Time to Diagnosis and Treatment for Ovarian Cancer and Associations with Outcomes: A Systematic Review

Abstract

Background:

Ovarian cancer is commonly diagnosed symptomatically at an advanced stage. Better survival for early disease suggests improving diagnostic pathways may increase survival. This study examines literature assessing diagnostic intervals and their association with clinical and psychological outcomes.

Methods:

Medline, EMBASE, and EmCare databases were searched for studies including quantitative measures of at least one interval, published between January 1, 2000 and August 9, 2022. Interval measures and associations (interval, outcomes, analytic strategy) were synthesized. Risk of bias of association studies was assessed using the Aarhus Checklist and ROBINS-E tool.

Results:

In total, 65 papers (20 association studies) were included and 26 unique intervals were identified. Interval estimates varied widely and were impacted by summary statistic used (mean or median) and group focused on. Of Aarhus-defined intervals, patient (symptom to presentation, n = 23; range [median]: 7–168 days) and diagnostic (presentation to diagnosis, n = 22; range [median]: 7–270 days) were most common. Nineteen association studies examined survival or stage outcomes with most, including five low risk-of-bias studies, finding no association.

Conclusions:

Studies reporting intervals for ovarian cancer diagnosis are limited by inconsistent definitions and reporting. Greater utilization of the Aarhus statement to define intervals and appropriate analytic methods is needed to strengthen findings from future studies.

Background

Ovarian cancer is the eighth most commonly diagnosed female cancer in the world, with over 300,000 women diagnosed in 2020.¹ Approximately, 70% of cases are diagnosed with advanced-stage disease which negatively impacts survival, from 5-year net survival of 80% for localized tumors to 30% for advanced disease.² Although trials of screening for asymptomatic women have not shown a mortality benefit,³ the differential survival for those diagnosed with early or advanced disease continues to suggest that improving diagnostic pathways may be one strategy to improve outcomes in the absence of screening.

Most women diagnosed with ovarian cancer present with nonspecific symptoms such as abdominal bloating, pain, fatigue, and altered bowel habits and urinary frequency.^4

–7 The nonspecific nature of these symptoms can lead to delays in seeking help, primary care referral, and diagnosis of ovarian cancer.⁸ Women with ovarian cancer have identified diagnostic delay as a significant issue,^9,10 which can negatively impact clinical and psychosocial outcomes.^11
–13 Faster time to diagnosis in symptomatic women is associated with reduced psychological distress and anxiety and increased satisfaction with care.¹³ However, a recent case-control study from Australia found an increased number of general practitioner appointments before diagnosis for women with ovarian cancer compared with controls, and 25% attended a hospital emergency department (ED) in the month before diagnosis, suggesting there are opportunities for more timely care for women diagnosed with this cancer.¹⁴

Previous literature reviews have examined ovarian cancer symptoms,^7,15 experience of diagnostic delay and factors influencing time to diagnosis,^8,15 and the association between delay and outcomes in gynecological cancers in general.¹⁶ The most recent systematic review to report on the relationship between diagnostic intervals and ovarian cancer outcomes included literature published almost a decade ago, up to November 2013.¹¹ Findings from this review were consistent with an earlier review for gynecological cancers in general: most studies show no association between time to diagnosis and ovarian cancer outcomes.¹⁶ Others have noted¹⁷ that most studies included in previous reviews were conducted before the development of clear recommendations for assessing and reporting diagnostic intervals, resulting in variation in interval definitions including their start and end points. The Aarhus Statement, published in 2012, provides standardized definitions of diagnostic intervals and the time points that mark their beginning and end.¹⁸ A recent systematic review of research investigating diagnostic intervals for colorectal cancer identified multiple limitations in the analytic approach used to examine relationships between intervals and outcomes, restricting interpretation of results.¹⁹ Key limitations identified by that review¹⁹ include arbitrary categorization of intervals assessed, lack of adjustment for potential confounders in analyses, and failure to use appropriate analytic methods to account for the waiting time paradox²⁰—where patients with shorter intervals have higher mortality which may reflect patients with alarm symptoms and more advanced disease being investigated more rapidly. No study included in earlier reviews investigating diagnostic intervals for ovarian cancer has attempted to account for the waiting time paradox. In addition, no previous review has attempted to summarize findings relating to the length of intervals associated with the diagnosis and treatment of ovarian cancer.

To address these gaps, this study aims to systematically review intervals to ovarian cancer diagnosis and treatment and provide an updated assessment of the relationship between the length of these intervals and clinical and psychological outcomes for ovarian cancer. Thus, the study provides insight into how quickly women are diagnosed and treated, and contemporary evidence of the nature of the relationship between interval length and ovarian cancer outcomes.

Method

The protocol was registered in PROSPERO (CRD42020207655). The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist was used for reporting.^21,22

Search strategy

Medline, EMBASE, and EmCare databases were searched by combining index terms and text words relating to ovarian cancer and time to diagnosis and treatment (see Supplementary Table S1 for search strategies for each database). Database searches were conducted for papers published between January 1, 2000 and November 1, 2021 with the search subsequently updated till August 9, 2022. Reference lists of included papers were searched to identify other potentially eligible papers.

Eligibility criteria

Eligible studies were original research articles of any study design published in English. Studies had to report a quantitative measure of at least one interval in the pathway from first symptom to first treatment for adults (aged 18 years or older) with a diagnosis of primary ovarian, peritoneal, or fallopian tube cancer. Studies were also included if they reported assessment of association between an interval measure and clinical outcome (survival, stage of disease), quality of life, or psychological distress.

Studies were excluded if the interval measure was not assessed quantitatively or if the study targeted children or adolescents (aged <18 years). Studies reporting on pregnancy-related cancers, screen-detected cancers, or assessing a biomarker for ovarian cancer were also excluded.

Data screening, extraction, and analysis

R.B., V.W., D.O.S., and S.D.-S. screened abstracts and full-text papers to assess eligibility against the criteria. Each abstract and article was assessed by two reviewers independently, with disagreements regarding eligibility resolved through discussion or referral to a third reviewer. A data extraction form was developed to extract the following information from each eligible article: author, year of publication, study design, aim, country of study, study period, eligibility criteria, study sample (including ovarian cancer type/subtypes, age of participants), sample size, interval definition, data source for time interval, and measures of length of interval. For papers assessing associations, measures of association with outcomes, statistical model used, and any confounders adjusted for were extracted. Data for each article were extracted by two reviewers with disagreements resolved through discussion. Review screening and extraction was conducted using Covidence.²³

Risk of bias assessment

The Aarhus statement 20-item checklist was used to assess the quality of reporting of interval data.¹⁸ This captured features such as how timepoints and intervals were defined, biases, complexity, and theoretical considerations for date or interval data, instrument characteristics for studies utilizing questionnaires for data collection, and clarity of data descriptions for studies using medical audit or databases. Items from the preliminary risk of bias in nonrandomized studies of exposures (ROBINS-E) tool, July 2017 version, were used to assess the risk of bias for association studies.²⁴ Additional items were assessed regarding time interval-specific factors, such as consideration of the wait-time paradox and adjustment for outliers,¹¹ before determining the overall risk of bias. We also assessed overall risk of bias of interval measurement based on the Aarhus checklist responses.¹⁸

Data synthesis

A narrative synthesis was conducted as meta-analysis was not possible owing to heterogeneity of study designs and outcome measures. For studies examining associations, we summarized the direction of the association in tables, similar to previous reviews.¹¹

Results

In total, 4,862 unique citations were identified and screened. After review of 110 full text papers, 65 papers from 58 separate cohorts (datasets) were eligible for inclusion (Fig. 1).^{4
–6,12,25

–85}

FIG. 1.

PRISMA study flow chart of papers through the screening and selection process. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Study characteristics

A summary of the key features of included articles are shown in Table 1, with details of individual studies in Supplementary Table S2.

Table 1.

Summary of Study/Article Characteristics

			Type of study
	All papers (n = 65)		Descriptive (n = 45)		Association (n = 20)
	n	%	n	%	n	%
Publication year
2000–2004	12	18	7	16	5	25
2005–2009	8	12	5	11	3	15
2010–2014	12	17	8	18	3	15
2015–2019	14	22	7	16	7	35
2020–2022	20	31	18	40	2	10
Region
United Kingdom	17	26	12	27	5	25
North America	16	25	12	27	4	20
Europe	13	20	9	20	4	20
Australia	6	9	3	7	3	15
Asia	6	9	3	7	3	15
International	3	5	2	4	1	5
Africa	2	3	2	4	0	0
South America	2	3	2	4	0	0
Study design
Cohort	43	66	31	69	12	60
Cross-sectional	13	20	9	20	4	20
Case-control	9	14	5	11	4	20
Data source (intervals)^a
Medical records	33	51	22	49	11	55
Patient self-report (survey or interview)	25	38	16	36	9	45
Registry	14	22	8	18	6	30
Administrative data	10	15	7	16	3	15
Clinician survey	5	8	3	7	2	10
Other (e.g., local hospital, survey)	1	2	1	2	0	0
Number of participants
<50	5	8	5	11	0	0
51–100	7	11	4	9	3	15
101–300	21	32	15	33	6	30
301–500	10	15	7	16	3	15
501-1,000	9	14	5	11	4	20
1,001–10,000	9	14	6	13	3	15
10,001–40,000	2	3	1	<.05	1	5
<40,000	2	3	2	4	0	0

Multiple responses will not add to 100%.

Most papers were published after 2015, and just under half were from the United Kingdom (22% of all studies) or North America (24%). Cohort designs were common and most papers included at least 100 cases. Stage of disease of participants was not reported in 16 papers (25%) however, when reported, disease stage at diagnosis was most commonly advanced (III/IV) (see Supplementary Table S2). Five main data sources were utilized for generating intervals, with medical records the most common (44% of 58 studies), followed by patient self-report survey or interviews (32%), registries (21%), and administrative data (14%). Of the studies that collected data from patients, response rates varied from 38%⁴ to 93%.⁸¹

Intervals

Thirty-five papers (54% of 65) reported only one interval, 22 reported data for 2–3 intervals, and eight reported on four or more intervals (Table 2). Interval duration was most commonly reported in days or months, and summarized as the mean or median, with the median being predominant (Table 2).

Table 2.

Summary of Interval Characteristics^a

	All papers n = 65		Association studies (n = 20)
	n	%	n	%
Number of intervals assessed
1	35	54	6	30
2	7	11	2	10
3	15	23	5	25
4	2	3	6	30
5	4	6	1	5
7	1	2	0	0
9	1	2	0	0
Interval unit of time measure
Days	35	54	11	58
Weeks	8	12	3	16
Months	16	25	4	21
Days and months	2	3	0	0
Weeks and months	3	5	1	5
Not reported	1	2	1	5
Interval length reporting
Continuous	40	64	10	50
Categorical	12	19	3	15
Continuous and categorical	12	19	6	30
Not reported	1	2	1	5
Interval summary statistics/measures^b
Mean (with/without SD)	21	33	4	21
Median (with/without IQR/IQI)	40	64	14	74
Proportions	27	43	9	47

Percents may not add to 100 due to rounding.

Multiple response allowed; will add to more than 100%.

IQR/IQI, interquartile range/interval; SD, standard deviation.

Twenty-six unique types of intervals were reported. Interval details for individual studies are provided in Supplementary Table S3, and a summary of the length of the nine most commonly reported intervals is provided in Figure 2. The number of studies reporting each of the six main Aarhus intervals are shown in Table 3. The most common intervals reported were the patient (symptom to presentation, n = 23) and diagnostic (presentation to diagnosis, n = 22) intervals. However, even for these intervals, definitions of start and endpoints varied across papers making it difficult to determine consistency. Patient-reported data were less common in studies reporting secondary care and referral to hospital intervals. The median interval based on patient-reported data varied more widely than that based on medical record/administrative datasets.

FIG. 2.

Diagnostic intervals for ovarian cancer reported across studies (studies reporting by subgroups excluded). Bars represent the interval length: orange bars represent intervals summarized as the mean, and blue bar represents intervals summarized as the median. First author, year, and country for each study is presented on the Y-axis. A star (*) indicates studies that included data from multiple countries.

Table 3.

Number of Papers Reporting the Six Most Common Intervals, Number Using Patient Data, and Range of Medians (in days) Reported for Patient Data and Medical Record/Administrative Data

Interval	Number of papers reporting interval	Number (%) using patient-reported data^a	Median range patient reported	Median range medical records/administrative data
First symptom to presentation (patient interval)	23	14 (61%)^b	9–120	7–14 (168 Nigeria)
First presentation to referral (Primary care interval)	10	4 (40%)	1–19	1–7 (112 Nigeria)
Referral to specialist/hospital visit	7	2 (29%)	6^c	3–24
Referral to treatment (secondary care interval)	7	1 (14%)	36^c	21–49
First presentation to diagnosis (diagnostic interval)	22	14 (64%)	14–270^d	7–137
First symptom to treatment (total interval)	5	2 (40%)	57–125	52.5^c

Studies may have also included data from medical records/registries and/or data from doctor surveys.

Only using patient-reported data.

Only one study presented relevant data.

For the patient interval, median time ranged from 7 days for patients surviving more than 18 months in a UK study⁴⁷ to 168 days in a Nigerian study⁶³ (Fig. 2). Median diagnostic interval length ranged from 7 days for Canadian patients presenting to an ED⁵¹ to 270 days for patients with borderline, early-stage disease from the United States⁸¹; whereas the mean ranged from 5.5 days in a Hong Kong study³⁴ to 165 days for US participants in an international study.⁶⁴ Mean and median time from presentation to referral, the primary care interval, was within 30 days for most studies.

Association studies

Twenty studies assessed associations between an interval and an outcome, with the vast majority (n = 19) exploring associations with survival and/or stage; whereas only one examined quality of life outcomes.¹² (Table 4 and Supplementary Table S4). Of the nine studies (45% of association studies) reporting multivariable analyses, six adjusted for variables on the causal pathway for survival, with control variables including stage of disease, treatment, and diagnostic pathway (see Supplementary Table S4). Six studies attempted to account for the wait time paradox in their analyses most commonly by stratifying analyses by stage or ED presentation (Table 4).

Table 4.

Summary of Association Study Analyses

	n	%
Exposure—interval^a	(n = 20)
Patient (first symptom to presentation)	8	40
Primary care (first presentation to referral)	3	15
First presentation to specialist/hospital appointment	1	5
Referral to specialist/hospital appointment	2	10
Referral to diagnosis	1	5
Secondary care (referral to treatment)	1	5
Diagnostic	10	5
Total diagnostic	7	35
Total interval (symptom to first treatment)	1	5
Health system (presentation to treatment)	1	5
Specialist to treatment	2	10
Other^b	4	20
Outcomes
Survival/Mortality	7	35
Stage	10	50
Survival/Mortality and stage	2	10
Quality of life	1	5
Exposure handling for analysis
Continuous	10	50
Categorical	5	25
Continuous and categorical	5	25
Analysis
Model/Test ^a
Time to event (e.g., survival)		(n = 9)
Relative survival	1	11
Cox regression	5	56
Multivariable logistic	1	11
Mann–Whitney U	1	11
Quantile regression	1	11
Kaplan–Meier	1	11
Unclear (t-test or Mann–Whitney U)	1	11
Stage at diagnosis ^a		(n = 12)
Mann–Whitney U	1	8
Chi-square	2	17
Kruskal–Wallis	2	17
Logistic regression	1	8
Ordinal logistic regression	1	8
Poisson regression	1	8
Wilcoxon rank sum	1	8
t-test	4	33
Log-rank	1	8
Quality of life		(n = 1)
Poisson regression	1	100
Univariate/Multivariate analysis		(n = 20)
Univariate	11	55
Multivariate	6	30
Univariate and multivariate model	3	15
Multivariate models for survival adjust for variable on causal pathway and/or unknown at exposure
Yes	6	30
No	2	10
No multivariate analyses or did not examine survival	12	60
Wait-time paradox: stratify analysis (e.g., by stage, ED) / nonlinear association (e.g., cubic spline) for survival outcomes^a		(n = 20)
Yes-stratify	5	25
Yes-non-linear exposure	2	10
No or did not examine survival	14	70

May not add to n or 100% as papers could report more than one analysis method.

Other intervals: Symptom onset to telling others; Time taken for gynecological-oncologist team to make a diagnosis and formulate a treatment decision; Symptom duration to imaging for disease; Decision to treat to starting treatment Total diagnostic = 1 article no association examined for this interval, though it is included in descriptive summary.

The diagnostic interval was the most common exposure variable, followed by the patient interval and total diagnostic interval. Stage at diagnosis was the most common outcome, with most (10 of 12 papers) categorizing stage as early (I/II) versus advanced (III/IV) for analysis. The study looking at quality of life assessed association with total interval (symptom to first treatment). Multivariate regression models were used in nine of the 20 papers. Most models adjusted for variables on the causal pathway between exposure and outcome or postexposure variables. For example, when examining the association between time from first symptom to diagnosis, and survival, some models included treatment type.

For the five intervals examined by at least three papers, the direction of effect estimates largely indicated no relationship between an interval length and outcome (Table 5). Findings were similar for papers that analyzed data with or without consideration of the wait-time paradox.

Table 5.

Summary of the Direction of Association for the Five Most Common Intervals

	Longer interval better	No statistical association	Longer interval worse	Nonlinear association
Patient interval
Survival		Naglea ^, b
		Thulesiusc ^, d:
		Kirwanc ^, d
Stage	Webbc ^, b:	Naglea ^, b
	Tokudac ^, d	Lime ^, d
	Naglea ^, b (stratified: serous)	Chana ^, d
		Olsenc ^, d
Diagnostic interval
Survival	Lovea ^, d	Altmane ^, b (stratified: early stage)		Altmane ^, b (stratified: late stage)
		Naglea ^, b
Stage	Lovea ^, d (stratified: non-ED)	Goffa ^, d (2000)	Webbc ^, b
		Lovea ^, d (stratified: ED)
		Chana ^, d
		Lime ^, d
		Naglea ^, b
		Olsenc ^, d
		Torringe ^, f
		Baune ^, d
First symptom to diagnosis (total diagnostic interval)
Survival		Dilleya ^, d
		Naglea ^, b
Stage	Naglea ^, b (stratified: serous)	Naglea ^, b
		Olsenc ^, d
		Webbc ^, b
		Chana ^, d
		Goff (2000)a ^, d
		Menczerc ^, d
Primary care interval
Survival		Hawardenc ^, d
		Kirwanc ^, d
Stage		Torringe ^, f
Referral to specialist/hospital appointment (Referral interval)
Survival	Kirwanc ^, d (univariate)	Kirwanc ^, d (multivariate)
		DiGirolamoe ^, b
Stage	Lovea ^, d (stratified: non-ED)	Lovea ^, d (stratified: ED)

ED, emergency department presentation.

Aarhus criteria: moderate risk of bias.

ROBINS-E serious risk of bias.

Aarhus criteria: serious/critical risk of bias.

ROBINS-E critical risk of bias.

Aarhus criteria low risk of bias.

ROBINS-E moderate risk of bias.

Quality of reporting and risk of bias

Reporting quality based on the Aarhus statement criteria was variable, with most papers defining intervals reasonably well, but few referring to theoretical frameworks for measurement and analysis (Supplementary Table S5). Risk of bias of association studies was largely low for participant selection and measurement of outcomes, but critical for confounding. This was largely due to lack of consideration of confounders, such as conducting univariate analysis alone, or adjustment for postexposure variables, for example, adjusting for treatment type or stage when examining the relationship between survival and time from first symptom to presentation. None of the association studies in Table 5 was judged to be low risk of bias using the ROBINS-E risk of bias tool. Five association studies^{28,30,36,50,79} in Table 5 were judged to be low risk of bias using the Aarhus criteria, with eight assessed as serious/critical of bias (Table 5).

Discussion

Our systematic review identified 65 papers reporting diagnostic or treatment intervals for women diagnosed with ovarian cancer from 58 studies, indicating substantial work in this area, particularly since 2015. Most papers assessed only one interval, most commonly the patient or diagnostic interval rather than examining the entire diagnostic process. Estimates for these two intervals varied, with median patient intervals ranging between 7 and 168 days, and median diagnostic intervals ranging between 14 and 270 days. Different start and end points used to determine intervals, variations in data collection methods, and variation in study populations makes combining data across studies difficult. Twenty studies examined the relationship between diagnostic intervals and patient outcomes, most commonly survival or diagnostic stage. Most found no association with these outcomes, which is in line with previous findings.¹¹ However, as most of the association studies were assessed as having poor methodological quality, strong conclusions in this area are difficult to draw.

With the majority of women diagnosed symptomatically with ovarian cancer not surviving 5 years,^3,86
–88 there is interest in understanding factors associated with time to diagnosis, with the expectation that earlier diagnosis may improve outcomes. This focus likely reflects the number of studies in our review assessing the patient interval, the time from first symptom to presentation. Our review found wide variation in this period, with longest intervals found for the two studies from African countries. As the authors of these papers note, reasons for delay are multifactorial and may be due to a lack of knowledge about ovarian cancer and symptoms among women and health care practitioners, use of traditional healers, and inadequacies in the health system.^63,69 Excluding these two studies, variation in the length of the patient interval was more modest. Preference for reporting means rather than medians is a key factor leading to greater estimates of delay across all intervals. As the distribution of responses for all intervals tends to be right skewed, using means to report interval data will tend to result in longer summary intervals. Median scores may therefore be more appropriate as summary measures of diagnostic intervals.

Patient interval estimates derived from surveys of women were generally longer and showed greater variation than those determined from medical record data. Multiple factors may contribute to these differences, including imperfect recall and hindsight bias in survey data where, once diagnosed, respondents review and reconsider the timing of symptom onset. Although multiple studies reported using trained and experienced interviewers to collect patient data through interview, none discussed use of methods to reduce imperfect recall in their data. In addition, medical record data may be influenced by the reporting practices of doctors that tend to round intervals to simpler categories of duration (e.g., weeks or months) and focus on red flag symptoms⁸⁹ that might also work to shorten the patient interval recorded in medical records. The difference may also be a consequence of the waiting time paradox,^20,79 where a subset of people with more advanced disease are more likely to present with alarm symptoms and are diagnosed more quickly than those presenting with less advanced disease. As survival is poorer for those with advanced disease, they may be underrepresented in patient surveys and therefore interval estimates from these surveys tend to exclude these more rapidly diagnosed, poorer prognosis cases. The wide range of patient intervals considered, and differences between estimates based on patient self-report and those derived from medical records, makes it difficult to draw conclusions, other than that it is difficult to assess this interval when symptoms are hard to characterize and often lack an “alarm” factor.⁸

Unlike for other cancers,^17,90 only four studies reported the interval between diagnosis and treatment. This finding most likely reflects that for many women, ovarian cancer is diagnosed as a consequence of surgery, hence diagnosis and first treatment occur simultaneously. Similar to reviews of diagnostic intervals for other cancers,¹⁷ we found a limited number of studies assessing the interval between referral and treatment. In contrast, 22 studies assessed the diagnostic interval (first presentation to diagnosis), which compares favorably to a recent systematic review on diagnostic intervals for colorectal cancer that found this interval to be reported in 10 of 55 studies.¹⁷ Although most studies assessing secondary care intervals used medical records/administrative data, the diagnostic interval was most commonly determined using patient reported data. Estimates of this interval were substantially longer when determined by patient report compared with medical record data. Median diagnostic intervals ranged from 2 to 13 weeks according to patient reports, but 1–19.5 weeks when determined through medical record/administrative data. Whether this difference reflects imperfect recall, or a wider range of experiences captured by the larger number of studies using patient reported data, is not clear. Multiple studies have suggested the referral pathway can be complex for at least a significant minority of ovarian cancer patients,^{14,30,45,51,66} with multiple factors including misdiagnosis, nonurgent referral, and referrals to nongynecological specialties associated with diagnostic delay.⁸ These complex pathways may be better captured by looking at the system interval described in the Aarhus statement which captures the time between the first investigation of the presenting problem to first treatment.¹⁸

Currently, there is a lack of consistent evidence regarding the impact of diagnostic delays on cancer outcomes for ovarian cancer. A 2015 review found only four studies had looked at associations between diagnostic intervals and patient outcomes, with stage and survival outcomes the focus.¹¹ That study found no evidence of an association between diagnostic delay and ovarian cancer outcomes but noted that conclusions were limited owing to no study accounting for the wait-time paradox. Our review identified 20 studies examining associations between diagnostic intervals and outcomes, suggesting increased interest in this area. However, similar to conclusions reached by that earlier review, most studies exploring this question were judged to have a moderate or higher risk of bias making it difficult to draw firm conclusions. Of the five studies we assessed as having low risk of bias on the Aarhus criteria,^{28,30,36,50,79} all found no association between the interval and outcome they examined. Confidence in the conclusions from these studies is strengthened by noting that four explicitly accounted for the wait-time paradox by either stratifying analyses by stage^28,36,50 and/or using nonlinear analytic models.^28,79 The low number of high-quality studies, and the limited number of outcomes explored suggests further work is needed before firm conclusions, a lack of association or otherwise, can be drawn.

Six of the association studies adjusted for potential mediators on the causal pathway between length of diagnostic intervals and cancer survival, particularly stage and treatment. The length of diagnostic intervals might affect stage and the treatment received, and both are strong determinants of survival. Adjusting for these variables will bias associations between diagnostic intervals and survival to the null.^91,92 Similar issues have been identified in the analyses examining associations between diagnostic intervals and survival outcomes for colorectal cancer.¹⁹ Disease stage and treatment should be considered as mediators rather than confounders in analyses, since confounders must affect diagnostic intervals rather than the other way round.⁹²

Recent findings of no long-term mortality benefit from ovarian cancer screening³ suggests that diagnosis of this cancer will continue to be through symptomatic presentation. Although the lack of an overall survival benefit from screening may suggest limited benefits of earlier diagnosis, it may also reflect limitations in current screening protocols as, despite multimodal screening resulting in a stage shift from stage IV to stage I, the majority of cancers detected in the screening arms were still stage III or IV at diagnosis.³ The authors note that the 10% reduction in stage III or IV cancers found in this trial is likely to be insufficient to reduce mortality given work relating to breast cancer screening suggests diagnosis of stage III and IV disease needs to reduce by at least 20% for mortality reductions.³ Despite the stage shift found for multimodal screening, mortality rates increased in this arm for Stage I cancers, suggesting that cancers with a poor prognosis were being detected at the earlier stage. Current treatment may be unable to improve outcomes for these cancers despite being found at an earlier stage of diagnosis.³ Advances in treatment of ovarian cancer may be needed to improve the outcomes for ovarian cancer, even those diagnosed with early-stage disease. Indeed as most cases in the UK screening trial were diagnosed and treated before widespread uptake of recent advances in ovarian cancer treatment, impact of these advances on outcomes is not known.³

This review has a number of strengths including the large number of papers included, making it the largest review to assess diagnostic intervals for ovarian cancer. We examined intervals across the diagnostic period, enabling an assessment of the intervals most and least commonly assessed. We reviewed a broad range of studies, including those that assessed intervals through patient report, medical record review, and/or administrative data sets. This enabled a comparison of the different results found using different data collection methods. Finally, we explored associations between intervals and outcomes and used the Aarhus Statement, designed to improve reporting of diagnostic intervals in research, to assess methodological quality. However, some limitations also need to be acknowledged. We did not attempt to conduct a quantitative summary of interval data owing to the heterogeneity of definitions used to define the start and end of time periods across studies. We focused on studies reported in English so papers from non-English journals are missing from our review. This means our results are likely to reflect the diagnostic experiences of people from high-income countries, which, as the findings from the two studies from African countries suggest, is likely to be different from low- and middle-income countries. Finally, while our search terms were broad and enabled the identification of a large number of papers, some papers may have been missed.

Conclusion

The lack of an effective screening program and improvements in survival for ovarian cancer have led to a focus on better understanding and identifying potential strategies to improve diagnostic pathways for this cancer. Our review shows most work has focused on understanding the patient interval and the diagnostic interval, reflecting interest in identifying strategies to improve both of these intervals. However, the wide variety of definitions used to identify different intervals, the mix of means and medians to summarize intervals and high prevalence of moderate-to-high risk of bias across studies and limitations in analytic methods mean it is difficult to develop meaningful conclusions. To enable a better overall understanding of the diagnostic intervals people with ovarian cancer experience, greater utilization of the Aarhus statement to define intervals and greater use of appropriate analytic methods, including use of nonlinear analytic methods to account for the waiting time paradox, are needed.

Registration

The protocol was registered in PROSPERO (CRD42020207655).

Footnotes

Acknowledgments

The authors thank Dr. Sabrina En-Hsun Wang for constructing in this article.

Authors’ Contributions

R.B.: Designed the work; developed and undertook literature searches; and undertook screening, data extraction, synthesis, and interpretation; and drafted the article. D.S.: Assisted with literature searches; screening; data extraction; and synthesis. S.D.S.: Assisted with literature screening; data extraction; and synthesis. J.E.: Contributed to study design; data interpretation; and drafting article. D.E.: Contributed to study design; data interpretation; and article drafting. R.L.M.: Contributed to study design; data interpretation; and article drafting. V.W.: Contributed to study design; screening; data extraction; synthesis and interpretation; and article drafting. All authors approved the final version and agreed to be accountable for all aspects of the work.

Author Disclosure Statement

All authors declare they have no competing interests to declare.

Funding Information

This work was supported by a NHMRC Partnership grant (APP1150012). JDE is supported by an NHMRC Investigator grant APP1195302; SDE is supported by an Alfred Deakin Postdoctoral Research Fellowship, RB is supported by a Victorian Cancer Agency Early Career Research Fellowship (ECRF20015).

Supplementary Material

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Table S4

Supplementary Table S5

References

Sung

, Ferlay

, Siegel

, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin, 2021; 71(3):209–249; doi: 10.3322/caac.21660

Matz

, Coleman

, Carreira

, et al. Worldwide comparison of ovarian cancer survival: Histological group and stage at diagnosis (CONCORD-2). Gynecol Oncol, 2017; 144(2):396–404; doi: 10.1016/j.ygyno.2016.11.019

Menon

, Gentry-Maharaj

, Burnell

, et al. Ovarian cancer population screening and mortality after long-term follow-up in the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS): A randomised controlled trial. Lancet, 2021; 397(10290):2182–2193; doi: 10.1016/s0140-6736(21)00731-5

Menon

, Weller

, Falborg

, et al. Diagnostic routes and time intervals for ovarian cancer in nine international jurisdictions; findings from the International Cancer Benchmarking Partnership (ICBP). Br J Cancer, 2022; 127(5):844–854; doi: 10.1038/s41416-022-01844-0

Goff

, Mandel

, Muntz

, et al. Ovarian carcinoma diagnosis. Cancer, 2000; 89(10):2068–2075.

Goff

, Mandel

, Melancon

, et al. Frequency of symptoms of ovarian cancer in women presenting to primary care clinics. JAMA, 2004; 291(22):2705–2712.

Bankhead

, Kehoe

, Austoker

. Symptoms associated with diagnosis of ovarian cancer: A systematic review. BJOG, 2005; 112(7):857–865; doi: 10.1111/j.1471-0528.2005.00572.x

Williams

, Murchie

, Bond

. Patient and primary care delays in the diagnostic pathway of gynaecological cancers: A systematic review of influencing factors. Br J Gen Pract, 2019; 69(679):e106–e111; doi: 10.3399/bjgp19X700781

Boban

, Downs

, Codde

, et al. Women diagnosed with ovarian cancer: Patient and carer experiences and perspectives. Patient Relat Outcome Meas, 2021; 12:33–43; doi: 10.2147/prom.S272688

10.

Jenkins

, Catt

, Banerjee

, et al. Patients’ and oncologists’ views on the treatment and care of advanced ovarian cancer in the U.K.: Results from the ADVOCATE study. Br J Cancer, 2013; 108(11):2264–2271; doi: 10.1038/bjc.2013.223

11.

Neal

, Tharmanathan

, France

, et al. Is increased time to diagnosis and treatment in symptomatic cancer associated with poorer outcomes? Systematic review. Full Paper. Br J Cancer, 2015; 112 Suppl 1(Suppl 1):S92–107; doi: 10.1038/bjc.2015.48

12.

Robinson

, Christensen

, Ottesen

, et al. Diagnostic delay, quality of life and patient satisfaction among women diagnosed with endometrial or ovarian cancer: A nationwide Danish study. Qual Life Res, 2012; 21(9):1519–1525; doi: 10.1007/s11136-011-0077-3

13.

Gagliardi

, Grunfeld

, Evans

. Evaluation of diagnostic assessment units in oncology: A systematic review. J Clin Oncol, 2004; 22(6):1126–1135; doi: 10.1200/jco.2004.06.057

14.

Yap

, Vassallo

, Goldsbury

, et al. Pathways to diagnosis of endometrial and ovarian cancer in the 45 and Up Study cohort. Cancer Causes Control, 2023; 34(1):47–58; doi: 10.1007/s10552-022-01634-2

15.

Jayde

, White

, Blomfield

. Symptoms and diagnostic delay in ovarian cancer: A summary of the literature. Contemp Nurse, 2010; 34(1):55–65; doi: 10.5172/conu.2009.34.1.055

16.

Menczer

. Diagnosis and treatment delay in gynecological malignancies. Does it affect outcome? Int J Gynecol Cancer, 2000; 10(2):89–94; doi: 10.1046/j.1525-1438.2000.00020.x

17.

Castelo

, Sue-Chue-Lam

, Paszat

, et al. Clinical delays and comparative outcomes in younger and older adults with colorectal cancer: A systematic review. Curr Oncol, 2022; 29(11):8609–8625; doi: 10.3390/curroncol29110679

18.

Weller

, Vedsted

, Rubin

, et al. The Aarhus statement: Improving design and reporting of studies on early cancer diagnosis. Br J Cancer, 2012; 106(7):1262–1267; doi: 10.1038/bjc.2012.68

19.

Drosdowsky

, Lamb

, Bergin

, et al. A systematic review of methodological considerations in time to diagnosis and treatment in colorectal cancer research. Cancer Epidemiol, 2023; 83:102323; doi: 10.1016/j.canep.2023.102323

20.

Tørring

, Frydenberg

, Hamilton

, et al. Diagnostic interval and mortality in colorectal cancer: U-shaped association demonstrated for three different datasets. J Clin Epidemiol, 2012; 65(6):669–678; doi: 10.1016/j.jclinepi.2011.12.006

21.

Moher

, Liberati

, Tetzlaff

, et al. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Int J Surg, 2010; 8(5):336–341; doi: 10.1016/j.ijsu.2010.02.007

22.

Page

, McKenzie

, Bossuyt

, et al. The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ, 2021; 372:n71; doi: 10.1136/bmj.n71

23.

Covidence systematic review software. 2023.

24.

Sterne

, Hernán

, Reeves

, et al. ROBINS-I: A tool for assessing risk of bias in non-randomised studies of interventions. BMJ, 2016; 355:i4919; doi: 10.1136/bmj.i4919

25.

Algera

, van Driel

, Slangen

BFM

, et al. Impact of the COVID-19-pandemic on patients with gynecological malignancies undergoing surgery: A Dutch population-based study using data from the ’Dutch Gynecological Oncology Audit’. Gynecol Oncol, 2022; 165(2):330–338; doi: 10.1016/j.ygyno.2022.02.013

26.

Alka

, Abraham

, Ashish

, et al. Primary peritoneal carcinoma and ovarian carcinoma - a ten-year comparative analysis. Klin Onkol, 2022; 35(1):63–71; doi: 10.48095/ccko202263

27.

Allgar

, Neal

. Delays in the diagnosis of six cancers: Analysis of data from the National Survey of NHS Patients: Cancer. Br J Cancer, 2005; 92(11):1959–1970.

28.

Altman

, Lambert

, Love

, et al. Examining the effects of time to diagnosis, income, symptoms, and incidental detection on overall survival in epithelial ovarian cancer: Manitoba Ovarian Cancer Outcomes (MOCO) Study Group. Int J Gynecol Cancer, 2017; 27(8):1637–1644; doi: 10.1097/IGC.0000000000001074

29.

Barrett

, Sharp

, Stapley

, et al. Pathways to the diagnosis of ovarian cancer in the UK: A cohort study in primary care. Bjog, 2010; 117(5):610–614; doi: 10.1111/j.1471-0528.2010.02499.x

30.

Baun

M-LL

, Jensen

, Falborg

, et al. Ovarian cancer suspicion, urgent referral and time to diagnosis in Danish general practice: A population-based study. Fam Pract, 2019; 36(6):751–757; doi: 10.1093/fampra/cmz013

31.

Bielik

, Bystricky

, Hoffmannova

, et al. Quality of life and ability to work in ovarian cancer patients in Slovakia. Neoplasma, 2020; 67(2):389–393; doi: 10.4149/neo_2020_190401N285

32.

Boac

, Xiong

, Apte

, et al. Adherence to practice guidelines is associated with reduced referral times for patients with ovarian cancer. Am J Obstet Gynecol, 2018; 218(4):436.e1-436–e7; doi: 10.1016/j.ajog.2018.01.015

33.

Bruce

, Huysman

, Bharucha

, et al. Impact of the COVID-19 pandemic on referral to and delivery of gynecologic oncology care. Gynecol Oncol Rep, 2022; 39:100928; doi: 10.1016/j.gore.2022.100928

34.

Chan

, Ng

, Lee

PWH

, et al. Symptoms, coping strategies, and timing of presentations in patients with newly diagnosed ovarian cancer. Gynecol Oncol, 2003; 90(3):651–656.

35.

Datta

, Fraser

, Burnell

, et al. Association of adult attachment with delays in accessing specialist care in women with ovarian cancer. J Psychosoc Oncol, 2022; 40(4):491–505; doi: 10.1080/07347332.2022.2025510

36.

Di Girolamo

, Walters

, Gildea

, et al. Can we assess Cancer Waiting Time targets with cancer survival? A population-based study of individually linked data from the National Cancer Waiting Times monitoring dataset in England, 2009-2013. PLoS One, 2018; 13(8):e0201288; doi: 10.1371/journal.pone.0201288

37.

Dilley

, Burnell

, Gentry-Maharaj

, et al. Ovarian cancer symptoms, routes to diagnosis and survival—Population cohort study in the ‘no screen’ arm of the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS). Gynecol Oncol, 2020; 158(2):316–322; doi: 10.1016/j.ygyno.2020.05.002

38.

Eggink

, Vermue

, Van der Spek

, et al. The impact of centralization of services on treatment delay in ovarian cancer: A study on process quality. Int J Qual Health Care, 2017; 29(6):810–816; doi: 10.1093/intqhc/mzx107

39.

Erondu

, Alberg

, Bandera

, et al. The Association between body mass index and presenting symptoms in African American Women with ovarian cancer. J Womens Health (Larchmt), 2016; 25(6):571–578; doi: 10.1089/jwh.2015.5359

40.

Funston

, Mounce

, Price

, et al. CA125 test result, test-to-diagnosis interval, and stage in ovarian cancer at diagnosis: A retrospective cohort study using electronic health records. Br J Gen Pract, 2021; 71(707):e465–e472; doi: 10.3399/BJGP.2020.0859

41.

Goenka

, Anandaradje

, Nakka

, et al. The “collateral damage” of the war on COVID-19: Impact of the pandemic on the care of epithelial ovarian cancer. Med Oncol, 2021; 38(11):137; doi: 10.1007/s12032-021-01588-6

42.

Hansen

, Vedsted

, Sokolowski

, et al. Time intervals from first symptom to treatment of cancer: A cohort study of 2,212 newly diagnosed cancer patients. BMC Health Serv Res, 2011; 11(101088677):284; doi: 10.1186/1472-6963-11-284

43.

Hawarden

, Russell

, Gee

, et al. Factors determining ultra-short-term survival and the commencement of active treatment in high-grade serous ovarian cancer: A case comparison study. BMC Cancer, 2021; 21(1):378; doi: 10.1186/s12885-021-08019-9

44.

Huepenbecker

, Sun

, Fu

, et al. Factors impacting the time to ovarian cancer diagnosis based on classic symptom presentation in the United States. Cancer, 2021; 127(22):4151–4160; doi: 10.1002/cncr.33829

45.

Jordan

, Francis

, Nelson

, et al. Pathways to the diagnosis of epithelial ovarian cancer in Australia. The Medical Journal of Australia, 2010; 193(6):326–330.

46.

Keeble

, Abel

, Saunders

, et al. Variation in promptness of presentation among 10,297 patients subsequently diagnosed with one of 18 cancers: Evidence from a National Audit of Cancer Diagnosis in Primary Care. Int J Cancer, 2014; 135(5):1220–1228; doi: 10.1002/ijc.28763

47.

Kirwan

JMJ

, Tincello

, Herod

JJO

, et al. Effect of delays in primary care referral on survival of women with epithelial ovarian cancer: Retrospective audit. BMJ, 2002; 324(7330):148–151.

48.

Koldjeski

, Kirkpatrick

, Swanson

, et al. Ovarian cancer: Early symptom patterns. Oncol Nurs Forum, 2003; 30(6):927–933.

49.

Lataifeh

, Marsden

, Robertson

, et al. Presenting symptoms of epithelial ovarian cancer. Aust N Z J Obstet Gynaecol, 2005; 45(3):211–214.

50.

Lim

, Mesher

, Gentry-Maharaj

, et al. Time to diagnosis of Type I or II invasive epithelial ovarian cancers: A multicentre observational study using patient questionnaire and primary care records. Bjog, 2016; 123(6):1012–1020; doi: 10.1111/1471-0528.13447

51.

Love

, Lambert

, Turner

, et al. Diagnostic and referral intervals for Manitoba women with epithelial ovarian cancer - the Manitoba Ovarian Cancer Outcomes (MOCO) study group: A retrospective cross-sectional study. CMAJ Open, 2017; 5(1):E116–E122; doi: 10.9778/cmajo.20160100

52.

Lyratzopoulos

, Saunders

, Abel

, et al. The relative length of the patient and the primary care interval in patients with 28 common and rarer cancers. Br J Cancer, 2015; 112 Suppl 1(Suppl 1):S35–40; doi: 10.1038/bjc.2015.40

53.

Mayba

, Lambert

, Turner

, et al. Examining the Selection Criteria of Neoadjuvant Chemotherapy Patients. J Obstet Gynaecol Can, 2018; 40(5):595–603; doi: 10.1016/j.jogc.2017.09.018

54.

McNally

, Wareham

, Flemming

, et al. The impact of the introduction of a fast track clinic on ovarian cancer referral patterns. Eur J Cancer Care (Engl), 2003; 12(4):327–330.

55.

Menczer

, Chetrit

, Sadetzki

, National Israel Ovarian Cancer G . The effect of symptom duration in epithelial ovarian cancer on prognostic factors. Arch Gynecol Obstet, 2009; 279(6):797–801; doi: 10.1007/s00404-008-0814-1

56.

Moterani

, Moterani

NJW

Jr , Candido Dos Reis

. Treatment delay and treatment pattern modifications among epithelial ovarian cancer patients during the COVID-19 pandemic: A retrospective cohort study. J Surg Oncol, 2022; 126(7):1155–1161; doi: 10.1002/jso.27048

57.

Nagle

, Francis

, Nelson

, et al. Reducing time to diagnosis does not improve outcomes for women with symptomatic ovarian cancer: A report from the Australian Ovarian Cancer Study Group. J Clin Oncol, 2011; 29(16):2253–2258; doi: 10.1200/JCO.2010.32.2164

58.

Noer

, Sperling

, Ottesen

, et al. Ovarian cancer and comorbidity: Is poor survival explained by choice of primary treatment or system delay? Int J Gynecol Cancer, 2017; 27(6):1123–1133; doi: 10.1097/IGC.0000000000001001

59.

Olsen

, Cnossen

, Green

, et al. Comparison of symptoms and presentation of women with benign, low malignant potential and invasive ovarian tumors. Eur J Gynaecol Oncol, 2007; 28(5):376–380.

60.

Paulino

, de Melo

, Silva-Filho

, et al. Panorama of Gynecologic Cancer in Brazil. JCO Glob Oncol, 2020; 6:1617–1630; doi: 10.1200/GO.20.00099

61.

Paulsen

, Kaern

, Kjaerheim

, et al. Symptoms and referral of women with epithelial ovarian tumors. Int J Gynaecol Obstet, 2005; 88(1):31–37.

62.

Price

, Spencer

, Zhang

, et al. Trends in time to cancer diagnosis around the period of changing national guidance on referral of symptomatic patients: A serial cross-sectional study using UK electronic healthcare records from 2006-17. Cancer Epidemiol, 2020; 69:101805; doi: 10.1016/j.canep.2020.101805

63.

Rabiu

, Akinola

, Adewunmi

, et al. Delays in presentation and management of ovarian cancer in Lagos, Nigeria. J Obstet Gynaecol, 2013; 33(3):305–308; doi: 10.3109/01443615.2012.753417

64.

Reid

, Bhatla

, Oza

, et al. The World Ovarian Cancer Coalition Every Woman Study: Identifying challenges and opportunities to improve survival and quality of life. Int J Gynecol Cancer, 2020; 31(2):238–244; doi: 10.1136/ijgc-2019-000983

65.

Robinson

, Christensen

, Ottesen

, et al. Socio-demographic factors, comorbidity and diagnostic delay among women diagnosed with cervical, endometrial or ovarian cancer. Eur J Cancer Care (Engl), 2011; 20(5):653–661; doi: 10.1111/j.1365-2354.2011.01259.x

66.

Robinson

, Ottesen

, Christensen

, et al. Diagnostic delay experienced among gynecological cancer patients: A nationwide survey in Denmark. Acta Obstet Gynecol Scand, 2009; 88(6):685–692; doi: 10.1080/00016340902971482

67.

Rubin

, Saunders

, Abel

, et al. Impact of investigations in general practice on timeliness of referral for patients subsequently diagnosed with cancer: Analysis of national primary care audit data. Br J Cancer, 2015; 112(4):676–687; doi: 10.1038/bjc.2014.634

68.

Smith

AJB

, Fader

. Association of the affordable care act with ovarian cancer care. Obstet Gynecol, 2020; 135(2):257–265; doi: 10.1097/AOG.0000000000003630

69.

Solomon

, Mulugeta

. Diagnosis and risk factors of advantage cancers in Ethiopia. J Cancer Prev, 2019; 24(3):163–172; doi: 10.15430/JCP.2019.24.3.163

70.

Soomers

VLMN

, Lidington

, Sirohi

, et al. The prolonged diagnostic pathway of young adults (Aged 25-39) with cancer in the United Kingdom: Results from the young adult cancer patient journey study. J Clin Med, 2021; 10(20):4646; doi: 10.3390/jcm10204646

71.

Sponholtz

, Mogensen

, Hildebrandt

, et al. Clinical impact of pre-treatment FDG-PET/CT staging of primary ovarian, fallopian tube, and peritoneal cancers in women. Acta Obstet Gynecol Scand, 2019; 99(2):186–195; doi: 10.1111/aogs.13726

72.

Spurgeon

, Barwell

, Kerr

. Waiting times for cancer patients in England after general practitioners’ referrals: Retrospective national survey. BMJ, 2000; 320(7238):838–839.

73.

Suh-Burgmann

, Alavi

. Detection of early stage ovarian cancer in a large community cohort. Cancer Med, 2019; 8(16):7133–7140; doi: 10.1002/cam4.2522

74.

Tate

, Martin

AGR

, Murray-Thomas

, et al. Determining the date of diagnosis–is it a simple matter? The impact of different approaches to dating diagnosis on estimates of delayed care for ovarian cancer in UK primary care. BMC Med Res Methodol, 2009; 9(100968545):42; doi: 10.1186/1471-2288-9-42

75.

Tate

, Nicholson

, Cassell

. Are GPs under-investigating older patients presenting with symptoms of ovarian cancer? Observational study using General Practice Research Database. Br J Cancer, 2010; 102(6):947–951; doi: 10.1038/sj.bjc.6605593

76.

Thulesius

, Lindgren

, Olsson

, et al. Diagnosis and prognosis of breast and ovarian cancer–a population-based study of 234 women. Acta Oncol, 2004; 43(2):175–181.

77.

Tokuda

, Chinen

, Obara

, et al. Intervals between symptom onset and clinical presentation in cancer patients. Intern Med, 2009; 48(11):899–905.

78.

Tone

, Boghosian

, Ross

, et al. Understanding the experience of canadian women living with ovarian cancer through the every woman StudyTM. Curr Oncol, 2022; 29(5):3318–3340; doi: 10.3390/curroncol29050271

79.

Torring

, Falborg

, Jensen

, et al. Advanced-stage cancer and time to diagnosis: An International Cancer Benchmarking Partnership (ICBP) cross-sectional study. Eur J Cancer Care (Engl), 2019; 28(5):e13100; doi: 10.1111/ecc.13100

80.

Vandborg

, Christensen

, Kragstrup

, et al. Reasons for diagnostic delay in gynecological malignancies. Int J Gynecol Cancer, 2011; 21(6):967–974; doi: 10.1097/IGC.0b013e31821d2770

81.

Vine

, Calingaert

, Berchuck

, et al. Characterization of prediagnostic symptoms among primary epithelial ovarian cancer cases and controls. Gynecol Oncol, 2003; 90(1):75–82; doi: 10.1016/S0090-8258(03)00175-6

82.

Vine

, Ness

, Calingaert

, et al. Types and duration of symptoms prior to diagnosis of invasive or borderline ovarian tumor. Gynecol Oncol, 2001; 83(3):466–471.

83.

Webb

, Purdie

, Grover

, et al. Symptoms and diagnosis of borderline, early and advanced epithelial ovarian cancer. Gynecol Oncol, 2004; 92(1):232–239.

84.

White

, Walton

, Kwedza

, et al. Variation in ovarian cancer care in Australia: An analysis of patterns of care in diagnosis and initial treatment in New South Wales. Eur J Cancer Care (Engl), 2022; 31(6):e13649; doi: 10.1111/ecc.13649

85.

Yawn

, Barrette

, Wollan

. Ovarian cancer: The neglected diagnosis. Mayo Clin Proc, 2004; 79(10):1277–1282.

86.

Trétarre

, Molinié

, Woronoff

, et al. Ovarian cancer in France: Trends in incidence, mortality and survival, 1980-2012. Gynecol Oncol, 2015; 139(2):324–329; doi: 10.1016/j.ygyno.2015.09.013

87.

Arnold

, Rutherford

, Bardot

, et al. Progress in cancer survival, mortality, and incidence in seven high-income countries 1995-2014 (ICBP SURVMARK-2): A population-based study. Lancet Oncol, 2019; 20(11):1493–1505; doi: 10.1016/s1470-2045(19)30456-5

88.

Faber

, Horsbøl

, Baandrup

, et al. Trends in survival of epithelial ovarian/tubal cancer by histology and socioeconomic status in Denmark 1996-2017. Gynecol Oncol, 2022; 164(1):98–104; doi: 10.1016/j.ygyno.2021.10.091

89.

Larsen

, Hansen

, Sokolowski

, et al. Agreement between patient-reported and doctor-reported patient intervals and date of first symptom presentation in cancer diagnosis - a population-based questionnaire study. Cancer Epidemiol, 2014; 38(1):100–105; doi: 10.1016/j.canep.2013.10.006

90.

Malalasekera

, Nahm

, Blinman

, et al. How long is too long? A scoping review of health system delays in lung cancer. Eur Respir Rev, 2018; 27(149); doi: 10.1183/16000617.0045-2018

91.

Schisterman

, Cole

, Platt

. Overadjustment bias and unnecessary adjustment in epidemiologic studies. Epidemiology, 2009; 20(4):488–495; doi: 10.1097/EDE.0b013e3181a819a1

92.

VanderWeele

. Principles of confounder selection. Eur J Epidemiol, 2019; 34(3):211–219; doi: 10.1007/s10654-019-00494-6

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.23 MB