With a range of both therapeutic and diagnostic indications, hysteroscopy is one of the gynecological proceduresmost commonly performed. Although both effective and generally safe, this intervention does carry a small risk of fluid overload, fluid absorption, and electrolyte imbalance (which can on rare occurrence be fatal). Absorption degree is affected by various factors, including type or length of procedure, intrauterine pressure, and the possibility of more fluid absorption as caused by operative hysteroscopy (versus diagnostic hysteroscopy, due to greater endomyometrial disturbance). Although fluid overload incidence is low (occurring in 0.1%-2% of cases), close monitoring can prevent excessive absorption of fluid. Because decreasing fluid absorption is a key gynecological concern, even at the cost of procedural noncompletion, various interventions with this absorption decrease have been recorded and described (including choice of anesthesia and preoperative hormonal medication). However, despite a wide array of studies and descriptions about these methods, varied conclusions about their efficacy in reduction of fluid absorption and their associated complications exist. This systematic retrospective review aims for an assessment of which interventions most effectively reduce absorption of fluid during hysteroscopy, as well as determining whether or not they improve the completion rate and safety of hysteroscopy. Types of studies included were randomized controlled trials (RCTs), prospective studies, and retrospective studies of women undergoing fluid-balance-reporting studies (including both a comparator and intervention arm). Eligibility for final inclusion also mandated that the studies be full-text, English-language studies. No-intervention control groups, alternative preoperative pharmacotherapy, placebo, and alternative anesthesia types were considered acceptable comparators. Seeing as standardization was lacking in both reporting and performance, studies primarily comparing surgical equipment or technique were also excluded. Two authors (J.D. and A.Z.) independently screened and fully reviewed materials with potential for inclusion. Covidence (www.covidence.org) was used in all data extraction, screening, and bias risk analyses. After initial screening, full-text assessment yielded 28 studies fully eligible for inclusion. Among these studies, the most commonly performed procedure proved to be hysteroscopic myomectomy, with endometrial resection next. Additional findings included glycine as the most common distending medium, a similarmean patient age across all studies (both intervention and control arms), and the determination of fluid deficit via use of a fluid-management system (true for over half of the studies). Results of the study included a great many findings, including (1) lower fluid absorption in the intervention arm of studies using GnRH agonists, (2) a significant reduction in fluid absorption after 4- to 6-week treatments using danazol (a finding recorded in 2 separate studies published by the same author), (3) no significant difference in fluid absorption levels using ulipristal acetate in myomectomies, (4) 2 studies recording no significant difference in fluid absorption levels when comparing general anesthesia and neuraxial anesthesia, with a third study finding a significant difference between these groups (propofol exhibiting a significantly lower deficit vs inhaled sevoflurane), (5) no detectable difference of fluid absorption between women receiving oxytocin infusion at the time of endometrial resection as compared with an intervention group, (6) insufficient study data to accurately analyze the effect of vasopressin injection on fluid absorption, (7) a significant reduction in fluid absorption using letrozole (in 1 study), (8) an overall reduction in operative time for women treated with letrozole (15 studies), (9) a reduction of overall fluid absorption following the use of danazol and GnRH agonists for completion of a hysteroscopy (14 studies), (10) a 10% decreased risk of experiencing a fluid deficit over 1000 mL following the use of local or neuraxial anesthetic, (11) fewer successful 1-step procedures for the pretreated groups as compared with the control group (1 study), and (12) an overall low risk of bias across the entire spectrum of studies and categories. Overall, the most commonly investigated preoperative treatment was with a GnRH agonist (11 studies of data), and these confirmed an effective reduction of fluid absorption. Many strengths of the study were found, including its design, its broad and thorough published literature review, its inclusion of varying interventions, and its liberal inclusion of various hysteroscopic procedures. More than half of the included studies were RCTs (by design), and most of the studies included had low bias risk and were high quality. In addition, institutions that published 2 ormore included studies were contacted to avoid inadvertent duplication of results within this review. Finally, the selected primary outcome was both easily measured and objective, and for providers of hysteroscopy, it was also clinically relevant. Limitations of the study include its small number of publications found on pharmacological interventions beyond type of anesthesia and GnRH agonists. The applicability of these findings was limited to common practice due to the lack of data for other hysteroscopy medications prescribed before treatment. Furthermore, as a 30-year time span was present for our studies; a minor temporal bias may have been introduced due to technological advancements in the field of hysteroscopy during this time. Selection bias may have been introduced via observational studies, despite the majority of studies being composed of RCTs. During these studies, more severe cases resulted in more likely intervention (as compared with control patients with milder cases). In addition, the ability to meta-analyze data was limited through a heterogeneous reporting of their study outcomes. Moreover, a certain degree of heterogeneity within the current review was contributed to via both the nonuse or use of fluid-management systems and pretreatment durations for various pharmacological agents. Finally, questionable clinical importance is gained from the range of 5 to 10 minute operative time or a 100-200 mL decrease in fluid absorption. Ultimately, the study found that both danazol and preoperative treatment with GnRH agonists were effective methods (to a modest degree) for reducing operative time and fluid absorption for a wide array of hysteroscopic procedures. Literature is lacking for high-quality research based on other better-tolerated and less costly interventions (ie, vasopressin, progestin therapy, and hormonal contraception).
