Abstract
Background: Ovarian cancer has a high mortality rate, especially in Denmark where mortality rates have been reported higher than in adjacent countries with similar demographics. This study therefore examined recent survival and mortality among Danish ovarian cancer patients over an 18-year study period.
Methods: This nationwide registry-based observational study used data from the Danish Gynecology Cancer Database, Danish Pathology Registry, and Danish National Patient Registry. All patients with ovarian cancer diagnosed between 1995 and 2012 were included in the study. The data sources were linked via the patients’ personal identification number and the analyses included data on cancer stage, age, survival, surgery status and comorbidity. The computed outcome measures were age-adjusted mortality rates and age-adjusted overall and relative survival rates for one and five years.
Results: We identified 9972 patients diagnosed with ovarian cancer in the period 1995–2012. The absolute one-year mortality rate decreased from 42.8 (CI 40.3–45.6) in 1995–1999 to 28.3 (CI 25.9–30.9) in 2010–2012, and the five-year mortality rate decreased from 28.2 (CI 27.0–29.5) in 1995–1999 to 23.9 (CI 22.9–25.0) in 2005–2009. After stratification by age, comorbidity and cancer stage, the decrease in one-year mortality was most substantial in the 65–74 year old age group 41.1 (CI 38.8–43.5) to 26.5 (CI 24.4–28.7) and for stage III 39.1 (CI 35.1–43.6) to 22.9 (CI 19.9–26.5) and stage IV 91.3 (CI 80.8–103.2) to 41.9 (CI 35.5–49.5). For overall survival, we showed an increase in one-year survival from 68% (CI 66–69%) in 1995–1999 to 76% (CI 74–78%) in 2010–2012 and an increase in five-year survival from 33% (CI 32–35%) in 1995–1999 to 36% (CI 34–38%) in 2005–2009. Relative survival showed similar increases through the period.
Conclusions: Ovarian cancer survival in Denmark has improved substantially from 1995 to 2012, bringing Denmark closer to the standards set by adjacent countries.
Ovarian cancer is the eighth most common cancer in women in Denmark with an average of 550 new cases per year, and a lifetime risk for Danish women of approximately 2%. Ovarian cancer has the highest mortality rate among gynaecological cancers, partly due to it developing without symptoms until advanced stage. Up to 70% of Danish ovarian cancer patients have stage II–IV at time of diagnosis [Citation1].
Previous international studies on cancer survival have repeatedly shown that cancer survival in Denmark is lower than in other Scandinavian countries and nations of comparable size and demographics [Citation2–4]. This is also the case for ovarian cancer [Citation5]. In order to improve cancer care and survival in Denmark, several national initiatives were initiated from 2001, e.g. implementation of national cancer patient pathways, centralisation of treatment to specialised surgical and oncological centres, and the establishment of the Danish Multidisciplinary Cancer Groups (DMCG.dk), which is a national organisation responsible for developing clinical guidelines and clinical databases for monitoring and research [Citation6]. For ovarian cancer, the national cancer patient pathway has reduced delays in diagnosis and treatment, and the centralisation of cancer care has reduced the number of hospitals performing surgery from 58 sites to first seven and now four highly specialised sites [Citation7,Citation8]. Furthermore, the Danish Gynecology Cancer Database (DGCD) began routine data collection on all Danish ovarian cancer patients in 2005 [Citation9].
This study aimed to evaluate survival and mortality among Danish patients with ovarian cancer diagnosed in the 18-year period from 1995–2012, and to evaluate temporal changes in the context of the national health policy initiatives undertaken since 2001. We report age-adjusted mortality rates and overall and relative survival after one and five years. Data was not only based on the DGCD, a national clinical database with prospective data collection of key clinical variables, but also on data from the Danish Pathology Register (DPR) [Citation10] and the Danish National Patient Registry (DNPR) [Citation11].
Material and methods
Setting and source population
Denmark has a population of 5.6 million. All residents have equal access to tax-financed, national healthcare and only a small fraction of healthcare services and medications are paid out-of-pocket. Moreover, all heathcare contacts and procedures carried out in the secondary hospital sector are registered in diverse administrative databases.
Data sources
Data on incident ovarian cancer were obtained from DNPR in the period 1995–2004 and additionally from DGCD from 2005 to 2012. Incident ovarian cancer was defined as first-time registration of ICD-10 code DC56 with a histologically verified ovarian cancer in DPR. Data on vital status was obtained from the Danish Civil Registration System (CRS) for overall survival. Data were linked via the unique person-identification number, which is assigned to all residents upon birth or immigration.
DGCD has prospectively collected data on Danish ovarian cancer patients since 1 January 2005 to present day. The database collects data on ovarian cancer (including borderline-type), primary peritoneal cancer, Fallopian tube cancer, uterine cancer (including atypical hyperplasia), cervical cancer, vulvar cancer and trophoblastic disease. Data parameters are registered online by participating gynaecology, and pathology departments throughout Denmark. In addition, data on nurse-related care and activities are also routinely registered.
All incident ovarian cancer patients diagnosed between 1995 and 2004 were identified from DNPR and patients diagnosed between 2005 and 2012 were identified from DGCD. Incident ovarian cancer was defined as first-time registration of the International Classification of Diseases, 10th Revision (ICD-10) code DC56 and with concurrent histological verification of ovarian cancer in DPR, which is essential for identifying borderline-type ovarian cancer. Patients <16 years old of age were excluded. DGCD uses unique criteria for gynaecological cancer diagnoses. As this study was intended not just to mirror other cancer registry-based studies, these unique criteria were used throughout the study period. These criteria involve the gynaecologist registering the specific diagnoses in several different components of the DGCD and the gynaecological pathologist registering the same specific diagnosis in the pathology DGCD form. Finally, the treating gynaecologist has to finalise the registration by evaluating and approving the final diagnosis. Agreement on the correct diagnose is furthermore discussed at the mandatory multidisciplinary team meetings (MDT). In cases of discrepancy between the DNPR and DGCD, the department involved was contacted in order to re-validate and correct the registration in the respective register with the incorrect registration. The treatment of gynaecologic cancer in DK has become increasingly centralised through the last decade to currently only four centres. Almost all ovarian cancer and borderline tumours are examined by experienced pathologists in a gynaecologic cancer team. No special requirements exist in the guidelines for second review. In cases of second review, this information is registered in DGCD. Our aim with this specialised expertise confined to few pathologists is to minimise this known misclassification.
Covariates
Data on cancer stage, comorbidity, age and surgery status were included in the analyses. Cancer stage was classified according to the International Federation of Gynecology and Obstetrics (FIGO) classification [Citation12]. Information on stage and surgery status was queried in combination with supplemental data from DPR (including borderline-type) and from DNPR. Comorbidity was classified according to the Charlson Comorbidity Index (CCI), which is a validated and widely used index for the classification of comorbidity with significant impact on survival [Citation13]. History of comorbidities were identified using ICD-8 (8th Revision) and ICD-10 codes corresponding to the CCI score [Citation14] and using data-linkage to the DNPR for patient histories up to 10 years prior to the cancer diagnosis date. The categories for comorbidity were none (CCI = 0), mild (CCI = 1), moderate (CCI = 2) and severe (CCI = 3), respectively. Finally, if a patient had a record of both mild and moderate-severe liver disease, scoring was only given for the moderate-severe liver disease and likewise for diabetes and diabetes with end organ damage.
Outcome measures and statistical analyses
We computed three outcome measures relating to the first-year and the first five-year period after diagnosis: (1) overall all-cause mortality rate per 100 person-years; (2) overall survival proportion; and (3) relative survival. All estimates were age-standardised according to the International Cancer Survival Standard cancer population weights (cluster 1) in accordance with international practice [Citation15]. Overall survival proportions were estimated using the Kaplan-Meier method with right censoring. Relative survival was estimated as the ratio of the observed survival of the patients to the expected survival, where the expected survival was estimated from the general Danish population, matched by gender, age, and calendar time, and using the Ederer I method. Data on the reference (i.e. cancer-free) population was obtained from Statistics Denmark [Citation16]. All estimates are presented with 95% confidence intervals. Statistical analyses were performed using SAS (v9.3, SAS Institute, Cary, NC).
Results
Patient characteristics
A total of 9972 patients were included in this study, with 5584 patients diagnosed between 1995 and 2004 (before DGCD was started), and 4388 diagnosed between 2005 and 2012. Among the 4388 ovarian cancer patients, 4013 were registered in DGCD, while an additional 375 cases, that met DGCG’ diagnosis criteria, were identified from DNPR and DPR. Median age of the patients increased from 62 years in 1995–1999 to 66 years in 2010–2012. Prevalence of any comorbidity (CCI >1) also increased during the study period from 10% in 1995–1999 to 20% in 2010–2012 (). The distribution of FIGO cancer stages remained the same, but there was a substantial decrease in the number of unclassified/missing cases ().
Table 1. Baseline characteristics of the study population (N = 9972).
Overall all-cause mortality rate per 100 person-years
The overall one-year mortality rate for ovarian cancer decreased from 42.8 (CI 40.3–45.6) in 1995–1999 to 28.3 (CI 25.9–30.9) in 2010–2012 (). Similarly, the overall five-year mortality rate decreased from 28.2 (CI 27.0–29.5) in 1995–1999 to 23.9 (CI 22.9–25.0) in 2005–2009. Decreases in mortality were also observed when stratified by age group, with the most substantial reduction in one-year mortality in those over 64 year olds; where the mortality rate decreased by 35% from 41.1 (CI 38.8–43.5) to 26.5 (CI 24.4–28.7) in the 65–74 year old age group, and declined from 82.2 (CI 77.8–86.8) to 46.8 (CI 43.3–50.5) in the over 75-year-old age group. When stratified by comorbidity, the decrease in mortality rate can be seen throughout the study period, but again with the largest differences in one-year mortality observed in patients with CCI ≤1, from 39.1 (CI 36.5–42.0) for CCI = 0 and 61.6 (CI 51.2–74.4) for CCI = 1 in 1995–1999 to 21.1 (CI 18.8–23.8) and 34.3 (CI 26.8–44.4) in 2010–2012. Stratification by FIGO stage showed a different trend with a more or less unchanged mortality rate for stages I and II, but with a substantial decrease in mortality for stages III and IV, and especially in the one-year mortality, where the reduction went from 39.1 (CI 35.1–43.6) to 22.9 (CI 19.9–26.5) for stage III, and from 91.3 (CI 80.8–103.9) to 41.9 (CI 35.5–49.5) for stage IV.
Table 2. One- and five-year age-adjusted mortality rate per 100 patient years (1995–2012).
Overall survival proportion
There was an increase in both one- and five-year overall survival (, and ). The one-year overall survival increased from 68% (CI 66–69%) in 1995–1999 to 76% (CI 74–78%) in 2010–2012. The five-year overall survival increased from 33% (CI 32–35%) in 1995–1999 to 36% (CI 34–38) in 2005–2009, during which the reorganisation of ovarian cancer treatment to specialised care sites was underway, but not completed. There was a consistent increase in survival for all age groups. The highest increase in one-year survival was observed in the over 75-years-age group, where survival increased from 46% (CI 41–50%) to 64% (CI 58–69%). The survival proportion by cancer stage differed depending on the stage and follow-up period. The one-year survival for stage I was constant throughout the time period, but the five-year survival increased from 70% (CI 66–74%) to 76% (CI 73–80%). Stages II, III and IV increased in one- and five-year survival through the calendar periods, with the most dramatic rises in one-year survival for stage III and IV from 69% (CI 66–72%) and 45% (CI 39–50%) in 1995–1999 to 80% (CI 76–83%) and 68% (CI 63–73%) in 2010–2012, respectively.
Figure 1. Expected survival and observed survival with 95% confidence limits (CI). Kaplan–Meier estimates of overall survival.
Figure 2. Age-standardised one- and five-year relative survival.
Table 3. One- and five-year age-adjusted overall survival for ovarian cancer (1995–2012).
The total number of patients with missing data on cancer stage declined over the study period, but the one-year overall survival for this group of patients decreased from 55% (CI 50–59%) to 44% (CI 36–52%). It was not possible to compute the five-year survival for the period 2005–2009, as not all patients diagnosed in this calendar period had completed their five-year follow-up at the time the analyses were performed.
When stratifying analyses by comorbidity, we observed an increase in overall survival, especially in the one-year survival in patients with CCI = 1, where survival increased for CCI = 1 from 56% (CI 49–53%) in 1995–1999 to 72% (CI 63–79%) in 2010–2012. Moreover, there was a 48% decrease in the total number of patients without comorbidity (CCI = 0) over the same time period.
Relative survival
The one- and five-year relative survival increased over time (, and ). The one-year relative survival increased from 69% (CI 67–71%) in 1995–1999 to 78% (CI 76–80%) in 2010–2012, whereas the five-year relative survival increased from 34% (CI 32–36%) in 1995–1999 to 37% (CI 35–38%) in 2005–2009. For the age groups, there was a slight increase in one- and five-year relative survival throughout the calendar periods for all age groups. In particular, there was a notable increase in one-year relative survival for the >75 year age group, from 49% (CI 44–54%) to 68% (CI 62–73%) and in five-year relative survival for the <45 year age group, from 64% (CI 58–70%) to 74% (CI 68–80%). The stage groups I–IV all showed an increase in one- and five-year relative survival, with the most substantial increases seen in one-year relative survival for stages II–IV and in five-year relative survival for stage II. The comorbidity subgroups all increased in one-year relative survival, whereas the five-year relative survival only increased for CCI = 0 and CCI ≥3.
Table 4. One- and five-year age-adjusted relative survival for ovarian cancer (1995–2012).
Discussion
Key findings
Overall there has been an increase in survival of Danish patients with ovarian cancer from 1995 to 2012. The increase was most substantial in one-year survival, especially in the subgroups of patients >75 years old, patients with stages III and IV, and patients without comorbidity.
Comparison and discussion
Our five-year relative survival rates were 34% in 1995–1999 and 37% in 2005–2009, which are similar to the results from NORDCAN, where the survival rates for Denmark in the same time periods were 30% and 38% [Citation2]. Our survival rates are still lower than Finland (37% and 45%), Sweden (41% and 45%) and Norway (37% and 42%) [Citation2]. In the UK, the five-year relative survival was 46% in 2010–2011 [Citation4], and for the USA, it was 42% in 1995 and 45% in 2005 [Citation17].
For the one-year relative survival, we found a rate of 69% in 1995–1999 and 78% in 2005–2009, which are also similar to the NORDCAN results for Denmark where they found 62% in 1994–1998 and 75% for 2009–2012. Again, these rates can be compared to the higher one-year relative survival rates in Finland (68% and 76%), Norway (69% and 77%) and Sweden (75% and 84%) for the same time periods, and to the UK (72% in 2010–2011) and the USA (73% in 1995 and 76% in 2009) [Citation2–4]. The reason for these differences in relative survival in countries with similar healthcare systems can be many. Later, we discuss the influence of appropriate staging and lymph node removal. However, diagnostic delay due to different attitudes and diagnostics practices among primary sector physicians has also been found to lead to lower survival. In Denmark, this is believed to be a major contributing factor to the poorer prognosis, at least up until 2010 [Citation18].
With regard to stage I cancer, Denmark has been reported to have a notably lower one-year relative survival compared to other countries [Citation3]. In this study, we also found a slightly lower one-year survival of 93% for 1995–1999 and 94% for 2005–2012, compared to Canada 97%, Norway 98% and the UK 97% [Citation19]. Our slightly lower one-year survival may partly be explained by the fact that Denmark, in contrary to other countries, did not recommend lymph node resection on stage I patients until 2005. Lymph node removal was nationally implemented in 2012, which is why only 25% of patients in 2005 and 55% of patients in 2011 had this procedure performed [Citation20]. Between 6% and 20% of patients are otherwise expected to have lymph node metastases at time of diagnosis, and in accordance with the FIGO classification, these patients should be reclassified as stage III [Citation12]. Therefore, patients without lymph node resection risk being staged wrongly and this can bias estimates towards lower survival. Consequently, the one-year survival rate in Denmark is expected to further increase in the low stages as lymph node resection becomes complete throughout the cohorts.
For stage III, the one-year survival increased throughout the time periods to 81% in 2010–2012, which is similar to the other countries in 2004–2007 (Canada 82%, Norway 78%, UK 70%) [Citation19]. Likewise for stage IV, the one-year survival increased to 70% in 2010–2012, which is notably better than other countries (Canada 57%, Norway 55% and UK 53%) [Citation19] may be due to a more precise staging in our cohort.
The basic surgical treatment with hysterectomy, bilateral salpingo oophorectomy and omentectomy and oncologic treatment was unchanged during the total period. However, the centralisation resulting in more radical extensive surgery with as example more peritoneal stripping and total omentectomy may also contribute to improved staging and hence change in total and stage-related survival. Staging has considerable impact on overall survival as survival estimates will generally be higher in patients with early localised stages. Staging has also been a measure of diagnostic delay. In Denmark, the higher proportion of women with stages III and IV observed in previous years has been interpreted as being due to delayed diagnosis compared to Australia, Canada and the UK (74% in Denmark compared to 60–70% in the other countries) [Citation19]. In this study, stage III patients constitute 46% of all staged patients in 1995–2004 and 47% in 2005–2012, and stage IV patients constitute 18% and 20%, respectively. This is in contrast to the other countries where stage III constitutes 38% and stage IV 23% of all patients, leading us to conclude that we may have improved our diagnostic delay. Finally, the increased proportion of stage IV cancers may be partly due to diagnostic advancements, e.g. PET/CT scanning, which could counterbalance with increased upstaging [Citation21].
The largest increase in survival was found in patients over 75 years old, which is consistent with changes in attitude towards treatment of the elderly during the study period. In the early time periods, there was a nihilistic attitude towards surgery and chemotherapy for elderly patients over 70 years old. Instead, these patients were often treated with less radical procedures and not offered adjuvant chemotherapy [Citation22].
Strengths and limitations
A defining feature of the Danish cancer clinical databases is high data completeness (i.e. minimum requirement of >90% coverage of the specific cancer population), clinically based prospective data capture, and an organisational infrastructure (i.e. DMCG.dk) led by clinical experts working closely with their respective patient population. Data validation and periodic evaluation of the national gynaecology clinical database, e.g. medical audit and annual indicator reports, are performed. Linkage of healthcare data at the individual patient level allows for population tracking and virtually complete follow-up of all residents [Citation23].
Unfortunately, we were unable to compute the five-year survival for the recent years. Also, data collection to DGCD began in 2005, making it still a relatively young clinical database. Therefore, in order to evaluate survival from 1995, we had to link data to DNPR and DPR with data from DGCD. However, the same inclusions criteria were used for the entire period. A comparison analysis of DGCG data with DNPR data from 2005 to 2012 showed that there were 7% more patients in the DNPR cohort for incident ovarian cancer, than when using DGCD alone. The additional 7% may be due to patients with other cancer diagnosis, i.e. erroneously registered with ovarian cancer in DNPR and not excludable in DPR. Others may be true missing patients that ought to be in DGCD. The coverage in DGCD for ovarian cancer patients varies between 91% and 97% through the years due to erroneous and missing registration from the non-gynaecological departments.
Misclassification of borderline tumours could also lead to biased results. Borderline tumours have far better survival, and may be incorrectly registered as malignant ovarian tumours. We accounted for this potential bias by defining strict inclusions/exclusion criteria for our cohort. Misclassification of borderline tumours is nevertheless still likely. Unfortunately, pathology record review and data validation of histology was beyond the scope and feasibility of this study.
Prevalence and incidence of comorbidity have increased during the last two decades and comorbidity might be a source of bias when comparing populations [Citation24]. First, registration practices might differ across countries and perhaps also over time. Second, different distributions of comorbidity in differing populations may account for some of the differences in survival. Women in Denmark have a high consumption of tobacco, which may lead to a higher level of comorbidity and lower overall survival. However, these lifestyle factors are only likely to account for survival differences for early stage ovarian cancer, as cancer-specific mortality is high for advanced stage cancer.
There are notable limitations when using CCI on data from DNPR, e.g. some of the diseases could be historical or well treated, and diseases treated in the primary sector go unreported in the registries. In this study, we wished to compare changes in comorbidity through the different calendar periods. Therefore, only data on comorbidities from DNPR was used for the Charlson Index.
Perspectives
Overall, we found increasing survival rates over time, likely due to the national initiatives in Denmark, centralisation of gynaecological oncological treatment [Citation8], introduction of MDT with more precise selection to correct primary more radical surgery or neoadjuvant chemotherapy and improved staging criteria. Yet, despite these recent improvements in survival in Denmark, there is need for further improvement and assessment of the disparities in diagnostic delay, registration practices, and other prognostic factors, such as staging, tumour biology, comorbidity and development of newer personalised treatments.
Conclusion
This study showed that survival for patients with ovarian cancer diagnosed and treated in Denmark has been increasing over time since 1995. Overall survival in the most recent years was similar to the countries we normally compare ourselves with, but remains lower than the best countries. The national initiatives that have been implemented within the last 10 years, e.g. cancer patient pathways, centralisation of treatment, national guidelines, DGCD, and medical advancements, will hopefully continue to increase the survival for these patients.
Funding information
This study was funded by the Danish Clinical Quality Improvement Program (Regionernes Kliniske Kvalitetsudviklingsprogram, RKKP).
Disclosure statement
The authors declare no conflicts of interests.
References
- Annual Report from The Danish Gynecologic Cancer Database – A nationwide clinical database for ovarian, endometrial and cervical cancer, year 2011. 2012; 1–97. [Online]. Available at: http://www.dgcg.dk/images/DGCD%20rsrapport%202011.pdf.
- Engholm G , Ferlay J , Christensen N , Kejs AMT , Johannesen TB , Khan S , et al. NORDCAN: Cancer Incidence, Mortality, Prevalence and Survival in the Nordic Countries, Version 7.0 (December 17, 2014). Association of the Nordic Cancer Registries. Danish Cancer Society. Available from: http://www.ancr.nu.
- Allemani C , Weir HK , Carreira H , Harewood R , Spika D , Wang X-S , et al . Global surveillance of cancer survival 1995–2009: analysis of individual data for 25 676 887 patients from 279 population-based registries in 67 countries (CONCORD-2). The Lancet 2015; 385: 977– 1010.
- Cancer Research UK . Ovarian Cancer Statistics [Internet]. Available from: http://www.cancerresearchuk.org/health-professional/cancer-statistics/statistics-by-cancer-type/ovarian-cancer. Accessed on 2015.
- Coleman MP , Forman D , Bryant H , Butler J , Rachet B , Maringe C , et al . Cancer survival in Australia, Canada, Denmark, Norway, Sweden, and the UK, 1995–2007 (the international cancer benchmarking partnership): an analysis of population-based cancer registry data. The Lancet 2011; 377: 127– 38.
- Probst HB , Hussain ZB , Andersen O. Cancer patient pathways in Denmark as a joint effort between bureaucrats, health professionals and politicians–a national Danish project. Health Policy Amst Neth 2012; 105: 65– 70.
- Bertelsen K. Tumor reduction surgery and long-term survival in advanced ovarian cancer: a DACOVA study. Gynecol Oncol 1990; 38: 203– 9.
- Fagö-Olsen CL , Høgdall C , Kehlet H , Christensen IJ , Ottesen B. Centralized treatment of advanced stages of ovarian cancer improves survival: a nationwide Danish survey. Acta Obstet Gynecol Scand 2011; 90: 273– 9.
- Petri AL , Høgdall C , Lidegaard O. Registration of primary ovarian cancer in Denmark. Ugeskr Laeger 2009; 171: 408– 11.
- Erichsen R , Lash TL , Hamilton-Dutoit SJ , Bjerregaard B , Vyberg M , Pedersen L. Existing data sources for clinical epidemiology: the Danish national pathology registry and data bank. Clin Epidemiol 2010; 2: 51– 6.
- Lynge E , Sandegaard JL , Rebolj M. The Danish National Patient Register. Scand J Public Health 2011; 39: 30– 3.
- Mutch DG , Prat J. 2014 FIGO staging for ovarian, fallopian tube and peritoneal cancer. Gynecol Oncol 2014; 133: 401– 4.
- Charlson ME , Pompei P , Ales KL , MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40: 373– 83.
- Thygesen SK , Christiansen CF , Christensen S , Lash TL , Sørensen HT. The predictive value of ICD-10 diagnostic coding used to assess Charlson comorbidity index conditions in the population-based Danish National Registry of Patients. BMC Med Res Methodol 2011; 11: 83.
- Corazziari I , Quinn M , Capocaccia R. Standard cancer patient population for age standardising survival ratios. Eur J Cancer Oxf Engl 1990 2004; 40: 2307– 16.
- Statistics Denmark homepage [Danmarks Statistik]. Available [in English] at: http://www.dst.dk/en.aspx. Accessed on 2015.
- SEER Reletaive Survival . Cancer of the Ovary (invasive). http://seer.cancer.gov/csr/1975_2012/results_merged/sect_21_ovary.pdf. Accessed on 2015.
- Rose PW , Rubin G , Perera-Salazar R , Almberg SS , Barisic A , Dawes M , et al . Explaining variation in cancer survival between 11 jurisdictions in the International cancer benchmarking partnership: a primary care vignette survey. BMJ Open 2015; 5: e007212.
- Maringe C , Walters S , Butler J , Coleman MP , Hacker N , Hanna L , et al . Stage at diagnosis and ovarian cancer survival: evidence from the International cancer benchmarking partnership. Gynecol Oncol 2012; 127: 75– 82.
- Svolgaard O , Lidegaard O , Nielsen MLS , Nedergaard L , Mosgaard BJ , Lidang M , et al . Lymphadenectomy in surgical stage I epithelial ovarian cancer. Acta Obstet Gynecol Scand 2014; 93: 256– 60.
- Risum S , Høgdall C , Loft A , Berthelsen AK , Høgdall E , Nedergaard L , et al . Does the use of diagnostic PET/CT cause stage migration in patients with primary advanced ovarian cancer? Gynecol Oncol 2010; 116: 395– 8.
- Maas H. a a M , Kruitwagen RFPM , Lemmens VEPP , Goey SH , Janssen-Heijnen MLG. The influence of age and co-morbidity on treatment and prognosis of ovarian cancer: a population-based study. Gynecol Oncol 2005; 97: 104– 9.
- Pedersen CB. The Danish Civil Registration System. Scand J Public Health 2011; 39: 22– 5.
- Weiss CO , Varadhan R , Puhan MA , Vickers A , Bandeen-Roche K , Boyd CM , et al . Multimorbidity and evidence generation. J Gen Intern Med 2014; 29: 653– 60.