Prognostic markers are listed in T§ble...5..
Table 5 Single prognostic factors for early detection of severity and prediction of early/late complications
Even experienced clinicians failed to identify on hospital admission 60 to 70 per cent of cases classified as severe episodes. Clinical indicators of severity, irrespective of the bacteriological status of necrosis, are based on impending remote organ failures and on the extent of retroperitoneal inflammation (abdominal mass, guarding and rebound tenderness, Cullen's sign (periumbilical staining), and the Grey-Turner sign (flank staining)). After 48 h, diagnostic accuracy rises to 80 per cent and yields similar results to scoring systems of prognostic criteria.
As 'prune-juice'-colored ascitic fluid is present in over half the patients with necrotizing pancreatitis who undergo early surgery, peritoneal fluid obtained by abdominal paracentesis may be sampled. By considering the smell, volume, and color of free ascitic fluid or the color of lavage fluid using a standard color chart, 90 per cent of eventual fatalities and 72 per cent of severe cases can be identified within 5 h of admission. In addition to its greater speed and availability, additional benefits of diagnostic peritoneal lavage are the correction of an erroneous clinical diagnosis of pancreatitis and selection of the patient for therapeutic lavage. There is an 0.8 per cent risk of visceral puncture and a lower sensitivity when compared with multiple laboratory criteria, particularly in patients with a gallstone etiology. Since it is based on early data, monitoring of the course of pancreatitis is impossible and accurate prediction of late complications poor.
Hypoxemia (PaO2 < 8 kPa (60 mmHg)) is a marker of severe attack and therapeutic need. Other useful biochemical indicators of severity include antiproteases (a2-macroglobulin and a1-antiproteinase), C-reactive protein, granulocyte elastase, phospholipase A 2, and trypsinogen activation peptides ( Gross efa/ 1993).
CT severity scoring systems correlate with the Ranson score, the length of hospital stay, the development of serious complications, and mortality ( Bajthazar eta/ 19.90). Although patients without regional necrosis had no mortality and a 6 per cent complication rate, mortality and morbidity rose to 29 per cent and 94 per cent respectively in those with more than 30 per cent necrosis. Similarly, patients with one or more peripancreatic fluid collections had a 14 per cent mortality and 54 per cent morbidity. Importantly, morbidity rose to 80 per cent in those with both necrosis and fluid collection visualized on initial CT. The likelihood that infected pancreatic necrosis or pancreatic abscess will develop is highest in patients with associated necrosis and fluid collections. As late development of necrosis may occasionally be demonstrated, this raises the question of optimal timing of the initial CT and justifies follow-up CT examinations in those with a protracted course. However, considerable variance between a mild clinical course and initial CT appearance exists in many patients, and so therapeutic implications are not straightforward.
Was this article helpful?