Professional Documents
Culture Documents
Complications of Hysterectomy
Daniel L. Clarke-Pearson, MD, and Elizabeth J. Geller, MD
Hysterectomy is the most common gynecologic procedure performed in the United States, with
more than 600,000 procedures performed each year. Complications of hysterectomy vary based
on route of surgery and surgical technique. The objective of this article is to review risk factors
associated with specific types of complications associated with benign hysterectomy, methods to
prevent and recognize complications, and appropriate management of complications. The most
common complications of hysterectomy can be categorized as infectious, venous thromboem-
bolic, genitourinary (GU) and gastrointestinal (GI) tract injury, bleeding, nerve injury, and vaginal
cuff dehiscence. Infectious complications after hysterectomy are most common, ranging from
10.5% for abdominal hysterectomy to 13.0% for vaginal hysterectomy and 9.0% for laparoscopic
hysterectomy. Venous thromboembolism is less common, ranging from a clinical diagnosis rate of
1% to events detected by more sensitive laboratory methods of up to 12%. Injury to the GU tract
is estimated to occur at a rate of 1–2% for all major gynecologic surgeries, with 75% of these
injuries occurring during hysterectomy. Injury to the GI tract after hysterectomy is less common,
with a range of 0.1–1%. Bleeding complications after hysterectomy also are rare, with a median
range of estimated blood loss of 238–660.5 mL for abdominal hysterectomy, 156–568 mL for
laparoscopic hysterectomy, and 215–287 mL for vaginal hysterectomy, with transfusion only being
more likely after laparoscopic compared to vaginal hysterectomy (odds ratio 2.07, confidence
interval 1.12–3.81). Neuropathy after hysterectomy is a rare but significant event, with a rate
of 0.2–2% after major pelvic surgery. Vaginal cuff dehiscence is estimated at a rate of 0.39%, and
it is more common after total laparoscopic hysterectomy (1.35%) compared with laparoscopic-
assisted vaginal hysterectomy (0.28%), total abdominal hysterectomy (0.15%), and total vaginal
hysterectomy (0.08%). With an emphasis on optimizing surgical technique, recognition of sur-
gical complications, and timely management, we aim to minimize risk for women undergoing
hysterectomy.
(Obstet Gynecol 2013;121:654–73)
DOI: http://10.1097/AOG.0b013e3182841594
VOL. 121, NO. 3, MARCH 2013 Clarke-Pearson and Geller Complications of Hysterectomy 655
to tissue trauma, or hematoma formation. An elevated tion and drainage of a hematoma or abscess. Specific
white blood cell count, although often present in the identification, evaluation, and management will be
face of infection, cannot be used as a confirmatory tool addressed separately for each category of infection.
because this may be attributable to the same noninfec-
tious etiologies such as elevated body temperature. The Antimicrobial Prophylaxis
management and treatment plan should be based on Perioperative antimicrobial prophylaxis is indicated for
the confirmation of an actual site of infection, rather all types of hysterectomy.11 First-generation cephalo-
than indirect signs such as fever or elevated white sporins are as effective as second-generation or third-
blood cell count. A 2009 Cochrane meta-analysis generation agents and are not only less expensive but
found that subtotal abdominal hysterectomy had also less likely to induce a bacterial b-lactamase
a decreased rate of febrile morbidity compared with response. Dose adjustment should be based on weight,
total abdominal hysterectomy (odds ratio [OR] specifically doubling the dose from 1 to 2 g at 80 kg or
0.43, confidence interval [CI] 0.25–0.75).10 Another more.11 If a penicillin allergy exists, alternatives include
Cochrane meta-analysis assessing route of hysterec- a combination of either Clindamycin or metronidazole
tomy found that laparoscopic hysterectomy (includ- plus either Gentamicin or Levofloxacin. Antibiotics
ing laparoscopically assisted vaginal hysterectomy) should be administered within 1 hour before the time
had a decreased rate of febrile morbidity compared of skin incision. Readministration is recommended for
with abdominal hysterectomy (OR 0.67, CI 0.51– surgical procedures lasting more than 3 hours or those
0.88).3 Vaginal hysterectomy also had fewer febrile with an estimated blood loss more than 1,500 mL.
episodes and unspecified infections compared with Continued use of “prophylactic antibiotics” beyond
abdominal hysterectomy (OR 0.42, CI 0.21–0.83). those administered in the operating room has no
There were no differences in infectious complications proven benefit.
between laparoscopic and vaginal hysterectomy.
When comparing subcategories of laparoscopic hys- Vaginal Cuff Cellulitis
terectomy, there were more febrile episodes with total One of the most common infectious complications
laparoscopic hysterectomy when compared with lap- after hysterectomy that is unique to this procedure is
aroscopically assisted vaginal hysterectomy (OR 3.77, vaginal cuff cellulitis. This usually occurs late in the
CI 1.05–13.51). hospital course or soon after discharge. The incidence
The evaluation of any postoperative patient ranges from 0% to 8.3% after hysterectomy.3 Symp-
suspected of having an infection should begin with toms can include fever, purulent vaginal discharge
a careful history and physical examination. Physical (to be distinguished from physiologic discharge), and
examination could include any of the following pelvic, abdominal, or low back pain. Examination
elements, based on the patient’s history: examination may reveal tenderness or induration of the vaginal
of the throat and chest for upper and lower respiratory cuff and purulent discharge.
tract infection including pharyngitis, bronchitis, and The bacteria associated with a cuff cellulitis are
pneumonia; examination of the abdomen for signs polymicrobial.9,11 Whereas cultures should be col-
of peritonitis or incisional infection; examination of lected if there is any purulent discharge, treatment
the lower extremity examination for phlebitis or should be started empirically. Antimicrobial therapy
thrombus; and possible genitourinary (GU) examina- can be either single-agent or multi-agent, depending
tion for signs of cystitis, pyelonephritis, cuff cellulitis, both on clinical presentation and clinician preference.
and pelvic hematoma. An early presentation (within a few days of surgery)
may be a sign of aerobic infection and can be treated
Laboratory Tests with a broad-spectrum penicillin with b-lactamase
Laboratory tests that may help guide diagnosis inhibitor or later-generation cephalosporin. If no clin-
include complete blood count with differential to look ical improvement is seen, a multi-agent regimen can
for a left shift, and urinalysis with culture if leukocytes be instituted, assuming anaerobic involvement, with
or nitrites are found. Cultures of the surgical site gentamicin and clindamycin. Metronidazole is an
should be performed with a low threshold of suspi- alternative to clindamycin, especially if one is con-
cion. Based on the findings of the physical examina- cerned about anaerobic resistance, but may necessi-
tion, imaging can be helpful in some cases, including tate the addition of ampicillin for Gram-positive
chest radiograph, abdominal radiograph, pelvic ultra- coverage. For oral therapy, a regimen of metronida-
sonography, or computed tomography (CT). The zole and fluoroquinolone will provide adequate
latter may be particularly helpful if planning aspira- coverage.
VOL. 121, NO. 3, MARCH 2013 Clarke-Pearson and Geller Complications of Hysterectomy 657
the incision site, clippers are vastly preferable to a razor. of bacteriuria with intermittent self-catheterization
Other techniques to reduce wound infection are pre- compared with suprapubic catheter use,21 Cochrane
operative antimicrobial body wash the night before meta-analyses have found no difference in the rate of
surgery and controlling postoperative blood glucose.12 urinary tract infection based on route or subtype of
If a procedure has been classified as dirty or infected, it hysterectomy.3,10
is often better to leave the incision open, either for Routine preoperative screening urinalysis, culture,
a delayed closure or to heal by secondary intention. or both are not recommended. One meta-analysis
In a Cochrane Review there was a decreased rate of found that perioperative urinalysis revealed abnormal-
abdominal wall or wound infection when comparing ities in 1–34.1% of cases but only led to changes in
laparoscopic with abdominal hysterectomy (OR 0.31, management in 0.1–2.8% of cases.22 The American
CI 0.12–0.77).3 Heart Association guidelines state that preoperative
evaluation may be indicated in procedures involving
Urinary Tract Infection the urinary tract.23
Infections of the urinary tract usually present early in
the postoperative period. The incidence of urinary Respiratory Infection
tract infection after hysterectomy ranges from 0% to Pneumonia can occur early in the postoperative period.
13.0%,3 but accounts for up to 40% of nosocomial In a patient without pulmonary disease, the incidence
infections.17 Signs and symptoms include low-grade of pneumonia after hysterectomy is low (0–2.16%),24
fever, frequency, urgency, and dysuria. Some patients The risk is significantly increased in patients with
may be asymptomatic. Examination may reveal underlying pulmonary disease. Symptoms include
suprapubic tenderness or tenderness to palpation of shortness of breath, chills, chest pain, cough, and puru-
the anterior vaginal wall, but again may be normal. lent sputum. Patients also may be asymptomatic.
Diagnosis is supported by a urinalysis revealing Examination may reveal decreased breath sounds,
nitrites and is confirmative with a midstream clean rales, hypoxia, tachycardia, and fever. Imaging in the
catch urine culture revealing at least 100,000 colony- form a chest radiograph can confirm and localize the
forming units of a single organism that is not consid- infection, whereas sputum cultures can provide organ-
ered skin flora or at least 100 colony-forming units on ism identification.
catheter specimen. The major causative agents for hospital-acquired
Therapy can be initiated with symptoms and a pneumonia include Haemophilus influenzae and Strepto-
suspicious urinalysis, but a urine culture should be coccus and Staphylococcus species, as well as E coli, and
sent to confirm infection and to determine antimicro- Enterobacter, Klebsiella, Proteus, and Serratia species.
bial sensitivities. Likely pathogens include facultative Treatment of these pathogens includes a second-
anaerobes originating from bowel flora. The most generation or third-generation cephalosporin, penicil-
common pathogen is Escherichia coli. Although more lin in combination with a b-lactamase inhibitor, or a
common in community-acquired urinary tract infec- fluoroquinolone. Alternatively, postoperative patients
tion, E coli is still the most common cause of hospital- presenting after discharge to home who are found to
acquired urinary tract infection, ranging from 30% to be minimally symptomatic with an otherwise normal
56% of all cases.18,19 examination likely have community-acquired pneu-
For an uncomplicated urinary tract infection, monia. Although Streptococcus pneumonia is the most
trimethoprim-sulfamethoxazole and nitrofurantoin common pathogen, consideration must be made for
are inexpensive and well-tolerated options. Nitrofur- possible resistance attributable to recent antibiotic use.
antoin reaches high concentrations in the urinary tract Thus, treatment should consist of either a fluroquino-
while having low serum concentrations, therefore lone or penicillin in combination with a b-lactamase
conferring few side effects on bowel and vaginal flora. inhibitor plus a macrolide or doxycycline.
It is not effective against Pseudomonas or Proteus spe- Another type of postoperative pneumonia that
cies. Fluoroquinolones provide another treatment bears consideration is aspiration pneumonia. These
option. This class is more expensive but may be con- cases often involve Staphylococcus species as well as
sidered first-line treatment if resistance to other agents Gram-negative aerobes. As a result, penicillin is not
is a concern. as effective. Agents such as piperacillin or ticarcillin
Prevention of a postoperative urinary tract infec- are indicated, in addition to a b-lactamase inhibitor.
tion is based on early removal of an indwelling catheter Alternatives include clindamycin plus either gentami-
and aseptic technique for clean intermittent catheteri- cin or ciprofloxacin, or an expanded-spectrum ceph-
zation.12,20 No difference has been shown in the rate alosporin plus metronidazole.
VOL. 121, NO. 3, MARCH 2013 Clarke-Pearson and Geller Complications of Hysterectomy 659
with unfractionated heparin, low-molecular-weight
Box 1. Venous Thromboembolism Risk Factors
heparins have more anti-Xa and less anti-thrombin
Increasing age* activity, leading to less effect on partial thromboplas-
Previous venous thromboembolism* tin time. An increased half-life of 4 hours results in
Malignancy*
Surgery* increased bioavailability when compared with unfrac-
Treatment for cancer: chemotherapy or radiation* tionated heparin and therefore allows the convenience
Duration of anesthesia* of dosing once per day dosing. Randomized con-
Nonwhite race* trolled trials in gynecologic surgery comparing low-
Venous insufficiency (leg edema, varicose veins)* molecular-weight heparin with unfractionated heparin
Hormone use: oral contraceptives, hormone replace-
ment therapy, selective estrogen receptor modulators† have shown no difference in the incidence of DVT or
Inherited or acquired thrombophilia† frequency of bleeding complications.36
Major lower extremity trauma† Reduction in venous stasis during and after gyne-
Pregnancy or postpartum period† cologic surgery has been investigated using either
Obesity† graded compression stockings or intermittent pneu-
Smoking†
Prolonged immobility or paresis† matic compression devices. If fitted properly, graded
Acute medical illness† compression stockings provide modest benefit in reduc-
Pulmonary or cardiac failure† ing the incidence of venous thromboembolism.37
Inflammatory bowel disease† Should the upper portion of the stocking roll down
Nephrotic syndrome† the leg, a tourniquet effect could actually increase stasis
Myeloproliferative disorders†
Paroxysmal nocturnal hemoglobinuria† and result in an increase in venous thromboembolism.38
Central venous catheterization† Intermittent pneumatic compression devices have
been shown to reduce the incidence of venous
* Data from Clarke-Pearson DL, DeLong ER, Synan IS, Coleman thromboembolism with little or no risk. The use of
RE, Creasman WT. Variables associated with postoperative intermittent pneumatic compression in the perioper-
deep venous thrombosis: A prospective study of 411
gynecology patients and creation of a prognostic model. ative period (intraoperatively and into the first post-
Obstet Gynecol 1987;69:146–50. operative day) has been shown to be effective in
† Modified from Geerts WH, Pineo GF, Heit JA, Bergqvist D,
patients at moderate risk. However, in patients at
Lassen MR, Colwell CW, et al. Prevention of venous
thromboembolism: the Seventh ACCP Conference on higher risk (women with gynecologic cancers), peri-
Antithrombotic and Thrombolytic Therapy. Chest operative use was found to be ineffective.39 In these
2004;126:338-400S. higher-risk gynecologic surgery patients, using inter-
mittent pneumatic compression in the operating room
conditions. In these trials, the heparin was administered and continuing its use throughout the hospital stay has
every 12 hours postoperatively.31–33 been shown to reduce venous thromboembolism by
However, in a prospective study that included approximately threefold.40
patients at higher risk with gynecologic malignancies, Two trials in gynecologic surgery directly compar-
the 12-hour regimen was found to result in no benefit ing low-dose unfractionated heparin or low-molecular-
when compared with the control group.34 A subse- weight heparin with intermittent pneumatic compres-
quent randomized trial found that administering the sion showed equivalent efficacy.41,42 The potential ben-
heparin every 8 hours postoperatively was effective in efit of combining intermittent pneumatic compression
significantly preventing venous thromboembolism in and a pharmacologic agent has been investigated in
a gynecologic oncology population.35 Similar findings general surgery and was found to be beneficial in
have been reported in other high-risk surgical popu- high-risk patients. Although there are no trials of “com-
lations and subsequent management guidelines advise bination prophylaxis” in gynecologic surgery, it is rea-
more “intense” low-dose heparin regimens in patients sonable to combine both a pharmacologic agent and
at higher risk.27 intermittent pneumatic compression in patients at very
Although low-dose heparin rarely has a measurable high risk for venous thromboembolism complications
effect on coagulation (prolonging partial thromboplastin (age older than 60 years, cancer, previous venous
time), an increase in bleeding complications—especially thromboembolism, or a combination of these).
wound and injection site hematomas—is reported. How-
ever, serious hemorrhagic complications are not Diagnosis and Management of
increased in patients receiving low-dose heparin. Venous Thromboembolism
An alternative to low-dose unfractionated heparin The clinical signs and symptoms of DVT (leg pain,
is low-molecular-weight heparin. When compared edema, erythema) are often misleading and may
VOL. 121, NO. 3, MARCH 2013 Clarke-Pearson and Geller Complications of Hysterectomy 661
trocar, bladder location should be considered and entry
should be placed sufficiently cephalad. This placement
can be confirmed laparoscopically by applying pres-
sure at the site of the proposed trocar placement before
making any incisions or with the use of a needle when
injecting a local anesthetic agent. During abdominal or
laparoscopic hysterectomy, it is important to mobilize
the bladder flap before uterine artery ligation and
cervical amputation to avoid inadvertent incorporation
of any bladder tissue into these incisions. Sharp
dissection will allow more careful identification of the
anatomic planes, especially if scarring is encountered
from previous cesarean delivery or other surgery.
During vaginal hysterectomy, sharp dissection at the
time of anterior peritoneal entry is recommended to
mobilize the bladder base and reduce the risk of injury.
Subsequent placement of a retractor in this vesicova-
ginal plane is recommended to elevate the bladder
away from further dissection.
The same principles of identification hold true for Fig. 1. Vascular and ureteral anatomy encountered during
hysterectomy. Illustration by John Yanson.
preventing ureteral injuries. The ureters pass over the
Clarke-Pearson. Complications of Hysterectomy. Obstet Gynecol
bifurcation of the common iliac vessels before diving 2013.
below the uterine artery and passing anterior and
lateral to the cervix50 (Fig. 1). Identification of the
ureter on the medial leaf of the broad ligament is but rather a thinning of the bladder wall. A full-
critical when operating in the pelvis. If not seen thickness injury will usually reveal leakage at the site
directly, a retroperitoneal approach can be used by of injury.
opening the para-rectal space. Maintaining surgical Ureteral injuries can be classified according to
principles such as avoiding aggressive surgical ped- type of injury—thermal, obstructive, or mechanical
icles and remaining medial to all previous surgical transection—as well as location in relation to the pelvic
pedicles when performing the hysterectomy also will brim. These classifications will help to determine what
help to prevent ureteral injury. In addition, if hemor- type of repair will ensure a tension-free anastomosis.
rhage is encountered in the pelvis, applying pressure
is more prudent than clamping and ligating when Identification of Injury
the surgical field is obscured. Preoperative ureteral Identification of a bladder injury can be obvious, with
stenting has not been shown to decrease the rate of extravasation of urine at the time of cystotomy.
ureteral injury during hysterectomy.51 However, intra- However, a small injury might go unrecognized.
operative stenting in cases in which identification of Retrograde injection of 300 mL of indigo carmine,
the ureters is not otherwise possible can be useful. methylene blue, or sterile milk into the bladder
through a Foley catheter should identify the presence
Type of Genitourinary Tract Injury of a bladder injury, the number of sites of injury, and
Injury to the bladder is often classified by location in the presence of any serosal or muscularis injuries
reference to the dome and the trigone, as well as the (which would appear as a thinning of the bladder wall
depth of injury, including whether it is full-thickness without actual leakage). Cystoscopy with a 30-degree
or isolated to the serosa and underlying muscularis. or 70-degree scope also is a useful adjunct in the
Injuries to the bladder that occur during hysterectomy diagnosis of a bladder injury. Again, at least 300 mL
are usually intraperitoneal and usually occur near the of saline should be used to distend the bladder walls
dome of the bladder. Bladder injuries that are near the sufficiently to uncover smaller defects. This method is
trigone may require more extensive evaluation to particularly useful during vaginal surgery when full
avoid concomitant ureteral injury during repair. inspection of the bladder is difficult. The finding of air
Injuries to the serosa or muscularis of the bladder in the Foley collection bag during laparoscopic
can be distinguished from full-thickness mucosal surgery is highly suspicious for a bladder injury and
bladder injuries in that there is no leakage of urine calls for careful inspection as described.
VOL. 121, NO. 3, MARCH 2013 Clarke-Pearson and Geller Complications of Hysterectomy 663
healing. A critical point is that the repair must be benefit to bowel preparation, even when bowel surgery
tension-free. If this step is not addressed, failure of the is anticipated.53,54 Oral antimicrobial administration
repair is likely. also is not necessary before hysterectomy because par-
To assure a tension-free repair, the location of the enteral antibiotics are administered.
injury dictates the type of anastomosis that will take Intraoperatively, the use of a nasogastric or
place. Injuries that occur in the proximal two-thirds of oralgastric tube also can help to decompress the
the ureter usually can be repaired using an end-to-end stomach and possibly prevent injury during the place-
reanastomosis with 4-0 absorbable suture. This is called ment of laparoscopic trochars. If performing laparo-
a ureteroureterostomy. If a large segment of ureter has scopic surgery, care should be taken not only with
been removed, it may be necessary to implant the placement of insufflating needles and trocars but also
damaged ureter into the lumen of the contralateral with manipulation of any laparoscopic instruments,
ureter. This is called a transureteroureterostomy. If especially electrosurgical instruments. These instru-
ureteral injury occurs in the distal one-third of the ments should be inspected for any signs of disruption
ureter, as is most common during gynecologic proce- of their protective insulation. This is true for abdominal
dures, the proximal end of the ureter can be tunneled and vaginal cases as well as laparoscopic surgery. If
through the full thickness of the bladder and affixed intra-abdominal adhesions are expected, an alternative
to the mucosal surface with a 3-0 absorbable suture. entry technique is a 2-mm to 5-mm incision at the
This is called a ureteroneocystostomy. If there is any Palmer point, followed by insufflation, trocar, and
concern for tension at the repair site, the bladder can be camera placement at this site to survey the lower
brought cephalad and fixed to the psoas muscle in what abdomen. An “open” entry technique also might be
is called a psoas hitch. considered. However, open entry for the initial port
With any of these repairs, a stent should be placed has not been shown to reduce the incidence of GI
in the ureter and a closed suction drain should be injury; it will reduce the incidence of vascular injury.
placed at the site of anastomosis. A catheter also
should be placed in the bladder to avoid urinary Type of Gastrointestinal Tract Injury
retention and back-flow of intravesicular pressure. There are three major types of bowel injury that can
These drains should be left in place for at least 1 week. occur during hysterectomy: thermal injury; direct
A CT scan with contrast also should be performed not mechanical damage; and indirect injury through inter-
only to confirm patency of the repair but also to look ruption of vascular supply. Thermal damage can occur
for any areas of stenosis. with all types of hysterectomy but will most often go
undiagnosed during laparoscopic surgery. This is most
Gastrointestinal Tract common with conductance of current from a monop-
Injury to the GI tract after hysterectomy ranges from olar instrument to the bowel during pelvic dissection,
0.1% to 1%, with estimates of 0.3% for abdominal and especially deep in the pelvis, at the cuff, or in the
0.2% for laparoscopic hysterectomy.8 The occurrence cul-de-sac, where visualization is not as clear.
of bowel injury with vaginal hysterectomy ranges Direct mechanical damage to the bowel can occur
from 0.1% to 1.0%.3,10 There was no difference in rate both with sharp and blunt instruments. This type of
of bowel injury based on subtype or route of hyster- injury most commonly occurs during adhesiolysis.
ectomy according to a 2009 Cochrane Review.3 Although more likely to be recognized at the time
of surgery, some serosal tears can go unnoticed.
Prevention of Gastrointestinal Tract Injury Mechanical injuries also can occur with placement of
Prevention of a bowel injury can be accomplished laparoscopic instruments.
through understanding the potential sites of harm as Vascular injuries can occur with interruption of
well as good surgical technique, such as gentle tissue the blood supply coursing through the mesentery to
handling, careful dissection along anatomic planes, the bowel. Such dissection rarely is indicated with
judicious use of electrosurgery, and compression of a benign hysterectomy, but the risk of devasculariza-
bleeding rather than clamping when the surgical field tion always should be recognized.
is obscured.
In the past, mechanical bowel preparation was Identification of Injury
advised. Advantages were thought to include improved Thermal injuries to the bowel can go unrecognized at
exposure in the pelvis with decompressed bowel and the time of surgery. However, inspection may reveal
deceased bacterial load if bowel injury were to occur. blanched spots on the serosal surface of the bowel. If
However, randomized trials have shown no clinical left unrepaired, a thermal injury often has a delayed
VOL. 121, NO. 3, MARCH 2013 Clarke-Pearson and Geller Complications of Hysterectomy 665
Prevention —depending on the situation and skills of the surgeon.
There are many variables that contribute to increased Additional bleeding may arise either when the bladder
bleeding complications, including obesity, poor visual is dissected from the cervix and upper vagina or when
exposure (attributable to endometriosis, adhesions, the vagina is incised in performing the colpotomy to
large masses), distorted anatomy, uterine fibroids, skill remove the uterus (also occurring at the initiation of
of the surgeon, volume of the surgeon’s practice, sur- a vaginal hysterectomy).
gical volume of the hospital, pharmacologic venous More serious bleeding occurs from injury to the
thromboembolism prophylaxis, concurrent use of aspi- plexus of pelvic veins, which are not easily isolated or
rin and other antiplatelet drugs, and over-the-counter dissected. The best means to manage bleeding from a
preparations (including, but not limited to, garlic, gin- venous plexus is direct pressure with a laparotomy
ger, ginko, and St. John’s wort).55 pack or sponge to immediately control the bleeding site.
Few data exist that prospectively describe the
most common sites of intraoperative bleeding at the Management
time of hysterectomy. The pelvis has a rich blood The surgeon then must assess the extent of the vascular
supply arising predominately from the external and injury and develop a plan of management, which
internal iliac arteries. Other arterial supply comes includes the following: obtaining better exposure by
from the inferior mesenteric artery (supplying the extending the incision, placing a different retractor, and
sigmoid colon) and the ovarian arteries. The venous adjusting lighting (including considering the use of
drainage from the pelvis is even more complex, with a headlamp); obtaining additional assistance and addi-
a network of collateral veins that drain through the tional suction; assuring that the patient is stable and
external iliac, internal iliac, inferior mesenteric, and that the anesthesia team has adequate blood products
ovarian veins. It is likely that the majority of significant available (additional intravenous lines for rapid blood
intraoperative bleeding comes from injury to these product infusion may be necessary); and securing
pelvic veins. necessary supplies that may be used to obtain hemo-
Understanding and identifying the vascular anat- stasis (suture, clips, and hemostatic agents).
omy of the pelvis and its relationship to other organs Once these steps have been taken, the area of
(ureter, bladder, rectum) are central to preventing bleeding is slowly exposed to ascertain the site and
hemorrhagic complications. This is even more impor- extent of the vascular injury. In many cases, direct
tant when the anatomy and anatomic land marks are pressure on a small venous bleeding point may
distorted by other pathology (Fig. 1). Being familiar coagulate on its own. If the bleeding site can be
with the retroperitoneal anatomy also helps protect identified and isolated, it may be clamped and suture-
vital structures and vessels in the course of a difficult ligated or clipped. Use of hemostatic agents contain-
dissection. Taking advantage of surgical planes that ing thrombin or coagulation products that come in
are avascular is key to preventing vascular injury solid, liquid, or powder form, combined with pres-
and in controlling bleeding. These spaces include par- sure, may be necessary to control diffuse bleeding
arectal space, paravesical space, rectovaginal septum from a venous plexus.
or space, cervico-vesicle plane, and retropubic space During significant bleeding, attention must be
(space of Retzius). paid to the patient’s overall status and the surgeon
must be certain that adequate blood products (includ-
Identification ing clotting factors) are being replaced. In most instan-
In the course of performing a hysterectomy (by ces, pressure on the bleeding site will stop bleeding
whatever technique or route), the primary vascular and allow time to replace blood and blood products.
supply to the uterus must be controlled. Basically, this The error we have seen too often is not recognizing
involves controlling the ovarian and uterine blood that the patient has become depleted of clotting fac-
supply. Loss of control of these vascular pedicles often tors in the course of trying to achieve hemostasis,
results in obvious bleeding, which usually can be resulting in additional diffuse bleeding from other
managed by clamping and ligating the pedicle. When sites. A general rule is to replace clotting factors
clamping bleeding vessels, especially the ovarian and (fresh-frozen plasma, cryoprecipitate, platelets) for
uterine vessels, care must be taken to identify the every 4 units of packed red blood cells administered,56
ureter as well as other vessels on the pelvic side wall. although newer trauma data recommend a 1:1 ratio
Suture ligature, hemostatic clips, and bipolar and for massive blood loss.57 For more difficult bleeding,
monopolar electrosurgery are all reasonable means ligation of the anterior division of the internal iliac
to obtain hemostasis—once vessels have been identified arteries has been successful in decreasing the arterial
VOL. 121, NO. 3, MARCH 2013 Clarke-Pearson and Geller Complications of Hysterectomy 667
Table 1. Less Common Nerve Injuries Associated With Hysterectomy
Prognosis depends on the severity of the injury. retractor) is the surgeon’s responsibility. For patients
As the nerve bundles go through degeneration and placed in the lithotomy position, a careful examination
subsequent regeneration, mild neuropathies may of the legs should confirm that the hips are not
resolve within days to weeks, whereas more severe hyperflexed or overly externally rotated, no matter
cases may take months to possibly years. It is important what type of surgical stirrups are used.63 A general rule
for the patient to understand that recovery may be is to line up the long axis of the lower extremity with
prolonged to prevent early feelings of fear and the umbilicus and the contralateral shoulder to ensure
frustration. Spontaneous resolution of most femoral the correct angle of external rotation of the hip. One
neuropathies is the rule rather than the exception. can assess for hyperflexion by palpating the popliteal
Prevention of a femoral nerve injury is accom- pulses. If the angle of flexion is great enough to com-
plished by good surgical technique. Care should be press the femoral artery in the inguinal canal, it is likely
taken with the placement of any self-retaining retrac- causing compression of the nerve as well.
tor. Examination of placement of the retractor blades
should be performed and periodically checked during Iliohypogastric and Ilioinguinal Nerves
the case to ensure that they are not directly lying on The iliohypogastric and ilioinguinal nerves also are at
the psoas muscle. For some retractors, such as an risk for injury during hysterectomy, albeit less com-
O’Conner-O’Sullivan retractor, the lateral blades may monly than the femoral nerve.58 These nerves emerge
be elevated off of the psoas muscle by placing towels from the T12 to L1 and L1 to L2 regions, respectively,
between the retractor and the abdominal wall. For and course through the muscles of the anterior abdom-
other retractors, elevation of the retaining ring off inal wall. Specifically, the iliohypogastric penetrates the
of the abdominal wall and selection of appropriately fascia of the internal oblique and the ilioinguinal pen-
sized retractor blades (such as with a Bookwalter etrates the fascia of the transverses abdominus. The
VOL. 121, NO. 3, MARCH 2013 Clarke-Pearson and Geller Complications of Hysterectomy 669
attention to positioning on the operating table and preferable to electrosurgery. When electrosurgery is
good surgical technique. A thorough understanding of used, techniques that reduce thermal spread should
the pathophysiology of injury, identification of dys- be used, including lower voltage, reduced tissue con-
function, and timely initiation of treatment are essen- tact, using a cutting (compared with coagulation) cur-
tial to optimize outcome and to minimize patient rent, and using other energy devices such as
anxiety when dealing with such an injury. Fortunately, a harmonic scalpel. There is no evidence that primary
most injuries are transient and will recover spontane- closure of the cuff is preferable to secondary granula-
ously, although this may take several weeks to tion in terms of preventing a dehiscence.67 There is no
months. Diligent counseling of patients not only after evidence that a two-layer closure is superior to a single-
an injury but also in the preoperative period is layer or open cuff.69 One study reported decreased
essential. dehiscence with use of a barbed suture during laparo-
scopic and robotic cuff closure.70 A retrospective anal-
VAGINAL CUFF DEHISCENCE ysis reported decreased dehiscence with transvaginal
Vaginal cuff dehiscence is a postoperative complica- suturing for total laparoscopic hysterectomy.71 Good
tion unique to hysterectomy. Although it is a rare surgical technique can help to decrease the risk of
complication, it can lead to serious morbidity. The dehiscence, including minimal use of electrosurgery
estimated incidence of vaginal cuff dehiscence is for hemostasis, end-to-end tissue approximation with
0.24% based on 10-year cumulative data, but it is incorporation of the all vaginal layers, 1-cm suture mar-
slightly higher (0.39%) when assessing more recent gins, and maintaining hemostasis.
years.64 Total laparoscopic hysterectomy has the high-
est rate of vaginal cuff dehiscence (1.35%) compared Presentation
with laparoscopic-assisted vaginal hysterectomy Vaginal cuff dehiscence usually occurs in the first
(0.28%; OR 4.9, CI 1.1–21.5), total abdominal hyster- several weeks to months after surgery but can be
ectomy (0.15%; OR 9.1, CI 4.1–20.3), and total vag- delayed up to several years (especially in postmeno-
inal hysterectomy (0.08%; OR 17.2, CI 3.5–75.9).65 pausal women). Median time to dehiscence has been
One estimate of cuff dehiscence after robotic total reported from 1.5 to 3.5 months.67,70 In one study,
hysterectomy based on 510 patients was 4.1%.65 How- median time to dehiscence was delayed up to 6.5
ever, there was no comparison with other routes. A months in the vaginal hysterectomy group compared
meta-analysis comparing vaginal, laparoscopic, and with 2.5 months for all other routes.64 Vaginal cuff
robotic cuff closure found that vaginal closure has dehiscence often presents with postcoital bleeding,
the lowest incidence (0.18%), followed by laparo- other vaginal spotting, or watery vaginal discharge.64,67
scopic (0.64%; OR 3.57, CI 1.54–8.33), and robotic If bowel evisceration has occurred, patients often have
(1.64%; OR 9.09, CI 3.86–25.0).66 The increased risk symptoms of pelvic pressure or a bulge. Any patient
seen with the laparoscopic-assisted routes is thought to suspected of having a cuff dehiscence requires a vaginal
be attributable to the unique use of monopolar elec- examination with visual inspection and manual palpa-
trosurgery to perform colpotomy with these modali- tion to determine cuff integrity. An abdominal exami-
ties and possible differences in closure methods and nation also should be performed to rule out an acute
techniques. abdomen from peritonitis, bowel injury, or obstruction.
If there is suspicion for occult bowel injury, CT imag-
Risk Factors and Prevention ing should be considered.
Several risk factors have been postulated for the risk of
cuff dehiscence, with direct trauma from sexual inter- Management
course being most common.67 Other risk factors Vaginal cuff dehiscence can be managed either conser-
include repetitive Valsalva maneuver (chronic cough, vatively or surgically. All dehiscences lead to exposure
constipation, obesity, straining), smoking, malnutrition, of the peritoneal cavity to vaginal flora and should be
anemia, diabetes, immunosuppression, menopausal managed with broad-spectrum antibiotics. Small or
status, previous pelvic surgery, previous vaginoplasty, partial dehiscences can be managed expectantly with
and corticosteroid use.68 Because the rate of dehiscence pelvic rest. Larger or complete dehiscences usually
is so low, it is difficult to truly understand how these require surgical closure. This should be accomplished
risk factors contribute to the overall incidence. As vaginally when possible. If bowel evisceration has
a result, there is little information regarding practice occurred, the bowel must be thoroughly inspected for
recommendations to prevent cuff dehiscence. Some injury. Although vaginal cuff dehiscence is rare, it can
hypothesize that “cold” incision of the vagina may be lead to serious morbidity. Prevention, early recognition,
VOL. 121, NO. 3, MARCH 2013 Clarke-Pearson and Geller Complications of Hysterectomy 671
27. Geerts WH, Pineo GF, Heit JA, Bergqvist D, Lassen MR, 43. Low-molecular-weight heparin in the treatment of patients with
Colwell CW, et al. Prevention of venous thromboembolism: venous thromboembolism. The Columbus Investigators. N
the Seventh ACCP Conference on Antithrombotic and Throm- Engl J Med 1997;337:657–62.
bolytic Therapy. Chest 2004;126:338–400S. 44. Gilmour DT, Dwyer PL, Carey MP. Lower urinary tract injury
28. Harkki-Siren P, Sjoberg J, Kurki T. Major complications of lapa- during gynecologic surgery and its detection by intraoperative
roscopy: a follow-up Finnish study. Obstet Gynecol 1999;94:94–8. cystoscopy. Obstet Gynecol 1999;94:883–9.
29. Ritch JM, Kim JH, Lewin SN, Burke WM, Sun X, Herzog TJ, 45. Walters MD, Karram MM. Urogynecology and reconstructive
et al. Venous thromboembolism and use of prophylaxis among pelvic surgery. 3rd ed. Philadelphia (PA): Mosby Elsevier; 2007.
women undergoing laparoscopic hysterectomy. Obstet Gyne-
46. Johnson N, Barlow D, Lethaby A, Tavender E, Curr E, Garry R.
col 2011;117:1367–74.
Surgical approach to hysterectomy for benign gynaecological dis-
30. Kakkar VV, Corrigan TP, Fossard DP, Sutherland I, Thirwell J. ease. The Cochrane Database of Systematic Reviews 2006, Issue 2.
Prevention of Fatal Postoperative pulmonary embolism by low Art. No.: CD003677. DOI: 10.1002/14651858.CD003677.pub3
doses of heparin. Reappraisal of results of international multi-
47. Gilmour DT, Das S, Flowerdew G. Rates of urinary tract injury
centre trial. Lancet 1977;1:567–9.
from gynecologic surgery and the role of intraoperative cystos-
31. Ballard RM, Bradley-Watson PJ, Johnstone FD, Kenney A, copy. Obstet Gynecol 2006;107:1366–72.
McCarthy TG. Low doses of subcutaneous heparin in the pre-
48. Carley ME, McIntire D, Carley JM, Schaffer J. Incidence, risk
vention of deep vein thrombosis after gynaecological surgery.
factors and morbidity of unintended bladder or ureter injury
J Obstet Gynaecol Br Commonw 1973;80:469–72.
during hysterectomy. Int Urogynecol J Pelvic Floor Dysfunct
32. Adolf J, Buttermann G, Weidenbach A, Gmeineder F. Optima- 2002;13:18–21.
tion of postoperative prophylaxis of thrombosis in gynaecology
49. Smith GL, Williams G. Vesicovaginal fistula. BJU Int 1999;83:
[in German]. Geburtshilfe Frauenheilkd 1978;38:98–104.
564–9.
33. Taberner DA, Poller L, Burslem RW, Jones JB. Oral anticoa-
50. Aronson MP, Bose TM. Urinary tract injury in pelvic surgery.
gulants controlled by the British comparative thromboplastin
Clin Obstetrics Gynecol 2002;45:428–38.
versus low-dose heparin in prophylaxis of deep vein thrombo-
sis. Br Med J 1978;1:272–4. 51. Kuno K, Menzin A, Kauder HH, Sison C, Gal D. Prophylactic
34. Clarke-Pearson DL, Coleman RE, Synan IS, Hinshaw W, ureteral catheterization in gynecologic surgery. Urology 1998;
Creasman WT. Venous thromboembolism prophylaxis in 52:1004–8.
gynecologic oncology: a prospective, controlled trial of low- 52. Rock JA, Jones HW. TeLinde’s operative gynecology. 10th ed.
dose heparin. Am J Obstetrics Gynecol 1983;145:606–13. Philadelphia (PA): Lippincott Williams & Wilkins; 2008.
35. Clark-Pearson DL, DeLong E, Synan IS, Soper JT, 53. Slim K, Vicaut E, Launay-Savary MV, Contant C, Chipponi J.
Creasman WT, Coleman RE. A controlled trial of two low-dose Updated systematic review and meta-analysis of randomized
heparin regimens for the prevention of postoperative deep vein clinical trials on the role of mechanical bowel preparation
thrombosis. Obstet Gynecol 1990;75:684–9. before colorectal surgery. Ann Surg 2009;249:203–9.
36. Borstad E, Urdal K, Handeland G, Abildgaard U. Comparison 54. Bretagnol F, Panis Y, Rullier E, Rouanet P, Berdah S,
of low molecular weight heparin vs. unfractionated heparin in Dousset B, et al. Rectal cancer surgery with or without bowel
gynecological surgery. II: reduced dose of low molecular preparation: the French GRECCAR III multicenter single-
weight heparin. Acta Obstet Gynecol Scand 1992;71:471–5. blinded randomized trial. Ann Surg 2010;252:863–8.
37. Scurr JH, Ibrahim SZ, Faber RG, Le Quesne LP. The efficacy of 55. Basila D, Yuan CS. Effects of dietary supplements on coagula-
graduated compression stockings in the prevention of deep vein tion and platelet function. Thromb Res 2005;117:49–53.
thrombosis. Br J Surg 1977;64:371–3.
56. Pacheco LD, Saade GR, Gei AF, Hankins GD. Cutting-edge
38. Clarke-Pearson DL, Jelovsek FR, Creasman WT. Thromboem- advances in the medical management of obstetrical hemor-
bolism complicating surgery for cervical and uterine malig- rhage. Am J Obstetrics Gynecol 2011;205:526–32.
nancy: incidence, risk factors, and prophylaxis. Obstet
57. Tan JN, Burke PA, Agarwal SK, Mantilla-Rey N, Quillen K. A
Gynecol 1983;61:87–94.
massive transfusion protocol incorporating a higher FFP/RBC
39. Clarke-Pearson DL, Creasman WT, Coleman RE, Synan IS, ratio is associated with decreased use of recombinant activated
Hinshaw WM. Perioperative external pneumatic calf compres- factor VII in trauma patients. Am J Clin Pathol 2012;137:566–71.
sion as thromboembolism prophylaxis in gynecology: report of
58. Irvin W, Andersen W, Taylor P, Rice L. Minimizing the risk of
a randomized controlled trial. Gynecol Oncol 1984;18:226–32.
neurologic injury in gynecologic surgery. Obstet Gynecol 2004;
40. Clarke-Pearson DL, Synan IS, Hinshaw WM, Coleman RE, 103:374–82.
Creasman WT. Prevention of postoperative venous thrombo-
59. Cardosi RJ, Cox CS, Hoffman MS. Postoperative neuropathies
embolism by external pneumatic calf compression in patients
after major pelvic surgery. Obstet Gynecol 2002;100:240–4.
with gynecologic malignancy. Obstet Gynecol 1984;63:92–8.
60. Alsever JD. Lumbosacral plexopathy after gynecologic surgery:
41. Clarke-Pearson DL, Synan IS, Dodge R, Soper JT, Berchuck A,
case report and review of the literature. Am J Obstetrics Gyne-
Coleman RE. A randomized trial of low-dose heparin and inter-
mittent pneumatic calf compression for the prevention of deep col 1996;174:1769–77.
venous thrombosis after gynecologic oncology surgery. Am J 61. Kvist-Poulsen H, Borel J. Iatrogenic femoral neuropathy sub-
Obstetrics Gynecol 1993;168:1146–53. sequent to abdominal hysterectomy: incidence and prevention.
42. Maxwell GL, Synan I, Dodge R, Carroll B, Clarke-Pearson DL. Obstet Gynecol 1982;60:516–20.
Pneumatic compression versus low molecular weight heparin 62. Quan D, Bird S. Nerve conduction studies and electromyogra-
in gynecologic oncology surgery: a randomized trial. Obstet phy in the evaluation of peripheral nerve injuries. Univ Pa
Gynecol 2001;98:989–995. Orthop J 1999;12:45–51.
VOL. 121, NO. 3, MARCH 2013 Clarke-Pearson and Geller Complications of Hysterectomy 673