The review concluded that robotic surgery offered limited advantages over classical surgical techniques with respect to short-term outcomes. Due to evidence limitations, the authors implied that their conclusions should be treated with caution.

To compare the safety and effectiveness of robotic surgery (DVSS), open surgery and conventional laparoscopic surgery in gynaecology.

MEDLINE, EMBASE and CINAHL were searched for primary studies from 2003. DARE, HTA, The Cochrane Library, Hayes Inc databases were searched for systematic reviews to October 2009. Search terms were reported. The bibliography of each retrieved article and a relevant systematic review was handsearched. Experts in surgery were contacted for relevant studies.

Studies with control groups comparing robotic surgery using the Da Vinci Surgical System (DVSS; Intuitive Surgical, Mountain View, Sunnyvale, California, USA) with open surgery or conventional laparoscopic surgery (CLS) in a gynaecological setting were eligible for inclusion. DVSS was the only robotic system cleared by the US Food and Drug Administration on the market. Studies were excluded from the meta-analysis if they did not provide quantitative data. Relevant outcomes varied with the surgery performed and included: complication rate; transfusions required; blood loss; conversion to another type of surgery or to open surgery; duration of surgery; number of lymph nodes, pelvic lymph nodes or aortic lymph nodes resected; percentage of positive margins; hospital stay length; time to return to work; numbers of pregnancies, miscarriages, ectopic pregnancies or intrauterine pregnancies; and mortality.

All studies except two were single-centre studies. Intervention recruitment period ranged from one to five years and control recruitment period ranged from less than one to 13 years. Surgery included: hysterectomy for the staging of endometrial cancer; radical hysterectomy for cervical cancer; hysterectomy for benign disease; myomectomy; sacrocolopexy; fallopian tube reanastomosis and adnexectomy. Similar numbers of studies compared DVSS to open surgery and DVSS to CLS; 23% of studies compared all three interventions. Where described, six studies used the same surgeon in the comparative groups and six studies used different surgeons. Most studies did not report whether patient management was identical in the comparative groups.

The authors did not report how many reviewers performed the study selection.

Methodological quality was assessed using the Guyatt quality checklist of clearly defined question, randomisation and blinding, type of follow-up, equality in patient management and comparable groups.

The authors did not report how many reviewers performed the quality assessment.

The number of events for each outcome was extracted to calculate odds ratios (OR) or risk differences (RD) with 95% confidence intervals (CI). Mean differences (MD) with 95% CIs were calculated for continuous variables. Data was extracted using a data collection form.

The authors did not report how many reviewers performed the data extraction.

Results were pooled when available from at least two studies using a fixed-effect model where heterogeneity was low and a random-effects model where heterogeneity was high. Between-study heterogeneity was determined using Χ² and I² where I²>50% indicated a high level of heterogeneity. For dichotomous variables, meta-analyses using odds ratios with 95% CIs were performed or where there were no events in one group or where events were reported in all patients in all groups, risk differences with 95% CIs were determined (Mantel-Haenszel).

For continuous variables, pooled mean differences with 95% CIs were calculated using inverse variance where subgroups had to be pooled. Where only medians were available they were used as estimates of means. For studies that reported ranges only and with 70 or fewer participants, standard deviations (SD) were determined using the method of Hozo et al. Meta-analyses were related to the type of surgery performed.

Where significant heterogeneity was detected, sensitivity analyses were performed to identify the study that contributed most to the heterogeneity. One study analysed obese participants separately and the authors analysed that group separately.

Twenty-two studies were identified (n=2,596 participants, range 15 to 322): 18 retrospective studies with historical controls (n=2,245, range 15 to 322) and four prospective studies (n=351, range 50 to 176). Sixteen studies provided quantitative data that could be used in the meta-analysis. All studies had a clear objective and were controlled. None of the studies were randomised or blinded. Most studies had adequate follow-up and used an intention-to-treat analysis. There were baseline differences between patient groups in approximately one half of the studies.

Hysterectomy for the staging of endometrial cancer (six studies): There were significantly fewer complications with DVSS versus open surgery (OR 0.22, 95% CI 0.13 to 0.39, I²=37%; four studies) and no significant difference for DVS versus CLS (three studies, I²=0%). There was significantly lower blood loss for DVSS versus open surgery (MD -152.1mL, 95% CI -184.2 to -120.0, I²=56%; four studies) and DVSS versus CLS (MD -76.0mL, 95% CI -142.4 to -9.5, I²=88%; four studies). There was a significantly lower need for blood transfusion for DVSS versus open surgery (OR 0.25, 95% CI 0.07 to 0.81, I²=0%; three studies) and for DVSS versus CLS (OR 0.24, 95% CI 0.09 to 0.64, I²=0%; three studies). There was a shorter hospital stay for DVSS versus open surgery (MD -2.68 days, 95% CI -3.53 to -1.84, I²=93%; four studies) and DVSS versus CLS (MD -0.17 days, 95% CI -0.28 to -0.06, I²=45%; four studies). Operation duration was significantly longer for DVSS versus open surgery (MD 89.3 minutes, 95% CI 51.7 to 126.8, I²=95%; four studies) with no significant difference for DVSS versus CLS (I²=85%, four studies). There was a greater risk of conversion to another type of surgery for DVSS versus open surgery (OR 11.5, 95% CI 1.4 to 95.0, I²=0%; two studies) but a reduced risk of conversion for DVSS versus CLS (OR 0.43, 95% CI 0.21 to 0.85, I²=0%; two studies). There was an increased number of resected lymph nodes for DVSS versus open surgery (MD 5.91, 95% CI 0.13 to 11.68, I²=90%; four studies) but no significant difference for DVSS versus CLS (I²=80%; four studies) and no significant differences for the number of pelvic lymph nodes or aortic lymph nodes resected for either comparison.

Radical hysterectomy for cervical cancer (seven studies): There was significantly lower blood loss for DVSS versus open surgery (MD -334.2mL, 95% CI -459.4 to -208.9, I²=90%; five studies) and DVSS versus CLS (MD -63.5mL, 95% CI -100.5 to -26.5, I²=0%; three studies). Similarly, there was a significantly lower need for blood transfusion for DVSS versus open surgery (OR 0.18, 95% CI 0.07 to 0.44, I²=0%; five studies) but not for DVSS versus CLS (I²=0%, two studies). There was a shorter hospital stay for DVSS versus open surgery (MD -2.05 days, 95% CI -2.80 to -1.29, I²=80%; five studies) but no significant difference for DVSS versus CLS (I²=51%; three studies) with no significant difference in operation duration for DVSS versus open surgery (I²=94%, five studies) or versus CLS (I²=68%, four studies). There were no significant differences for complications for DVSS versus open surgery (I²=0%, five studies) or CLS (I²=11%, four studies), need for conversion to another type of surgery (DVSS versus open surgery I²=0%, two studies and DVSS versus CLS, I²=0%, three studies) and numbers of lymph nodes resected (DVSS versus open surgery I²=86%, five studies and DVSS versus CLS, I²=94%, three studies).

Fallopian tube reanastomosis (two studies both comparing DVSS versus open surgery): Duration of operation was significantly longer for DVSS versus open surgery (MD 46.9 minutes, 95% CI 34.7 to 59.0, I²=0%). There was a significantly earlier return to work for DVSS versus open surgery patients (MD -16.0 days, 95% CI -19.6 to -12.4, I²=0%). There were no significant differences for hospital stay, complications, blood loss, pregnancy rate, miscarriage and ectopic and uterine pregnancies. Both studies reported increased costs for robotic surgery.

Myomectomy (three studies): There was significantly lower blood loss for DVSS versus CLS (MD -72.4mL, 95% CI -133.2 to -11.5, I²=44%; two studies) and no significant difference for duration of surgery (I²=76%, two studies). One study compared DVSS with open surgery with significantly lower duration of operation, lower blood loss, shorter hospital stay and increased costs for DVSS versus open surgery.

Individual studies: A large study of hysterectomy for benign disease found significantly longer duration of operation, shorter duration of hospital stay and lower blood loss for DVSS versus CLS. A study of sacrocolopexy found significantly longer duration of operation, shorter hospital stay and lower blood loss, but a higher risk of postoperative fever for DVSS versus open surgery. Adnexectomy: The only significant outcome was a longer duration of surgery for DVSS versus CLS.

Other individual results and results for a subgroup of obese patients were reported.

Two studies of fallopian tube reanastomosis reported increased costs for DVSS versus open surgery of USD $1,446 and an overall $2,000 plus $300 for each newborn. One study of myectomy reported significant increased costs of $18,000 for DVSS versus open surgery (p<0.001).

Robotic surgery offered limited advantages over classical surgical techniques with respect to short-term outcomes. The clinical outcomes should be interpreted with caution owing to the methodological quality of the studies.

The review addressed a well-defined question in terms of participants, interventions, study design and relevant outcomes. Relevant databases were searched, It was unclear whether any language restrictions were applied and some studies could have been missed. The authors were Spanish, which implied that both English and Spanish papers would be included. Unpublished studies were considered. Publication bias was not assessed. Study quality was assessed using suitable criteria. No relevant information was reported on whether efforts were made to avoid error and bias in the review process. Relevant study details were reported, but few details of length of follow-up and loss to follow-up were given and there were no details of participant age.

Statistical heterogeneity was assessed and there was evidence for heterogeneity for some outcomes. The statistical methods used for the meta-analyses seemed appropriate. Relevant sensitivity analyses were performed. The authors were concerned that many of the historical cohort studies recruited different patient groups at different times and the techniques evolved with time. Few of the included studies were prospective.

Potential limitations in the review process, the limited quality of included studies and the authors' own reservations mean the conclusions should be treated with caution.

Practice: The authors considered that some of the significant differences in blood loss and hospital stay identified were not large enough to be of clinical importance. Robotic surgery with DVSS was evolving and was unlikely to replace open surgery or CLS in gynaecology in the near future.

Research: The authors identified a need for longer-term studies to weigh the benefits and drawbacks of DVSS for different indications. They identified a need for more rigorous study design to control for bias. Costs should be investigated.

The report was supported by public funds and produced within a collaborative framework of the Quality Plan for the National Health System produced by the Ministry of Health and Social Policy under an agreement between Carlos III Institute of Health, Ministry of Science and Innovation and Lain Entralgo Agency. The report was based on an HTA report financed by the Spanish Ministry of Health.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.