BACKGROUND: Minimally invasive procedures to treat infected necrotizing pancreatitis (INP) are gaining popularity. The proportion of patients suitable for this approach remains unknown. METHODS: Preoperative computed tomography (CT) scans were reviewed from 106 consecutive patients who had surgery for INP between 2000 and 2003 in 11 Dutch hospitals. Collections related to the pancreas were classified according to their distance from the left abdominal wall. Five radiologists judged 'accessibility' for drain placement and the likelihood that there was a fluid component that would drain ('drainability'). Agreement between radiologists was determined. RESULTS: CT scans of 80 (75 per cent) patients were available (59 men; age range 29-80 years). The median interval between hospital admission and preoperative CT scan was 20 days. In 55 (69 per cent) patients, the lateral border of the collection was less than 5 cm from the left abdominal wall. Placement of a drain was deemed feasible in 67 (84 (range 77-89) per cent) patients; mean(s.d.) kappa 0.428(0.096). In 45 (56 per cent) patients, a drain could be placed through the left retroperitoneum. In 43 (54 (range 49-82) per cent) patients, collections were judged to contain a drainable fluid component. Interobserver agreement on 'drainability' was poor, mean(s.d.) kappa 0.289(0.101). CONCLUSION: Most peripancreatic collections in INP were considered accessible to a minimally invasive approach.
BACKGROUND: Minimally invasive procedures to treat infected necrotizing pancreatitis (INP) are gaining popularity. The proportion of patients suitable for this approach remains unknown. METHODS: Preoperative computed tomography (CT) scans were reviewed from 106 consecutive patients who had surgery for INP between 2000 and 2003 in 11 Dutch hospitals. Collections related to the pancreas were classified according to their distance from the left abdominal wall. Five radiologists judged 'accessibility' for drain placement and the likelihood that there was a fluid component that would drain ('drainability'). Agreement between radiologists was determined. RESULTS: CT scans of 80 (75 per cent) patients were available (59 men; age range 29-80 years). The median interval between hospital admission and preoperative CT scan was 20 days. In 55 (69 per cent) patients, the lateral border of the collection was less than 5 cm from the left abdominal wall. Placement of a drain was deemed feasible in 67 (84 (range 77-89) per cent) patients; mean(s.d.) kappa 0.428(0.096). In 45 (56 per cent) patients, a drain could be placed through the left retroperitoneum. In 43 (54 (range 49-82) per cent) patients, collections were judged to contain a drainable fluid component. Interobserver agreement on 'drainability' was poor, mean(s.d.) kappa 0.289(0.101). CONCLUSION: Most peripancreatic collections in INP were considered accessible to a minimally invasive approach.
Authors: Hjalmar C van Santvoort; Marc G Besselink; Thomas L Bollen; Erik Buskens; Bert van Ramshorst; Hein G Gooszen Journal: World J Surg Date: 2007-06-16 Impact factor: 3.352