Table 1 Characteristics of included studies (n = 32)

Aletti et al.

Retrospective medical record audit

All women diagnosed with FIGO stage IIIC primary epithelial ovarian or tubal carcinoma from January 2000 to December 2003 and January 2006 to December 2007

To evaluate the effectiveness of a quality improvement program on surgical quality and the rates of specific procedures.

Data were analysed from a 3-year period prior to the quality improvement (QI) program implementation, and a 3-year period following the implementation. Complete debulking increased from 31% in the pre-QI program to 43% post QI program.

Aletti et al.

Review of NCCN guidelines

NCCN guidelines and previously published quality indicators

To describe existing surgical quality indicators for early and advanced ovarian cancer in relation to the recent NCCN guidelines.

No changes or additions recommend to the indicators. Highlighted the need for the implementation of quality control programs, with the focus on improving surgical outcomes for women with ovarian cancer.

Bristow et al.

Retrospective population based registry analysis

Adult women who underwent a surgical procedure for ovarian cancer in Maryland, USA, from 1 July 2000 to 30 June 2008

To investigate patterns of primary surgical care for ovarian cancer in Maryland according to surgeon and hospital volume.

Surgeon volume was defined as: low ≤4 cases/yr., intermediate 5–9 cases/yr., high ≥10 cases/yr. Hospital volume was defined as: low ≤9 cases/yr., intermediate 10–19 cases/yr., high ≥20 cases/yr. Annual surgeon volume was statistically significant for association with risk of in hospital death. Hospital volume was not found to be statistically significant for association with in hospital death.

Bristow et al.

Retrospective population based registry analysis

Adult women who underwent a surgical procedure for ovarian cancer in Maryland, USA, from 1 July 2000 to 30 June 2009

To evaluate the impact of surgeon and hospital case volume on in-hospital mortality, extent of surgery, length of stay and hospital cost of care.

Surgeon volume was defined as: low < 10 cases/yr., high ≥10 cases/yr. Hospital volume was defined as: low < 20 cases/yr., high ≥20 cases/yr. Surgery performed by a high volume surgeon was associated with a 69% reduction in risk of in-hospital death. Surgery at a high volume hospital was a significant and independent predictor of cytoreductive surgery.

Bristow et al.

Retrospective population based registry analysis

Women over 18 yrs. with FIGO stage IIIc/IV epithelial ovarian cancer from 1996 to 2005 in the USA

To examine and quantify the effect of hospital procedure volume on overall care and processes for women with epithelial ovarian cancer.

Hospital volume was defined as: low < 9 cases/yr., intermediate 9–20 cases/yr., high 21–35 cases/yr., very high > 35 cases/yr. 5-year survival for patients treated at hospitals completing < 21 cases/yr. was significantly poorer than for those treated at hospitals completing > 21 cases/yr.

Bristow et al.

Retrospective population based registry analysis

Cases of epithelial ovarian cancer reported to the California Cancer Registry from January 1999 to December 2006.

To validate NCCN ovarian cancer guideline adherence as a quality process in California, USA.

37.2% of patients received NCCN guideline adherent care. Receipt of non-NCCN guideline based care led to more than a 30% increase in ovarian cancer related death.

Chan et al.

Retrospective population based registry analysis

Women residing in Northern California diagnosed with stage IC/II primary invasive epithelial ovarian cancer from 1 January 1994–31 December 1996.

To investigate factors associated with women under 55 years of age with early stage ovarian cancer not receiving chemotherapy based on standard treatment guidelines.

78.5% of patients with early stage disease received adjuvant chemotherapy. Non-receipt of adjuvant chemotherapy was found to be more likely for those living in poor neighbourhoods, have low grade cancers and be less likely to have seen a gynaecological oncologist.

Cliby et al.

Retrospective population based registry analysis

Women with invasive epithelial ovarian cancer diagnosed between 1 January 1998 and 31 December 2008 in the USA

To evaluate the patterns of ovarian cancer care in the USA to define the influence of patient and institutional factors on overall survival, including the independent relationship between volume and outcomes.

Annual hospital volume was ranked in quartiles: 1–7, 8–16, 17–28 and > 29 cases/yr. 43% of patients received NCCN guideline adherent care. Non-receipt of NCCN guideline care was associated with worse overall survival (HR 1.4, 95% CI 1.36–1.45) with overall survival best in centres with > 29 cases/yr. (HR 0.91, 95% CI 0.86–0.96).

Dodge et al.

Survey

Gynaecologists performing gynaecologic surgery in Ontario

To quantify the gap between current practice in surgical staging of ovarian cancer with Canadian clinical practice guidelines.

44.3% of participants indicated they would complete surgical staging in line with the Canadian clinical practice guidelines. 81% of gynaecological oncologist’s vs 41.5% of non-oncologists indicated they would complete optimal staging.

Erickson et al.

Retrospective medical record audit

Women diagnosed with stage IC/II epithelial ovarian cancer between 2004 and 2009 treated in University of Alabama health system

To examine the reasons preventing patients from receiving NCCN guideline adherent care.

78.5% of patients received NCCN guideline adherent care. All patients were seen by a gynaecological oncologist, removing the bias of physician specialty. Most common reason for not receiving NCCN guideline adherent care was comorbidity.

Galvan-Turner et al.

Retrospective population based registry analysis

Women with stage I-IV epithelial ovarian cancer diagnosed in California between 1 January 1996 and 31 December 2006

To evaluate the feasibility of developing an observed-to-expected (O/E) ratio of adherence to NCCN guidelines as a risk-adjusted hospital measure of quality ovarian cancer care, correlated with disease specific survival.

Care at high O/E hospitals was associated with significant improvement in ovarian cancer specific survival when compared to intermediate O/E and low O/E hospitals. It was found to be feasible to develop a risk adjusted hospital measure of quality.

Harter et al.

Retrospective local quality assurance database analysis

Women with stage IIB-IV ovarian cancer who received surgery between 1997 and 2008 at HSK Hospital, Germany.

To improve the quality of surgery, specifically increase optimal cytoreduction, a quality management program was introduced in 2001. This study reports the outcome of this quality management program in women with advanced ovarian cancer.

Complete resection rates increased from 33 to 62%, with postoperative residuals ≤1 cm increased from 65 to 86%, with residual disease > 1 cm decreasing from 35 to 14%. There was a significant improvement in overall survival from 26 months (1997–2000), to 37 months (2001–2003), to 45 months (2004–2008).

Hodeib et al.

Retrospective population based registry analysis

Women with stage I/II invasive epithelial ovarian cancer diagnosed in California between 1 January 1996 and 31 December 2006

To investigate the impact of socioeconomic status and other demographic variables on adherence to NCCN guidelines in patients with stage I/II disease.

24% of patients received NCCN guideline adherent care. 16% of patients in the lower socioeconomic group received NCCN guideline adherent care.

Ivanova et al.

Retrospective population based registry analysis

Women in Bulgaria with ovarian cancer diagnosed from 2009 to 2011

To investigate if low hospital volume contributes to the number of cases with unspecified morphologic characteristics, which has been assumed as a possible indicator of suboptimal care.

Hospital volume was defined as: low < 30 cases/yr., high ≥30 cases/yr. 53% of patients were treated in low volume hospitals. Low volume vs high volume hospitals had higher number of cases with unspecified grade (27.7% vs 14.3%) and unspecified stage (37.9% vs 27.4%).

Kommoss et al.

Retrospective and prospective local institution clinical tumour registry analysis

Women with stage IA-IIIA ovarian cancer who received surgery between 1997 and 2007 at HSK Hospital, Germany.

To improve the quality of surgery, specifically increase optimal cytoreduction, a quality management program was introduced in 2001. This study reports the outcome of this quality management program in women with early ovarian cancer.

Women receiving standard surgery increased from 27 to 88.5% following the implementation of the quality management program.

Lee et al.

Retrospective medical record audit

Patients with stage I epithelial ovarian cancer treated surgically from January 1991 to December 2010 at Seoul National University Hospital

To evaluate the effects of NCCN guideline adherence on survival outcomes in early stage epithelial ovarian cancer.

Guideline adherent surgical staging was completed in 26.7% of patients, of these 100% received guideline adherent chemotherapy. Difference in disease-specific survival between the two groups was not statistically significant. Recurrence-free survival showed a statistically significant improvement in the guideline-adherent group.

Liang et al.

Retrospective medical record audit

Consecutive patients who underwent primary surgical staging/cytoreduction for ovarian cancer by 6 gynaecologic oncology providers at the Ohio State University from January 2010 and December 2012.

To evaluate compliance at a single National Cancer Institute-designated Comprehensive Cancer Centre with the 8 ovarian cancer quality indicators proposed by the Society of Gynaecology.

60% of patients were completely staged, with lymphadenectomy being the most frequently omitted staging procedure. Bilateral para-aortic lymph node dissection was excluded in 23.6% and bilateral pelvic lymph node dissection excluded in 18.2% of stage I-IIIB cases. 51.8% of optimally debulked stage III patients received intra-peritoneal chemotherapy within 42 days of cytoreduction. Of note, there were documented reasons for decision not to proceed with intra-peritoneal chemotherapy in all but 2 cases. Compliance was reasonable for the other quality indicators.

Mandato et al

Retrospective medical record audit

All patients with a diagnosis of epithelial ovarian cancer in Emilia-Romagna Hospitals, Italy, from 2007 to 2010

To investigate the management and treatment of epithelial ovarian cancer in Emilia-Romagna Hospitals.

Hospital volume was defined as: low ≤10 cases/yr., medium 11–20 cases/yr. and high ≥21 cases/yr. 46% of patients were treated in a high volume hospital. Complete cytoreduction was achieved in 20.1% of patients with stage III-IV disease. Patients treated in high volume hospitals presented a significantly lower risk of dying compared to patients treated in medium and low volume hospitals.

Marth et al.

Retrospective population based registry analysis

Women with ovarian cancer treated in Austria from 1999 to 2004 from participating Austrian gynaecological departments.

To evaluate factors predicting overall survival, with consideration of department volume.

Hospital department volume was defined as: small ≤23 patients/yr., large ≥24 patients/yr. Survival was longer at large vs small departments, 5-year survival of 69% vs 61% (p = 0.01). Department size was found to be an independent predictor of survival (HR 1.39 for treatment in small departments).

Phillips et al.

Retrospective medical record audit and literature review

Patients diagnosed with stage III/IV advanced ovarian cancer from 16 August 2007 to 3 February 2014 at the Pan-Birmingham Gynaecological Cancer Centre, UK.

To evaluate the effect of the denominator on survival of the total advanced ovarian cancer cases identified through a systematic literature, as well as data from the Pan-Birmingham Gynaecological Cancer Centre, UK.

Reporting the denominator in studies in advanced ovarian cancer is important to correctly interpret the outcomes being reported. Of the 18 studies identified, 2 reported on their total patient cohort, with no studies reporting overall survival for their total patient cohort. Data from the medical record audit demonstrated decreasing overall survival as treatment became less aggressive. Median overall survival for the total patient cohort was 30.2 months.

Phippen et al.

Retrospective medical record audit

Patients diagnosed with epithelial ovarian cancer, fallopian tube cancer, or primary peritoneal cancer between 2002 and 2010 at Brooke Army Medical Centre.

To evaluate the optimal cytoreduction rate, NCCN guideline adherence rate and surgical outcomes at a low volume institution.

85.4% of patients received NCCN guideline adherent surgery. Compliance rate in line with reports from high volume centres from this low volume centre.

Querleu et al.

Literature review and expert consensus

Development of quality indicators based on standards of practice and expert consensus.

Development of quality indicators for France by the French Society of Gynaecologic Oncology.

Three structural, eight process and two outcome quality indicators were developed. The paper highlighted the need to implement the indicators through a quality assurance program and pose that their implementation would result in a positive impact in survival for women with ovarian cancer in France.

Querleu et al.

Literature review and expert consensus

5 initiatives publishing quality indicators for advanced ovarian cancer surgery were identified. A multidisciplinary International Development Group was established to develop the quality indicators.

To develop quality indicators for advanced ovarian cancer surgery, carried out by the European Society of Gynaecologic Oncology (ESGO).

10 quality indicators were developed through this evidence based and consensus process. The indicators are categorised as structural, process or outcome. ESGO aims to implement their quality assurance program for advanced ovarian cancer using these indicators, with achievement targets applied to each indicator.

Shakeel et al.

Retrospective population based registry analysis

Women with an ovarian cancer diagnosis in Canada (excluding Quebec) from 2004 to 2012.

To evaluate the quality of surgical care for women with ovarian cancer in Canada by assessing surgical volume and surgeon speciality on short-term postoperative outcomes.

Hospital volume was defined as: low 1–27 procedures/yr., intermediate 29–28 procedures/yr., high 99–201 procedures/yr. Surgeon volume was defined as: low 1–4 procedures/yr., intermediate 5–23, high 24–57 procedures/yr. In-hospital mortality rate in 2004 was 1.3%, declining to 0.74% in 2012. Increasing age and comorbidity were significant predictors of in-hospital mortality. Hospital surgical volume was a significant predictor of reduced risk of in-hospital mortality and failure-to-rescue, but longer length of stay. Surgeon volume significantly predicted an increased risk of major complications and prolonged length of stay. Ovarian cancer surgery in Canada is performed by a large number of surgeons with low surgical volume/yr.

Sijmons et al.

Retrospective medical record audit

Women diagnosed between 1991 and 1997 with stage IA, IB or IC ovarian cancer in the central region of the Netherlands.

To assess compliance with current surgical staging and adjuvant local treatment guidelines and overall survival for patients with early stage epithelial ovarian cancer.

32.8% of patients received optimal surgical staging, chemotherapy guidelines were followed in 100% of grade 1, and 74.6% of grade II and III patients. 5-year overall survival in the optimally staged cohort was 97.6, and 68.5% in the non-optimally staged cohort.

Sobrero et al.

Retrospective medical record audit

Residents of the Piedmont Region of Italy diagnosed with ovarian cancer in 2009

To investigate whether ovarian cancer in the Piedmont Region is being managed according to current local, evidence-based clinical guideline; to identify determinants of lack of adherence to guidelines; and to evaluate the association between adherence to clinical guidelines and survival.

35.2% of patients received guideline adherent surgery, with 87.8% of patients receiving guideline adherent chemotherapy. Receipt of guideline adherent chemotherapy was associated with a significant reduction in all-cause mortality (HR, 0.5; 95% CI,0.28–0.89)

Uppal et al.

Retrospective population based registry analysis

Women with a diagnosis of high-grade serous ovarian carcinoma undergoing primary cytoreductive surgery from 2004 to 2013.

To evaluate the role of 30-day readmission rate as an indicator of quality of care in ovarian cancer surgery.

Hospital case volume was defined as: ≤ 10 cases/yr., 11–20 cases/yr., 21–30 cases/yr., > 31 cases/yr. Higher volume hospitals had higher 30-day readmission rates, but had significantly lower 30 (OR 0.69, 95%CI 0.50–0.96) and 90-day mortality (OR 0.74, 95%CI 0.60–0.91). Higher volume hospitals had higher re-admission rates, but also had higher NCCN guideline adherence, and improve 5-year overall survival.

Verleye et al.

Literature review and expert consensus

The PUBMED database was searched, with journal papers as well as guidelines and standards of care reviewed. Details of the search results were not reported.

To develop a list of process quality indicators for staging laparotomy in stage I-IIIA ovarian cancer, and debulking laparotomy in stage IIIB-IV ovarian cancer for the European Organisation for Research and Treatment of Cancer (EORTC-GCG).

5 indicators were proposed for staging laparotomy (stage I-IIIA), with 6 indicators proposed for debulking surgery (stage IIIB-IV). These proposed indicators will need to be evaluated for feasibility, validity and reliability.

Warren et al.

Retrospective population based registry analysis

Women in the USA diagnosed with ovarian cancer in 2002 and 2011

To evaluate population based trends in ovarian cancer treatment and survival, with a focus on NCCN guideline adherent care, assessing patient and provider characteristics and estimating trends in 2-year cause-specific survival.

In stage II ovarian cancer the percent of women who received guideline adherent surgery increased from 18.3% in 2002 to 31.7% in 2011, in stage II from 48.9% in 2002 to 56.7% in 2011, with stage IV remaining stable at 34% for the two time periods. Receipt of guideline adherent chemotherapy rose significantly over the time period, with at least 70% of women receiving multiagent chemotherapy in 2011.

Wright et al.

Retrospective population based registry analysis

Women in the USA aged 65 years or older with ovarian or uterine cancer who underwent surgery from 2000 to 2007.

To estimate trends in hospital volume and referral patterns for women with uterine and ovarian cancer.

The median hospital volume remained constant throughout the study period at 2 cases per 2 years. During the study period there was market concentration, where a similar number of patients undergo the procedure over time, with the number of hospitals decreasing, with increasing number of procedures being completed in higher volume hospitals. Overall, women with gynaecologic cancers continue to be treated at very low volume centres.

Wright et al.

Retrospective population based registry analysis

Women in the USA aged 18–90 with ovarian cancer and underwent oophorectomy from 1998 to 2009.

To examine the influence of failure to rescue as a source of variation in mortality for patients with ovarian cancer.

Hospital volume was defined as: low 1–36 procedures/yr., intermediate 36.1–53 procedures/yr., high > 53 procedures per year. The overall complication rate was 22.8%, with the failure to rescue rate being 6.2%, decreasing from 8.7% in 1998 to 6% in 2009. The adjusted failure to rescue rate was 48% higher at low compared to high volume hospitals.

Wright et al

Retrospective population based registry analysis

Women in the USA diagnosed with invasive epithelial ovarian cancer from 2004 to 2013

To examine if is lower volume hospitals that comply with NCCN quality metrics can achieve outcomes similar to high volume hospitals.

Five quality metrics were defined based on NCCN guidelines, with adherence to the metrics identified, reported on and aligned to hospital volume. Compliance with the quality metrics increased with hospital volume. For each volume category 2-year adjusted survival increased with adherence to the quality metrics, from 61.4% in the low volume with low quality metric compliance hospitals to 78.6% at the highest volume and high quality metric compliance hospitals.