Hysteroscopy is safe and effective and can greatly reduce morbidity and enhance recovery compared with conventional open procedures. Furthermore, it can minimise the inherent risks of general anaesthesia and hospital admissions because many hysteroscopic procedures can now be delivered in non-anaesthetised women in an outpatient setting. The overall complication rate of hysteroscopic intervention is estimated to be less than 1% [1, 2], of which half is due to serious complications, namely fluid overload, uterine perforation and upper genital tract haemorrhage. Complications occur during access and entry into the uterine cavity or during the diagnostic or operative procedure. Complication rates are higher with operative hysteroscopy (0.95%) compared with purely diagnostic procedures (0.13%) . These risks are highest with more complex hysteroscopic surgery such as myomectomy and adhesiolysis and lowest with purely diagnostic procedures.
Hysteroscopy is safe and effective and can greatly reduce morbidity and enhance recovery compared with conventional open procedures. Furthermore, it can minimise the inherent risks of general anaesthesia and hospital admissions because many hysteroscopic procedures can now be delivered in non-anaesthetised women in an outpatient setting. The overall complication rate of hysteroscopic intervention is estimated to be less than 1% [1, 2], of which half is due to serious complications, namely fluid overload, uterine perforation and upper genital tract haemorrhage. Complications occur during access and entry into the uterine cavity or during the diagnostic or operative procedure. Complication rates are higher with operative hysteroscopy (0.95%) compared with purely diagnostic procedures (0.13%) . These risks are highest with more complex hysteroscopic surgery such as myomectomy and adhesiolysis and lowest with purely diagnostic procedures. Counter-intuitively, the experience and caseload of the surgeon does not appear to be associated with reduced rates of complications, but this is because such surgeons are undertaking the more difficult hysteroscopic surgical procedures [1, 3].
As with any medical intervention where adverse events are possible, it is essential to counsel the patient thoroughly and take fully informed consent. In particular, the benefits, objectives, and alternatives to the proposed surgery should be discussed. Realistic post-operative expectations should be set, including the possibility of further surgeries with incomplete resection of pathology such as fibroids and adhesions. Potential risks of the procedure should be raised, including fluid media complications and the serious complications of trauma, thermal injury and haemorrhage, which may require additional corrective surgery such as laparoscopy, laparotomy or hysterectomy.
Complications may be categorised broadly into general surgical complications and specific complications particular to the operation. General complications are more likely to occur if patient preparation is inadequate. Risks of such complications (associated with anaesthesia, allergies, suboptimal patient experience, etc.) can be mitigated with careful diagnostic workup, including taking a thorough clinical history, physical examination and relevant investigations such as pelvic ultrasound. In this way, relevant patient factors, co-morbidities, anatomic variations, underlying pathologies and preferences are accounted for to inform the surgical approach, including the treatment setting. Examples could include:
Appreciating a woman’s anxieties, prior gynaecological or life experiences, and preferences when counselling about treatment settings, analgesia and anaesthesia
Considering anatomic factors and genital tract pathologies (e.g. pre-existing conditions such as cervical stenosis from cervical surgery, acquired uterine abnormalities, such as distorted uterine anatomy or adhesions from a previous endometrial ablation) when selecting the size and type of instrumentation, the need for cervical preparation and the optimal mode of anaesthesia
Identifying the possibility of a pyometra in an elderly woman with intrauterine fluid and/or purulent vaginal discharge, and deferring procedures so that antibiotics can be administered, reducing the risk of systemic infection
Accounting for a woman’s elevated body mass index (BMI) when considering the mode of anaesthesia; an outpatient setting with or without local anaesthesia, or a regional anaesthetic for more a complex hysteroscopic procedure, is often safer than a general anaesthetic.
While there are few contraindications to a procedure being undertaken in an outpatient setting, some co-existing medical complaints, especially cardiac conditions, need careful management and liaison with a cardiologist and anaesthetist. This includes women with Fontan’s circulation, where the consequences of an induced vasovagal reaction (one of the commonest ‘minor’ side effects of outpatient hysteroscopy) may be more serious, because they may not compensate adequately for a sudden drop in blood pressure.
9.4 Specific Operative Complications
a. Prevention. Mitigate risks by carefully selecting cases, preparing patients and conducting proficient surgery. This requires adequate training, experience and an appropriate caseload, as well as quality equipment including optics, stacks, technologies and fluid management systems.
b. Recognition. Reduce morbidity by being aware of potential problems, having a high index of suspicion for adverse events and promptly identifying complications.
c. Management. Optimise clinical outcomes by carrying out a thorough clinical assessment and instituting effective remedial treatment and post-operative surveillance.
Hysteroscopic surgery can induce lower and upper genital tract trauma from manoeuvres needed to instrument the uterine cavity or the operative procedure itself. The most common traumatic complications are cervical laceration, uterine perforation (including injury to intra-abdominal viscera and thermal burns) and haemorrhage.
A cervical laceration may occur during the dilatation phase of any hysteroscopic procedure. This is most likely to occur when there is cervical stenosis, perhaps due to previous cervical surgery or post-menopausal state, especially in women not having had a vaginal delivery, or the elderly. The cervix may be lacerated by a tenaculum placed to help stabilise the cervix, especially when the cervix is atrophic after the menopause or where greater cervical dilatation is required, such as for the insertion of a 9 mm resectoscope, where significant traction and counter-traction can be necessary.
Be aware of patient factors and of instrumentation as outlined above that may make cervical trauma more likely, and adopt a careful, gentle and meticulous approach. Consider the use of vaginoscopic techniques to ensure continuous visual entry, especially where there is stenosis of the external cervical os; 5 Fr hysteroscopic scissors or bipolar electrodes can be used to open up the blockage allowing a miniature hysteroscope to advance along the cervical canal. While gonadotrophin-releasing hormone analogues (GnRHa) may improve visualisation during transcervical resection of the endometrium or fibroids and reduce intraoperative blood loss and fluid overload [4, 5], they do make the cervix atrophic and harder to dilate, increasing the risk of cervical trauma . Some practitioners advocate cervical ‘preparation’ to ease passage of the hysteroscope, reduce the need for mechanical cervical dilatation and reduce genital tract trauma. Cervical ‘ripening’ agents include prostaglandins such as misoprostol, the progesterone antagonist mifepristone and osmotic dilators (laminaria tents or synthetic materials). Although the data do not show any benefit prior to diagnostic hysteroscopy, for women undergoing operative hysteroscopy where cervical dilatation beyond 5 mm diameter (Hegar 5) is predicted, these agents may be beneficial, especially in pre-menopausal women [7, 8]. A Cochrane review of randomised controlled trials evaluating cervical priming before operative hysteroscopy found that cervical lacerations and false passages were reduced with the use of such agents, although misoprostol was associated with side effects including pre-operative pain and vaginal bleeding . Moreover, the possible benefits of laminaria need to be weighed against the inconvenience of their insertion and retention for one to two days.
Bleeding, with or without a tear or avulsion of cervical or vaginal tissue, is usually apparent once the tenaculum or vulsellum forceps are removed. Greater degrees of tissue trauma occur when the forceps are inadvertently detached when applying significant degrees of tissue counter-traction or with repeated detachments.
A cervical laceration can usually be managed conservatively with simple tamponade achieved by applying pressure using a cotton wool ball or gauze (‘swab on a stick’) because in most cases the bleeding is minor and short-lasting. However, if bleeding continues then treatment options include the application of a topical haemostatic agent (e.g. Monsel’s (ferric sulfate) solution), compression with an atraumatic haemostat (e.g. Rampley sponge-holding forceps) applied for up to 15 minutes, or placing a suture. The latter is usually necessary where cervico-vaginal tissue has been avulsed.
Uterine perforation may be caused at a number of different times during the hysteroscopic procedure. This complication is usually minor, but on occasion can have major consequences if associated with haemorrhage, intra-abdominal mechanical injury, thermal visceral injury, or infection. The factors that increase the risk of uterine perforation are uterine anomalies, infection, recent pregnancy, previous intrauterine surgery, an acutely anteverted or retroverted uterus, distorted uterus with fibroids, and post-menopausal status [1, 10]. Uterine perforation can occur during any intrauterine procedure but is most commonly associated with termination of pregnancy . The Royal College of Obstetricians and Gynaecologists (RCOG) reports an average incidence of uterine perforation during hysteroscopy at 0.002–1.7% ; others have reported an incidence of 1.6% . Many perforations are in the body of the uterus and are small, causing very little damage or bleeding. However, there is always a risk that abdominal organs could be damaged by whatever instrument caused the perforation.
Data from the largest published series evaluating complications associated with hysteroscopy estimated the incidence of uterine perforation to be 0.13% and 0.76% for diagnostic and operative hysteroscopy, respectively . For operative hysteroscopy, the occurrence of uterine perforation for hysteroscopic adhesiolysis (4.5%) was more than 10 times as great as that for hysteroscopic polypectomy (0.4%).
Be aware of patient factors and the instrumentation as outlined above that may increase the likelihood of uterine trauma, particularly full-thickness uterine perforation. Adopt a a careful, gentle and meticulous approach. Avoid routine blind cervical dilatation and blind insertion of instruments such as polyp forceps. Appreciate the angle of the distal lens of the endoscope so alignment during insertion is correct, and use fluid management systems to maintain a clear operative field. Only hysteroscopic surgeons with sufficient expertise in these procedures and an appropriate caseload should undertake resection of FIGO type 2 fibroids and hysteroscopic adhesiolysis for Asherman’s syndrome. Consider additional guidance with ultrasound, fluoroscopy or laparoscopy for these more difficult cases. Pre-operative endometrial downregulation is a subject of debate but, for the more difficult cases, up to three months of gonadotrophin-releasing hormone analogue (GnRHa) administration can improve visualisation by suppressing the endometrium and reducing the volume of submucosal fibroids . If blind, global endometrial ablation technologies are being used, then it is essential to adhere to the manufacturer’s instructions for use.
A false passage or ‘track’ may be seen where the hysteroscope has been inadvertently buried into the underlying endometrium or myometrium. This can be seen as a narrow, blind-ending channel that does not seem to follow the usual expected contours of the cervix and uterus and often bleeds. The first sign of a full-thickness uterine perforation is often the feeling of an instrument (usually a dilator or mechanical instrument such as polyp forceps introduced blindly) going ‘beyond’ the previously sounded uterine cavity length, or a sudden loss of tissue resistance. A perforation, whether caused by blind instrumentation or during hysteroscopic surgery, appears as a dark and usually circular defect because light from the hysteroscope is absorbed beyond the confines of the uterus and into the abdominal cavity. If the hysteroscope is advanced through the perforated uterine wall, a network of filmy ‘cobwebs’ of areolar tissue is seen as the serosal surface is breached. The yellow fat of bowel epiploica or omentum confirms the perforation. Where visualisation is compromised from bleeding deep within the myometrium (e.g. during electrical resection of fibroids) or distorted anatomy (e.g. during adhesiolysis) then the first indication of a perforation can be the sudden collapse of the uterus and loss of vision together with a sudden, large loss of distension media with a rapid accumulation of fluid deficit recorded on an automated fluid management system (Figure 9.1).
Managing a uterine perforation depends on the type of procedure being undertaken, when it occurred and what instrument caused the trauma. If the instrument is 5 mm or narrower and no energy has been used, such as a small hysteroscope or dilator, then the likelihood of significant bleeding or inadvertent visceral injury is negligible (the exception being in women with dense intra-abdominal adhesions, usually from previous bowel surgery). A single dose or course of a broad-spectrum antibiotic, observation for an hour or so and an explanation to the patient is all that is required. Patients should be advised to contact the hospital directly in the event of increasing pain, bleeding, general malaise or fever.
Bleeding and visceral injury are more likely where a larger instrument has caused the perforation, especially if blind attempts at mechanical grasping of tissue have taken place or hysteroscopic scissors, tissue removal systems or activated electrical instruments have been used. In such circumstances, a laparoscopy must be performed. A urinary catheter should be placed to check not only for the urinary volume as there may be shock, but also for haematuria, which could indicate bladder trauma. Careful inspection of the uterus and surrounding organs is required; bleeding from the uterine laceration may require a suture, which should be placed laparoscopically. If there is considerable bleeding, for example into the broad ligament, then laparotomy and assistance from a general or vascular surgeon may be required. If perforation occurs during the activation of a resection loop then there should be prompt consultation with a general surgeon as full inspection of the bowel will be needed and this is likely to require laparotomy. The risk of bowel injury during perforation is 0.2–0.8% [15, 16]. A stable patient can be discharged the following day with advice that they should recover without any problems, but if they have abdominal pain, a fever or begin to feel unwell then they need to contact the hospital and will require readmission. For those who have had a perforation and associated injuries to other intra-abdominal structures, surgical repair, intravenous antibiotics and close observation will be necessary.
The most dangerous and unpredictable uterine perforations are those associated with diathermy. Delayed thermal injury causing tissue necrosis and visceral perforation may occur, so women should be made aware of possible delayed presentations. Vaginal electrical burns can occur during hysteroscopy when diathermy is used, but the incidence is low .
Haemorrhage is often associated with genital tract trauma, but in the absence of this it arises from the uterine surface, especially during endometrial resection or resection of a fibroid. The estimated incidence ranges from 0.61% to 6.9% [13, 16]. In the absence of uterine perforation, management will depend on the severity of the bleeding.
Adopt a technique that ensures gentle tissue handling under clear visualisation. The use of GnRHa may aid visualisation and avoidance of inadvertent tissue trauma [4, 5, 18, 19]. Two randomised controlled trials have supported the use of dilute intracervical vasopressin for decreasing bleeding and total blood loss at the time of surgery [20, 21].
Profuse intrauterine or vaginal bleeding is obvious. Continued ‘trickling’ is more insidious, but where there is ongoing bleeding, especially with unstable vital signs, then measures as for profuse bleeding should be undertaken. Concealed, intra-abdominal bleeding can occur if uterine perforation is not recognised. This is more likely after resectoscopic surgery, and deterioration in vital signs (hypotension, tachycardia, low oxygen saturation) and the woman’s clinical condition (pallor, pain, abdominal tenderness) should alert the clinician to concealed intra-abdominal bleeding.