sterile necrosis is treated supportively.

Due to the increased risk of complications, infection, and organ failure, patients with pancreatic necrosis should be cared for in a monitored setting. The goal is to prevent infection by minimizing intravenous lines, avoiding parenteral nutrition, and providing enteral nutrition. Typically, patients should be kept NPO with enteral nutrition for 3 to 4 weeks before an attempt is made to orally feed them. At this time, if the patient is unable to tolerate oral feeding, debridement may be necessary.

How should I manage patients with pancreatic necrosis?
Preventing infected necrosis: the role of antibiotics

Once sterile pancreatic necrosis exists, prevention of infection is of paramount importance. The presence of infected necrosis may necessitate surgical debridement, although there is a general trend to manage these patients with targeted antibiotics if they are clinically stable to allow the necrosis to be demarcated from surrounding tissues and to liquefy. This procedure allows less invasive therapies of drainage and debridement, including endoscopic approaches.

Surgical intervention, if necessary in patients with infected necrosis, increases the morbidity and mortality rate in patients with acute pancreatitis. The surgical management of infected necrosis is an issue that is typically addressed after the first or second week of managing patients with acute pancreatitis. During the first week, the vast majority of patients with necrosis have sterile necrosis. Surgical intervention during this time is avoided.

The origin of the bacteria leading to pancreatic infection is unclear. However, the current consensus is that infection of the necrosis occurs by either direct transmural spread or by the transmigration of bacteria from the colon. Other sources include intravenous lines, including those used for parenteral nutrition.

In an attempt to decrease pancreatic infection, initial trials with ampicillin in the 1970s showed a lack of efficacy when ampicillin was given prophylactically. It has been shown that only a few antibiotics – including imipenem, quinolones, and metronidazole – penetrate pancreatic necrosis. Subsequently, a prospective, randomized trial comparing imipenem to placebo in the prevention of infected necrosis showed a significant decrease in septic complications. This study was followed by several other trials demonstrating decreased morbidity and mortality in patients with necrotizing pancreatitis who are treated with antibiotics within 72 hours of admission.

Multiple reviews, including a Cochrane review in 2004, concluded that pancreatic-penetrating antibiotics were useful in patients with necrotizing pancreatitis. Based on these initial unblinded studies, most clinicians began the widespread use of antibiotics in patients with necrotizing pancreatitis with the belief that infectious necrosis would be avoided.

Several new, larger, multicenter, randomized double-blinded trials have changed our opinion regarding the use of antibiotics in sterile necrosis. Dellinger and colleagues (2007) performed a double-blinded, placebo-controlled randomized study set in 32 centers in North America and Europe. One hundred patients were equally randomized into two groups: meropenem (1 g intravenously every 8 hours) or placebo within 5 days on the onset of symptoms. The medication was continued for 7 to 21 days. This study demonstrated no statistically significant difference between the treatment groups for pancreatic or peripancreatic infection, mortality, or requirement for surgical intervention. Based on these two latest studies, in the absence of biliary sepsis or obvious pancreatic and peripancreatic infection, the routine use of antibiotics to prevent infection of pancreatic necrosis is not warranted.

Enteral versus parenteral nutrition in severe acute pancreatitis

The physical stress of acute pancreatitis leads to a catabolic state that promotes nutritional deterioration in the setting of a systemic inflammatory response. Adequate supply of nutrients may play an important role early in the management of patients. However, the early use of total parenteral nutrition (TPN) in patients with acute pancreatitis has not been shown to be beneficial. TPN requires a break in the mucosal barrier for delivery, leading to an increased incidence of infection. Several early studies found that enteral nutrition will reduce septic morbidity in conditions such as trauma and thermal injury. Early enteral nutrition through a nasojejunal tube maintains the integrity and function of the intestinal barrier while providing adequate nutrition.

In patients with severe acute pancreatitis, enteral feeding is safe and as effective as parenteral nutrition. Enteral nutrition attenuates the acute inflammatory response and improves disease severity in acute pancreatitis. Several randomized prospective studies comparing nasojejunal and parenteral nutrition have shown a decrease in morbidity and mortality in patients given enteral nutrition early in the course of disease. By providing nutrients and altering the bacterial flora, a significant decrease in the development of infected pancreatic necrosis was observed. There was a consensus among clinicians involved in the trials that the trials demonstracted decreased infectious complications, length of stay, significant cost savings, and decreased mortality.

In patients with acute pancreatitis, the use of enteral nutrition has been delayed by the old belief that pancreatic rest is required to prevent complications. This reasoning appears untrue. Although the nasojejunal route has been used in several trials, a nasogastric route may also be safe. In a comparison of patients with severe acute pancreatitis randomized to nasojejunal feeding versus naso-gastric feeding, there was no apparent difference in safety, pain score, narcotic requirements, morbidity, and mortality. Although further study is needed regarding the timing of initiating enteral nutrition in patients with severe acute pancreatitis, it is clear that enteral nutrition is preferred to parenteral nutrition. Enteral nutrition should utilize an elemental or semi-elemental formula.

Infected necrosis

Approximately one-third of patients with necrotizing pancreatitis develop infected necrosis. The infection usually becomes clinically apparent after 10 to 14 days of illness. Most patients with infected necrosis have systemic toxicity, such as fever and leukocytosis. Almost half of the patients with infected necrosis have persistent organ failure. In a patient suspected of having infection of pancreatic necrosis, a CT-guided, fine needle aspiration of the pancreatic or peripancreatic bed should be done. The gram stain alone of this aspirate, if handled correctly, often (approaching 95% sensitivity) will establish the presence or absence of infection.

In the absence of infection, sterile necrosis is treated conservatively. Infected necrosis is treated by targeting microbes with pancreatic-penetrating antibiotics, such as carbapenems, quinolones in combination with metronidazole, or high-dose cephalosporins. If a patient cannot undergo CT-guided aspiration and infection is strongly suspected, antibiotics should be started (not for the purpose of preventing infection but for the purpose of treating strongly suspected infection). Drainage – whether surgically, endoscopically, or radiologically – should be timed based on the patient’s condition.

In the past, the diagnosis of infected necrosis led to prompt surgical intervention. This had been the standard of care until recently. In a retrospective review of 53 patients whose median time to surgery was 28 days, mortality decreased 22% when necrosectomy for infected necrosis was delayed. After reviewing 11 studies that included 1136 patients, the author also found a significant correlation between the timing of surgery and mortality. It appears that postponing necrosectomy in stable patients with antibiotics until 30 days after initial hospital admission is associated with decreased mortality.

The concept that infected pancreatic necrosis requires prompt surgical debridement has also been challenged by multiple reports and case series showing that antibiotics alone can lead to resolution of infection and, in select patients, the avoidance of surgery altogether. In one report, of 28 patients given antibiotics for the management of infected pancreatic necrosis, 16 patients avoided surgery. There were two deaths among the patients who underwent surgery and two deaths among the patients who were treated with antibiotics alone. Thus, in this report, more than half of the patients were successfully treated with antibiotics, and the mortality rate in both the surgical and nonsurgical group was similar.

The current consensus is that the initial management of infected necrosis for clinically stable patients should be a 2- to 4-week course of antibiotics prior to surgery to allow the inflammatory reaction to become better managed. At this time, the disease will present with a well-defined wall and liquefied material within it; the decision and method of drainage can then be considered.

For patients with symptomatic WOPN, a combined multimodality approach that bringings together both minimally invasive surgery with endoscopic drainage appears to be more effective, safer, and results in a shorter hospitalization. Although further study is needed, the concept that urgent surgery is required in patients found to have infected necrosis is no longer valid.

How should I monitor patient with acute pancreatitis?

Mostpatients tolerate aggressive hydration well. Patients with renal, pulmonary, and cardiac disease may require more focused care in an ICU. Although uncommon, volume overload must also be identified early. If apatient develops shortness of breath secondary to acute respiratory distress syndrome (ARDS) rather than volume overload, the hydration rate should be continued and intubation with mechanical ventilation should be applied as needed. When it is unclear whether respiratory compromise is secondary to ARDS or cardiac disease (congestive heart failure), intracardiac pressure measurements and/or a bed side echocardiogram should be obtained to guide resuscitation.

What’s the evidence?

Banks, PA, Freeman, ML. “Practice guidelines in acute pancreatitis”. Am J Gastroenterol. vol. 101. 2006. pp. 2379-400.

Forsmark, CE, Baillie, J. “AGA Institute technical review on acute pancreatitis”. Gastroenterology. vol. 132. 2007. pp. 2022-44.

Tenner, S. “Initial management of acute pancreatitis: critical decisions during the first 72 hours”. Am J Gastroenterol. vol. 99. 2004. pp. 2489-94.

Wu, B, Johannes, R, Sun, X, Conwell, D, Banks, P. “Early changes in blood urea nitrogen predict mortality in acute pancreatitis”. Gastroenterology. vol. 137. 2009. pp. 129-35.

Gardner, T, Vege, SS, Pearson, R, Chari, S. “Fluid resuscitation in acute pancreatitis”. Clin Gastroenterol Hepatol. vol. 6. 2008. pp. 1070-6.

Tenner, S. “Fluid resuscitation in severe acute pancreatitis”. Clin Gastroenterol Hepatol. 2010. pp. 16-19.

Fan, ST, Lai, EC, Mok, FP. “Early treatment of acute biliary pancreatitis by endoscopic papillotomy”. New Engl J Med. vol. 328. 1993. pp. 228-32.

Dellinger, EP, Tellado, JM, Soto, NE. “Early antibiotic treatment for severe acute necrotizing pancreatitis: a randomized, double blind, placebo controlled study”. Ann Surg. vol. 245. 2007. pp. 674-83.

Eatock, FC, Chong, P, Menezes, N. “Nasogastric feeding in severe acute pancreatitis may be practical and safe”. Am J Gastroenterol. vol. 100. 2005. pp. 432-9.

Vege, SS, Chari, ST, Clain, JE. “Severe acute pancreatitis”. JAMA. vol. 291. 2004. pp. 2865-8.