CHAPTER 40: Intraoperative Considerations

Gynecologic surgery is used to treat a broad spectrum of underlying pathology. As a result, the list of surgical procedures is extensive, but in general, techniques maximize tissue healing and patient recovery. Successful outcomes depend on appropriate patient and procedure selection, sound intraoperative technique, and preparation for possible complications.

Many anesthetic options are available for patients undergoing gynecologic procedures and include general anesthesia, regional analgesia, or local paracervical blockade with or without conscious sedation. These anesthetic techniques are provided by clinicians who are skilled with their placement and capable of managing their side effects. Thus, paracervical blockade and intravenous sedation may be provided by gynecologists. General and regional anesthesia typically are delivered and managed by anesthesiology staff.

Anesthesia selection for gynecologic surgery is complex and influenced by the procedure planned, extent of disease, patient comorbidities, and personal preferences of the patient, anesthesiologist, and surgeon. Last, the providing hospital or clinic may further define options based on their practicing norms and availability of personnel or equipment. For example, an outpatient gynecology clinic may be equipped to provide paracervical blockade or intravenous conscious sedation, but may lack sophisticated equipment or expertise required for regional or general anesthesia.

In all cases, both the anesthesia provider and the surgeon communicate regarding patient and surgery progress and are prepared for potential problems. Difficult patient intubation may complicate general anesthesia, and regional anesthetic procedures may lead to higher than anticipated levels of blockade and respiratory muscle dysfunction. Cases using paracervical blockade may be complicated by inadequate levels of anesthesia, or conversely by anesthetic toxicity. Conscious sedation may also fail to provide adequate analgesia, or alternatively may lead to respiratory depression. Thus, no procedure is free of risk, and contingency plans for each should be in place.

Paracervical Block

Paracervical block is used most commonly during first-trimester pregnancy evacuation but also may be selected for cervical ablative or excisional procedures, transvaginal sonographically guided oocyte retrieval, and in-office hysteroscopy. Some studies have also described preemptive analgesia with paracervical block for vaginal hysterectomy (Long, 2009; O’Neal, 2003). Paracervical blockade is often combined with nonsteroidal antiinflammatory drugs (NSAIDs) or intravenous conscious sedation or both. Conscious sedation may be achieved with several agents, but intravenous midazolam (Versed) and fentanyl (Sublimaze) is a frequent combination (Lichtenberg, 2001).

Technique

The cervix, vagina, and uterus are richly supplied by nerves of the uterovaginal plexus (Fig. 38-13). Also known as Frankenhäuser plexus, this plexus lies within the connective tissue lateral to the uterosacral ligaments. For this reason, paracervical injections are most effective if placed immediately lateral to the insertion of the uterosacral ligaments into the uterus (Rogers, 1998). Thus, divided doses are given at the 4 and 8 o’clock positions at the cervical base (Fig. 40-1).

In most cases, total doses of 10 mL of 0.25-percent bupivacaine, 1-percent mepivacaine, or 1- or 2-percent lidocaine may be administered (Cicinelli, 1998; Hong, 2006; Lau, 1999). However, specific calculation of a maximum safe dose for each patient before injection is recommended (Dorian, 2015). The toxic dose of lidocaine approximates 4.5 mg/kg (Table 40-1). For a 50-kg woman, this would equal 225 mg. Thus, if a 1-percent lidocaine solution is used, the calculated allowed amount would be: 225 mg ÷ 10 mg/mL = 22.5 mL. Of note, for any drug solution, 1-percent = 10 mg/mL.

Anesthesia is presumed to result from pharmacologic nerve conduction blockade by the local anesthetic agent (Chanrachakul, 2001). The injection itself may have an immediate anesthetic effect by swelling surrounding tissue and exerting mechanical pressure on nerves to disrupt neural transmission (Phair, 2002; Wiebe, 1995). Addition of epinephrine to these solutions leads to local vasoconstriction, which enhances analgesia quality, prolongs duration of action, and decreases ­toxicity. Thus, higher maximum doses may be used. Return of neural function is spontaneous as the drug is metabolized.

In general, increased doses of local anesthetics may lead to clinically significant conduction blockade within the central nervous system (CNS) and heart. Signs range from drowsiness, tinnitus, perioral tingling, and visual disturbances to confusion, seizure, coma, and ventricular arrhythmia. Monitoring patients for the subtle symptoms of CNS toxicity is important because the therapeutic-to-toxic ratios are often narrow with these agents.

When toxicity develops, cardiac effects are potentiated by acidosis, hypercapnia, and hypoxia. Thus, treatment typically includes intravenous access, adequate oxygenation, and seizure control. A benzodiazepine such as diazepam (Valium) given intravenously is effective anticonvulsant therapy (Naguib, 1998). For treatment, diazepam, 2 mg/min, is administered until seizures stop or a total dose of 20 mg is delivered.

Intrauterine Instillation

Injection of local anesthetic solutions through a catheter into the uterine cavity has been reported to safely lower pain scores in women undergoing in-office hysteroscopy or endometrial biopsy (Cicinelli, 1997; Trolice, 2000). The presumed mechanism is anesthetic blockade of nerve endings within the endometrial mucosa. Studies have used 5-mL doses of 2-percent lidocaine or of 2-percent mepivacaine. For first-trimester abortion procedures, Edelman and coworkers (2004, 2006) evaluated instillation of 5 mL of 4-percent lidocaine combined with paracervical blockade. However, for this indication, a significant number of women reported symptoms attributed to lidocaine toxicity.

Postoperative Pain

Anesthesiologists are employing multimodal strategies intraoperatively to reduce postoperative pain. Gabapentin and ketorolac are now in common use (Alayed, 2014; De Oliveira, 2012). The transversus abdominis plane (TAP) block has been studied in abdominal and laparoscopic hysterectomy with promising results (Carney, 2008; De Oliveira, 2014). The surgeon may also improve postoperative analgesia by implanting suprafascial wound soaker catheters to administer local anesthesia (Iyer, 2010; Kushner, 2005). Additionally, local infiltrative analgesia, using a long-acting medication such as liposomal bupivacaine, may be injected into the incision by the surgeon (Barrington, 2013).

Communication between all members of the team is vital to the success of an operation and avoidance of patient harm. The Joint Commission established the Universal Protocol for Preventing Wrong Site, Wrong Procedure, and Wrong Person Surgery (Joint Commission, 2009). This protocol encompasses three components: (1) preprocedural verification of all relevant documents, (2) marking the operative site, and (3) completion of a “time out” prior to procedure initiation. The “time out” requires attention of the entire team to assess that patient, site, and procedure are correctly identified. Important interactions also include introduction of the patient care team members, verification of prophylactic antibiotics, anticipated procedure length, and communication of anticipated complications such as potential for large blood loss. Additionally, requests for special instrumentation are addressed preoperatively to prevent potential patient compromise that may accompany lacking an instrument at the time it is needed. Breakdowns in communication are common across pre-, intra-, and postoperative phases of care and are linked to adverse events and patient harm (Greenberg, 2007; Nagpal, 2010). Specifically, the transfer of a patient to a new care team or new location has been identified as a time particularly vulnerable to communication breakdowns (Greenberg, 2007).

A gynecology resident may sometimes feel that the role of assistant is unimportant. However, an experienced surgeon knows the critical difference that good assistance can provide. Assistants should anticipate surgeon needs and aid smooth progress of the operation. Therefore, an assistant must be familiar with the planned procedure’s steps, relevant anatomy, and clinical patient details.

Maintaining exposure by proper retraction and keeping the operative field clear of obstruction are primary functions. Laparotomy sponge or suction use is timed to avoid interfering with the surgeon, and a sponge is used to blot rather than wipe. Immediate pressure is placed on bleeding surfaces until the situation can be assessed systematically. Clamps are released slowly to avoid tissue slippage. Attention must be fixed on the procedure. Thus, if music or conversation is distracting, they are avoided.

Anesthetized patients who undergo prolonged gynecologic procedures are at risk for peripheral neuropathy of their upper or lower extremities. These neuropathies are uncommon, and cited incidences approximate 2 percent of gynecologic cases (Cardosi, 2002). Neurologic deficits typically are mild, transient, and resolve spontaneously. Infrequently, prolonged or permanent disability may result.

During gynecologic surgery, lower extremity injuries can involve nerves of the lumbosacral plexus. Mechanisms of injury include surgical nerve transection, rupture following increased stretch, or nerve ischemia. Ischemia may result from compression of perineural vessels during prolonged or pronounced nerve stretch or compression. Although any patient may develop postoperative neuropathy, higher rates are noted in patients who smoke, who have anatomic abnormalities, or who are thin, diabetic, or alcoholic. Use of self-retaining retractors and prolonged surgical duration are additional factors (Warner, 2000).

Symptoms reflect functional loss of the affected nerve. Motor loss typically manifests as muscle weakness, whereas sensory loss may be noted as anesthesia, paresthesia, or pain in the nerve’s sensory distribution (Fig. 40-2 and Table 40-2). Therefore, a detailed neurologic examination allows clinical identification of most peripheral neuropathies. Electrodiagnostic testing is indicated if motor function is diminished (Knockaert, 1996). Generally, electromyography is most useful after a 2- to 3-week delay to permit denervational changes to fully develop within affected muscles (Winfree, 2005).

Treatment will vary depending on whether motor or sensory function is affected. If motor function is impaired, neurologic consultation is typically warranted. Physical therapy begins immediately to minimize contracture and muscle atrophy. Alternatively, for those with only mild sensory losses, observation for return of function is reasonable. For those with pain, treatments may include oral analgesics, gabapentin, biofeedback, and serial trigger point injection with local anesthetics.

Laparotomy

Femoral Nerve

This nerve perforates the psoas muscle early in its course and passes medially beneath the inguinal ligament before exiting the pelvis. It then enters the femoral triangle to lie lateral to the femoral artery and vein. This nerve can be compressed anywhere along its course but is particularly susceptible within the body of the psoas muscle and at the inguinal ligament. Improper placement of a self-retaining retractor is the most common cause of surgical femoral nerve injury, and rates following abdominal hysterectomy may reach 10 percent (Fig. 40-3) (Goldman, 1985; Kvist-Poulsen, 1982). In affected women, the patellar reflex is usually absent in addition to impaired sensory and motor function.

In prevention, lateral retractor blades are selected and positioned such that only the rectus abdominis muscle and not the psoas muscle is retracted (Chen, 1995). The retractor blades are evaluated when placed, to confirm that they are not resting on the psoas muscle. For thin patients, folded laparotomy towels may be placed between the retractor rim and skin to elevate blades away from the psoas muscle. Importantly, a small percentage of cases occur when a retractor has not been used.

Genitofemoral and Lateral Femoral Cutaneous Nerve

The genitofemoral nerve pierces the medial border of the psoas muscle and traverses below the peritoneum on this muscle’s surface. Similar to the femoral nerve, the genitofemoral nerve may suffer injury during psoas muscle compression (Murovic, 2005). In addition, this nerve may be injured during removal of a large pelvic mass adhered to the sidewall or during pelvic lymph node dissection (Irvin, 2004).

The lateral femoral cutaneous nerve appears at the lateral border of the psoas major muscle just above the crest of the ilium. It courses obliquely across the anterior surface of the iliacus muscle and dips beneath the inguinal ligament laterally as the nerve exits the pelvis. This nerve may also be compressed or be injured during dissections (Aszmann, 1997). Painful neuropathy specifically involving the lateral femoral cutaneous nerve carries the specific name meralgia paresthetica.

Transverse Incisions

Nerve injury during transverse abdominal entry is common and typically involves the ilioinguinal and iliohypogastric nerves or less frequently, genitofemoral nerve branches. The ilioinguinal and iliohypogastric nerves emerge through the internal oblique muscle approximately 2 to 3 cm inferomedial to the anterosuperior iliac spine (Whiteside, 2003). The iliohypogastric nerve extends a lateral branch to innervate the lateral gluteal skin. An anterior branch reaches horizontally toward the midline and runs deep to the external oblique muscle. Near the midline, this nerve perforates the external oblique muscle and becomes cutaneous to innervate the superficial tissues and skin in the region above the symphysis pubis. The ilioinguinal nerve extends medially to enter the inguinal canal and innervates the lower abdomen, labia majora, and upper thigh.

These are sensory nerves, and fortunately, most skin anesthesia or paresthesias that follow their injury resolves with time. Accordingly, injuries frequently are underreported by both patients and clinicians. Less often, pain can begin immediately or many years later and is usually sharp and episodic and radiates to the upper thigh, labia, or upper gluteal region. Later, sensations may become chronic and burning, as described in Chapter 11. To avoid compromising these nerves, a surgeon ideally avoids extending the fascial incision beyond the lateral border of the rectus abdominis muscles (Rahn, 2010).

Pelvic Sidewall Dissection

The obturator nerve pierces the medial border of the psoas muscle and extends anteriorly along the lesser wall of the pelvis. The obturator nerve exits through the obturator foramen. Lymph node dissection, tumor excision, or endometriosis resection performed at the pelvic sidewall may injure the obturator or genitofemoral nerves. Moreover, the obturator nerve also can be injured during dissection within the space of Retzius during some urogynecologic procedures.

Dorsal Lithotomy

This surgical position is used for vaginal, laparoscopic, and hysteroscopic surgeries. It is modified and described as standard or low lithotomy positions (Fig. 40-4). Dorsal lithotomy may be associated with injury to several nerves derived from the lumbosacral plexus, including the femoral, sciatic, and peroneal nerves. For example, compression and ischemic injury of the femoral nerve beneath the rigid inguinal ligament can follow prolonged sharp flexion, abduction, and external hip rotation in dorsal lithotomy (Fig. 40-5) (Ducic, 2005; Hsieh, 1998). Ideal positioning as shown can minimize these injuries.

The sciatic nerve, derived from the lower sacral plexus, exits the pelvis through the greater sciatic foramen. It extends down the posterior thigh and branches into the tibial nerve and common peroneal nerve above the popliteal fossa. The sciatic and common peroneal nerves are anatomically fixed at the sciatic notch and head of the fibula, respectively. For this reason, sciatic nerve injury may reflect impaired function of the entire sciatic nerve or only the common peroneal division. Sciatic nerve stretch injury can develop if a patient’s hips are placed in sharp flexion or pronounced external rotation or both. Moreover, even an appropriately positioned patient may be injured if a surgical assistant during vaginal surgery leans against the thigh and creates extreme hip flexion.

The common peroneal nerve, now termed the common fibular nerve, originates above the popliteal fossa and crosses the lateral head of the fibula before it descends down the lateral calf. At the lateral fibular head, this nerve is at risk for compression against leg stirrups. Therefore, the addition of cushioned padding or patient positioning that avoids pressure at this point is warranted (Philosophe, 2003).

Brachial Plexus

This plexus derives from the ventral rami of C5-T1, traverses the neck and axilla, and supplies the arm and shoulder. Positioning injuries can follow hyperextension of the upper extremity, for example, when the arm is positioned at an angle to the body that exceeds 90 degrees. Additionally, even in situations in which the arm has been positioned appropriately, inadvertently leaning against the arm or placing the patient in steep Trendelenburg position may push the extremity into hyperextension. With injury, either motor or sensory function can be lost (Warner, 1998). Peripheral ulnar neuropathies can also develop by external compression if the arm is placed at the patient’s side. Padding the elbow may help avoid this (Warner, 1998).

In women for whom laparotomy is selected, an ideal abdominal incision allows rapid entry, affords adequate exposure, permits early ambulation, promotes strong wound healing, does not compromise pulmonary function, and maximizes cosmetic results. These criteria form the foundation in choosing the best incision for each patient. In gynecology, opening the abdomen typically is achieved using a midline vertical incision or one of three low transverse incisions, the Pfannenstiel, Cherney, or Maylard incisions.

Midline Vertical Incision

This incision is used frequently if access to the upper abdomen and generous operating space are required. It can be extended up and above the umbilicus and thus is preferred when the preoperative diagnosis is uncertain. Moreover, simple midline anatomy allows quick entry into the abdomen and low rates of neurovascular injury to the anterior abdominal wall (Greenall, 1980; Lacy, 1994). Moreover, because of decreased midline vascularity, Nygaard and Squatrito (1996) recommend this incision in patients who have coagulopathy, decline transfusion, or are administered systemic anticoagulation.

Its greatest disadvantage stems from increased tension on the incision when abdominal muscles contract. For this reason, compared with transverse incisions, midline vertical incisions are associated with higher rates of fascial dehiscence and hernia formation and poorer cosmetic results (Grantcharov, 2001; Kisielinski, 2004). Additionally, patients who have repeat vertical incisions for gynecologic indications tend to develop more adhesive disease than with low transverse incisions (Brill, 1995).

Transverse Incisions

These incisions are used commonly in benign gynecologic surgery, provide several advantages, and are illustrated in the atlas. They follow Langer lines of skin tension and thus offer superior cosmetic results. They also carry low rates of incisional hernia (Luijendijk, 1997). Moreover, their placement in the lower abdomen is associated with decreased postoperative pain and improved pulmonary function compared with midline vertical incisions. Of low transverse incisions, Pfannenstiel incision is typically the simplest to perform, and for this reason, it is selected most often.

Despite these advantages, transverse incisions have limitations. These incisions limit access to the upper abdomen and offer smaller operating space compared with midline incisions. This is especially true of the Pfannenstiel incision and results from narrowing of the surgical field by intact rectus abdominis muscle bellies, which straddle the incision.

Consequently, Cherney and Maylard incisions were developed to overcome this restriction, and to some degree, they do improve exposure. The Cherney incision releases the rectus abdominis muscle at its inferior tendinous insertion. This approach affords greater exposure of pelvic organs and access to the space of Retzius. The Cherney incision may also be used if a Pfannenstiel incision has already been initiated, but then additional exposure is required.

The Maylard incision transects the rectus abdominis muscle and provides substantial operative space. However, it is technically more difficult because isolation and ligation of the inferior epigastric arteries are required. The incision is used infrequently because of concerns regarding operative pain, decreased abdominal wall strength, longer operating times, and increased febrile morbidity. Randomized studies, however, have not supported these concerns (Ayers, 1987; Giacalone, 2002). This incision is avoided in patients whose superior epigastric vessels have been interrupted and in those with significant peripheral vascular disease who may rely on the inferior epigastric arteries for lower-extremity collateral blood supply.

Incision Creation

Entry into the abdomen begins with scalpel incision of the skin, and scars are excised to improve wound healing and cosmetic results. Although an electrosurgical blade may be used to incise the skin, faster healing and improved appearance in general follow scalpel incision (Hambley, 1988; Singer, 2002b).

For the remaining layers, scalpel or electrosurgical blade may be selected, with no differences in short- or long-term wound healing with either (Franchi, 2001). However, in evaluating surgical bleeding and postoperative pain, Jenkins (2003), in his review, noted an advantage with electrosurgical blade use. Regardless of type of incision or instrument used, adherence to proper technique is emphasized: obtaining meticulous hemostasis, minimizing devitalized tissue, and avoiding dead space creation.

Following laparotomy, closure of a laparotomy incision must address the peritoneum, fascia, subcutaneous layer, and skin. Wound closure may be broadly categorized as either primary or secondary. With primary closure, materials are used to approximate tissue layers. In closure by secondary intention, wound layers remain open and heal by a combination of contraction, granulation, and epithelialization. Secondary closure is used infrequently in gynecologic surgery and typically is indicated if tissues planned for closure contain significant infection. The option of delayed primary closure is also available in these situations once infection has cleared.

Optimal closure of a laparotomy incision is the subject of much debate. Most data stem from general surgery and gynecologic oncology studies on midline abdominal incision closure and from research on cesarean delivery techniques. Ideally, closure avoids wound infection, adhesion formation, dehiscence, and hernia or sinus tract formation; minimizes patient discomfort; yet preserves cosmesis to the extent possible.

Peritoneum and Fascia

The peritoneum provides no abdominal wall strength, and closure of this layer has been suggested to prevent adhesions between the anterior abdominal wall and adjacent organs. However, evidence is conflicting, and several studies have shown that nonclosure of the peritoneum compared with closure decreases operating time without increasing adhesion formation, wound complications, or infection (Franchi, 1997; Gupta, 1998; Tulandi, 1988). However, few well-done randomized controlled trials have assessed long-term adhesion formation. Accordingly, closure of the visceral or parietal peritoneum is often provider dependent. Without closure, this layer typically regenerates within days following surgery (Lipscomb, 1996).

Thus in many cases, the first tissue closed is fascia. Many studies have supported the use of a continuous running-stitch closure of abdominal incisions compared with interrupted closure of the fascia (Colombo, 1997; Orr, 1990; Shepherd, 1983). Continuous closure usually is faster and associated with comparable rates of dehiscence, wound infection, and hernia formation. Suture material selection tends to favor delayed-absorbable suture compared with nonabsorbable. Delayed-absorbable sutures appear to afford adequate wound support yet lead to less pain and lower rates of sinus tract formation (Carlson, 1995; Leaper, 1977; Wissing, 1987). However, nonabsorbable suture is considered if a hernia is identified or if the incision has cut through previously placed mesh. A 0-gauge or no. 1 suture is suitable for closure of most fascial incisions. Sutures are placed approximately 1 cm apart and 1.2 to 1.5 cm from the fascial edge. Little additional security is attained beyond 1.5 cm (Campbell, 1989). Stitches ideally appose fascial edges and allow tissues to swell postoperatively without cutting through fascia or causing avascular necrosis.

Subcutaneous Adipose Layer and Skin

Collections of blood and fluid serve as potential accelerants to bacterial growth. For this reason, to decrease rates of hematoma or seroma, investigators have evaluated subcutaneous layer suture closure or drains. In those with layers less than 2 cm thick, most studies have found no advantage to either practice. However, wound infection and fat thickness are the greatest risk factors for subcutaneous layer dehiscence (Soper, 1971; Vermillion, 2000). For patients with subcutaneous layers 2 cm or more thick, closing the subcutaneous layer is effective prevention (Gallup, 1996; Guvenal, 2002; Naumann, 1995). The ideal suture and technique for closure of this layer are unknown, but efforts ideally close dead space yet minimize suture burden and inflammatory reaction. A 2-0 gauge plain gut suture is one suitable choice.

Skin may be closed effectively with staples, subcuticular suturing, wound tape, or tissue adhesive. Thus, in most instances, surgeon preference influences closure method. Technically, the incision line is approximated without skin tension, and subcutaneous adipose or deep dermal suturing may assist with carrying tension loads.

Of options, the running subcuticular suture is placed by taking horizontal bites through the dermis on alternating sides of the wound using absorbable suture (Fig. 40-6). Delayed-absorbable material such as polyglactin (Vicryl) or poliglecaprone (Monocryl) in a fine gauge, such as 3-0 or 4-0, is suitable. Advantages include decreased cost, effective skin approximation, and no required suture removal. However, this method typically requires the greatest amount of time and technical expertise.

Automatic stapling devices are favored because of their fast application and secure wound closure. However, they do not allow as meticulous a closure as sutures, and wounds requiring accurate approximation of tissue are not ideal candidates for staple closure (Singer, 1997). Staples may be uncomfortable, may be associated with discomfort during removal, and require the patient to return for staple removal.

Before stapling, the wound edges are everted, preferably by a second operator. If the edges of a wound invert or if one edge rolls under the opposite side, a poorly formed, deep, noticeable scar will result. Additionally, pressing too hard against the skin surface with the stapler is avoided to prevent placing the staple too deep and causing ischemia within the staple loop. When placed properly, the crossbar of the staple is elevated a few millimeters above the skin surface (Lammers, 2004). Staples are removed in a timely fashion to avoid leaving “track mark” scarring.

Of topical skin adhesives, octyl-2-cyanoacrylate (Dermabond) is applied as a liquid and polymerizes to a firm, pliable film that binds to the epithelium and bridges wound edges (Fig. 40-7). It can be used for closure of skin incisions that carry minimal tension such as laparoscopy trocar or transverse laparotomy incisions, or as an adjunct protective layer in larger incisions. Tissue adhesives achieve results similar to those for traditional sutures (Blondeel, 2004; Singer, 2002a).

Following approximation of deeper incision layers, the adhesive is applied in three thin layers above apposed skin edges. The adhesive extends at least ½ cm on each side of the apposed wound edges. Placement of the liquid between skin edges is avoided because the adhesive may retard healing (Quinn, 1997). Although 30 seconds between layers for drying is required, application is fast. Moreover, adhesives create their own dressing and appear to afford some antibacterial protection (Bhende, 2002). The adhesive sloughs in 7 to 10 days. Showering and gentle washing of the site are allowed, but swimming is discouraged. Petroleum-based products on the wound can decrease adhesive tensile strength and are avoided.

The primary indication for tape closure is a superficial straight laceration under little tension. Thus, closure of laparoscopy trocar sites or laparotomy incisions in which deep layer closure has brought skin edges into close proximity are suitable cases. Moreover, skin edges ideally are thoroughly dry for proper adhesion. Thus, tape may not be appropriate for a wet or oozing wound, for concave surfaces such as the umbilicus, for areas of significant tissue tension, or for areas of marked tissue laxity.

Tape closure is fast, inexpensive, and associated with high patient satisfaction scores. Tapes typically are removed by the patient 7 to 10 days following surgery. They may also be used after staple removal to provide additional strength, as wounds have regained only approximately 3 percent of their final strength at 1 week. Adhesive tape strips are applied in a parallel, nonoverlapping fashion after coating the entire application area with adjuvant adhesive such as tincture of benzoin (Katz, 1999). Importantly, skin blistering may develop if tape is stretched excessively taut across the wound (Lammers, 2004; Rodeheaver, 1983).

Scalpel and Blades

Surgical instruments have been designed to extend the capability of a surgeon’s hands and thus are crafted to retract, cut, grasp, and clear the operative field. Tissue types encountered in gynecologic surgery vary, and accordingly, so too do the size, fineness, and strength of the tools used.

Of these tools, typical surgical blades used in gynecologic surgery are pictured in Figure 40-8 and include number 10, 11, 15, and 20 blades. Function follows form, and larger blades are used for coarser tissues or larger incisions, whereas a no. 15 blade is selected for finer incisions. The acute angle and pointed tip of a no. 11 blade can easily incise tough-walled abscesses for drainage, such as those of the Bartholin gland duct.

With a correct scalpel grasp, a surgeon can direct blade movement. Fingers may be positioned either to straddle the scalpel, termed the “power grip,” “violin grip,” or “bow grip,” which maximizes the use of the knife belly. Alternatively, the scalpel is held like a pencil, termed the “pencil grip” or “precision grip” (Fig. 40-9). With the no. 10 and no. 20 blades, the scalpel is held at a 20- to 30-degree angle to the skin and drawn firmly along the skin toward the surgeon using the arm with minimal wrist and finger movement. This motion aids cutting with the full length of the scalpel belly and avoids burying the tip. The initial incision penetrates the dermis, and the scalpel remains perpendicular to the surface to prevent skin edge beveling. Firm and symmetrical lateral skin traction keeps the incision straight and helps avoid multiple tracks and irregular skin edges.

The no. 15 and 11 blades, in contrast, are typically held using the pencil grip to make fine, precise incisions. With the no. 15 blade, the scalpel is held approximately 45 degrees to the skin surface. Fine knife dissection is best controlled using the fingers, and the heel of the hand can be stabilized on adjacent tissue. The no. 11 blade scalpel is ideal for stab incisions and is held upright at nearly 90 degrees to the surface. Creating tension at the skin surface is important as it reduces the amount of force required for penetration. Omission of this can result in uncontrolled penetration of underlying structures. To lengthen the incision, a gentle in-and-out sawing motion is used.

Scissors

These are used commonly to divide tissues, and modification in blade shape and size allows their use across various tissue textures (Fig. 40-10). For correct positioning, the thumb and fourth ­finger are placed within the instrument’s rings, and the index ­finger is set against the crosspiece of the scissors for greater control. This “tripod” grip allows maximum shear, torque, and closing forces to be applied and provides superior stability and control. In general, surgeons cut away from themselves and from dominant to nondominant sides.

The fine blades of Metzenbaum or iris scissors are used routinely to dissect or define natural tissue planes such as dividing thin adhesions or incising peritoneum or vaginal epithelium. During dissection, traction on opposing poles of the tissue to be dissected typically simplifies the process, and a small nick is often necessary to enter the correct tissue plane. The blades are closed and inserted between planes, following the natural curves of tissues being dissected (Fig. 40-11). The blades are opened and then withdrawn. After turning both wrist and blades 90 degrees, the surgeon reinserts the lower blade, and tissues are divided. When dissecting around a curve, the scissors follow the natural curve of the structure. Dissection proceeds in the same plane to avoid burrowing into the structure or deviating away and toward unintended adjacent tissues.

For thicker tissues, sturdier scissors such as curved Mayo scissors are used. Similarly, Jorgenson scissors have thick blades and tips that are curved at a 90-degree angle. These are used commonly to separate the vagina and uterus during the final steps of hysterectomy. Suture-cutting scissors have blunt, flat blades and are reserved for this function to avoid dulling tissue scissors.

Needle Holders

These may be straight or curved, and commonly, one with straight, blunt jaws is chosen during routine tissue approximation and pedicle ligation. Needles ideally pierce tissues perpendicularly. Thus, in most cases, the needle holder grasps a needle at a right angle and at a site approximately two-thirds from the needle tip.

Alternatively, some needle holders, such as the Heaney needle holder, are curved and aid needle placement in confined or angled areas. If a curved holder is used, the needle is grasped similarly, and the inner curve of the holder typically faces the needle swage (Fig. 40-12).

Traditionally, the needle holder is held with the thumb and fourth finger in the rings. The greatest advantage of this grip is the precision afforded. The spring tension of the handles is relieved from the lock in a controlled fashion, thereby releasing and regrasping the needle more precisely. Alternatively, with the “palmar grip,” the needle holder is held between the ball of the thumb and the remaining fingers, and no fingers enter the instrument rings. This grip allows a simple rotating motion for driving curved needles through an arc. Its greatest advantage is the time saved during continuous suturing, as the needle can be released, regrasped, and redirected efficiently without replacing fingers into the instrument rings. Disadvantageously, this grip has the potential to lack precision during needle release. When unlocking the needle driver, release of the spring lock should be smooth and gradual. This avoids an abrupt release, which may suddenly pop the handles apart with potential for awkwardness, loss of needle control, and tissue injury.

Tissue Forceps

Forceps function to hold tissue during cutting, retract tissue for exposure, stabilize tissue during suturing, extract needles, grasp vessels for electrosurgical coagulation, pass ligatures around hemostats, and pack sponges. Forceps are held so that one blade functions as an extension of the thumb and the other as an extension of the opposing fingers. Alternate grips may appear awkward and limit the full range of wrist motion, leading to suboptimal instrument use.

Heavy-toothed forceps, such as the Potts-Smith single-toothed forceps, Bonney forceps, and Ferriss-Smith forceps, are used when a firm grasp is more important than gentle tissue handling (Fig. 40-13A). These tools are often used to hold fascia for abdominal wound closure.

Light-toothed forceps, such as the single-toothed Adson, concentrate force on a tiny area and give more holding power with less tissue destruction. These are used for more delicate work on moderately dense tissue such as skin. Nontoothed forceps, also known as smooth forceps, exert their grip through serrations on the opposing tips (Fig. 40-13B). They are typically used for delicate tissue handling and provide some holding power with minimal injury. DeBakey forceps are another type of smooth forceps originally designed as vascular forceps but can be occasionally used for other delicate tissues. In contrast, the broader, shallow-grooved tips of Russian forceps and Singley forceps may be preferred if a broader or thicker area of tissue is manipulated.

Retractors

Abdominal Surgery Retractors

Clear visualization is essential during surgery, and retractors conform to body and organ angles to allow tissues to be pulled back from an operative field. In gynecology, retractors may be grouped broadly as self-retaining or handheld and as vaginal or abdominal.

During abdominal surgery, retractors that by themselves hold abdominal wall muscles apart, termed self-retaining, are used commonly. Styles such as the Kirschner and ­O’Connor-O’Sullivan contain four broad, gently curved blades and retract in four directions. Blades pull the bladder caudally and the anterior abdominal wall muscles laterally and cephalad. The Balfour retractor retracts in three directions but can be made to retract in four with the addition of an upper arm attachment. Alternatively, ring-shaped retractors such as the Bookwalter and Denis Browne styles offer greater variability in the number and positioning of retractor blades. However, these styles usually require more time to assemble and place. With all of these retractors, deep or shallow blades can be attached to the outer metal frame according to the abdominal cavity depth. As discussed earlier, blades should be shallow enough to avoid femoral nerve compression.

In addition to these metal bladed styles, several disposable retractors consist of two equal-sized plastic rings connected by a cylindrical plastic sheath. One ring collapses into a canoe shape that can be threaded through the incision and into the abdomen. Once inside the abdomen, it springs again to its circular form. The second ring remains outside. Between these rings, the plastic sheath spans the thickness of the abdominal wall and creates 360-degree retraction. As shown in Figure 44-8.7, Alexis or Mobius brands can be ideal for minilaparotomy, but sizes are also available for laparotomy.

Handheld retractors may be used in addition to or in place of self-retaining styles. These instruments allow retraction in only one direction but can be placed and repositioned quickly. The Richardson retractor has a sturdy, shallow right-angled blade that can hook around an incision for abdominal wall retraction (Fig. 40-14). Alternatively, Deaver retractors have a gentle arching shape and conform easily to the curve of the anterior abdominal wall. Compared with Richardson retractors, they offer increased blade depth and are often used to retract bowel, bladder, or anterior abdominal wall muscles. A Harrington retractor, also called a sweetheart retractor, has a broader tip that also effectively holds back bowel.

In certain instances, such as during suturing of the vaginal cuff, a thin, deep retractor blade, termed a malleable retractor, may be required to retract or protect surrounding organs. Also called a ribbon retractor, this tool is a long, relatively flexible metal strip that can be bent to conform to various body contours for effective retraction. Narrow and wider sizes are available. These also may be used to cover and protect underlying bowel from needle-stick injury during abdominal wall closure.

For smaller incisions, the preceding retractors are too large, and those with smaller blades such as the Army-Navy retractor or S-retractor are selected. S-retractors offer thinner, deeper blades, whereas the sturdier blades of the Army-Navy style allow stronger retraction (Fig. 40-15). A Weitlaner self-retaining retractor may also be used for minilaparotomy incisions.

Vaginal Surgery Retractors

The vaginal walls can be separated using several self-retaining models. The Gelpi retractor has two narrow teeth that are placed distally against opposing lateral vaginal walls and is most appropriate for perineal procedures (Fig. 40-16A). The Rigby retractor, with its longer blades, effectively separates lateral vaginal walls, whereas a Graves speculum, shown in Figure 1-6, holds apart anterior and posterior walls. An Auvard weighted speculum contains a long, single blade and ballasted end, which uses gravity to pull the posterior vaginal wall downward (Fig. 40-16B).

The degree of retraction offered by vaginal self-retaining retractors, however, at times may be limited. Therefore, handheld retractors are often required to augment or replace these instruments. Handheld retractors used in vaginal surgery include the Heaney right-angle retractor, the narrow Deaver retractor, and the Breisky-Navratil retractor (Figs. 40-17). Additionally, during vaginal procedures, the cervix often must be manipulated. Lahey thyroid clamps offer a secure grip during vaginal hysterectomy, but their several sharp teeth can cause significant trauma. These are therefore less than ideal in patients in whom the cervix will remain. In these patients, in whom curettage or laparoscopy is performed, a single-toothed tenaculum can afford a firm grip but with less cervical damage (Fig. 40-18).

Tissue Clamps

Retraction is a fundamental requirement during most gynecologic surgery. As a result, various shapes, sizes, and strengths of clamps have been created to manipulate the different tissues encountered. For example, the smooth, cupped jaws of a Babcock clamp are ideal for gentle elevation of fallopian tubes, whereas the serrated teeth of the Allis and Allis-Adair clamps can provide a fine, firm grip on covering epithelia or serosa during dissection (Fig. 40-19).

Clamps are also used to occlude vascular and tissue pedicles during organ excision. Hemostats and Mixter right-angle clamps have small, slender jaws with fine inner transverse ridges to atraumatically grasp delicate tissue, especially vessels (Fig. 40-20).

Heavier clamps are required to grasp and manipulate stiffer tissues such as fascia and include Pean (also termed Kelly) and Kocher (also termed Ochsner) clamps. These clamps have finely spaced transverse grooves along their inner jaws to minimize tissue slippage. They may be straight or curved to fit tissue contours and like Kocher clamps, may contain a set of interlocking teeth at the tip for additional grip security. Another choice, the ring forceps, has large open circular jaws with fine transverse grooves. These effectively grasp broad, flat surfaces. Additionally, a folded gauze sponge can be placed between its jaws and used to absorb blood from the operative field or gently retract tissues.

Ligaments that support the uterus and vagina are fibrous and vascular. Thus, a sturdy clamp that resists tissue ­slippage from its jaws is required during hysterectomy. Several clamps, including Heaney, Ballantine, Rogers, Zeppelin, and Masterson clamps, among others, are effective (Fig. 40-21). The thick, durable jaws of these clamps carry deep, finely spaced grooves or serrations arranged either transversely or longitudinally for secure tissue grasping. Additionally, some contain a set of interlocking teeth at the tip or heel or both. Although this modification improves grip, it also may increase tissue trauma. More acutely angled clamps are typically selected when available operating space is cramped.

Securing tissue pedicles may be accomplished using a variety of suturing techniques (Fig. 40-22). A single tie alone may be placed around the pedicle. In addition, a second distal transfixing suture can be placed to minimize dislodgement of the suture by vessel pulse pressures or pedicle manipulation.

Suction Tips

During gynecologic surgery, bleeding, peritoneal fluids, pus, ovarian cyst contents, and irrigants may obscure the operating field. Accordingly, choice of suction tip typically is dictated by the type and amount of fluid encountered. Adson and Frazier suction tips are fine bore and are useful in shallow or confined areas and when little bleeding is noted (Fig. 40-23).

Alternatively, a Yankauer suction tip offers a midrange-sized tip and is used commonly in general gynecology cases. However, if a larger volume of fluid or blood is expected, then a Poole suction tip may be desired. Its multiple pores allow continued suction even if some are obstructed with clot or tissue. In addition to removing large volumes of fluid quickly, this suction tip’s sieved sheath may be removed. The thinner-bore inner suction cannula can then be used for finer suctioning. Larger-bore Karman suction cannulas are used for products of conception evacuation and are discussed in Section 43-16.

These are foundational tools of tissue approximation, vessel ligation, and wound closure. They are crafted in various strengths, shapes, and sizes to meet surgical needs. Appropriate selection can profoundly affect wound healing and patient recovery. Thus, surgeons should be familiar with their characteristics and most appropriate applications.

Needles

The ideal surgical needle pierces tissue with ease, with minimal tissue damage, and without bending or breaking. Tissues differ in their density and location, and thus needles are designed with variable sizes, shapes, and tips. The anatomy of a needle is simple: each contains a tip, body, and site of suture attachment (Fig. 40-24). For most gynecologic cases, the suture and needle used are attached as a continuous unit, which is described as swaged. This contrasts with needles that have eyes for suture threading. Swaged needles may be firmly secured to the suture and require cutting at the end of suturing. Alternatively, controlled-release, or “pop-off,” swaged needles detach from the suture with a brisk tug. Controlled-release needles are often used when securing vascular pedicles or placing interrupted sutures. Continuous running suturing typically requires a swaged needle without the controlled-release feature.

In certain urogynecologic procedures, such as abdominal sacrocolpopexy, a double-armed suture is often chosen. This suture contains identical swaged needles at each of its ends. This design enables surgeons to suture distant tissues with different ends of the suture before approximating them.

Descriptors of needle size and shape are noted in Figure 40-24. Of these, needle radius, circle configuration, and gauge more frequently influence surgical selection. For example, a needle should be large enough to pass completely through the tissue and exit far enough to allow the needle holder to be ­repositioned on the end of the needle at a safe distance from the tip. Repeated grasping of the needle tip leads to a dulled tip. A dulled tip subsequently leads to difficult tissue penetration and greater tissue trauma.

For thicker tissues, a larger radius and gauge are warranted. For confined surgical spaces, a needle with smaller radius and greater circle configuration typically is required. Thus, for most gynecologic procedures, a three-eighths or one-half circle configuration is used. For some urogynecologic operations, a five-eighths circle configuration is preferred.

The tip should allow passage of the needle through tissue with the smallest amount of tissue damage. Those with tapered points are used for suturing thin tissues, such as peritoneum (Fig. 40-25). Alternatively, cutting needles are preferred for denser tissue such as fascia and ligaments. Cutting points have sharp edges laterally and a third sharp edge extending either toward or away from the needle’s inner curve. A conventional cutting needle features the third cutting edge on the inside curve and provides shallower tissue bites. In contrast, reverse cutting needles have the third cutting edge directed away from the inner curve of the needle and are used for particularly tough tissues.

Sutures

Sutures maximize wound healing and tissue support and are categorized by their biologic or synthetic derivation, their filamentous structure, and their ability to be degraded and reabsorbed (Table 40-3). Other qualities are described in the subsequent paragraphs.

Sutures such as catgut, silk, linen, and cotton are derived from biologic sources. As a group, biologic sutures produce the greatest tissue reaction and have the lowest tensile strength profile. Accordingly, most suture materials currently used in gynecologic surgery are synthetic.

Of synthetic materials, the number of strands that make up a given suture defines it as either monofilament or multifilament. Monofilament suture is constructed as a single strand, whereas multifilament suture contains multiple strands that are braided or twisted. Monofilament sutures have lower friction coefficients and therefore pull more easily through tissues, and as a result, they create less tissue injury. As a group, they tend to incite less tissue reaction. Moreover, braid crevices are absent, and bacteria therefore are less likely to adhere (Bucknall, 1983; Sharp, 1982). However, monofilament sutures are in general less pliant for knot tying and, if nicked by instruments, are more prone to break.

The diameter of a suture reflects its size and is measured in tenths of a millimeter (Table 40-4). A midpoint diameter size is designated as 0, and as suture diameter increases above this, arabic numbers are assigned. For example, no. 1 catgut is thicker than 0-gauge catgut. As suture diameter decreases from this midpoint, 0s are added. By convention, an arabic number followed by a 0 also may be used to reflect the total number of 0s. For example, 3-0 suture also may be represented as 000. Moreover, 3-0 suture is greater in diameter than 4-0 (0000) suture.

Ideally, the appropriate suture caliber is fine enough to limit tissue damage during placement and minimize subsequent tissue reaction yet provide ample tensile strength to support and approximate involved tissues. Defined as the amount of weight necessary to break a suture divided by its cross-sectional area, tensile strength is an important characteristic for suture selection. The tensile strength of material chosen should approximate the strength of the tissues being sutured.

Tensile strength is lost at different rates among suture types. Materials that have lost most of their tensile strength by 60 days following surgery are considered to be absorbable (Bennett, 1988). Absorbable suture is destroyed enzymatically or hydrolyzed, whereas nonabsorbable suture persists and ultimately is encapsulated. Ideally, absorbable suture material remains throughout wound healing but no longer. Logically, individual tissue healing characteristics typically dictate whether short- or long-term sutures are required. Accordingly, nonabsorbable material plays a greater role in pelvic floor reconstruction procedures, whereas absorbable suture is used routinely in general gynecologic surgery.

All sutures, when placed within tissue, will incite inflammation. This response mirrors the total amount of suture placed and the suture’s chemical composition (Edlich, 1973). In general, lower inflammatory responses are elicited by monofilament structure compared with multifilament, and synthetically derived compared with natural fiber (Sharp, 1982).

The ease of fluid to wick from the wet end of a suture to its dry end defines its capillarity. A suture’s fluid absorption ability describes the amount of fluid it absorbs when immersed. Both properties are presumed to have an impact on the access of contaminating bacteria. Increased capillarity and fluid absorption ability greatly increase the amount of bacteria similarly absorbed (Blomstedt, 1977). In general, multifilament sutures, even those with coatings, display greater capillarity compared with synthetic monofilament sutures (Geiger, 2005).

The ability of a material to return to its prior length following stretch defines its elasticity. For tissues in which swelling or movement is expected postoperatively, a suture with increased elasticity is preferred because it will stretch rather than cut into approximated tissues. Memory defines the ability of material to return to original form following deformation. Sutures with greater memory tend to untie more easily during knot tying.

Knots

The surgical knot is the weakest link in a tied suture loop, and the force necessary to break a knotted suture is less than that to break an individual suture strand. Knot failure can lead to serious complications such as bleeding, hernias, and wound dehiscence (Batra, 1993; Trimbos, 1984). Thus, an understanding of knots is essential.

A surgical knot consists of: a loop, which maintains tissue apposition, and a knot, composed of several throws snugged against each other. A single throw is formed when one strand is wrapped around the other one time, and when wrapped twice, a double throw is created (Zimmer, 1991). This double weave forms the basis of a surgeon’s knot. In characterizing knots, each throw is given a numerical description, in which single throws are designated as number 1, and double throws as number 2. If successive throws are identical, a multiplication sign is placed between the numbers. If throws mirror one another, then an equal sign is used. Thus, a square knot is described as 1 = 1; granny knot, 1 × 1; and square surgeon’s knot, 2 = 1 (Fig. 40-26). Alternative nomenclature schemes exist, but understanding the basic principles of knot construction is more clinically relevant than these descriptive definitions (Dinsmore, 1995).

Flat and Sliding Knots

Surgical knots can have flat or sliding configurations. Flat configurations include square, granny, and surgeon’s knots. To construct a flat square knot, forehanded and backhanded throws are alternated, and the suture strands are pulled with equal tension in opposite directions but in the same plane. In addition, the suture strands or the hands may have to cross with each throw to ensure that the knot lies flat.

In contrast, sliding knots, also termed slip knots, are characterized as identical, nonidentical, and parallel. They are created when unequal tension is applied to the strands, such as during one-hand knot tying. Sliding knots are useful in situations when flat square knotting is difficult or cumbersome, such as in the deep pelvis or vagina (Ivy, 2004b). In general, sliding knots have been shown to have a higher failure rate than that of flat knots (Hurt, 2005; Schubert, 2002).

Identical sliding knots are created by holding one strand constantly under tension and repeating identical tying maneuvers with the other hand. Unfortunately, these identical sliding knots carry a high failure rate and are not recommended for general use (Schubert, 2002; Trimbos, 1984, 1986). Nonidentical sliding knots are formed when a suture strand is held under constant tension, and one hand alternates forehanded and backhanded tying around this strand (Trimbos, 1986). This knot is the most frequent and practical knot used for vaginal surgery. Although these knots can unravel, additional throws can greatly improve their security (Ivy, 2004a; Trimbos, 1984; van Rijjsel, 1990). A loop-to-strand variation of the nonidentical sliding knot is performed by holding the final loop of a continuous suture line taut, while alternate throws are made around the loop with the remaining single strand. Scant data support the security of this knot type in gynecologic surgery. When completed with monofilament suture, these knots carry high failure rates (Hurt, 2005).

Finally, with a parallel sliding knot, the suture strand under tension is alternated with each throw, causing alternate throws to slide down the other strand each time. Existing studies show this knot to be strong and reliable (Ivy, 2004b; Trimbos, 1986).

Surgical Knot Effectiveness

The effectiveness of surgical knots depends mainly on two parameters: initial loop security and knot security. Loop security describes the ability to maintain a tight suture loop around the tissue as the initial knot throws are placed (Lo, 2004). Suture loops that are initially loose will fail to secure tissues no matter how tightly the knot is tied and will result in ineffective knots, colloquially termed “air knots” (Burkhart, 1998). Three ways to optimize loop security include: maintenance of tension on both strands during tying, use of an initial surgeon’s throw, and slip knots (Anderson, 1980). If slip knots are placed initially, they can be converted to square knots or reinforced with a square knot once the pedicle or vessel is secured. Importantly, upward tension on both strands deep within a body cavity should be limited. Excessive force can avulse the pedicle or cause the suture loop to pull completely off.

For knot security, the tension with which a given throw is tied is the most important. A knot laid down tightly under great tension is less likely to slip than a knot with the same configuration but with more throws tied loosely (Gunderson, 1987).

The number and type of knots required to secure various suture materials vary. Qualities such as elasticity and memory often direct these recommendations. In general, multifilament sutures are easier to handle and display less memory, whereas synthetic monofilament suture or multifilament sutures with coatings have increased memory and may hold a knot poorly. For most sutures, four to six throws appears to be adequate, but the exact number depends on the type of suture and whether a flat or sliding knot is formed. Up to a point, additional throws provide more security to a knot, but this benefit must be balanced against the corresponding elevated infection risk from increased knot volume (van Rijssel, 1990).

Semantically, electrosurgery differs from electrocautery, although the terms are often incorrectly interchanged. With electrocautery, electric current passes through a metal object, such as a wire loop, with internal resistance. Passage of current heats the loop, which then may be used surgically. The flow of current is limited to the metal being heated, and no current enters the patient. In contrast, electrosurgery directs the flow of current to the tissues themselves and produces localized tissue heating and destruction. As a result, electric current must pass through tissues to produce the desired effect (Amaral, 2005). The electrosurgical circuit contains four main parts: the generator, the active electrode, the patient, and the return electrode. Electrosurgery may be broadly categorized as monopolar or bipolar depending on the proximity of these two electrodes.

Monopolar Electrosurgery

Electric current is the flow of electrons through a circuit (Fig. 40-27). Voltage is the force that drives those charges around the circuit. Impedance is the combination of resistance, inductance, and capacitance that alternating current meets along the way (Morris, 2006). In monopolar electrosurgery, the return electrode in clinical use is the grounding pad. Current therefore flows: (1) from the generator, which is the source of voltage, (2) through the electrosurgical tip to the patient, the source of impedance, and then (3) onto the grounding pad, where it is dispersed. Current leaves the pad to return to the generator, and the circuit is completed (Deatrick, 2015). In electrosurgery, tissue impedance converts electric current into thermal energy that causes tissue temperatures to rise. It is these thermal increases that create electrosurgery’s tissue effects.

Surgical Effects

Differing tissue effects are created by altering the manner in which current is produced and delivered. First, altering the current wave pattern can affect tissue temperatures. For example, the high-frequency continuous sinusoidal waveform produced with cutting current creates higher tissue temperatures than that with coagulation current (Fig. 40-28). Second, the extent to which current is spread over an area, also termed current density, alters the rate of heat generation (Fig. 40-29). Thus, if current is concentrated onto a small area, such as a needle-tip electrode, greater tissue temperatures are generated than if delivered over a wider area, such as an electrosurgical blade. In addition to current density, voltage can alter tissue effects. As voltage increases, the degree of thermal tissue damage similarly increases. And finally, the qualities and impedance of the tissues themselves affect energy transfer and heat dissipation. For example, water has low electrical impedance and liberates little heat, whereas skin with its greater impedance generates significantly higher tissue temperatures (Amaral, 2005).

With electrosurgical cutting, a continuous sine wave of current is produced. The flow of high-frequency current typically is concentrated through an electrosurgical needle or blade and meets tissue impedance. Sparks are created between the tissue and electrode, intense heat is produced, cellular water vaporizes, and cells in the immediate area burst. Tissues are cut cleanly, and there is minimal coagulum production. As a result, few vessels are sealed, and minimal hemostasis accompanies electrosurgical cutting.

In contrast, coagulation current does not produce a constant waveform. Less heat is produced than with cutting current. However, tissue temperature still rises sufficiently to denature protein and disrupt normal cellular architecture. Cells are not vaporized instantly, and cellular debris remains associated with wound edges. This coagulum seals smaller blood vessels and controls local bleeding (Singh, 2006).

Variations in the percentage of time that current is flowing can create electrosurgical effects that contain both cutting and coagulating features. Such blended currents are used commonly in gynecologic surgery. In most cases, selection of specific percentages of cutting and coagulation current is affected by surgeon preference and type of tissues encountered. Thinner vascular tissue may be best suited for a blend with less active current time, whereas denser avascular tissues may require a greater percentage of active current.

Patient Grounding

As discussed earlier, current is concentrated at the electrode tip and enters the patient at a small site. Current follows the path of least resistance and exits the body through a grounding pad that is designed to have a large surface area, high conductivity, and low resistance. Dissipation across this large surface area allows current to leave the body without generating significant tissue temperatures at the exit site.

However, patient burns may result if current is concentrated through a return electrode. Clinically, this may occur if a grounding pad is partially dislodged. In this setting, the surface area is decreased, and the exiting current concentration and the tissue temperatures at the exit site rise. In addition, patient jewelry, metal candy cane stirrups, or other surfaces with high conductivity and low resistance may serve as a return electrode. In such cases, patients may be burned by concentrated current exiting through these small contact sites. Ideally, grounding pads are firmly adhered to a relatively flat body surface that is near the operative field. Thus, in most gynecology procedures, grounding pads are placed along the lateral upper thigh.

Patients with pacemakers, implantable cardioverter-defibrillators (ICDs), or other electrical implants require special precautions. Stray electrosurgical current may be interpreted as an intracardiac signal by an implanted device and lead to pacing changes. In addition, myocardial electrical burns may result from conduction of current through the pacing electrode rather than through the grounding pad (Pinski, 2002). Accordingly, for patients with these devices, preventive recommendations include pre- and postoperative cardiology consultation; use of minimal monopolar electrosurgical current settings or substitution with bipolar electrosurgery or harmonic scalpel; continuous cardiac monitoring; contingency plans for arrhythmias; and close proximity of active and return electrodes (Crossley, 2011).

Bipolar Electrosurgery

This form of electrosurgery differs from monopolar electrosurgery in that the tip of a bipolar device contains both an active electrode and a return electrode. For this reason, a distant grounding return pad is not required. Coagulation current is concentrated on tissues grasped between the electrodes, and tissue must remain between them. If tissue slips from between the tips, then active and return electrodes contact and create a short circuit. Coagulation will not occur (Michelassi, 1997). Bipolar electrosurgery uses only coagulation current and lacks cutting capability. However, it is useful for vessels coagulation and also is used during laparoscopic sterilization to coagulate fallopian tubes (Section 44-2).

Argon Beam Coagulation

With argon beam coagulation (ABC), radiofrequency energy is transferred to tissues through a jet of inert argon gas to create noncontact monopolar electrothermal coagulation. Additionally, the gas jet clears blood and tissue debris during coagulation. Advantages of ABC include the ability to coagulate broad surface areas and larger vessels (Beckley, 2004). In gynecologic surgery, ABC is used most often during ovarian staging cases in which extensive debulking may be required.

Sound waves are mechanical waves that transport energy through a medium. Those above audible range are described as ultrasound or ultrasonic. In medicine, ultrasound waves that are applied at low levels such as those used in diagnostic sonography are harmless. However, if higher power levels are used, then mechanical energy transferred to the affected tissues is strong enough to produce cutting, coagulation, or tissue cavitation.

Of tools, the ultrasonic scalpel, also known as a Harmonic scalpel, has a tip that vibrates at high frequency. This allows the surgical device to be used effectively for both cutting and coagulating during laparotomy or laparoscopy (Gyr, 2001; Wang, 2000). The vibrating tip transfers mechanical energy to tissues. Mechanical energy breaks hydrogen bonds and generates heat within tissues to denature protein and form a sticky coagulum that produces hemostasis. A balance between cutting and coagulation is controlled by three factors: power levels, tissue tension, and blade sharpness. Higher power level, greater tissue tension, and a sharp blade will lead to cutting. Lower power, decreased tissue tension, and a blunt blade will create slower cutting and greater hemostasis (Sinha, 2003).

Used most commonly in laparoscopic surgery, the ultrasonic scalpel serves as an alternative to suture ligation, electrosurgical coagulation, laser, and stapling or clipping devices. However, only a few studies have compared the clinical effectiveness of this method with other methods of hemostasis (Kauko, 1998).

Cavitational ultrasonic surgical aspiration (CUSA) is another ultrasound-based tool. The ultrasonic surgical aspirator hand-piece contains three main components: a high-frequency vibrator, which transfers the ultrasonic energy to tissues; irrigation tubing, which directs cooling saline to the tip; and a suction system, which draws tissue up to the tip for contact with the vibrator and which also clears away tissue fragments and irrigant. Ultrasound energy can be used to raise tissue temperatures dramatically and thereby disrupt tissue architecture by a process termed cavitation. For cavitation, the CUSA tip produces mechanical waves that create heat and vapor pockets around cells in tissues with high water content such as adipose, muscle, and carcinoma. Collapse of these pockets leads to disruption of cell architecture (Jallo, 2001). Affected tissues are removed subsequently by suction aspiration. However, tissues containing less water and higher contents of collagen and elastic fibers, such as blood vessels, nerves, ureters, and serosa, are more resistant to damage (van Dam, 1996).

In gynecology, CUSA has a limited surgical role. It may be used effectively in the treatment of vulvar intraepithelial neoplasia and bulky condyloma acuminata (Section 43-28). It can also speed cytoreductive ovarian cancer surgery (Aletti, 2006; Deppe, 1988; Robinson, 2000; van Dam, 1996).

Autologous Donation and Cell Salvage

Although the risk of hemorrhage accompanies most gynecologic procedures, certain factors are associated with higher rates of bleeding and are assessed prior to surgery. Specifically, obesity, the presence of a large pelvic mass, adhesions such as those from endometriosis or pelvic inflammatory disease, cancer or prior radiation, and coagulation dysfunction have been linked with an increased risk of hemorrhage. For those identified to be at risk, intraoperative red cell salvage or preoperative autologous blood donation can be considered.

Red blood cell (RBC) salvage machines (Autolog; Cell Saver) collect, filter, and centrifuge blood lost during surgery. RBCs are heavier and are separated from plasma and smaller blood components during centrifugation and are then reinfused into the patient. Anticoagulants such as heparin or citrate are added to prevent clotting (Karger, 2005). Salvage efficiencies approximate 60 percent with good technique. However, vacuum levels, suction tip size, and thoroughness of salvaging efforts can affect this value. For example, turbulence destroys RBCs. Thus, suction tips with greater diameters and lower suction force can minimize hemolysis (Waters, 2005). Additionally, laparotomy sponges can be rinsed in sterile saline to maximize RBC removal. The RBC-containing saline then is suctioned into the salvage device for processing. The filtering systems in these devices have limitations. Accordingly, RBC salvage is not appropriate for contaminated cases or those in which malignancy, hemostatic agents, or amnionic fluid may be present (Waters, 2004).

Another approach for cases with expected hemorrhage involves preoperative autologous donation. Thus, to avoid potential transfusion reaction or blood-borne infection, a patient may elect to donate her own blood for personal use approximately once a week for 3 to 5 weeks preceding surgery. Patient hemoglobin levels should be greater than 11.0 g/dL before each donation. Moreover, units are not collected within 72 hours before surgery. This allows intravascular volume to be replenished by the patient and units to be processed by the blood bank (Goodnough, 2005). Disadvantageously, this process has been associated with preoperative anemia secondary to donation, more liberal transfusion, transfusion reaction following clerical error, volume overload, and bacterial contamination of blood products during processing (Henry, 2002; Kanter, 1996, 1999).

Improved blood banking safety has accompanied a decline in preoperative autologous donation (Brecher, 2002). Moreover, for most gynecologic cases, the risk of transfusion is low. For these reasons, autologous donation typically is reserved for selected instances in which the risk of transfusion is significant, such as radical hysterectomy or surgery for patients with coagulopathies. Additionally, patients with rare blood phenotypes in whom acquisition of compatible blood may be difficult may benefit from autologous donation.

Proper Surgical Method

In many instances, proper surgical technique can minimize vascular injury and hemorrhage. Thus, prior to ligation, vessels ideally have excess connective tissue removed sharply in a process called skeletonizing. Additionally, tissue clamps selected for grasping a vascular pedicle are large enough to contain the entire pedicle in the distal portion of the clamp. Large pedicles that force excess tissue toward the clamp’s heel carry greater risk of tissue slipping from the heel and bleeding. Once secure, sutures placed on vascular pedicles are not to be used for traction because the risk of avulsing the suture or vessel increases.

Tying an intact vessel at two places along its length before tissue cutting is considered in certain situations. This technique may be appropriate if a vessel is on tension or if space for a clamp is limited, such as when the ureter or bowel is in close proximity. A window is created below the vessel and ties are passed beneath the vessel before doubly ligating and dividing it (Fig. 40-30).

Steps of Hemorrhage Management

A methodical approach to intraoperative hemorrhage is critical to minimize patient morbidity. If an isolated vessel is clearly identified, then grasping it with a hemostat, vascular clamp, or fine forceps may allow ligation, electrosurgical coagulation, or vascular clip application.

In contrast, venous bleeding in the pelvis is typically from a venous plexus and rarely stems from a single vessel. A pelvic venous plexus contains thin-walled veins. Accordingly, indiscriminate clamping, suturing, clipping, and electrosurgical coagulation can cause further laceration and bleeding. However, if other vulnerable structures have been retracted and protected, a few shallow stitches that incorporate the bleeding area can be placed using fine absorbable suture.

If these initial efforts are unsuccessful and significant hemorrhage continues, the bleeding site is compressed with fingertips, sponge stick, or laparotomy sponges. Anesthesia staff is informed of events to allow for additional monitoring. Nursing staff is also informed as additional resources may be required, such as specialized instruments, suture, clips, and blood products. Fluid resuscitation is individualized depending on the degree of hemorrhage and other patient factors described later.

Adequate exposure typically aids control of bleeding. The operative field is assessed and increased as needed by extending a vertical incision cephalad, converting a Pfannenstiel incision to a Cherney incision, adding retractors, or converting a vaginal or laparoscopic approach to laparotomy. A second suctioning system may be required, and appropriate suture or clips are made available before removing pressure. Additional dissection of avascular planes around the bleeding site may improve isolation and ligation of a lacerated vessel. Furthermore, nearby vulnerable structures such as the bladder, ureter, or other vessels are identified and protected. After these steps, the surgeon may remove the tamponading pressure to assess the location, amount, and character of bleeding and to formulate the most appropriate technique for controlling it.

Vessel Ligation

As noted, bleeding vessels may be tied, clipped, or coagulated. Advantages to suture ligation include low cost and effectiveness over a broad range of vessel diameters. However, knot tying in general is time-consuming, is difficult in narrow spaces, and can be associated with ligature slippage or breakage. Small vessels may be ligated by a free-tie suture placed around the heel and tip of a vascular clamp and then secured with knots (see Fig. 40-23A). Alternatively, surgeons often prefer to secure larger vascular pedicles with two separate sutures. The first ligature is a free tie placed around the toe and heel of a vascular clamp and tied. The second ligature is distal to the first and typically incorporates a bite through the tissue pedicle (see Fig. 40-23B, C). Such transfixion of the ligature to the pedicle decreases the risk that it will slip off the pedicle’s end. Importantly, this second ligature is placed distal to the first to avert hematoma formation if a vessel is pierced during transfixion.

Alternatively, titanium clips seal vessels by direct compression. They are used more commonly during gynecologic onco­logy cases and offer the advantage of speed. However, clips are expensive, require surgical dissection of the vessel prior to application, and may dislodge from a vessel. Their use in routine gynecology is limited by these factors and surgeon preference.

Electrical and ultrasound energy also may be used to seal vessels. Ultrasonic coagulating shears (Autosonix; Harmonic scalpel; Sonosurg) and electrosurgical bipolar vessel sealing clamps (Enseal; LigaSure) transfer energy that denatures vascular collagen and elastin. Some of these seal vessels up to 7 mm in diameter (Heniford, 2001). Thermal spread for these devices is comparable and averages 2.5 mm (Harold, 2003). These tools are particularly useful for laparoscopic surgeries, in which knot tying is time-consuming.

Local Topical Hemostats

These topical products may be placed on bleeding sites where ligature or vessel coagulation is not possible or has been ineffective. They are most effective in controlling low-pressure bleeding, such as from veins, capillaries, and small arteries. Commercially available materials are categorized as mechanical hemostats, active hemostats, flowable hemostats, and fibrin sealants (Table 40-5). Some liquid hemostats deliver topical thrombin or thrombin plus fibrinogen and thereby induce clot formation. Others provide combined effects that, in sum, create direct pressure against wound surfaces, entrap platelets, promote platelet aggregation, and serve as a scaffold on which clot can organize.

Although effective, these agents do have disadvantages. They should not be introduced intravascularly or used with RBC salvage machines. Packing agents tightly into bony foramina is avoided because these agents can swell and cause neurologic dysfunction or pressure necrosis. Moreover, they are not placed within skin edges because they may retard edge reapproximation. They may serve as an infection nidus and thus may not be appropriate in grossly infected tissue (Baxter Healthcare, 2014; Pfizer, 2014). Few data support the use of one agent over another. Selection typically is dictated by surgeon preference and availability in the operating room.

Tranexamic Acid

Intravenous administration of tranexamic acid, an effective antifibrinolytic, has been widely investigated in trauma surgery (Hunt, 2015). It is a synthetic lysine derivative, which blocks the conversion of plasminogen to plasmin, as illustrated in Figure 8-12. Massive hemorrhage may be complicated by coagulopathy and uncontrollable microvascular hemorrhage. Tranexamic acid, 1 g given intravenously, may be of benefit in this setting in concert with pelvic pressure and blood component administration.

Pelvic Artery Embolization or Pelvic Packing

As described in Chapter 9, embolization similar to that used to treat symptomatic leiomyomas can be used to occlude either the internal iliac artery or the uterine artery. This technique has been described in the management of hemorrhage in both gynecologic and obstetric cases.

In other cases, for patients with persistent heavy bleeding despite attempts at control, pelvic packing with gauze and termination of the operation may be warranted. Rolls of gauze are packed against the bleeding site to provide constant local pressure. Typically, 24 to 48 hours later, if the patient is stable and bleeding appears to have stopped clinically, packing is removed. Some surgeons recommend leaving one end of the gauze outside the wound. After administration of general anesthesia, packing is pulled slowly through a small opening left in the incision. Alternatively, entire gauze rolls may be packed into the abdomen and removed during a second laparotomy (Newton, 1988).

Internal Iliac Artery Ligation

The internal iliac artery, also called the hypogastric artery, contains anterior and posterior divisions. Its anterior division supplies blood to central pelvic viscera (Fig. 38-12). Occlusion of the internal iliac artery decreases mean blood flow by 48 percent in branches distal to ligation, which in many cases slows hemorrhage sufficiently to allow identification of specific bleeding sites (Burchell, 1968). Fortunately, the female pelvis has extensive collateral circulation, and the internal iliac artery shares arterial anastomoses with branches of the aorta, external iliac artery, and femoral artery. For this reason, ligation of the internal iliac’s anterior division can be performed without compromise to pelvic organ viability. Several studies have described normal postligation fertility in these women (Demirci, 2005; Nizard, 2003).

To perform ligation, the round ligament is divided, and the pelvic sidewall peritoneum lateral to the infundibulopelvic ligament is incised cephalad. Identification of the internal iliac artery is essential because ligation of the common or external iliac arteries will have vascular consequences to the lower extremity. Once the internal iliac artery is located, a Mixter right-angle clamp is placed under the vessel at a point 2 to 3 cm distal to the bifurcation of the common iliac artery. Two free ties of no. 1 or 0 absorbable suture are passed beneath the artery and then secured (Fig. 40-31). The artery is ligated but not transected. If the internal iliac is ligated at this site, its posterior division theoretically should be spared (Bleich, 2007). Care is required in passing instruments beneath the artery because the thin-walled internal iliac vein is easily lacerated.

Specific Sites of Bleeding

Infundibulopelvic Ligament

During or after ligation of this vascular pedicle, a lacerated ovarian vessel may retract into the retroperitoneum to create a hematoma. In most cases, isolation of the bleeding vessel is required to halt hematoma expansion. For this, the pelvic sidewall peritoneum lateral to the ureter and the hematoma is opened, and the incision is extended cephalad to the upper pole of the hematoma. The incision in the peritoneum may be carried up the white line of Toldt. This line, found on the right and the left, is the peritoneum’s posterior reflection over the mesenteric attachment site of the ascending or descending colon to the posterior abdominal wall. The upper pole of the hematoma is identified by a return to normal vessel caliber above the hematoma. The ovarian vessels are identified, and a closed Mixter right-angle clamp is placed beneath them. A tie on a pass then is threaded beneath and used to ligate these vessels. If large, the hematoma then is evacuated to minimize infection risk (Tomacruz, 2001). In rare cases in which vascular or ureteral anatomy is unclear, an ovarian artery may require ligation as proximal as its aortic origin below the renal arteries (Masterson, 1995).

Space of Retzius and Presacral Venous Plexus

The space of Retzius, also called the retropubic space, is often entered during urogynecologic procedures and contains important vascular structures such as the venous plexus of Santorini, the obturator vessels, and the aberrant obturator vessel (Fig. 38-24). Approximately 2 percent of tension-free vaginal tape procedures are complicated by bleeding in this space (Kolle, 2005; Kuuva, 2002). In most instances, bleeding is controlled with pressure or suturing.

In contrast, the presacral venous plexus can be lacerated by dissection or suturing during sacrocolpopexy (Fig. 38-23). Cut vessels may retract into the vertebral bone, and problematic bleeding can follow. Management is described in full in Section 45-17.

Major Pelvic Vessels

High-volume pelvic vessels include the internal, external, and common iliac vessels, the inferior vena cava, and aorta. These may be lacerated during tumor removal, endometriosis excision, or laparoscopic trocar placement.

Initially following large vessel injury, pressure is applied for several minutes. Although gynecologic surgeons may attempt to repair these injuries, excessive delay in obtaining vascular surgery assistance often leads to greater blood loss (Oderich, 2004). Therefore, in many instances, pressure is applied, a vascular surgeon is consulted for repair, blood products are made available, and exposure is maximized. If a large vessel is punctured by a trocar or needle during laparoscopic entry, the instrument should remain in place to act as a plug while preparations for repair are made.

As discussed earlier, internal iliac artery ligation does not lead to ischemia of central pelvic organs due to collateral blood supply. However, injury to the external or common iliac arteries requires repair to maintain blood supply to the lower extremity. Consultation with a vascular surgeon may be indicated depending on the degree of laceration and surgeon skill. Maneuvers that may extend the injury are avoided until appropriate assistance is available.

If repair is undertaken, familiarity with this vascular anatomy is essential. On the left, the common and external iliac arteries remain lateral to their respective veins. On the right, however, the common iliac artery runs medial to the vein.

These arteries can be repaired by placing vascular clamps 2 to 3 cm proximal and distal to the tear, then closing the defect with a continuous suture line using monofilament synthetic 5-0 suture (Gostout, 2002; Tomacruz, 2001). The proximal clamp is removed first to allow air and debris to exit the suture line, and then the distal clamp is removed.

Parametrial and Paravaginal Vessels

During obstetric and gynecologic surgery, vessels supplying the uterus and vagina, especially venous plexuses, can be lacerated. At times, bleeding may not be easily identified and controlled by direct pressure, suturing, or clips. In these extreme situations, ligation of the internal iliac artery, which is a main source of blood supply to the pelvis, may decrease pooling of blood and afford a better opportunity to find a bleeding source. Alternatively, if resources are available, pelvic artery embolization is effective in controlling pelvic hemorrhage. Despite these techniques, in rare persistent situations, pelvic packing and termination of surgery may be indicated.

The intraoperative and immediate postoperative periods are viewed as resuscitative phases, with a goal of attaining relative euvolemia. With acute hemorrhage, priorities include control of additional losses and replacement of sufficient intravascular volume for tissue perfusion and oxygenation. In hypoperfused areas, progressive failure of oxidative metabolism with lactate production leads to worsening systemic metabolic acidosis and eventual organ damage. To avoid these effects, resuscitation begins with an assessment of the patient’s clinical status, calculation of total blood volume, and estimation of blood loss.

Clinical Assessment

Total blood volume for an adult approximates 70 mL/kg, and thus a 50-kg woman’s calculated blood volume is 3500 mL. Of this volume, 15 percent can be lost by most patients with no changes in arterial pressure or heart rate. A 15-percent blood loss can be roughly calculated by multiplication of a patient’s weight in kilograms by 10. Thus, for a 50-kg woman, a 15-percent loss approximates 500 mL.

With losses of 15 to 30 percent (500 to 1000 mL for a 50-kg woman), tachycardia and narrowing of the pulse pressure are seen (Table 40-6). Peripheral vasoconstriction leads to pale, cool extremities and poor capillary refill. In unanesthetized patients, there may be mild confusion or lethargy. In most women with normal preoperative hemoglobin levels, this amount of blood loss requires fluid volume replacement, but RBC transfusion typically is not required. Greater losses, however, lead to worsening perfusion, hypotension, and tachycardia. In these cases, blood transfusion in combination with fluid resuscitation typically is indicated (Murphy, 2001).

During surgery, blood collects in suction canisters and laparotomy sponges. Although calculations from these sources provide surgeons with an approximation, blood loss estimates typically are low, and inaccuracy increases as the length and extent of a procedure increases (Bose, 2006; Santoso, 2001). Additionally, a hematocrit may be measured to assess hemorrhage. However, hematocrit values typically lag true losses, and values may reflect only the degree of hemorrhage. For example, following a blood loss of 1000 mL, hematocrit levels typically fall only 3 volume percent in the first hour but usually show an 8 volume percent drop at 72 hours (Schwartz, 2006). Hemorrhage leads to global tissue hypoxia, anaerobic metabolism, and lactate production. Thus, elevated serum lactate levels can be a helpful marker. Blood gas analysis can also provide a rapid estimate of the serum base deficit. Hemorrhage severity can be predicted using the following stratification of these base deficits: 2 to –5 (mild hemorrhage), –6 to –14 (moderate hemorrhage), and –15 or less (severe hemorrhage). If patients continue to have dropping base deficits despite aggressive resuscitation, ongoing hemorrhage is a concern (Davis, 1988).

Fluid Resuscitation

If hypovolemia is identified, fluid resuscitation begins with crystalloid solutions. If hypotension and tachycardia are present, rapid replacement is warranted, and 1 or 2 liters, as indicated, may be infused over several minutes. Normal saline and lactated Ringer solutions are the two crystalloids used commonly, and their composition is described in Chapter 42. For moderate hemorrhage, both perform equally well as fluid replacements (Healey, 1998).

Although crystalloids have an immediate effect to expand intravascular volume, a portion will extravasate into extracellular tissues. Thus, in the setting of hemorrhage, crystalloid volume is administered in a 3:1 ratio to blood lost (Moore, 2004). Clinically, urine output of 0.5 mL/kg/per hour or 30 mL or more per hour, heart rate less than 100 beats per minute, and systolic blood pressure greater than 90 mm Hg may be used as general indicators of volume improvement. If rapid crystalloid infusion fails to correct hypotension or tachycardia, then RBC transfusion usually is prudent.

In addition to or as an alternative to crystalloid solutions, colloids may be used for volume expansion. These fluids have higher molecular weights than crystalloids. As a result, a greater portion remains intravascular and is not lost to extracellular extravasation. Despite this perceived advantage, studies comparing survival rates when crystalloids or colloids are administered find no superiority with colloids but greater expense (Perel, 2013).

Intraoperative fluid management strategies broadly fall into categories of liberal (sometimes thought of as fixed volume), restrictive, or goal directed. Of these, evidence from colorectal and trauma surgery is now more supportive of restrictive management. Less bowel edema, quicker return of bowel function, and fewer pulmonary complications are all purported benefits (Chappell, 2008; Joshi, 2005). Restrictive strategies generally use colloid to replace blood loss in a 1:1 ratio, unless red cell transfusion is indicated. Crystalloids are then used to replace urine and insensible losses 1:1. In contrast, liberal strategies rely on large volumes of crystalloid. Last, goal directed therapy refers to using a monitoring device (such as an arterial line) and administering fluids to achieve a goal (such as maximizing stroke volume). Patients with severe comorbid conditions undergoing major procedures may benefit from this strategy (Chappell, 2008).

Red Blood Cell Replacement

Clinical Assessment

The decision to administer RBCs is complex and must balance the risks of transfusion with needs for adequate tissue oxygenation. Assessment includes hemoglobin level, vital signs, patient age, risks for further blood loss, and underlying medical conditions, especially cardiac disease. These needs will vary depending on the clinical setting. Accordingly, no specific hemoglobin threshold dictates when RBCs are administered. Consensus guidelines suggest that in those without significant cardiac disease, transfusion to a hemoglobin level above 10 g/dL is rarely indicated (Carless, 2010). If hemoglobin levels acutely drop to 6 g/dL, transfusion almost always is required (Madjdpour, 2006). Hemoglobin levels between 6 and 10 g/dL are more problematic, and patient factors and risk for continued hemorrhage dictate therapy (American Society of Anesthesiologists, 2015). In one randomized study of 838 critically ill patients, one group of euvolemic patients received transfusion when their hemoglobin levels fell below 7 g/dL. These individuals fared better than those transfused at an earlier threshold (hemoglobin below 10 g/dL), excepting those with significant cardiac disease (Hébert, 1999).

Transfusion

When the possible need for transfusion is present, an order for a type and screen informs blood bank personnel that blood products may be required and initiates two tests to characterize a patient’s RBCs. The first evaluation, termed typing, mixes commercially available standardized controls with a patient’s blood sample to determine her ABO type and Rh phenotype. The second test, or screen, combines a patient’s plasma sample with control RBCs that express clinically significant RBC antigens. If a patient has formed antibodies to any of these specific RBC surface antigens, then agglutination or hemolysis of the sample is seen. However, if blood is needed immediately and a full screen is not possible, then ABO type-specific blood or O-negative blood may be used. Typing and screening require approximately 45 minutes for completion and are valid for 3 days in patients who do receive transfusion. In those who are not transfused, the validity is considerably longer and typically is determined by individual blood banks. Alternatively, an order to type and crossmatch blood products alerts blood bank personnel to designate specific units of blood solely for one individual’s use. Those specific units are tested against the patient’s for specific antigen reactions.

Previously, whole-blood transfusion was used commonly to provide RBCs, coagulation factors, and plasma proteins. This largely has been replaced by component therapy. Packed RBCs are the primary product used for most clinical situations and are prepared by removing most of the supernatant plasma during centrifugation. One unit of packed RBCs contains the same red cell mass as 1 unit of whole blood at approximately half the volume and twice the hematocrit (70 to 80 percent). One unit of packed RBCs raises the hematocrit approximately 3 volume percent in an adult or increases the hemoglobin level of a 70-kg individual by 1 g/dL (Table 40-7) (Gorgas, 2004). With severe hemorrhage that is anticipated to require ≥10 RBC units, massive transfusion protocols that combine units of packed RBCs, platelets, and plasma in 1:1:1 ratios are effective (McDaniel, 2014).

Complications

Despite numerous tests for compatibility, adverse reactions to blood products can develop. An acute or delayed hemolytic transfusion reaction, febrile nonhemolytic transfusion reaction, allergic reaction, infection, or associated lung injury are among these.

First, acute hemolytic transfusion reaction involves acute immune-mediated hemolysis usually from destruction of transfused RBCs by patient antibodies. This most commonly results from ABO incompatibility. Symptoms begin within minutes or hours of transfusion and may include chills, fever, urticaria, tachycardia, dyspnea, nausea and vomiting, hypotension, and chest and back pain. In addition, these reactions can lead to acute tubular necrosis or disseminated intravascular coagulopathy, and treatment is directed to these serious complications.

If acute hemolysis is suspected, transfusion is halted immediately. A sample of the patient’s blood is sent with the remaining donor unit for evaluation in the blood bank. In patients with significant hemolysis, laboratory values will be altered. Specifically, serum haptoglobin levels will be lowered; serum lactate dehydrogenase and bilirubin levels will be increased; and serum and urine hemoglobin levels will be elevated. Serum creatinine and electrolyte levels and coagulation studies additionally are ordered. To prevent renal toxicity, diuresis is prompted with intravenous crystalloids and administration of furosemide or mannitol. Alkalinization of urine may prevent precipitation of hemoglobin within the renal tubules, and therefore, intravenous bicarbonate also may be given.

In contrast to acute hemolytic transfusion reaction, delayed hemolytic transfusion reactions may develop days or weeks later. Patients often lack acute symptoms, but lowered hemoglobin levels, fever, jaundice, and hemoglobinemia may be noted. Clinical intervention typically is not required in these cases.

Febrile nonhemolytic transfusion reaction is characterized by chills and a greater than 1°C rise in temperature and is the most common transfusion reaction. Blood transfusion typically is stopped to exclude a hemolytic reaction, and treatment is supportive. For patients with a previous history of febrile reaction, premedication with an antipyretic such as acetaminophen prior to transfusion is reasonable.

Last, an allergic reaction can follow an antibody-mediated response to donor plasma proteins. Urticaria alone may develop during transfusion and typically is not associated with serious sequelae. The transfusion does not need to be stopped, and treatment with an antihistamine, such as diphenhydramine (Benadryl) 50 mg orally or intramuscularly, usually suffices. Rarely, an anaphylactic reaction may complicate transfusion, and treatment follows that for classic anaphylaxis (Table 27-2).

Infectious complications associated with packed RBC transfusion are uncommon. The risk for transmission of HIV and hepatitis B and C virus has diminished over the past decade, and bacterial contamination now stands as a greater infection risk. In addition, emerging infection concerns include transmission of West Nile virus, transfusion-transmitted virus (TTV), hepatitis G, Epstein-Barr virus, and Creutzfeldt-Jakob disease (Luban, 2005).

Transfusion-related acute lung injury (TRALI) is an infrequent but serious complication of blood component therapy that is similar clinically to acute respiratory distress syndrome (ARDS). Symptoms develop within 6 hours of transfusion and may include extreme respiratory distress, frothy sputum, hypotension, fever, and tachycardia. Noncardiogenic pulmonary edema with diffuse bilateral pulmonary infiltrates on chest radiography is characteristic (Toy, 2005). Treatment of TRALI is supportive and focuses on oxygenation and blood pressure support (Silliman, 2005; Swanson, 2006).

Platelet Replacement

For patients with moderate hemorrhage, RBC transfusion typically is sufficient, but for patients with severe hemorrhage, platelet transfusion also may be indicated. Platelets may be acquired from a single individual during plateletpheresis and are termed single-donor platelets. Alternatively, platelets may be derived from random units of whole blood and are referred to as random-donor platelets.

Fewer platelets are harvested from a unit of whole blood compared with the amount removed during donor plateletpheresis. Specifically, a single-donor platelet dose contains at least 3 × 10¹¹ platelets in 250 to 300 mL of plasma, and this approximates the dose from six random-donor platelet concentrates. Each random-donor platelet concentrate contain 5.5 × 10¹⁰ platelets suspended in approximately 50 mL of plasma. Each concentrate transfused should raise the platelet count by 5 to 10 × 10⁹/L, and the usual therapeutic dose is one platelet concentrate per 10 kg of body weight. Five to six concentrates provide a typical adult dose. Donor plasma must be compatible with recipient erythrocytes because a few RBCs are invariably transfused along with the platelets. Only platelets from D-negative donors should be given to D-negative recipients.

Surgical patients with bleeding usually require platelet transfusion if the platelet count is less than 50 × 10⁹/L and rarely require therapy if it is greater than 100 × 10⁹/L (American Society of Anesthesiologists, 2015). With counts between 50 and 100 × 10⁹/L, the decision to provide platelet transfusion is based on a patient’s risk for additional significant bleeding.

Factor Replacement

Fresh-frozen plasma (FFP) is one option for factor replacement and is prepared from whole blood or by plasmapheresis. It is stored frozen, and approximately 30 minutes are required for it to thaw. One unit contains all coagulation factors, including 2 to 5 mg/mL of fibrinogen in 250 mL of volume. Recommended FFP dosing is 15 mL/kg. Fresh-frozen plasma is used commonly as first-line hemostatic therapy in massive hemorrhage because it replaces multiple coagulation factors. It is considered in a bleeding woman with a fibrinogen level below 1.0 g/L or with abnormal prothrombin and partial thromboplastin times.

Another option, cryoprecipitate, is prepared from fresh-frozen plasma and contains fibrinogen, factor VIII, von Willebrand factor, factor XIII, and fibronectin. Cryoprecipitate was developed and used originally for treatment of hemophilia A and von Willebrand disease. However, specific factor concentrates are now available for these disorders, and thus, the clinical indications for cryoprecipitate are limited. Fresh-frozen plasma provides all coagulation factors and is favored in severe hemorrhage over cryoprecipitate. However, cryoprecipitate is an excellent source of fibrinogen and may be indicated if fibrinogen levels persist below 1.0 g/L despite administration of fresh-frozen plasma, such as in disseminated intravascular coagulopathy (DIC). The dose of cryoprecipitate is usually 2 mL/kg of body weight, and each unit contains approximately 15 mL volume. One unit typically increases the fibrinogen level by 10 mg/dL (Erber, 2006).

Lower Urinary Tract

Sound anatomic knowledge, adequate operating exposure, meticulous technique, and surgical experience are essential to avoid surrounding organ injury during pelvic surgery. The lower gastrointestinal and urinary tracts are closely related to the female reproductive organs, and disease processes, anatomic distortion, and adverse operating conditions can increase their injury risk.

Iatrogenic damage to the lower urinary tract is common, and up to 75 percent of ureter or bladder injuries sustained during gynecologic surgery occur during hysterectomy (Walters, 2007). Most injuries have no antecedent risk factors, but high-risk elements are ideally sought preoperatively. These include compromised visibility from large pelvic masses, hemorrhage, pregnancy, obesity, inadequate incision, suboptimal retraction, and poor lighting. Additionally, scarring or anatomic distortion from cervical and broad ligament leiomyomas, malignancy, endometriosis, pelvic organ prolapse, and prior pelvic infection, surgery, or radiation are risks (Brandes, 2004; Francis, 2002).

Patients who sustain surgical injury to the bladder or ureter suffer significantly greater morbidity. In one case-control study, women with injury to the lower urinary tract during abdominal hysterectomy had significantly greater operative time, estimated blood loss, blood transfusion rates, febrile morbidity, and postoperative stay length than their respective controls (Carley, 2002).

Bladder Injury

Cystotomy is common and complicates approximately 0.3 to 11 per 1000 benign gynecologic surgeries, especially urogynecologic procedures and hysterectomy (Gilmour, 2006; Mathevet, 2001). In sum, depending on the procedure, the bladder may be at greater risk during: (1) initial abdominal entry when incising the anterior parietal peritoneum, (2) dissection within the space of Retzius, (3) vaginal epithelium dissection during anterior colporrhaphy, or (4) hysterectomy when dissecting in the vesicocervical space, entering the anterior vagina, or suturing the vaginal cuff. With hysterectomy, bladder injury traditionally has been associated more often with the vaginal hysterectomy, but some data suggest that laparoscopic procedures pose the greatest risk (Francis, 2002; Frankman, 2010; Harris, 1997). Preventatively, clear identification of the bladder, gentle retraction, meticulous surgical technique, sharp dissection, and maintenance of a drained bladder intraoperatively are standard principles.

Cystotomy is suspected if a Foley bulb, bloody urine, or urine leaking into the operative field is seen. During laparoscopy, the Foley bag may also distend with gas from the pneumoperitoneum. For diagnosis, retrograde instillation of sterile milk through a catheter confirms injury and delineates its full extent. This is superior to methylene blue and indigo carmine dyes, as infant formula does not stain surrounding tissues and is readily available. In addition, small defects can be difficult to identify and repair if the tissues surrounding the defect become dye stained. Prior to repair, cystoscopy is indicated for any bladder base injury to assess ureteral patency. In addition, the full extent of injury can be defined, and the bladder can be evaluated for additional injuries or intravesical sutures. If bladder distension cannot be maintained during cystoscopy or if the patient is not in dorsal lithotomy, the ureteral orifices can also be evaluated grossly through the cystotomy site. If the cystotomy is small, suprapubic teloscopy, which is described in Chapter 45, is also an option, or the cystotomy can be extended to allow evaluation.

Repair during the primary surgery is preferred and lowers risks of later vesicovaginal fistula formation. Principles of repair include injury delineation; wide mobilization of surrounding tissues; tension-free, multilayered, watertight closure; and adequate postoperative bladder drainage (Utrie, 1998). Suture identified in the bladder is cut, as persistence can lead to cystitis, stone formation, or both. Needle-stick and subcentimeter lacerations can be managed conservatively. Larger defects may be closed in two or three layers with a running stitch using 3-0 absorbable or delayed-absorbable suture (Fig. 40-32). The first layer inverts the mucosa into the bladder, and subsequent layers reapproximate the bladder muscularis and serosa. In the area of the trigone, the ureters are typically stented first, and the repair may be performed with interrupted sutures to avoid ureteral kinking (Popert, 2004). Postoperatively, continuous bladder drainage is continued for 7 to 10 days (Utrie, 1998). Evidence is conflicting regarding the use of prophylactic antibiotics for the expected duration of catheterization, and thus remains at the provider’s discretion.

Urethral Injury

The female urethra is rarely injured during gynecologic surgery, but cystoscopy, urethral diverticulum repairs, antiincontinence operations, and possibly anterior colporrhaphy are at-risk procedures. Repair is completed with 3-0 or 4-0 absorbable suture in an interrupted fashion and in multiple layers, if possible. Similar to cystotomy, a Foley catheter is typically placed postoperatively for 7 to 10 days, with antibiotic prophylaxis provided at the surgeon’s discretion (Francis, 2002).

Ureteral Injury

This is uncommon in benign gynecologic surgery, and the incidence approximates 0.2 to 7.3 per 1000 surgeries. For hysterectomy, the highest rate of ureteral injury is linked with laparoscopic hysterectomy, and the lowest with vaginal hysterectomy (Gilmour, 2006). Other associated procedures include operations for pelvic organ prolapse, incontinence, malignancy, or endometriosis (Patel, 2009; Utrie, 1998).

The ureter is 25 to 30 cm long, and its anatomy is described in Chapter 38. Gynecologic ureteral injury typically occurs in the distal third and includes transection, ligation, kinking, and crushing (Brandes, 2004; Utrie, 1998). Trauma to the outer sheath can also disrupt ureteral blood supply. Of these, the ureter more often is transected or kinked, and each accounts for approximately 40 percent of injuries. During hysterectomy, the most common trauma site is at the level of the uterine artery and accounts for 80 percent of injuries (Ibeanu, 2009). The ureter is also vulnerable near the pelvic brim during adnexectomy and at the distal uterosacral ligaments. Mechanisms of injury include clamping or suturing with the ureter poorly visualized. Thermal insult or devascularization may lead to stricture or leak.

Injury prevention measures include preoperative risk evaluation and if indicated, intravenous pyelography (IVP) or computed tomography (CT). Ureteral stenting assists with intraoperative recognition but does not necessarily prevent injury. As with bladder injury, the best prevention is sound intraoperative technique and direct visualization of the peristalsing ureter. Also, the ureter may be felt to “snap” if palpated and stretched along its course on the broad ligament’s medial leaf. However, vessels, adipose tissue, and peritoneal folds can mimic this.

Diagnosis

Iatrogenic injury ideally is diagnosed early, as immediate repair is associated with improved outcomes and less patient morbidity (Neuman, 1991; Sakellariou, 2002). Damage may be seen directly or identified during intraoperative cystoscopy. Intravenous administration of indigo carmine or methylene blue can aid cystoscopic evaluation, with observation of blue-stained urine from the ureteral orifices. This is described fully in Section 45-1 of the atlas. However, use of the latter may increase given current indigo carmine shortages. With these dyes, blue effluent is usually seen in 5 to 10 minutes. Failure to see dye after 30 to 40 minutes mandates further evaluation with either IVP or ureteral catheterization. Unfortunately, normal-appearing findings at cystoscopy do not guarantee ureteral integrity, as nonobstructive, partially obstructive, or late ureteral injuries may be unrecognized.

Diagnosing injury shortly after surgery is challenging, as patient symptoms may be attributable to other causes. Thus, thorough patient evaluation and a high clinical suspicion are crucial. Renal damage may begin 24 hours after obstruction and can be irreversible in 1 to 6 weeks (Walter, 2002). Symptoms usually develop about 48 hours after surgery, and fever, abdominal pain, flank pain, and watery discharge may be among these. Findings can include leukocytosis, elevated blood urea nitrogen level, and ileus. Prolonged skin or vaginal drainage suggests a urinary leak, and high creatinine levels in these fluids are diagnostic of urine. Serum creatinine measurement may or may not be helpful. In one retrospective study of 187 patients, a 24-hour postoperative change <0.3 mg/dL from preoperative levels had a specificity of 98 percent and negative predictive value of 100 percent in confirming bilateral ureteral patency. Because an increase >0.2 mg/dL has been associated with obstruction, the authors recommended repeating a creatinine measurement and renal imaging for persistent elevation above this level (Walter, 2002). With elevated serum creatinine levels, calculation of the fractional excretion of sodium (FENa) or assessment of urine sodium levels may also help clarify the renal injury source as prerenal, intrarenal, or postrenal, as described in Chapter 42.

Sonography, CT, or magnetic resonance (MR) imaging will help identify hydronephrosis, urinoma, or abscess. Lack of contrast in the distal ureter on delayed CT images confirms total obstruction (Armenakas, 1999). IVP can also help localize injury. However, IV contrast can be nephrotoxic, and thus CT with contrast may be a less than ideal choice for those with already elevated creatinine levels. Retrograde pyelography with fluoroscopic guidance and attempted retrograde ureteral stent placement can be considered in cases where suspicion remains high, and IVP is contraindicated or has equivocal findings. All of these imaging modalities can be used to diagnose injury in both the early and late postoperative periods.

Treatment

The best repair method depends on the location, extent, time from surgery, and mechanism of injury. Expert assistance from a urogynecologist, gynecologic oncologist, or urologist may be prudent. The ureter can be repaired by stenting, reimplantation, or end-to-end reanastomosis. For low-grade sheath injuries from clamping or suturing, removal of the insult and stent placement may be sufficient. For incomplete obstruction or injury identified postoperatively, stenting alone can resolve injuries in up to 80 percent of cases. For more extensive injury, either reimplantation or reanastomosis is performed (Utrie, 1998).

Reimplantation, namely, ureteroneocystotomy, is preferred for injuries within 6 cm of the bladder. Uncommonly with this, if the ureter is short, a psoas hitch, that is, mobilizing the bladder and attaching it to the psoas muscle tendon, may be necessary to bridge the gap and relieve tension on the repair. An alternative to the psoas hitch is a Boari flap. In this procedure, the bladder ipsilateral to the injury is mobilized, and a pedicle of anterior bladder wall is fashioned into a tube to bridge to the ureter.

For injuries greater than 7 cm from the bladder, ureteral reanastomosis, that is, ureteroureterostomy, is preferred. Rarely, transureteroureterostomy is needed for a more proximal injury or one in which the bladder cannot be mobilized. With this procedure, the injured ureter is tunneled across and connected to the healthy ureter.

Little evidence guides the decision for reoperation in the early postoperative period. Intraoperatively, tissues are in their best condition, and the likelihood for successful repair is great. However, most iatrogenic injuries are recognized after a delay and tend to be complex (Brandes, 2004). In general, reexploration within the first few days appears to be well tolerated, leads to good outcomes, and is not technically difficult (Preston, 2000; Stanhope, 1991). Firm recommendations regarding reoperation beyond this early postoperative period are lacking, but reexploration 2 to 3 weeks after initial surgery is difficult due to inflammation, fibrosis, adhesions, hematoma, and distorted anatomy (Brandes, 2004).

For delayed diagnoses, retrograde stenting is unsuccessful in 50 to 95 percent of cases and recommended only for certain low-grade injuries (Brandes, 2004). Occasionally, an antegrade stent can be placed percutaneously, which will avoid the need for laparotomy, provided there is no ureteral leak or stricture. More extensive damage, such as complete transection, cannot be easily stented and is more appropriately repaired by definitive surgery. When diagnosis is significantly delayed, urinary diversion with a percutaneous nephrostomy (PCN) and later repair is preferred. For some low-grade lesions, such as ligation with absorbable suture, proximal urinary diversion by PCN may allow spontaneous healing without further surgery. In addition, PCN diversion may be used as a temporizing measure for patients temporarily unfit for surgery (Preston, 2000).

Universal Cystoscopy

Lower urinary tract injury is poorly detected by direct visualization, and rates range from 7 to 12 percent for ureteral trauma and approximate 35 percent for bladder damage (Vakili, 2005). To increase early diagnosis, universal intraoperative cystoscopy has been advocated, and detection rates are near 96 percent (Ibeanu, 2009; Vakili, 2005; Visco, 2001). Proponents argue that the procedure is cost-effective, carries minimal risk, and prevents both postoperative morbidity and liability. Opponents cite overall low rates of injury, imperfect detection rates, increased costs, credentialing problems, and a need for ­training (Patel, 2009). Using a decision analysis model, one study estimated that routine cystoscopy was cost-effective when ureteral injury rates were above 1.5 percent for abdominal hysterectomy and 2 percent for vaginal and laparoscopically assisted hysterectomy (Visco, 2001).

Cystoscopy is currently indicated for urogynecologic procedures, but there are no strict recommendations for other routine gynecologic procedures, including hysterectomy (American College of Obstetricians and Gynecologists, 2013; Patel, 2009). At present, the decision remains at the surgeon’s discretion. Some have elected selective cystoscopy, or cystoscopy restricted to patients with risk factors or when intraoperative events make injury more likely.

Bowel Injury

Injury to the bowel infrequently complicates gynecologic surgery, and rates are <1 percent (Harris, 1997; Makinen, 2001). A traumatic breach during dissection is the most common, particularly if the bowel wall is abnormally fixed by adhesions (Mathevet, 2001; Maxwell, 2004). Additional risks include reduced organ mobility from Crohn disease or diverticulitis, laparoscopic trocar or Veress needle insertion, diathermy use, and anterior abdominal wall entry during laparotomy.

For the gynecologic surgeon, prevention and injury recognition help avoid serious postoperative sequelae. Strict adherence to surgical principles with sharp dissection for adhesions, gentle tissue handling, adequate exposure, light retraction, and sparing use of diathermy near hollow organs is key. Entering through prior abdominal incisions, dissection proceeds methodically in layers. Alternatively, a separate incision or extension of the existing one to an area that has not been previously opened can be considered. After any extensive pelvic dissection, the bowel is systematically inspected along its entire length to detect serosal defects and unrecognized perforation. At suspected sites, the bowel is scrutinized for mucosal eversion and content leakage. Evaluation is gentle to avoid additional damage.

Management of enterotomy depends on the site and size of injury, surgeon skill, degree of blood supply compromise, and time of recognition. With the small intestines, serosal defects may be either left alone or reinforced with small-gauge absorbable suture (Maxwell, 2004). Short small-intestine enterotomies may be repaired in layers using fine absorbable suture. During repair, rubber-shod clamps are placed across the intestinal lumen on either side of the wound to prevent content spill. To avoid narrowing of the bowel lumen, the suture line should lie transverse to the normal axis of the intestine (Stanton, 1987). Postoperative antibiotics are typically not required.

Large-bowel injuries increase the risk of fecal peritonitis, sepsis, and poor wound healing. Serosal defects and small lacerations may be managed similarly to those of the small intestine. For more extensive injuries or fecal soiling that may require resection, diversion, or complicated repair, consultation with a gynecologic oncologist or colorectal surgeon is often indicated. Broad-spectrum antibiotic prophylaxis is provided for the next 24 hours in these cases. In general, for both small and large intestinal injuries, early feeding is acceptable and not associated with repair site complications (Fanning, 2001).

Rectal injury occurs most commonly during vaginal surgery, especially during posterior colpotomy or posterior colporrhaphy. These injuries are typically midline, extraperitoneal, and less than 2 cm. Postmenopausal status, prior posterior colporrhaphy, or pathology that obliterates the cul-de-sac or limits organ mobility increases injury risk (Hoffman, 1999; Mathevet, 2001). Prevention centers on careful examination under anesthesia to detect cul-de-sac fullness or uterine immobility, sharp dissection with the aid of a guiding rectal finger, and vasoconstricting agent use to reduce obscuring operative field bleeding.

Rectal injury may be extraperitoneal or intraperitoneal, and rectal examination will typically detect the injury and delineate its borders. Minor intraperitoneal injuries with minimal or no contamination can be repaired primarily in layers as described previously. Larger injuries with gross soiling may require expert consultation. Broad-spectrum antibiotic prophylaxis is provided for the next 24 hours in these cases.

Low extraperitoneal rectal injury during vaginal surgery in a healthy patient can be repaired primarily and rarely requires a diverting colostomy or abdominal repair. Repair is accomplished transvaginally using two to three layers of fine absorbable suture. The peritoneum may be used as an additional layer for injuries near the peritoneal reflection. During repair, a digit in the rectum exposes the defect, tissues surrounding the defect are mobilized, the site is copiously irrigated, and appropriate antibiotic prophylaxis is provided for 24 hours (Hoffman, 1999). In general, small (<2 cm) rectal injuries recognized and repaired at the time of vaginal surgery tend to heal well without complications or fistula formation (Mathevet, 2001). Diet can be advanced as tolerated, but a stool softener is recommended once the patient is taking solid foods (Hoffman, 1999).