Unfortunately, complications may occur during surgery. One of the surgeon’s goals is to reduce the risk for complications. The preoperative preparation, surgical techniques and postoperative care of the surgical patient focuses on maximizing benefits while minimizing risks. An understanding of the types of complications possible is critically important for the surgeon and may allow the patient to better understand surgical decisions made during a case.
(1) Dilatation of the cervix:
The cervix must be dilated in order to enter the hysteroscope into the uterine cavity. Many resectoscopes have an outer sheath diameter of about 9 mm (so that cervical dilatation using mechanical dilators must be at least this amount). It is optimal to avoid overdilatation of the cervix since leakage of the distending media through the cervix and around the hysteroscope (especially under pressures of about 100-150 mm Hg) then becomes possible.
Some cervical canals are difficult to negotiate with dilators. Different dilators have a variable amount of curvature to choose from. It is possible to perforate the lower uterine segments during dilatation. Clinical situations in which perforation is more common include the pregnant uterus, fibroid uterus, uterus of a women exposed to DES in utero, uterus after exposure to prostaglandins for cervical ripening, and infected uterus. Many cases of uterine perforation occur at the onset of dilatation and the subsequent dilators then continue to open the perforation site.
Occasionally, a rent in the lower uterine segment occurs during dilatation. It is thought that rapid dilatation or a difficult dilatation involving a stenotic inflexible cervix may enhance the frequency of these tears. It is possible for a tremendous amount of distending media to become intravasated through these rents and into the large vessels of the lower uterine region if they are transected.
Cervical incompetence following hysteroscopic surgery is rarely reported but theoretically possible. The cervix is composed of a tough fibroconnective tissue and smooth muscle. Closure of the internal os of the cervix is the general rule even following manual dilatation of up to 15 mm.
The pressure maintained in the uterine cavity may (but generally should not) exceed both the venous and the arterial pressures so that active blood flow from transected vessels may not become apparent until the uterus is deflated. At lesser pressures, bleeding can be identified and usually controlled. If there is excessive bleeding following destructive procedures such as endometrial ablation then this is frequently controlled by tamponade using an inflated foley catheter balloon (10-30 mL for up to 16 hours) in the uterus. Sometimes the excessive flow can be controlled with estrogen hormonal therapy (if due to a denuded endometrial lining).
(3) Excessive intravasation of distending media or CO2 gas:
Whenever vessels are transected during hysteroscopic surgery and either fluid or gas is entered into the uterine cavity under pressure there is a possibility of intravasation (entry of these substances into the circulation).
I use D5W (5% Dextrose in Water) almost exclusively for my resectoscopic surgery. Major complications with this solution are very rare. In fact, I have found no reports in the world literature of major morbidity or mortality with the use of D5W at hysteroscopy. Possible complications include water intoxication (a reduction in serum osmolality) with a dilutional reduction in sodium concentration, volume overload (when the circulating volume in the vascular system exceeds the ability of the heart to adequately pump this volume and the excess fluid typically begins to collect in the tissues of the lungs), hypothermia (significant reduction in body temperature) if room temperature solutions are used without warming the patient with devices like a “Bair Hugger,” and hyperglycemia (significant excess in circulating glucose concentration that may not be rapidly metabolized if the patient has insulin resistance or diabetes mellitus).
The major complication that most hysteroscopic surgeon's focus on avoiding is water intoxication. The risk of water intoxication from D5W in a healthy woman with normal renal function is very low, since the kidneys can typically produce in excess of 1000cc of dilute urine per hour in response to a decrease in serum osmolarity.
Following hysteroscopic surgery, there is a chance of adhesion (scar) formation. When I use significant electrocoagulation within the uterine cavity I provide the infertility patient with intraoperative estrogen IV (25 or 50 mg of Premarin) and at least a 30 day course of higher dose Premarin postoperatively (1.25 mg or preferably 2.5 mg if tolerated).
(5) Burn injury to the bowel:
When resectoscopic electrosurgery is performed in the area of the uterine ostia (near the entry site of the fallopian tubes) there is a chance of thermal injury to adjacent tissue outside the uterine cavity. This is because the uterine wall in these regions is very thin and heat from the cautery can travel through the uterine wall and burn adjacent bowel.
Endometritis is uncommon after operative hysteroscopy and antibiotics are usually not “routinely” given. I however have a very low threshold for the decision to use antibiotics since their potential benefits outweigh their risks when exposure to infection occurs.
(1) Verres needle and Trocar injuries:
Most complications from laparoscopic surgery have been reported to occur at the time of Verres needle or Trocar placement.
The primary concern of the surgeon when entering these tools is the possibility of inadvertent laceration of a major blood vessel that may not be recognized immediately since the tools used are entered blindly into the abdomen. Several procedural techniques minimize this risk, which actually occurs very rarely in the hands of an experienced laparoscopist. A high index of suspicion also allows for a more rapid identification of injury to a major vessel.
If the Verres needle is placed directly into a major vessel and this is not recognized then insufflation of the vessel with CO2 gas may result in a massive pulmonary (gas) embolism and cardiovascular compromise or collapse.
Trocar injuries may involve the major vessels. More commonly the trocar injures the bowel during the (blind) insertion. To minimize either of these events, I always invert the umbilicus, hold the abdominal wall up (to maximize distance from the deep structures such as bowel and vessels), and enter the instruments into the abdomen just past the inner lining of the abdominal wall. I have not had any such injuries using this technique.
(2) Vascular injury in abdominal wall and great vessels of the pelvis and abdomen
Vascular injuries appear to account for about 30-50% of laparoscopic trauma, but injury to the major vessels is very uncommon (less than 1 in 1000). Injury to the vessels in the abdominal wall may involve the
- Deep epigastric vessels (inferior or superior epigastric vessels). The epigastric artery originates from the external iliac artery near its transition to the femoral artery and the deep inferior epigastric artery then lies adjacent (lateral) to the obliterated umbilical ligaments (usually easy to see on direct laparoscopic inspection of the inner abdominal wall) beneath the lateral margin of the rectus abdominus muscle. Entry of additional trocars lateral to these vessels reduces vascular injury to these vessels within the abdominal wall.
- The superficial epigastric vessels. These vessels course”on top” of the rectus abdominus muscle and are usually able to be visualized with transillumination using the laparoscope as the source of light within the pelvis.
- The deep and superficial circumflex iliac vessels. These vessels course lateral to the epigastric vessels, usually lateral to any typical placement site for a trocar.
If there is an injury to one of these deep abdominal wall vessels with significant bleeding seen from the site of the trocar the exact location of the bleeding site should be identified (if possible) by carefully moving or rotating the trocar so as to identify the precise site of the injury. Then bipolar cautery can be used to try to immediately control the bleeding. The epigastric artery should be coagulated above and below the site of injury. This is not always possible and a second manipulation that can be useful is tamponade (using a Foley catheter that is passed through the 5 mm port, inflation of the balloon, pulling the balloon tightly against the inner abdominal wall and holding it in place with a (Kelly) clamp. After a few minutes (about 10-15 minutes) the balloon may be released and the bipolar used once again if the bleeding has significantly decreased. If this fails to control bleeding, a figure of eight suture may be placed through and through the abdominal wall to control the bleeding (with tying on the outer skin of the abdominal wall). If all of these techniques fail then a cut down for exploration and tying off the vessels may be required.
If there is injury to one of the superficial vessels of the abdominal wall then pressure on the skin usually controls the flow. If external pressure is not successful then ligature is required. At the end of the case the sites should be reevaluated after removal of the trocars since the release of tamponade from the trocar may allow reactivation of bleeding.
Injury to the aorta, inferior vena cava or common iliac vessels can be life threatening. Immediate laparotomy with the assistance of a vascular surgeon is usually indicated.
(3) Urinary tract injury:
Injury to the urinary tract (bladder or ureters) is uncommon in laparoscopic surgery involving the pelvis, however, endometriosis or infection can certainly involve these structures intimately and the possibility of injury is important to recognize.
Bladder perforation with a Verres needle or the midline lower abdominal trocar is possible. This risk is dramatically reduced if a foley catheter is placed into the bladder at the onset of the case (with confirmation of urine flow since occasionally the lubricating gel plugs the catheter). Treatment may be a double layer closure (water tight) with a low reactive long half life suture like Vicryl or PDS, placement of a foley catheter for 7 days and antibiotics.
Bladder injury during resection of endometriosis or adhesiolysis (especially if there is a history of prior bladder surgery or cesarean section) is not common and depends on the extent of pathology being resected and the experience of the surgeon.
Ureteral injury is not common, most often involving the ureters as they course near the uterosacral ligaments. The most common cause is the use of cautery (either unipolar or bipolar) in the area and either direct or lateral thermal damage. If not immediately recognized, the diagnosis is suggested by development of flank pain, (unilateral) pelvic pain, fever, leukocytosis and peritonitis. If there is an intraoperative concern about ureteral injury, 5 cc of indigo carmine can be injected IV by the anesthesiologist and this should appear in the Foley bag within about 10 minutes. Postoperative diagnosis is usually guided with intravenous pyelography.
(4) Gastrointestinal injury:
Injury to the bowel is most often caused by the Verres needle or Trocar when blindly inserted. A Verres needle injury may be quite small and remain unrecognized. These injuries may involve the small bowel, large bowel or stomach. Electrosurgical or laser injury is relatively uncommon.
Extensive enterolysis for small bowel adhesions may result in injury in as many as 25% of the cases. These injuries often are not recognized. If recognized, the repair of small bowel injuries consists of a surgical consultation, repair with 3-0 or 4-0 silk (or PDS) suture (tapered SH needle), and nasogastric tube.
Large bowel injury is probably most often caused by the Verres needle and may go undiagnosed a large percentage of the time. Foul smelling gas is a characteristic sign. If the hole is small (needle size), expectant management is usually recommended.
Trocar injury to the large bowel usually occurs when there are dense adhesions from the large bowel to the inner anterior abdominal wall. Occasionally, these injuries are only identified when the trocar is removed under direct observation since they may be through and through the walls of the bowel. Laparotomy and repair by a general surgeon is indicated, leaving the trocar in situ for identification of the site of injury.
Most large bowel injuries are due to mechanical trauma as adhesions from the rectosigmoid bowel are freed from the cul de sac in the presence of dense endometriosis or chronic inflammation. Superficial injury may be able to be managed expectantly but deep injury requires surgical repair.
Stomach injuries may be more common following difficult endotracheal intubation since the stomach may be filled with gas. These injuries usually occur with subsequent (blind) entry of the Verres needle or Trocar into the inflated stomach. Placement of a nasogastric tube may reduce this possibility.
Reactivation of a dormant infection during laparoscopic surgery is possible but uncommon. Whenever surgery is performed in the presence of an active infection or is to repair the pelvis that has been damaged (by chronic infection) antibiotics should be used liberally.
Available Case Reports: