Emergency Room Information For Bariatric Surgery Patients
***Note: Please first read General Principles which immediately follows this introduction.***
The following information is for the use of Emergency Room Physicians, Surgeons, and others seeing our Roux-en-Y gastric bypass (RNY GBP) patients on an urgent basis.
Though the vast majority of patients generally do well after surgery, some will present with serious medical or surgical problems, and even problems functional in origin may ultimately have grave consequences. Abdominal pain, nausea, vomiting, and dehydration occur with a certain frequency following Bariatric Surgery, most often in the first month or two after surgery. Specific approaches to these and other problems will be listed below.
After reviewing our section on General Principles, which should be taken into consideration first with any Bariatric Patient, please feel free to skim down to the following sections as they apply to your patient.
Adjustable Gastric Band related problems of the Band Too Tight, Slip, Port Infection, Band Erosion
Michael A. Todd, MD, FACS
Please address these issues first:
2 or More Liters of IV Fluids - If IV hydration is required, please give at least 2 liters, usually of LR. This will meet immediate needs, and usually will leave a reserve to sustain the patient while they resume oral hydration at home.
Laboaratories - CBC and CMP will usually cover most issues involved with weight loss surgery. Consider an amylase and lipase when evaluating patients with severe abdominal pain.
Imaging Studies - An abdominal/pelvic CT will usually the best study to help sort out Bariatric postoperative issues, such as leak, perforation, abscess, obstruction, or stones. Plain films are usually of limited value and are hard to interpret. Upper GI can help with perforation or leak issues, or obstruction, but does not yield information on the status of the bypassed stomach and small bowel, and can leave contrast that confounds later studies such as CT. Abdominal Ultrasound is helpful for gallstone patients, but this diagnosis can usually be made with a CT, which also helpful to evaluate other disease entities.
Vitamin Assessment/Cocktail – Vitamin depletion easily occurs in Bariatric patients for a variety of reasons, from noncompliance to lack of formulation efficacy, and may be difficult to ascertain clinically. Please assess whether the patient has been keeping up with their vitamin and mineral supplements. Unless you can be convinced that they are doing so on a regular basis, and their ER complaints are mild and they do not require an IV, please request the following laboratory assays: B12, B1, Vit D, Folate, and PTH, and then administer 1 amp MVI, 100 mg Thiamine, 2 mg Folate IV, and give 1000 mcg B12 IM.
Hypokalemia - Bariatric patients often develop an inexplicable hypokalemia in the first several months after surgery, and for longer periods in some patients. Treat with IV and oral supplementation as needed in the ER. Outpatients are usually well treated with 10 meq KCL capsules, # 100, 1 bid for repletion, and 1 qd for maintenance. The capsules can also be opened and the granules sprinkled on food.
Ask About NSAID Use - Ask patients if they are taking NSAID’s. Patients taking NSAID’s other than Tylenol may present with dramatic abdominal pain nonsurgical abdominal pain that is well treated with a proton pump inhibitors and Carafate. Paitents taking large amounts of Tylenol of course can have abnormal liver function tests.
Notify the Bariatric Surgeon/Call for Consultation - The task of first care physicians evaluating our Bariatric patients in the ER or urgent care setting in large part involves distinguishing serious issues from the otherwise simple but symptomatic. Our Bariatric Center staff can often help streamline an otherwise confusing and lengthy visit. Please call us at (907) 929-4263 about any of our patients seen in an emergency room or urgent care, please send us copies of your notes, and please refer our patients back to us for further care. Please also feel free to call us for advice on Bariatric patients who similarly present, but have had their surgery elsewhere. .
Vomiting Reflex - A frequent problem that troubles almost all of our patients sometime after surgery. Patients vomit after eating something that didn’t agree with them, after eating too much, or after eating too quickly. Once they start vomiting they may have difficulty stopping. Patients with Vomiting Reflex can present with sharp, aching epigastric pain, dehydration, and near emotional collapse. These symptoms may last hours to days, and is aggravated by attempts at taking anything by mouth, especially solid foods. Emesis usually consists of whatever has been ingested, or oral secretions which can have a whitish & foamy character.
Treatment includes reassurance, judicious anti-emetics (Phenergan 25 mg po or pr q4-6h, Zofran orally disintegrating tabs 4mg sl q 6h), and gentle attempts at po, starting with small quantities of simple comfort foods that the individual patient tolerates best, such as water, sugar free popcicles, tea, broth, and crackers. Patients may have to stay on this "Beginning" diet for several days, and should continue it for a day after symptoms resolve.
Patients referred to the ER for IV hydration are best treated with at least 2 liters of fluid, vitamin assessment and cocktail, and Phenergan IV or IM. Encourage patients to try to control these episodes on their own by understanding that they do occur from time to time and are usually self limited, and by otherwise avoiding irritating foods or eating behaviors, and by learning to switch to the "Beginning" diet at the first signs of trouble.
Dehydration - Dehydration is a common problem in Bariatric patients, especially in the first month following surgery. It may occur without vomiting in patients having difficulty with the dietary and emotional changes following the surgery. Encouraging them to resume po is sufficient treatment in many patients, including having drink a bottle of water in the exam room while visiting with sympathetic staff. Patients may require IV hydration, best done with at least 2 liters of IV fluid, vitamin assessment and cocktail, and CBC and metabolic panels drawn to rule out electrolyte disturbances and as part of the general assessment.
Food Impaction -This actually occurs much less frequently that expected. Patients present with emesis, pain, and with relatively sudden inability to keep anything down. Patients should be asked if they have tried to take meats or other solid foods in the first days or weeks after surgery, when they should be on only liquids or pureed, but food impaction can occur at any time after surgery. Emesis may be persistent and dark colored, stained by the impacted material. Initial evaluation includes UGI, which will reveal a complete obstruction at the gastrojejunostomy. Treatment includes referral for endoscopic disimpaction versus a trial of Adolph’s or other meat tenderizer, one tbsp. in a glass of water. Judgement is necessary in this regard as there is concern for tenderizer induced esophageal injury.
Stenosis - In our experience anastomotic stenosis usually occurs at least 3 weeks postoperatively. Stenosis usually presents with a progression of symptoms, beginning with several days of difficulty with solids, followed by difficulty with pureed foods, then inability to take liquids. It may be confirmed with UGI, with subsequent EGD whenever adequate stomal patency is not well documented, but some strictures are incomplete and not clearly documented with UGI contrast material, yet tight enough to be symptomatic with solids or thick liquids. Treatment involves dilatation up to 1.2 cm, followed by Carafate and a proton pump inhibitor, with a repeat dilatation one week later to insure good patency.
Abdominal Pain - Abdominal pain occurs frequently in bariatric patients, though it decreases in frequency after the first year. It may occur not only in the epigastrium or upper quadrants, but also in the lower quadrants, and suprapubic area, which is somewhat difficult to understand in patients who have undergone an upper abdominal procedure.
Abdominal pain may be associated with emesis, though not usually with fevers. It may be of relatively sudden onset, and quite severe. We have seen relatively few obstructions or abscesses as confirmed causes of our patients presenting with abdominal pain, and when these have occurred they were for the most part otherwise readily apparent.
Postulated etiologies for idiopathic episodes of abdominal pain include atony of the bypassed stomach, spasm of the Roux limb, traction of the tube gastrostomy site against the abdominal wall, and renal stones. The latter may be a frequent cause of lower abdominal or flank type pain.
Two frequently presenting known causes, Ulcers/NSAID Use, and Costochondritis, are discussed in following sections. Abdominal pain has also been a chief complaint of patients presenting in emotional collapse after trying to deal with the many changes required after Bariatric Surgery. This is usually elucidated by a thorough and prolonged visit and a judicious work-up, and is well treated with much reassurance that the patient is actually on tract (is not having an unexpected amount of difficulty with the postoperative recovery and adaptation to the postoperative lifestyle), and with a subsequent support group meeting.
Patients presenting with significant abdominal pain should be evaluated as indicated with CBC, CMP, amylase, lipase. Abdominal/Pelvic CT with oral and IV contrast seems to be the best imaging study, and will render information regarding leaks, abscess formation, distention of the bypassed stomach and/or biliary limb, obstruction, or kidney stone (see emergency section).
In the majority of cases no cause will be determined for the pain, and it will not recur, though it may have initially been quite severe and have prompted a 24 hour admission for observation and pain relief. Judicious narcotics and anti-emetics usually are very helpful while waiting for the pain to resolve. Patients presenting with a deep aching LUQ abdominal pain may have atony of the bypassed stomach, and a few have responded remarkably well to Reglan.
Ulcers/NSAID Use - Abdominal pain in our patients may be related to inflammation of the upper GI tract from the use of proscribed NSAID’s. A patient may have arthritis that continues to trouble them postoperatively, despite impressive weight loss. Such patients may then resort to Ibuprofen in any of its many forms, or other NSAID’s, even though these are known to frequently cause GI upset.
With or without NSAID use patients may develop ulceration and inflammation of the esophagus, of the gastric pouch, or of the anastomosis between the pouch and Roux limb (gastrojejunostomy). The pain is typically epigastric, radiates to the back and sides, and is deep, sharp, and burning. Initial work-up may include an UGI, though the best study will be EGD. Current treatment includes cessation of all NSAID’s, other than Tylenol, AcipHex, 20 mg bid or Nexium 40 mg bid, # 60, and Carafate, 2 - 8 0z bottles, 1 gram (10cc) 1 hour before meals and qHS, with a referral to gastroenterology for EGD.
Costochondritis - Patients frequently present with complaint of epigastric pain that on examination is actually just off to either side (usually the left), and is exactly reproducible by firm palpation of the rib margin. This almost always resolves with a 1 or 2 injections (staged weeks apart) of 1cc Celestone and 2cc 0.5% Marcaine.
Wound Infection/Fluid collections - (also discussed in Surgical Problems)- Wound infections and fluid collections may be somewhat difficult to determine in patients with generous subcutaneous tissues, whose wounds will not close rapidly after they are opened for a negative diagnostic exploration. Purulent discharge is of course clear evidence. Patients complaining of the "feeling of something in their wound" should be evaluated with ultrasound. Be sure to ask the Radiologist if there is a subcutaneous collection of fluid seen on the "negative" CT done for abdominal pain. Treatment of course is incision and drainage and antibiotics, though the entire wound may not have to be opened to drain a hematoma or seroma, infected or not.
Adjustable Gastric Band Too Tight - Patients with adjustable gastric bands may have them adjusted too tightly. This may either be accidental, or intentional as part of a somewhat zealous but misguided effort for additional weight loss. Patients will say that they had fluid added to their band either days or even months before, and have been unable to take solids or even liquids other than water. They will also presents with significant reflux, at times even involving night time aspiration, dysphagia.
Evaluation involves an UGI to assess for complete obstruction, placement of the band, and esophageal dilatation. Patients presenting with dehydration should be assessed with laboratory studies and considered for IV hydration.
Treatment involves taking fluid out of the band. This can be done with a noncoring port access needle in consultation with a weight loss surgeon, and may best be done under fluoroscopy. The simplest approach may be to hydrate the patient, and then to refer them to the weight loss surgeon or experienced radiologist the next day.
Adjustable Gastric Band Slip - Adjustable gastric bands are intended to be precisely located just below the GE junction, with just a bit of gastric tissue located above it. Three should not be a sizeable pouch, above it, usually meaning less than the size of a vertebral body as seen on UGI, and the pouch should not flop over the band on either side. If the pouch is larger than expected, or flops over the band, the patient, until stated otherwise, has a presumed slip.
Slips usually present with a certain degree of obstruction, from difficulty taking solids to complete obstruction with intractable nausea and vomiting and abdominal pain. If minimally to moderately symptomatic, they can be hydrated and referred to a Bariatric surgeon for timely followup. Slips involving severe abdominal pain and elevated WBC may be complicated by gastric ischemia, and require urgent referral and evaluation for urgent/emergent band removal.
Adjustable Gastric Band Port Infection - Band ports may become infected. Such infections are typical, with redness, swelling, and tenderness in the area of the port, and an elevated WBC. Aspiration from the area around the port may yield pus, and CT may show inflammation around the catheter tract.
Treatment can be initiated with IV and oral antibiotics, and usually patients can be referred for outpatient followup with a Bariatric surgeon.
Adjustable Gastric Band Erosion - Infrequently, gastric bands can erode into the stomach. Erosions do not usually present with catastrophic spillage of gastric contents, as the band is usually well encapsulated. Rather, the patient presents with days to weeks if epigastric and left upper quadrant pain, an elevated WBC, and at times local signs suggesting a port infection. Imaging studies are usually not diagnostic, but EGD will show the band having eroded into the stomach.
Stable patients may be referred for outpatient followup. Compromised or septic patients should be considered for immediate explantation, with immediate consultation with an available weight loss surgeon.
The following information is presented to help surgeons preparing to operate, either emergently or electively, on patients who have undergone Roux-en-Y Gastric Bypass (RNY GBP) at our Center. References will be made to our experiences operating on patients who received Bariatric surgery elsewhere and who have returned toAlaska. Other than abdominal wall hernias, our patients have only infrequently required re-exploration for problems specifically related to the index RNY GBP, with the perhaps 2 to 4 (of our almost 2000 patients as of 2011) presenting annually with perforated anastomotic ulcers, and perhaps 1 to 2 presenting every year with Peterson defect related internal hernias, with almost none of these occurring in patients done after we began to tightly close this defect in 2006.
Operative Anatomy- Patients who received surgery at our Center between September of 2000 and June of 2002 generally underwent a 7 cm vertical pouch, retrocolic/retrogastric, RNY gastric bypass with Roux and Bypassed limbs of approximately 100 cm each, and with concurrent cholecystectomy/ intra operative cholangiogram, wedge liver biopsy, and tube gastrostomy.
Most patients received a hand sewn gastrojejunostomy anastomosis (g-j) with a 6 cm, green #5 Ethibond band placed over the lower 3rd of the pouch, and a stapled jejunojejunostomy anastomosis (j-j). 3 patients operated on or before January 2001 received a short pouch with a stapled anastomosis, and no band. After June of 2002 the vast majority of our open RNY GBP patients have had their Roux limbs passed in an antecolic/antegastric fashion. Approximately 6 patients have not been given a tube gastrostomy, principally for anatomic reasons. As of June 2003 we also began to perform laparoscopic RNY GBP’s. These patients have for the most part an antecolic/antegastric configuration, stapled g-j and j-j anastomoses, and Roux and biliary limbs of 80-100 and 40-50 cm, respectively. Laparoscopic patients operated on after 2004 for the most part have 5 to 7 cm long pouches, and are antecolic/antegastric, with 35 cm biliary limbs and 95 cm roux limbs. Laparoscopic patients after 2008 or so have very long pouches, sometimes 10 to 12 cm in length.
The Peterson defect between the Roux limb and transverse colon has been closed for laparoscopic patients since summer 2006, and for the infrequent open patient since 2009.
At the time of exploration, should the g-j anastomosis need to be evaluated, consider placing a light source in the pouch itself, to help localize it in the dense adhesions encountered. In our patients the green Ethibond band around the lower 3rd of the pouch is a useful landmark, and the anastomosis can be palpated just distal to it.
General Comments on Revision Surgery-
Revision bypass surgery is demanding to those experienced in the field, and even more so to the non-bariatric general surgeon tasked to care for these patients on an emergency basis. Avoid it when possible. Bowel may appear compromised, only to revive when given the time to do so after the obstruction is relieved or the hernia reduced. Give the bowel sufficient time (15 to 30 minutes?) to regain color. When bowel viability is still in question, strongly consider closing and returning in 24 to 48 hours to see if the bowel has recovered, and then only excise the minimal length required. The above is especially true when dealing with problems of the Roux limb. Whenever possible, avoid revising the entire Roux limb, as dissecting out and revising the g-j can be difficult. Again, relieve the obstruction or reduce the limb, and later excise only what definitely appears non-viable.
Patients post RNY GBP are often in a delicate nutritional balance, and resection of any appreciable length of bowel may leave them with inadequate digestive capability. If for any reason, such as an Oncologic procedure, vascular accident, or obstruction in the pelvis from adhesions due to GYN surgery, should there be a need for excision of much of the common channel (small bowel below the j-j), consider shortening the Roux limb at the same time by relocating the j-j more proximally on the Roux limb. This will place more bowel in the common channel, which carries both ingested nutrients and digestive juices from the liver and pancreas, and is the segment where most absorption of nutrients may occur. Use care not to shorten the Roux limb to less than 30 cm, and not less than 50 cm if possible, as shorter lengths may lead to bile reflux into the gastric pouch.
Whenever our patients are seriously ill or may shortly become seriously ill and are planned to undergo surgery, strongly consider placing a g-tube for postoperative nutrition. Their ability to regain weight by eating is limited in the best of circumstances, and may be very difficult for them to achieve when debilitated. Do not feed via the g-tube for at least the first 3 days following surgery, as the tract may take several days to seal adequately from the peritoneal cavity. A g-tube tube, along with a carefully placed NG tube directed through the g-j anastomoses and 10 to 15 cm into the Roux limb, may be critical in decompressing the GI tract of RNY GBP patients who undergo subsequent surgery, and who may then have a prolonged postoperative ileus.
Bowel Obstructions- Obstructions occur generally in 1 to 2 percent of Bariatric patients. The location of the obstruction depends on the initial Bariatric procedure performed, and may include twists, stenosis, scarring, or adhesions at the retrocolic/retrogastric tunnel, mesocolic window, or jejunojejunostomy, or internal hernias through the mesocolic window or under or around the Roux limb (Peterson Defect). Obstructions may involve the Roux limb, the bypassed limb (afferent limb syndrome), the common channel (bowel distal to the j-j), or combinations of these.
When evaluating patients for suspected obstructions CT with IV and oral contrast is often the best study to obtain, as it will give information on obstruction of the Roux limb and common channel as well as the bypassed, afferent limb. UGI may be used to evaluate the Roux limb and common channel, with ultrasound then done to evaluate for a distended, bypassed stomach.
Obstructions may be difficult to diagnose. Incomplete obstructions may cause venous congestion to the point of bowel compromise, but without obstructing the bowel lumen or flow of contrast flow. Obstructions involving all or parts of the bypassed limb are difficult to evaluate, as they do not involve the alimentary stream, and thereby are not elucidated with oral contrast studies per se. Diagnosis is made more difficult by the not infrequent presentation of RNY GBP patients with nonsurgical abdominal pain, or by the chronic and recurrent nature of symptoms in patients who present with internal hernias that may incarcerate intermittently and reduce spontaneously. Patients presenting with pain out of proportion to presentation not relieved with time or medication should be considered for exploratory surgery.
Greater sac obstructions or those occurring caudad to the omentum are usually relatively easy to locate and treat by generally accepted techniques. Obstructions may occur at the mesocolic window due to scarring of the mesocolon itself, and are usually diagnosed when a gentle-probing finger cannot be passed through the mesocolic window next to the bowel. These may be difficult to approach from below and easier to approach from the lesser sac. After widely dividing the tight, obstructing sclerotic band usually seen, I chose to suture the edges of the defect down to the bowel serosa to prevent subsequent internal herniation, unless the lesser sac area is so filled with adhesions that no potential cavity exists to herniate into.
Obstructions occurring above the mesocolic defect may occur due to dilatation and elongation of the bowel in the retrogastric tunnel, which then twists on itself, causing a closed loop obstruction. These may be treated by simply reducing the bowel below the mesocolic window, and then fixing it to the mesocolon. Dilatation, elongation, and twisting of the Roux limb may actually be caused by a scarring and obstruction at the mesocolic window, and both sites of obstruction should be sought for and treated.
When treating an obstruction of the bypassed, (biliary, or afferent) limb at the j-j anastomosis, consider simply performing an additional side-to-side anastomosis of the bypassed limb onto the Roux limb or common channel, rather than taking down the j-j and redoing it.
Abscesses are usually associated with leaks. Both occur in a few percent of patients, generally 1 to 2. I suspect they occur more frequently in patients with histories of chronic Prednisone or methotrexate use, or in those with compromised tissues due to collagen vascular disease or advanced age, as these seem to be the patients with the most friable tissue. Leaks and abscesses usually present in the first week after surgery, and as such will usually be cared for by the operative team. We have had one patient present with an abscess and leak 2 weeks after surgery, which may have been related to a fall. We also had one patient present 2 years post op, with a large, perforated g-j stomal ulcer associated with NSAID use.
Infusing the pouch with air or methylene blue at the time of exploration is often helpful for locating leaks, or for assessing the integrity of any repair.
The leaks we have encountered have usually been related to the staple line located at the top corner of the pouch, at what was the angle of His, and are thought to be due to failure of the staple line. If no leak can be identified at the time of exploration, and an abscess has occurred in the area of the pouch, attempt to oversew or staple reinforce this area. Expect the leak to recur, and in all cases widely drain the area. Place a g-tube to be used for postoperative nutrition while waiting for any subsequent leak to close.
Leaks may occur from the g-j. These are best treated with direct closure and advancing the jejunum over the repair. We used this technique for the patient who had the perforated stomal ulcer 2 years postoperatively, and it worked well. Again, expect the leak to recur, though to a lesser degree, and drain and place a g-tube.
This patient went on to develop a subdiaphragmatic abscess. This was well treated with percutaneous drainage rather than open exploration due to dense adhesions in the area.
Troublesome is the posterior anastomotic leak that cannot be readily located or visualized, and that may best be treated with wide drainage and TPN or tube feeds.
Leaks may occur in the bypassed portion of the stomach, around the g-tube, at the j-j anastomosis, and at other sites. General Surgery principles apply to revision and closure techniques. Place a g-tube to decompress and/or feed via the bypassed stomach.
Wound Infections- Wound infections occur in perhaps 5% of our patients or less. This occurrence may be less than that reported elsewhere, and I believe is in part due to the cooler temperatures here in Alaska, but may also be due in part to our use of colonic bowel preps, although this is controversial. Wound infections, as usual, seem to be related to seromas, hematomas, or stitch abscesses.
Of note in the morbidly obese patient who has generous subcutaneous tissue is the phenomenon of the persistent seroma. This seems to be caused by apparent liquifaction of adjacent adipose tissue that continues as long as any component of the seroma is present. Keep this phenomenon in mind with any fluid collection or drainage that apparently will not go away, and drain these widely.
We used looped #1 Panacryl for fascial closure until approximately October 2002, at which time it went off the market, and we switched to #1 looped PDS. We appreciated Panacryl for the apparent low incisional hernia rate, which its use seemed to result in (? Less than 5%), but several patients may have reacted to it with chronic stitch granulomas or recurrent abscesses. Consider this when seeing a patient who presents with recurrent wound infections (1 patient), or nodular masses deep to the incision scar (2 patients).
Bleeding/Anemia- Several of our patients have presented with postoperative anemia, usually due to inadequate iron supplementation, stomal ulceration, or metrorrhagia. It is mentioned here only to postulate that a difficult to approach bleeding gastric pouch stomal ulcer may best be treated by visualizing the bleeding site via EGD, and using the scope to guide sutures placed from outside the pouch to control the bleeding, thus avoiding the need to take down or open the anastomosis.
Abdominal Wall Hernias- Many patients present to us for consideration for Bariatric Surgery with preexisting abdominal wall hernias, and our rate of postoperative incisional hernias is less than 5%. Whenever possible, we defer repair of preexisting hernias so that we are not placing mesh at the time of the gastric bypass when the GI tract is of necessity opened, though we of course will repair epigastric ventral hernias which lay in the area of the incision. We believe all hernias, preexisting and those detected postoperatively (at times only after patients have lost considerable amounts of weight) are best repaired at approximately one year postoperatively, as this will increase the rate of success of the hernia repair in general, and will allow for consideration of concurrent panniculectomy for patients so inclined.
Hernias should not be deferred once they become symptomatic. Be wary of the umbilical hernia that was chronically incarcerated with omentum, and that reduces spontaneously once the patient begins to lose weight, as it may then become incarcerated with bowel. Also be wary of the non-symptomatic hernia that the patient casually says they have had to occasionally reduce without difficulty for years without apparent difficulty, as the behavior of the hernia may change suddenly postoperatively.
Plastic Surgery Revisions- Our patients frequently request plastic surgery revisions of excessive skin postoperatively. Panniculectomy and breast-lift type procedures are the ones most frequently sought after, with revisions of excessive tissue on the arms ("wings") and thighs requested as well. These procedures are best done after achieving the expected 70% excessive body weight loss (70% EBW), and also after waiting at least one year postoperatively. Patients will often present demanding skin revision procedures before these goals are achieved, and will require much counseling to understand why they must wait.