Shortening the Endometriosis Diagnostic Pathway: A Narrative Review and Research Agenda

Abstract

Endometriosis is often diagnosed several years after the first onset of symptoms; however, the duration of this delay may vary depending on the population studied, the assessment starting point, and the timing of diagnosis. This narrative review, which also utilises a structured evidence mapping methodology, examined interventions aimed at ensuring timely diagnosis across six measurable process domains and distinguished process changes from indirect outcomes. Targeted, iterative searches were conducted in PubMed/MEDLINE, guideline databases and reference lists up to 2 August 2026. The review includes 61 sources published between 1996 and 2026, including 23 empirical reports published between 2015 and 2026 that were mapped. The search was not systematic, the screening was not repeated, and no formal bias assessment was carried out. The identified literature provided limited controlled evidence that interventions effectively reduced diagnostic delays; none of the reviews considered diagnostic accuracy, patient outcomes, service demand, cost or equity measures, or timing. Whilst educational initiatives may increase knowledge in the short term, the effectiveness of symptom assessment tools and emerging diagnostic techniques varies across the selected populations. Imaging methods may improve the detection and characterisation of endometriosis; however, a comparative historical analysis indicates changes in the diagnostic pathway and management approach, without quantifying the time elapsed. Current guidelines do not mandate laparoscopy for every patient before establishing a provisional diagnosis and treatment plan. We propose a framework—combining a structured symptom assessment, differential diagnosis, appropriate imaging, diagnostic tests when clinically justified, a safety net, and post-referral follow-up—whose validity has not yet been established. Pragmatic comparative studies are required to determine whether coordinated diagnostic pathways provide earlier, more appropriate and, ultimately, more beneficial care.

Plain Language Summary

Endometriosis may take years to recognise because symptoms are normalised, recur across consultations or are investigated without a clear route to further care. We reviewed research on education, symptom questionnaires, digital tools, clinical diagnosis, imaging, service redesign and new tests. Most studies assessed knowledge, feasibility or diagnostic accuracy rather than the time needed to reach appropriate care. Among the evidence identified, few measured the interval targeted by an intervention, and none considered timeliness alongside safety, patient outcomes, workload, cost and equity. Current guidance supports a working clinical diagnosis and treatment without requiring surgery in every patient, although uncertainty and alternative diagnoses still require careful assessment. We propose a framework for testing: structured symptom enquiry, appropriate first-line imaging, parallel investigation and treatment, follow-up after negative imaging, and referrals tracked to completion. The framework has not been validated as a package. The goal is not speed alone, but timely care that provides a clear diagnosis and useful management plan without avoidable harm.

Key Messages

• Diagnostic accuracy, awareness and feasibility do not show that a diagnostic pathway has become shorter or better.

• Few identified studies measured the interval targeted by an intervention; none assessed timeliness alongside appropriateness, patient outcomes, resource use, cost and equity.

• An actionable diagnosis is an author-proposed endpoint: a documented and communicated working, imaging-supported or surgical diagnosis linked to an agreed plan and safety-net.

• The proposed pathway, service framework and outcomes require co-design, consensus and prospective evaluation.

Introduction

Endometriosis affects an estimated 10% of women and girls of reproductive age worldwide and may present with dysmenorrhoea, non-menstrual pelvic pain, deep dyspareunia, dyschezia, urinary symptoms, fatigue or infertility.^1,2 None of these features is diagnostic on its own. Symptoms overlap with adenomyosis, pelvic floor dysfunction, irritable bowel syndrome, bladder pain syndrome and other causes of persistent pelvic pain, while superficial peritoneal disease may not be visible on routine imaging. Difficulties can therefore arise at several points: symptoms may be normalised, help-seeking postponed, clinical suspicion left undocumented, referral delayed, imaging performed without an endometriosis-directed protocol, or a diagnostic label withheld until surgery.

The familiar estimate of a seven- to ten-year delay is directionally useful but scientifically imprecise. A systematic review of 17 observational studies reported estimates from 0.3 to 12 years, with substantial variation in geography, recruitment, recall period and diagnostic definition.³ A social-ecological scoping review reported a mean of about 6.8 years across heterogeneous studies,⁴ and a later systematic review again found that patient, clinician and health-system factors interact within an evidence base dominated by retrospective self-report.⁵ Among 837 respondents from 11 Dutch hospitals, the median total interval was seven years (interquartile range 2–15); the median patient and general-practice intervals were each one year, while the specialist interval was less than one year.⁶ An intervention should therefore be judged against the part of the pathway it can plausibly change, not against time accumulated before it begins.

Diagnostic endpoints have also changed. Several earlier studies stopped the clock at surgical or histological confirmation,^7–9 whereas a later US survey included both surgical and non-surgical routes.¹⁰ Surveys continue to report long intervals. Although within-cohort trends in Australia and Aotearoa New Zealand suggest that the clinician-related component may have shortened in some settings,^11–13 a newer Aotearoa New Zealand survey of 1024 respondents found a mean self-reported interval of 9.7 years (standard deviation 7.1).¹⁴ These retrospective observations are not contradictory—the overall interval can remain long while particular components improve—but they cannot identify which concurrent changes were responsible. Qualitative studies also show that elapsed time alone misses repeated dismissal, menstrual stigma, fragmented responsibility and the burden of self-advocacy.^15,16

Current guidance no longer requires laparoscopy before every diagnosis. ACOG supports a clinical diagnosis as a basis for empirical treatment; ESHRE recommends imaging and reserves laparoscopy for selected circumstances; NICE recommends transvaginal ultrasonography for suspected disease even when examination is normal; and Australian and Canadian guidance also recognises clinical, imaging and surgical routes.^17–21 This shift can remove avoidable procedural waiting, but it also changes the endpoint being measured. A pathway may appear faster because the endpoint has moved from surgery to a working diagnosis, even if access to appropriate management has not improved.

This review asks which interventions have measured change in a prespecified interval of the endometriosis diagnostic pathway, in which populations, and with what consequences. It distinguishes elapsed-time outcomes from knowledge, feasibility, diagnostic performance and other proxies. It then offers an author-proposed pathway framework, candidate outcomes and a research agenda. The emphasis is on appropriate timeliness: the shortest pathway compatible with diagnostic safety, informed choice and access to useful management, rather than speed alone.

Review Approach

Design, Scope and Search

This article is a narrative review with structured evidence mapping. It was designed to distinguish direct evidence of pathway change from proxy outcomes and to identify priorities for implementation research; it was not designed as a systematic review or formal evidence-grading exercise.

PubMed/MEDLINE and the official repositories or publication pages of ACOG, ESHRE, NICE, RANZCOG, SOGC and WHO were searched through 2 August 2026. No start-date limit was imposed. Searches were targeted and iterative. They combined endometriosis with terms relating to diagnostic delay, interval, pathway, journey, timeliness, recognition, referral, education, awareness, questionnaires, symptom checkers, decision support, clinical diagnosis, ultrasonography, magnetic resonance imaging, biomarkers, artificial intelligence, fast-track services, navigation, policy and implementation. Reference lists and forward citations of recent systematic and scoping reviews were also examined to identify landmark reports and studies whose titles did not use delay-related terminology.

We prioritised guidelines, evidence syntheses, comparative interventions, implementation evaluations, diagnostic-impact studies, prospective diagnostic-accuracy studies, health-economic models and research that clarified patient experience or inequity. A source was considered relevant when it addressed an intervention or observation that could alter symptom recognition, help-seeking, clinical suspicion, access to further assessment, diagnostic decision-making or initiation of management. Purely therapeutic studies were outside scope unless the diagnostic strategy determined access to treatment.

The final narrative cites 61 sources published between 1996 and 2026. For transparency, we charted 23 key empirical reports published between 2015 and 2026 that directly informed claims about recognition, intervention performance or pathway change. This is a study-level map of the empirical reports discussed in the review, not a claim of exhaustive identification or a systematic included-study corpus.

Selection, Charting and Synthesis

Retrieval and selection were purposive rather than exhaustive. We did not maintain a PRISMA-style screening log, perform duplicate screening or data extraction, undertake a formal risk-of-bias assessment, or conduct meta-analysis. Sources were grouped by pathway position and intervention class. Interpretation considered study design, setting, population, comparator or reference standard, endpoint, whether the outcome represented direct pathway change or a proxy, and the principal limitation relevant to pathway inference. The limitations are descriptive and should not be read as formal risk-of-bias judgements. Because the search was targeted, absence-of-evidence statements are bounded to the sources identified. We therefore use formulations such as “we did not identify” rather than claiming that no relevant study exists. The search strategy and selection process also mean that the cited-source count must not be interpreted as a systematic study yield.

Author-Developed Outcome-Reach Map

We grouped evidence by the furthest outcome along the diagnostic pathway that a study measured:

• Stage 0: barrier description, concept development or co-design.

• Stage 1: technical performance, prediction-model development or diagnostic accuracy.

• Stage 2: knowledge, attitudes, acceptability, usability or feasibility.

• Stage 3: an intermediate process outcome, such as referral completion, waiting time, diagnostic yield, consultation count or pathway attrition.

• Stage 4: a prespecified time-to-diagnosis or time-to-treatment outcome.

• Stage 5: comparative real-world effectiveness that also considers harms, equity, cost and sustainability.

These author-assigned categories describe outcome reach, not methodological quality, certainty of evidence or risk of bias. A rigorous diagnostic-accuracy study may remain at Stage 1 because it cannot establish pathway impact, whereas a weak uncontrolled study may measure a Stage 4 outcome. Assignment was interpretive and the categories have not been validated.

Basis of the Proposed Frameworks

The segmented pathway, actionable-diagnosis criteria, service framework and candidate outcomes were developed from the cited literature and the authors’ interpretation. They have not undergone direct patient or stakeholder co-design, Delphi consensus or prospective validation. Published qualitative evidence informed the discussion of patient priorities,^15,16 but formal co-design remains necessary before any of these proposals is treated as a standard.

Defining Diagnostic Endpoints and Intervals

Terminology Used in This Review

The diagnostic pathway is the linked sequence from symptom onset and recognition through healthcare contact, diagnostic assessment, communication of a diagnostic decision and initiation or revision of agreed management. A diagnostic interval is the elapsed time between two explicitly named milestones. Time to diagnosis is therefore meaningful only when its starting point and endpoint are stated. We reserve diagnostic delay for avoidable or clinically unwarranted prolongation rather than using it as a synonym for any long observed interval.

Diagnostic recognition means that endometriosis has been documented as a credible possibility that warrants assessment, treatment or referral, without necessarily being adopted as the working explanation. A working or clinical diagnosis is a provisional clinical judgement that can support shared decisions while uncertainty remains. An imaging-supported diagnosis is based on recognised features identified using an appropriate protocol. Surgical visualisation is the macroscopic assessment of lesions at operation, while histopathological confirmation applies only to sampled tissue. Histopathology can confirm sampled lesions and help identify mimics, atypia or associated neoplasia. A negative sample does not necessarily exclude disease elsewhere or a missed or unrepresentative lesion.^17,18,22 These terms should not be used interchangeably. Disease phenotyping describes lesion site, extent and organ involvement; it is important for complex care and operative planning but is not a prerequisite for every initial working diagnosis.

The Author-Proposed “Actionable Diagnosis” Endpoint

For pathway evaluation, we define an actionable diagnosis as a documented and communicated working, imaging-supported or surgical diagnosis that is sufficiently credible to guide shared decision-making and an appropriate management or further-assessment plan. This is an author-proposed operational endpoint, not a validated or consensus definition. At minimum, the record should show:

1. a compatible symptom pattern and its functional effect;

2. proportionate assessment of urgent and credible competing diagnoses;

3. the diagnostic basis, route and degree of uncertainty;

4. explicit communication with the patient and an agreed management or further-assessment plan; and

5. a review point and route for escalation if symptoms persist, the diagnosis becomes doubtful or further phenotyping is needed.

The diagnostic route must be recorded. Visual suspicion at laparoscopy and histopathological confirmation answer related but different questions; imaging can support diagnosis and map some phenotypes; and a working clinical diagnosis may be appropriate before either procedure. Surgery remains relevant when symptoms persist despite empirical management, imaging is negative or equivocal, another diagnosis is suspected, disease mapping or operative treatment is required, or a patient chooses surgery after informed discussion.^17–21

Shorter is not invariably better. Additional time may be clinically appropriate when symptoms are non-specific, red flags or competing diagnoses require investigation, specialist mapping is needed, symptoms evolve, or a patient prefers staged assessment. Such time should be explained and documented. Pathway evaluation should therefore pair timeliness with appropriateness, safety, patient preference and outcomes.

Six Measurable Intervals

Figure 1 uses seven milestones to define six intervals, and Table 1 sets out their starting points, modifiable delays and candidate measures. These are measurement points rather than a mandatory linear route. A working diagnosis may be made in primary care or general gynaecology before specialist referral, treatment may begin before imaging or diagnostic confirmation, and later assessment may revise the diagnosis or management plan. Each intervention should be judged against the interval it is intended to change.

Figure 1.

Author-proposed framework for six measurable intervals in the endometriosis diagnostic pathway. Seven milestones define six intervals: I1, symptom appraisal, from symptom onset to recognition that symptoms warrant care; I2, help-seeking, from recognition to first relevant healthcare contact; I3, first-contact recognition, from first contact to endometriosis being recorded as a possibility with an initial plan; I4, access to further assessment, from the initial plan or referral to the next appropriate diagnostic assessment; I5, diagnostic assessment, from that assessment to a documented and communicated actionable diagnosis; and I6, diagnostic action, from the communicated diagnosis to the start or revision of agreed management. These are measurement points, not a required linear route. “Actionable diagnosis” is an author-proposed operational endpoint and may be a working clinical, imaging-supported or surgical diagnosis; the basis, route and certainty should be recorded. Histopathology confirms sampled lesions and can help identify mimics, atypia or associated neoplasia. An intervention intended to shorten a pathway should reduce its prespecified target interval without unacceptable effects on diagnostic appropriateness, patient outcomes, workload, cost or equity. Balancing measures include alternative and revised diagnoses, false reassurance, overdiagnosis, unnecessary procedures, referral burden and opportunity cost. Patient-centred outcomes include validation, informed choice, appropriate treatment, symptoms, function and quality of life. Equity should be assessed using locally relevant dimensions, which may include age, race and ethnicity, socioeconomic position, geography, disability, language, insurance status and gender identity. Blue boxes indicate early pathway milestones, green indicates further diagnostic assessment, and amber indicates diagnosis and management. Source: created by the authors.

Table 1.

Six Measurable Intervals in the Endometriosis Diagnostic Pathway

Interval	Start and End	Common Modifiable Source of Delay	Candidate Intervention and Measure
1. Symptom appraisal	Symptom onset to recognition that symptoms warrant attention	Normalisation of severe menstrual pain; stigma; limited menstrual-health literacy	Inclusive education and clear functional thresholds; record onset and recognition separately.
2. Help-seeking	Recognition to first relevant healthcare contact	Cost, access, embarrassment, previous dismissal or poor care	Navigation and accessible first contact; measure the patient interval and reasons for postponement.
3. First-contact recognition	First contact to recorded recognition, initial plan or referral	Fragmented symptoms; repeated symptomatic prescribing; anchoring on gastrointestinal, urinary or psychological explanations	Structured history, longitudinal record prompts, clinician education and safety-netting; measure consultations, alternatives considered and plan quality.
4. Access to further assessment	Initial plan or referral to appropriate further assessment	Unclear criteria, rejected referrals, repeated routine imaging, queues and geographical barriers	Advice-and-guidance, transparent criteria, tracked referral and navigation; measure rejection, completion, waiting time and inequity.
5. Diagnostic assessment	Further assessment to a documented and communicated actionable diagnosis	Variable imaging expertise; sequential investigation; routine surgical gatekeeping; unresolved uncertainty	Appropriate clinical assessment, endometriosis-directed US, selective MRI, shared decision-making and selective surgery; record diagnostic route, certainty and time.
6. Diagnostic action	Communicated diagnosis to the start or revision of agreed management	A label without access to treatment; repeated reassessment; poor continuity	Immediate or revised care plan, named route for follow-up and multidisciplinary access when needed; measure time to management and patient-reported experience.

Notes: These intervals are measurement constructs rather than a mandatory linear route. A working diagnosis or treatment may occur before specialist review, imaging or surgical confirmation.

Abbreviations: MRI, magnetic resonance imaging; US, ultrasonography.

Where Delay Arises

Symptom Appraisal and First-Contact Recognition

Dysmenorrhoea is common, and endometriosis is only one possible cause. Suspicion becomes more appropriate when pain is severe, progressive, function-limiting, resistant to first-line treatment, or accompanied by deep dyspareunia, dyschezia, cyclical urinary or bowel symptoms, infertility, fatigue or a family history.^17–21 Adolescents may have no useful reference point for a typical period; families and schools may regard absence as an unavoidable part of menstruation; and menstrual stigma may discourage disclosure.^15,16 Education is therefore more useful when it teaches functional thresholds and routes to care than when it simply names the disease.

Primary care often encounters fragments of the presentation over time. Dutch GPs and trainees described uncertainty about when symptoms become sufficiently suggestive, limited access to expertise and a preference for linked guidance, decision support, education and regional networks.²³ British GPs similarly described diagnostic complexity and the difficulty of balancing earlier referral against unnecessary referral.²⁴ Electronic records may contain earlier signals. In a nested case-control study, pain and menstrual symptoms recorded within the same year were associated with later endometriosis diagnosis (odds ratio 6.5, 95% confidence interval 3.9–10.6); lower gastrointestinal symptoms recorded close to gynaecological pain showed a similar association (odds ratio 6.1, 95% confidence interval 3.6–10.6).²⁵ These associations could inform decision support, but that study did not test a prompt or measure time saved.

Access, Referral and Capacity

Recognition will not improve timeliness if referrals are rejected, imaging is repeated without a clear question, travel is impractical or specialist queues are already full. Equally, referral of every person with dysmenorrhoea to a tertiary centre would be neither feasible nor clinically appropriate. We therefore propose distributing diagnostic capability across primary care, general gynaecology, imaging and specialist services, with escalation guided by symptom burden, diagnostic uncertainty, suspected deep disease, organ involvement, fertility needs and treatment response. The precise service model will vary by health system and requires evaluation rather than assumption.

The Effect of the Diagnostic Endpoint

When diagnosis is defined only by histology, the wait for laparoscopy becomes part of the diagnostic interval even after a credible working diagnosis has changed care. A definition that is too loose creates the opposite problem: a symptom label may be counted as success despite poor communication, inadequate consideration of alternatives or an ineffective plan. Pathway studies should therefore report the first actionable diagnosis, its basis and degree of certainty, and any later imaging, surgical or histological refinement.

Evidence for Pathway Interventions

Across the sources identified, controlled evidence that interventions shorten endometriosis diagnostic pathways was limited. A recent scoping review described several retrospective or cross-sectional comparisons as interventions, but none prospectively implemented and evaluated a pathway change.⁴ Few studies measured the particular interval an intervention was designed to change, and we did not identify an evaluation that assessed timeliness together with diagnostic appropriateness, patient outcomes, service demand, cost and equity. Table 2 presents the key empirical reports informing this conclusion.

Table 2.

Key Empirical Reports Informing the Structured Evidence Map

Citation; Setting and Design	Population or Sample	Intervention, Exposure or Comparator	Endpoint and Pathway-Relevant Finding	Principal Limitation for Pathway Inference
Soliman et al;10 USA; cross-sectional online survey	638 respondents with self-reported physician-diagnosed endometriosis	Diagnosis by an obstetrician–gynaecologist vs another clinician	Mean consultation-to-diagnosis interval was 21.5 vs 40.3 months	Recall and selection bias, industry involvement and non-randomised exposure; not a comparison of clinical vs surgical diagnosis.
de Kok et al;23 Netherlands; qualitative focus groups	22 GPs and 13 GP trainees	Views on guidance, education and diagnostic support; no comparator	Participants favoured a combined guideline, flowchart, education and regional network	Elicited preferences rather than testing an intervention or time outcome.
Dixon et al;24 England; qualitative interviews	GPs discussing suspected endometriosis in primary care	Clinical reasoning and referral experience; no comparator	Described uncertainty, competing explanations and referral decisions	Qualitative perceptions cannot establish frequency or pathway effect.
Burton et al;25 Scotland; nested case-control EHR study	366 cases; 1453 population controls and 610 symptomatic controls	Composite symptom and consultation patterns vs matched controls	Several patterns were associated with later diagnosis up to three years beforehand	Retrospective associations; no prompt was implemented and no interval was changed.
Bush et al;26 New Zealand; school programme and clinic audit	School students and specialist clinic attendees	Menstrual-health/endometriosis education; historical programme comparisons	Increased awareness and younger specialist attendance were reported	No contemporaneous control and no symptom-onset-to-diagnosis outcome.
Wahl et al;27 Canada; cluster-randomised classroom trial	Four classes; 71 students enrolled and 48 supplied paired data	One 60-minute virtual programme; two intervention vs two waitlist classes	Short-term knowledge improved by 1.56 points on a six-point measure	Small sample, brief follow-up and no linkage to healthcare or diagnosis.
Khan et al;28 Japan; educational intervention	Young women receiving a targeted education programme	Before-and-after knowledge and awareness assessment	Knowledge and awareness improved after the programme	Uncontrolled proxy outcome; no care-seeking or diagnostic interval.
Ellis et al;29 Aotearoa New Zealand; cross-sectional survey	657 respondents with self-reported endometriosis	Awareness at symptom onset vs no awareness	Median self-reported interval was nine vs 11 years	Retrospective association; cannot establish that education shortened the pathway.
Abd Elrahim et al;30 Egypt; quasi-experimental school study	300 female technical secondary-school students	Education and referral activity; pre/post assessment	32 were referred and eight were reported as confirmed on medical consultation at three months	Diagnostic criteria and elapsed-time outcome were not reported; no comparative diagnostic endpoint.
Frayne et al;31 Australia; mixed-methods pilot	12 GPs and 111 patients	RATE tool used during consultations; no concurrent comparator	Tool was acceptable and prompted discussion, investigation and follow-up	Did not establish yield, false-positive decisions, workload or elapsed time.
Fauconnier et al;32 France; diagnostic study	105 cases and 197 controls	Patient-completed questionnaire at low- and high-risk thresholds	Sensitivity/specificity were 98%/70% and 62%/98%, respectively	Selected, unmatched case-control cohort; thresholds were not tested as a pathway intervention.
Mishra et al;33 Australian cohorts; prediction-model development and validation	4,005 development, 1335 internal-test and 4077 separate-cohort participants	SAFE score; model discrimination across cohorts	Separate-cohort AUC was 0.71; sensitivity 46.7% and specificity 85.7% at score ≥2	Separate cohort was also Australian; limited sensitivity for exclusion and no prospective management-impact test.
Peven et al;34 digital clinical-vignette study	24 fictitious, clinician-written reproductive-health vignettes	Digital symptom checker suggestions vs vignette diagnoses	Endometriosis sensitivity was 81.8% and specificity 84.6%	Developer-funded/employee-authored vignette study; no real users, behaviour, access or downstream decisions.
Luz and Ferreira Lima;35 Brazilian primary care; single-arm formative study	60 participants across 10 primary-care units	Eight-week EndoConnect implementation; no controlled comparator	Feasibility, engagement and referral signals were reported	Single-arm, short-term study; no causal time-to-diagnosis or patient-outcome evidence.
Xu et al;36 health-economic modelling study	Simulated symptomatic cohort	Digital symptom checker pathway vs modelled usual care	Projected a 4.36-year reduction, QALY gain and cost savings	Company-funded model partly used company data; findings depend on assumptions rather than observed care.
Fraser et al;37 Canada; retrospective imaging review	40 surgically confirmed cases	Routine vs expert-guided TVUS	Sensitivity was 25% with routine and 78% with expert-guided examination	All participants had disease, so specificity could not be estimated; no time or management outcome.
McClenahan et al;38 retrospective analysis of a tertiary imaging cohort	467 participants; data collected prospectively in 2018–2019 before surgery	Complete four-step IDEA protocol vs basic first step; visual laparoscopy reference	Sensitivity increased (0.94 vs 0.65) while specificity decreased (0.58 vs 0.78)	Surgery-enriched specialist spectrum, retrospective analysis and accuracy outcome; no pathway-time evaluation.
Markwei et al;39 USA; single-centre historical before-and-after cohort	444 new adults with pelvic pain suspicious for deep endometriosis, 2013–2020	Endometriosis-specific MRI protocol introduced in 2017 vs earlier practice	Diagnostic laparoscopy decreased twofold and medical management rose by 21.1% points	Historical and secular confounding; no elapsed-time, patient-outcome, equity or resource-use evaluation.
Gastineau et al;40 France; retrospective single-centre before-and-after study	Patients with moderate or severe disease diagnosed before or after the 2022 national plan	Pre-plan vs post-plan periods	Median reported delay was 5.7 vs 4.2 years; difference not significant	Recall, case mix, secular change and limited power restrict causal inference.
Bendifallah et al;41 France; prospective 17-centre diagnostic study	971 symptomatic participants; disease prevalence 77%	Salivary microRNA signature vs composite imaging/surgical reference	Sensitivity was 97.3% and specificity 94.1%	Enriched referral cohort, composite reference and no diagnostic-impact or time outcome.
Noar et al;42 selected diagnostic cohort; open-label study	Symptomatic cohort with only two disease-negative participants	Gastrointestinal myoelectrical activity biomarker vs mixed reference standards	High age-stratified diagnostic accuracy was reported	Specificity was imprecise; small threshold-development samples, mixed references, limited replication and relevant commercial interests.
Wong et al;43 case-control development and validation study	298 participants, including a balanced independent retrospective validation cohort of 80	Multi-omic HerResolve assay vs case/control status	Validation AUC 0.944, sensitivity 80% and specificity 97.5%	Case-control spectrum and industry involvement; no prospective primary-care or management-impact evaluation.
Gibbons et al;44 single-centre exploratory Phase 2 study	19 evaluable participants, including only two disease-negative participants	99mTc-maraciclatide SPECT-CT vs surgery	Imaging and surgery were concordant in 16 of 19 participants	Very small, industry-involved open-label study; false-positive performance, access, cost and pathway effect remain uncertain.

Notes: This table maps the key empirical reports discussed in the narrative and is not an exhaustive systematic-study inventory. The limitations are descriptive, not formal risk-of-bias judgements.

Abbreviations: AUC, area under the curve; EHR, electronic health record; GP, general practitioner; IDEA, International Deep Endometriosis Analysis; QALY, quality-adjusted life-year; RATE, Raising Awareness Tool for Endometriosis; SAFE, Simplified Adolescent Factors for Endometriosis; SPECT-CT, single-photon emission computed tomography-computed tomography; TVUS, transvaginal ultrasonography.

Education, Structured Consultation and Decision Support

Education addresses symptom normalisation and may improve recognition. The New Zealand school programme linked menstrual-health education with referral to appropriate care. Programme and clinic audits suggested greater awareness and younger specialist attendance, but lacked a contemporaneous control group and did not measure the interval from symptom onset to diagnosis.²⁶ In a small cluster-randomised trial involving four school classes (two intervention and two waitlist), a 60-minute virtual programme improved short-term knowledge by 1.56 points on a six-point measure (95% confidence interval 1.12–2.00); 48 of 71 enrolled students provided paired data.²⁷ A targeted programme for young women also improved awareness and knowledge.²⁸ In a 2026 Aotearoa New Zealand survey of 657 respondents, awareness of endometriosis at symptom onset was associated with a shorter median self-reported interval (nine vs 11 years), but the retrospective cross-sectional design cannot establish that education caused the difference.²⁹ In an Egyptian quasi-experimental school study, 32 students were referred and eight were reported as confirmed on medical consultation after three months, although diagnostic criteria and elapsed-time outcomes were not reported.³⁰ None of these studies measured a prespecified pathway interval in a comparative implementation evaluation.

Structured consultation tools may help clinicians assemble symptoms that have appeared across several visits. In an Australian mixed-methods pilot, 12 GPs used the Raising Awareness Tool for Endometriosis with 111 patients. The tool prompted discussion, investigation and follow-up and was considered acceptable, but diagnostic yield, elapsed time, false-positive decisions and downstream workload were not established.³¹ Qualitative studies support clearer guidance, training and regional advice networks,^23,24 yet they do not show that these components change pathway times.

Electronic prompts could use longitudinal patterns rather than a single symptom. The record associations reported by Burton and colleagues are plausible candidates for such prompts,²⁵ but their clinical effect remains untested. Any trial should examine appropriate suspicion and referral, missed competing diagnoses, alert burden and differential performance across coding practices and population groups.

Symptom Questionnaires, Prediction Scores and Digital Tools

Patient-completed tools can make a fragmented history easier to see. In a diagnostic study of 302 participants, a low-risk questionnaire threshold had 98% sensitivity and 70% specificity, while a high-risk threshold had 62% sensitivity and 98% specificity.³² The thresholds may assist triage, but performance in a selected diagnostic cohort cannot be assumed in adolescents, primary care or lower-prevalence populations.

The Simplified Adolescent Factors for Endometriosis (SAFE) score was developed in 4005 participants, internally tested in 1335 and evaluated in a separate Australian cohort of 4077. The area under the curve fell from 0.79 internally to 0.71 in the separate cohort; at a score of at least two, sensitivity was 46.7% and specificity 85.7%.³³ A low score would therefore be unsuitable for reassurance.

Digital symptom checkers face a similar translation problem. A developer-funded study using 24 fictitious clinical vignettes reported sensitivity of 81.8% and specificity of 84.6% for suggesting endometriosis, but vignettes are more orderly than real consultations and do not measure access or user behaviour.³⁴ An eight-week, single-arm Brazilian primary-care study of EndoConnect with 60 participants across 10 units reported feasibility and engagement rather than causal pathway change.³⁵ A company-funded Markov model, partly informed by company data, projected a 4.36-year reduction in diagnostic time with a digital symptom checker, alongside gains in quality-adjusted life expectancy and lower costs.³⁶ These are modelled outcomes that depend on assumptions about uptake, repeated use, algorithm performance and clinician response. More broadly, a scoping review found that the measurement properties of many screening measures were incompletely characterised and that evidence of clinical utility was limited.⁴⁵ A review of artificial intelligence reached a similar practical conclusion: tool performance is often reported, whereas prospective evidence of clinical impact remains scarce.⁴⁶

Working Clinical Diagnosis and Empirical Management

The move away from obligatory laparoscopy before treatment is a major change in the pathway. ACOG states that symptoms and clinical assessment can support a diagnosis sufficient to begin empirical treatment, although the recommendation is based on low-quality evidence.¹⁷ ESHRE, NICE, RANZCOG and SOGC also support clinical and imaging routes, investigation and treatment in parallel, and selective use of surgery.^18–21 An industry-supported expert call to action argued that withholding a diagnosis until operative confirmation could delay useful care.⁴⁷

This guideline convergence supports practice change, but it is not comparative evidence that one service model shortens a defined interval. A cross-sectional US online survey associated diagnosis by an obstetrician–gynaecologist with a shorter mean consultation-to-diagnosis interval than diagnosis by another clinician (21.5 vs 40.3 months), but it did not compare clinical with surgical diagnostic routes and is vulnerable to recall, case-mix and access differences.¹⁰ A working diagnosis should therefore remain open to revision and should include review, safety-netting and access to further assessment.

Ultrasonography and Magnetic Resonance Imaging

The term “ultrasonography” covers different examinations. Routine pelvic transvaginal ultrasonography provides a basic pelvic assessment and may identify endometriomas or other signs, but detection varies. Endometriosis-directed ultrasonography uses a structured protocol, relevant compartment assessment and dynamic manoeuvres performed by a trained operator. Expert mapping ultrasonography is used to define the location and extent of deep disease and support complex or operative planning. Magnetic resonance imaging is a complementary option when transvaginal ultrasonography is unsuitable or inconclusive, when local expertise favours MRI, or when further mapping is required. Performance depends on the operator, protocol, phenotype, lesion site, reference standard and clinical spectrum.

NICE and ACOG recommend ultrasonography as first-line imaging for suspected endometriosis, with transabdominal examination when a transvaginal examination is unsuitable or unacceptable.^17,19 Neither a normal ultrasound nor MRI reliably excludes superficial peritoneal disease, so negative imaging should not automatically close the pathway.^17–21

Selected studies show what greater expertise and standardisation may achieve. In a retrospective review of 40 surgically confirmed cases, routine transvaginal ultrasonography had 25% sensitivity compared with 78% for an expert-guided examination; because every participant had surgically confirmed disease, specificity could not be estimated.³⁷ The International Deep Endometriosis Analysis group proposed a systematic four-step examination and common terminology,⁴⁸ while the Society of Radiologists in Ultrasound recommended additions to routine pelvic ultrasound that include posterior-compartment assessment, uterine and ovarian positioning and the sliding sign.⁴⁹ An international consensus supports expert transvaginal ultrasonography and MRI specifically for pelvic deep disease and presurgical mapping, with performance varying by site and operator.⁵⁰ In a retrospective analysis of prospectively collected 2018–2019 data from a tertiary, surgery-enriched cohort of 467 participants, the complete four-step IDEA protocol had higher sensitivity but lower specificity than the basic first step, illustrating the trade-off between greater detection and false-positive findings.³⁸

The accuracy and consensus studies concern detection and mapping, particularly of ovarian and deep disease, in selected settings; alone, they do not establish shorter diagnostic intervals. Surgical enrichment, verification bias and variability in operator skill limit transferability,^51,52 and expert imaging may itself become a source of delay if capacity is limited. One single-centre historical before-and-after cohort of 444 adults found that introducing an endometriosis-specific MRI protocol was associated with a twofold reduction in diagnostic laparoscopy and a 21.1-percentage-point increase in medical management.³⁹ This study showed a change in diagnostic route and treatment, but it did not measure elapsed diagnostic time, patient outcomes, equity or resource use; secular change cannot be excluded.

Service Redesign and Policy

Rapid-access clinics, navigation, advice-and-guidance, one-stop assessment and regional networks are plausible ways to address fragmentation. We did not identify a controlled evaluation showing that one of these service models shortened a prespecified interval. They should therefore be introduced with a logic model, capacity planning and prospective measurement rather than presented as established solutions.

A retrospective single-centre evaluation of the French national plan compared moderate and severe disease diagnosed before and after February 2022. Median reported delay was 5.7 years before and 4.2 years afterwards, a difference that was not statistically significant.⁴⁰ Recall, case mix, secular trends and limited power restrict causal interpretation. The study shows why publication of policy should not be treated as evidence of implementation; workforce, imaging capacity, referral agreements and routine measurement also matter.

Biomarkers and Emerging Diagnostics

Biomarkers offer the prospect of a non-surgical result, but pathway value cannot be inferred from diagnostic accuracy alone. A prospective 17-centre French study of a salivary microRNA signature enrolled 971 symptomatic participants and reported 97.3% sensitivity and 94.1% specificity against a composite imaging or surgical reference.⁴¹ The population had a 77% disease prevalence and was drawn from referral settings. Performance, clinical utility and cost in primary care remain uncertain, and the assay does not map lesion location or surgical complexity. A gastrointestinal myoelectrical activity test also reported high age-stratified accuracy in a selected open-label study, but only two symptomatic participants were disease-negative; specificity was therefore imprecise, thresholds were derived in small samples, reference standards were mixed and the investigators disclosed relevant commercial interests.⁴²

At the search cut-off, NICE had published draft early-use guidance proposing that Endotest and EndoSure could be used for three years while evidence is generated in a defined primary-care pathway after normal examination and negative, inconclusive or unsuitable imaging.⁵³ The technologies were not proposed as standalone tests. This provisional approach differs in purpose from routine diagnostic recommendations and does not establish clinical effectiveness.

Other technologies remain earlier in development. The multi-omic HerResolve assay reported an area under the curve of 0.944, sensitivity of 80% and specificity of 97.5% in a balanced, independent retrospective validation cohort within a 298-participant case-control study.⁴³ In the exploratory DETECT Phase 2 study, ^99mTc-maraciclatide SPECT-CT was concordant with surgery in 16 of 19 evaluable participants, but only two participants were disease-negative.⁴⁴ Both studies had industry involvement. These estimates require independent prospective testing in the intended population, followed by diagnostic-impact, economic and equity evaluation.

An Author-Proposed Framework for Prospective Implementation and Evaluation

The following framework combines guideline-consistent clinical practices with author-proposed service components. It has not been validated as a package and should not be interpreted as an evidence-proven intervention for shortening diagnosis. Services that adopt it should treat it as a learning-system intervention, with local adaptation, capacity planning, prospective measurement and prespecified balancing outcomes.

Guideline-Consistent Clinical Principles

• Ask about functional effect as well as symptom presence, including school or work absence and cyclical bowel, urinary and sexual symptoms.

• Assess urgent and credible competing diagnoses proportionately, recognising that conditions may coexist.

• Offer appropriate analgesic or hormonal treatment while arranging investigation or referral; response to treatment should not be used as a diagnostic test.

• Provide competent first-line imaging, with the examination route and protocol chosen according to clinical need and patient acceptability.

• Explain the limits of negative imaging and define the next review or escalation point when symptoms persist.

• Use a working clinical diagnosis when justified, communicate uncertainty and retain access to imaging, surgery or other assessment when needed.^17–21

Proposed Service Components

• Record the diagnostic basis, route and certainty, together with the date on which the diagnosis was communicated.

• Track referrals to acceptance and completion, including rejected, redirected or incomplete referrals.

• Match the level of assessment to complexity. Primary care and general gynaecology may manage common presentations; expert imaging and specialist services should remain accessible for suspected deep disease, organ involvement, refractory symptoms, complex fertility questions, surgery or unresolved uncertainty.

• Monitor segmented intervals, diagnostic appropriateness, patient experience, treatment access, workload, cost and equity, and revise the pathway when improvement in one part creates pressure elsewhere.

This tiered arrangement is a proposal for local testing, not a universal organisational model. Health-system structure, workforce, geography and patient preference will determine how its functions are distributed.

Potential Harms, Capacity and Equity

Premature Closure and False Reassurance

A faster pathway can still provide poor care if pelvic symptoms are attributed too quickly to endometriosis. Pelvic floor dysfunction, adenomyosis, gastrointestinal and bladder disorders, central sensitisation and other conditions may coexist or offer a better explanation. A working diagnosis should therefore preserve credible alternatives and be reviewed when the clinical course is atypical or management does not help. The Endometriosis Treatment Italian Club position statement’s caution that more intervention is not invariably better applies to repeated testing and surgery as well as treatment.⁵⁴

Negative tests create a related risk. Superficial disease may not be visible on routine ultrasonography or MRI, and low-risk symptom scores may have insufficient sensitivity for exclusion. Reports, referral policies and patient information should state what a negative result can and cannot rule out. Diagnostic studies should follow false-negative and indeterminate groups rather than report discrimination alone.^51,52

Capacity and Opportunity Cost

More sensitive tools may increase referrals, repeat consultations, imaging and surgery unless triage and treatment capacity expand. The economic question concerns the incremental cost per appropriate diagnosis and the budget effect of the whole pathway, rather than the price of a test in isolation. Longer queues for people with organ-threatening disease or other causes of pelvic pain are genuine balancing harms. Diagnostic delay has been associated with substantial healthcare use and cost,⁵⁵ but modelled savings should be distinguished from observed effects.

Equity

Access to surgery, imaging, referral and clinician recognition shapes the apparent prevalence and presentation of endometriosis across racial and ethnic groups.⁵⁶ In a large US electronic-record analysis, matched Black patients had a median coded interval of 1.34 years compared with 0.67 years for White patients; Hispanic patients also had longer intervals than non-Hispanic patients.⁵⁷ These records do not measure symptom onset and residual confounding is likely, but they identify inequities that a new pathway could reproduce.

Symptoms beginning before the age of 20 have been associated with longer diagnostic intervals, and specialist studies continue to report delayed assessment in adolescents and young adults.^58,59 Rurality, socioeconomic disadvantage, disability, language, migration status, insurance and gendered assumptions may affect several transitions. Services should select priority equity measures with the communities they serve, stratify major process outcomes and examine whether improvement in the average conceals deterioration in a less well-served group.

A Candidate Outcome Framework for Pathway Studies

Endometriosis already has an internationally agreed core outcome set for treatment research.⁶⁰ Diagnostic-pathway studies need complementary measures of time, appropriateness, experience, implementation and system effects. Table 3 sets out candidate domains for that purpose. This author proposal has not undergone core-outcome-set methods, patient and stakeholder co-design, Delphi consensus or prospective validation. It is intended as a starting point for formal agreement rather than a minimum standard.

Table 3.

Candidate Outcomes for Evaluating Endometriosis Diagnostic-Pathway Interventions

Domain	Candidate Outcomes	Reporting Considerations
Time	The intervention-relevant interval; symptom onset to prespecified diagnosis; first relevant presentation to diagnosis; diagnosis to agreed management	Define every milestone and diagnostic route; report median, IQR and longer-tail measures; distinguish patient-reported and record-derived dates.
Segmented pathway	Recognition to help-seeking; first presentation to recognition/plan; plan or referral to further assessment; assessment to communicated diagnosis; diagnosis to management	Report the segment targeted, consultation count, emergency visits, rejected referrals, attrition and censoring.
Diagnostic appropriateness	Credible alternative or revised diagnoses; false-positive, false-negative and indeterminate results; repeat imaging; laparoscopy; complications	Prespecify reference and follow-up; retain imaging-negative and screen-negative groups; consider independent adjudication.
Patient-centred outcomes	Feeling heard and validated; shared decisions; diagnostic confidence; pain, function, quality of life, distress and school/work participation	Use validated measures where available; collect baseline and follow-up data; report intervention burden and acceptability.
Treatment and continuity	Time to appropriate analgesic/hormonal treatment, fertility care, pelvic-health physiotherapy, pain service or surgery; named clinician; pathway completion	Distinguish care offered, accepted and received; record changes in management and diagnostic confidence.
Implementation	Reach, adoption, fidelity, acceptability, feasibility, cost, penetration and sustainability	Describe components and context; assess maintenance beyond initial implementation where feasible.
Health system and economics	Waiting time, queue length, referral volume, staff time, direct/indirect cost, cost per appropriate diagnosis, QALYs and budget impact	Include downstream tests, consultations and treatments, not only intervention price.
Equity and balancing outcomes	Prespecified priority-group outcomes; missed alternatives, anxiety, overdiagnosis, unnecessary procedures and opportunity cost	Examine differential missingness and performance; report absolute and relative differences without treating group categories as biological explanations.

Abbreviations: EHP, Endometriosis Health Profile; IQR, interquartile range; QALY, quality-adjusted life-year.

Time outcomes should be tied to explicit milestones and reported as medians and interquartile ranges, with the longer tail of the distribution also shown. A shorter interval should be interpreted alongside diagnostic revisions, missed alternatives, treatment access, patient experience and resource consequences. The route to diagnosis and whether dates are patient-reported or record-derived should always be stated.

Research Priorities

Evaluate Pathways Rather Than Isolated Products

A high-priority study would be a pragmatic multicentre cluster-randomised or stepped-wedge evaluation of a primary-to-specialist pathway. Possible components include clinician education, a structured history or record prompt, access to endometriosis-directed ultrasonography, explicit escalation criteria, a working diagnosis with safety-netting and support for incomplete referrals. The comparator should be usual care with the same outcome ascertainment. The primary time outcome should correspond to the interval the intervention is designed to change; diagnostic appropriateness, patient experience, specialist demand, cost and equity should be prespecified key outcomes. Factorial or adaptive designs could help identify which components add value where the setting and sample size allow.

Require Diagnostic-Impact Evidence

Before a symptom score, artificial-intelligence model, biomarker or molecular-imaging test enters routine care, it should undergo independent validation in the intended population, followed by a prospective study showing how results alter appropriate clinical decisions. Comparative pathway and economic evaluations should then assess time, outcomes, downstream testing, resource use and equity. Thresholds must reflect the intended role - rule-out, rule-in or triage - and the consequences of false results. Industry funding does not invalidate a study, but protocol registration, independent analysis, complete reporting and access to data and code become particularly important.

Use Endpoints That Remain Interpretable as Practice Changes

Clinical diagnosis, imaging quality, coding and public awareness can change during a study. Trials should define the endpoint in advance, report every diagnostic route and follow participants long enough to identify revised diagnoses. Routine-data studies should separate symptom codes, first documented recognition, imaging findings, communicated diagnosis and surgical or histological confirmation. Where randomisation is not feasible, controlled interrupted time-series or difference-in-differences designs across regions are more informative than an uncontrolled before-and-after audit.

Co-Design What Counts as Success

Published qualitative evidence suggests that patients value being believed, receiving a clear explanation and gaining access to useful care.^15,16 Clinicians must also avoid premature closure and unnecessary procedures. Formal consensus work should define acceptable uncertainty and appropriate actionability with patients, primary-care clinicians, gynaecologists, radiologists, pathologists, methodologists and health-system leaders. Adolescents, ethnically diverse communities, rural populations, disabled people, and transgender and gender-diverse people should be involved from the outset. Table 4 summarises study designs that could address the main evidence gaps.

Table 4.

High-Priority Studies for Closing the Evidence Gap

Question	Preferred Design	Decisive Outcomes
Does a coordinated primary-care pathway improve timeliness?	Multicentre cluster-randomised or stepped-wedge pragmatic trial	Target interval, time to management, diagnostic appropriateness, patient experience, referral burden, cost and equity.
Does school education change real care?	Cluster trial with consented linkage to healthcare data and longer follow-up	Help-seeking interval, appropriate consultation/referral, anxiety, school participation and time to management.
Does an EHR prompt improve recognition without alert fatigue?	Cluster trial embedded in primary-care records	Appropriate recognition/referral, missed alternatives, alert burden, clinician uptake and subgroup performance.
Does endometriosis-directed first-line imaging improve the pathway?	Regional controlled rollout or randomised imaging-network strategy	Imaging wait, repeat scans, diagnostic route, planning, false reassurance, target interval and cost.
Does a biomarker or digital tool add clinical utility?	Prospective randomised diagnostic-impact trial in the intended setting	Appropriate management change, time, false results, downstream referrals/procedures, patient outcomes, cost and fairness.
Does national policy reduce avoidable delay?	Controlled interrupted time series or difference-in-differences across regions	Segmented intervals, capacity, diagnostic route, treatment initiation, longer-tail delay and inequity.

Synthesis and Limitations

Most empirical reports identified in this review measured knowledge, acceptability, diagnostic performance or modelled benefit. These outcomes are useful for development, but they cannot establish whether patients move through an appropriate pathway sooner or receive better care. Current guideline convergence supports clinical assessment, competent first-line imaging, empirical management and selective surgery.^17–21,61 The comparative effectiveness of any particular service configuration remains uncertain.

A strength of this review is the consistent separation of pathway outcomes from proxy measures, together with explicit attention to diagnostic appropriateness, capacity, cost and equity. Its limitations are substantial. The search relied on one bibliographic database, targeted guideline repositories and citation chaining. Retrieval was iterative rather than exhaustive, selection and charting were not duplicated, and no formal risk-of-bias or certainty assessment was performed. Relevant implementation, education and health-services research may therefore have been missed, particularly unpublished work or studies not indexed in PubMed. The study-level table describes the empirical reports selected for discussion and is not a systematic inventory.

Populations, settings, starting points, diagnostic endpoints and reference standards varied too widely for quantitative synthesis. Selection of sources and assignment to the author-developed outcome stages required judgement; the stages describe what was measured, not study quality. Most delay studies include people who eventually received a diagnosis, thereby excluding those still undiagnosed and introducing recall and selection bias. Transferability across health systems and underserved populations is uncertain. Finally, the actionable-diagnosis criteria, segmented pathway, service framework and candidate outcomes are author proposals. They require patient and stakeholder co-design, formal consensus and prospective validation. Statements about evidence that was not found apply only to this targeted search.

Conclusion

Current guidelines no longer require laparoscopy before a working diagnosis in every patient. In the sources identified, however, controlled evidence that an intervention shortens its intended interval remained limited. Educational studies reported knowledge gains, and symptom tools, expert imaging and emerging diagnostics showed potential in selected settings. One historical imaging-protocol comparison altered the diagnostic route and management without measuring elapsed time. We found no evaluation that considered timeliness alongside diagnostic appropriateness, patient outcomes, resource use and equity. These findings are an evidence map, not a definitive estimate of effectiveness.

We propose, rather than validate, a coordinated framework comprising structured symptom assessment, parallel investigation and treatment, competent first-line imaging, follow-up after negative imaging, a working diagnosis when appropriate and referral followed through to completion. Pragmatic comparative studies should measure the interval targeted and report diagnostic revision, treatment initiation, patient experience, service demand, cost and equity. The goal is timely access to a clinically sound explanation and useful care.

Funding Statement

The authors received no financial support for the research, authorship or publication of this article.

Generative Artificial Intelligence

During preparation of this manuscript, the authors used OpenAI ChatGPT (version GPT-5) and OpenAI Codex (version GPT-5) solely to support English-language editing and reference formatting, with the aim of improving clarity, consistency and conformity with journal style. These tools were not used to generate or analyse patient data or to make scientific interpretations. The authors independently reviewed and verified all AI-assisted material and every cited reference, confirm that all manuscript content is original and accurate, and take full responsibility for the integrity of the manuscript, including the accuracy of its references. The authors also reviewed the applicable terms of use for both tools and confirm that their use is suitable for publication.

Abbreviations

ACOG, American College of Obstetricians and Gynecologists; AUC, area under the curve; EHR, electronic health record; EHP, Endometriosis Health Profile; ESHRE, European Society of Human Reproduction and Embryology; GP, general practitioner; IDEA, International Deep Endometriosis Analysis; IQR, interquartile range; MRI, magnetic resonance imaging; NICE, National Institute for Health and Care Excellence; QALY, quality-adjusted life-year; RANZCOG, Royal Australian and New Zealand College of Obstetricians and Gynaecologists; RATE, Raising Awareness Tool for Endometriosis; SAFE, Simplified Adolescent Factors for Endometriosis; SOGC, Society of Obstetricians and Gynaecologists of Canada; SPECT, single-photon emission computed tomography; SRU, Society of Radiologists in Ultrasound; TVUS, transvaginal ultrasonography; US, ultrasonography; WHO, World Health Organization.

Data Sharing Statement

Data sharing is not applicable because no datasets were generated or analysed.

Ethics Approval and Informed Consent

Not applicable. This narrative review analysed published literature and did not involve human participants, identifiable human data, human tissue or animals.

Consent for Publication

Not applicable. The article contains no identifiable individual data, images, videos or recordings.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors report no financial or non-financial competing interests in this work.