In reply
We would like to thank Dr Fujita for his comments and support of our data on outcomes of patients with pancreatic fistulas following pancreaticoduodenectomy.1 Regarding the issue of drain placement and maintenance, we routinely place 2 flat, 7-mm Jackson-Pratt drains, which are kept to continuous bulb suction. We placed 1 behind the pancreatic and biliary anastomosis and the other in front. These drains are maintained with daily “stripping” and documentation of the nature and consistency of the fluid (eg, serous, milky, cloudy, purulent, bloody). Once the output slows, they are removed. Unfortunately, it is challenging to know whether the output has slowed because the anastomosis is no longer leaking or because the drain is either occluded or malpositioned. The latter option is the likely explanation for what we have termed occult fistulas. Those patients who develop an abscess or collection from an undrained fistula undergo interventional radiologic-guided drainage. These are typically 10 F pigtail catheters that are flushed daily with 5 to 10 mL of saline.