A RECENT commentary by Australian authors, published in the BMJ, which proposed overdiagnosis of polycystic ovary syndrome (PCOS), drew on very limited evidence and did not engage affected women.

Along with related media attention, the BMJ article arguably adds unnecessary confusion to a field in which researchers, clinicians and patient advocacy groups have worked together internationally to deliver constructive, evidence-based messages aiming to provide clarity for women with PCOS and their health care providers. We have brought together patient advocates, health care providers and academics to counter this largely non-evidence based perspective, drawing on evidence regarding PCOS prevalence, diagnosis, complications and distress. We also outline how improvement in health care and support for women with PCOS will be aided by initiatives to develop the first international evidence-based guideline and comprehensive knowledge translation program for PCOS.

Background

PCOS has long been recognised as the primary cause of oligomenorrhoea, amenorrhoea and anovulatory infertility, and is associated with higher risks of metabolic complications. It is now known that PCOS also has psychological impacts and is associated with pregnancy complications. Examples include a higher prevalence and severity of anxiety and depression symptoms and higher prevalence and earlier age of onset of type 2 diabetes mellitus. Overweight and obesity worsen PCOS and lifestyle management is a first-line management strategy for improving reproductive and metabolic features.

The original National Institutes of Health (NIH) PCOS diagnostic consensus required oligoovulation or anovulation and clinical or biochemical hyperandrogenism with exclusion of other aetiologies (Zawadzki J, Dunaif A. Diagnostic criteria for polycystic ovary syndrome: towards a rational approach. In: Dunaif A, Givens J, Haseltine F, Marrian G, editors. Current Issues in Endocrinology and Metabolism: Polycystic Ovary Syndrome Boston, USA: Blackwell Scientific; 1992. p. 377-384). These criteria are consistently recommended in adolescence, where inaccuracy and inappropriateness of vaginal ultrasound is well recognised. However, in adult women, the internationally endorsed diagnostic criteria (including by the NIH) are the Rotterdam criteria, which require two of three characteristics: oligoovulation or anovulation, clinical or biochemical hyperandrogenism and polycystic ovaries on ultrasound, with exclusion of other causes. The Rotterdam criteria give rise to multiple PCOS phenotypes, which in a clinical context reflect a woman’s individual symptoms, experience and potential PCOS complications. The advent of multiple phenotypes has generated controversy, and a greater understanding of the natural history of PCOS phenotypes is a major research priority. It is in the context of broadened diagnostic criteria and multiple phenotypes that Copp and colleagues argued that PCOS is overdiagnosed and recommended a delayed approach to diagnosis. Yet, considerable evidence and patient experience show that under-recognition is the overwhelming challenge, and that women with PCOS want prompt diagnosis and education (see Box below).

PCOS prevalence

Copp and colleagues suggest that the introduction of the Rotterdam criteria has driven a dramatic increase in prevalence from 4–6% to 21%, arguing that this increase provides evidence of overdiagnosis. This selective reporting captures extreme estimates from diverse and uncomparable populations. The estimate of 21% is drawn from a study involving Aboriginal and Torres Strait Islander women. This study reported a very high prevalence (15%) under the original NIH criteria, a marginal increased prevalence with the application of modified Rotterdam criteria, and recognised the primary drivers of prevalence were ethnicity and obesity. Insulin resistance is involved in the aetiology of PCOS and factors that affect insulin resistance, such as ethnicity and increasing prevalence of obesity, will contribute to variability in estimates of PCOS prevalence. Also, rather than overstating the impact of the Rotterdam criteria on prevalence, we cite the recent systematic review in this area, showing a more modest variation in prevalence: 6% (95% confidence interval, 5–8%) from studies applying the NIH criteria and 10% (95% confidence interval, 8–13%) from studies applying the Rotterdam criteria. We contend that variation in reported PCOS prevalence is multifactorial and cannot be used as direct evidence of overdiagnosis.

Diagnosis

On the basis of proposed overdiagnosis, Copp and colleagues recommend “a slower, stepped or delayed approach to diagnosis to optimise benefits and reduce harm from disease labelling”. Yet, there is no direct evidence of overdiagnosis, or of harm from diagnosis. On the contrary, evidence suggests that PCOS is under-recognised and that delays cause harm. Sivayoganathan and colleagues reported that 34% of women who were later confirmed to have PCOS did not receive a diagnosis despite attending infertility, gynaecology, endocrinology or dermatology clinics with PCOS clinical features. In a primary care setting, 50% of probable cases didn’t have a recorded diagnosis. March and colleagues reported that in a community setting, screening showed that 68% of women with PCOS were undiagnosed. While the latter is suggestive of under-recognition of PCOS, it is important to note that this study relied on community-based screening case finding, where symptoms may be minimal. Consistent with these findings, current guidelines do not recommend community-based screening for PCOS. More compelling are the clinic-based studies, such as Sivayoganathan and colleagues, where women present with PCOS symptoms to services that commonly manage PCOS and, yet, a considerable proportion do not receive a diagnosis. It is also important to acknowledge that PCOS diagnostic features represent a continuum, just as they do in diabetes or hypertension and establishing ideal cut-offs for each feature is challenging. Furthermore, inconsistent application of the diagnostic criteria may drive misdiagnosis, and may be more likely to occur in women with polycystic ovarian morphology on ultrasound, but no other features of PCOS. These issues will not be addressed by claiming overdiagnosis and recommending delayed diagnosis, but rather by research, consistent practice based on evidence-based guidance and education.

It was also suggested that women with non-hyperandrogenic PCOS phenotypes may experience unnecessary distress if diagnosed, potentially leading to inaccurate perceptions of risk of long term metabolic complications. While research to further characterise the natural history of PCOS and long-term health outcomes is essential across phenotypes, it is notable that insulin resistance is present in the overwhelming majority of both lean and overweight young women with PCOS, with current evidence suggesting little impact of different diagnostic criteria. A recent review, which incorporated systematic review evidence, concluded that PCOS status confers cardiometabolic and diabetes risk, and differences across phenotypes are largely obesity related. Current guidelines therefore recommend PCOS diagnosis across all phenotypes and cardiometabolic screening for all women with PCOS, and until evidence shows risk divergence by phenotype, these guidelines should direct diagnosis and risk assessment.

Distress

Cautious diagnosis is the current internationally recommended approach for adolescents, for whom a PCOS diagnosis is not recommended within 2 years of menarche and diagnostic ultrasound is not advised. Diagnosis in adolescence requires both oligoovulation and hyperandrogenism, with follow-up to confirm persistence of clinical features longer term. However, suggesting a delayed approach to diagnosis for all women is inappropriate. There is no evidence that women with PCOS would prefer a delayed diagnosis, but there is compelling, clear and consistent evidence from affected women around the world showing unacceptable delays in diagnosis with adverse impact.

Women currently report seeing multiple health care providers about their PCOS symptoms before receiving a diagnosis. Women have to raise their concerns on multiple occasions (here and here) and experience frustrating delays prior to diagnosis (here, here and here). Delayed diagnosis is associated with anxiety and depression symptoms. In a longitudinal, community study of over 11 000 women, those with PCOS reported more psychological distress than those without PCOS. This distress was equally high in the year before the diagnosis as afterwards, suggesting that distress was related to PCOS features rather than the diagnosis label. Establishing a diagnosis is important to women and many report feeling relief at this time. However, receiving a diagnosis can also lead to anxiety or lack of control in the absence of appropriate support and information. Women’s initial source of information is their health care provider, who can provide personally relevant information, enabling better management of the condition. To propose delaying diagnosis based on unfounded arguments of distress with diagnosis and on whether long term metabolic features of PCOS may be limited in some phenotypes disregards the clearly reported experiences of poor diagnosis experienced by women with PCOS.

Stephen Hawking recently noted, in relation to health, that: “For a scientist, cherry-picking evidence is unacceptable. When public figures abuse scientific argument, citing some studies but suppressing others to justify policies they want to implement for other reasons, it debases scientific culture. One consequence of this sort of behaviour is that it leads ordinary people to not trust science at a time when scientific research and progress are more important than ever”.

We would likewise contest that for academics to selectively present evidence to propose a contention that disregards compelling evidence on patient experience is unacceptable. Also, to suggest simply treating symptoms without providing women with a diagnosis, or opportunity for education and consideration of often neglected features of PCOS, is challenging in the current era of shared decision making and health care as a partnership between health care providers and patients.

Moving forward

The time has come for consistent, evidence-based approaches to PCOS, developed in partnership with women and health care providers. A better understanding of PCOS features in adolescents and of the long term health outcomes across the PCOS phenotypes is essential. However, in the absence of evidence of overdiagnosis, or of adverse impact of diagnosis, and in the presence of evidence of under-recognition and unacceptable and distressing delays in diagnosis, we contend that suggestions to delay PCOS diagnosis in adult women and treat without a diagnosis are inappropriate and serve only to increase confusion.

In partnership with women with PCOS, we are leading a unique international PCOS initiative seeking to address challenges in diagnosis and inconsistency in care. Built on international engagement of thousands of health care providers and women, the first international evidence-based guideline for PCOS will be launched in mid-2018 and will be accompanied by an extensive implementation and translation program. This initiative is led by the National Health and Medical Research Council funded Australian Centre for Research Excellence in PCOS, in partnership with the European Society of Human Reproduction and Embryology and the American Society for Reproductive Medicine, with 31 other societies and patient advocacy groups internationally.

The initiative will be evaluated for impact on measures of direct importance to women and health care providers. Research priorities will emerge, including around the diagnostic criteria, along with a case for action for greater research investment in this previously neglected condition. We contest that this is a more constructive and appropriate approach that acknowledges the current distress and suboptimal experience widely reported by women with PCOS and aims to improve diagnosis, experiences, health care and outcomes for women with PCOS.

Melanie Gibson-Helm is an NHMRC research fellow at the Monash Centre for Health Research and Implementation.

Professor Robert Norman is Professor Reproductive – Periconceptual Medicine at the Robinson Research institute at the University of Adelaide.

Professor Helena Teede is an endocrinologist, Professor of Women’s Health, and Director of the Monash Centre for Health Research and Implementation, and an NHMRC Practitioner Fellow. All correspondence to helena.teede@monash.edu

 

Box

by Veryan McAllister, President of the Polycystic Ovary Syndrome Association of Australia

PCOS overdiagnosis claim "not based on evidence", say experts - Featured Image

The assertion from Copp and colleagues that PCOS is an over diagnosed condition, where patients receive no benefit from diagnosis, and advise delaying diagnosis and starting treatment anyway is extremely upsetting to read from a patient and patient advocacy perspective. This is the opposite to international evidence from women affected by PCOS and appears to seek to grandstand an opinion without substance.

As President of the patient support group Polycystic Ovary Syndrome Association of Australia, and as a patient who had to see multiple health care providers, my own diagnosis was delayed 6 years and caused extreme anxiety. Sadly, my story is far from isolated. The labelling of the disease did not provide distress, rather having a name to the problem helped me to realise it wasn’t fiction; there were steps I could take to control it. While this is but one patient story, it is supported by evidence on women’s experiences from around the world.

Ten years ago, no-one talked about PCOS, and there was very little research. The biggest challenges we have faced as patients were people talking about the condition, having consistent and timely diagnosis, and access to appropriate management of symptoms. We have lobbied for funding and worked closely with researchers and health care professionals as members of the PCOS Australian Alliance and the Centre for Research Excellence in PCOS. We partnered in the development of the original Australian guidelines and now work closely, with other support groups from around the world — as members of the National Health and Medical Research Council funded initiative to foster international engagement, guideline development and evidence translation. This work is far from opinion-based and is founded on evidence-based research, together with patient experience and health professional experts, and is a constructive and exciting opportunity to improve health care for women with PCOS and ensure they receive appropriate information. We support the major need for increased research in partnership with patients, noting that PCOS is very neglected in research funding.

Patient groups are working hard to increase awareness of PCOS and overcome social taboos around the symptoms; the Copp and colleagues article did not assist our aims and did not consider or consult with women affected with PCOS. We aim for more awareness, which will increase diagnosis (and we welcome this), hand in hand with ongoing research to improve diagnosis and education to reassure and inform women about PCOS, which will be enabled through the international PCOS initiative currently underway.

 

To find a doctor, or a job, to use GP Desktop and Doctors Health, book and track your CPD, and buy textbooks and guidelines, visit doctorportal.

 

4 thoughts on “PCOS overdiagnosis claim “not based on evidence”, say experts

  1. Tessa Copp says:

    Our paper aimed to investigate areas of potential overdiagnosis of PCOS, and consider the evidence and uncertainty surrounding diagnosis of this condition (2). It is widely acknowledged that the Rotterdam criteria resulted in the inclusion of milder phenotypes within the broadened spectrum of PCOS (3,4). At least 8 studies have examined the prevalence of PCOS diagnosis in head to head comparisons from the same study populations and have shown a substantially increased prevalence using the Rotterdam criteria versus other criteria (see Table 4 of our paper)(2). The systematic review quoted by Gibson-Helm and colleagues compares the three sets of criteria using studies conducted in different populations (5). These summary estimates of prevalence are likely to be confounded by the ethnicity and age of the different populations, and therefore may not be reliable for understanding the impact of changes in the diagnostic criteria on prevalence.

    PCOS is a complex and heterogeneous syndrome on a continuum of severity, and in our paper we call for a more tailored, patient-centred approach to diagnosis that weighs up the benefits and harms for each individual woman, wherever her experience sits on this spectrum. We do not recommend delaying treatment for symptoms that are distressing, nor do we suggest delayed diagnosis for women with severe forms of PCOS. We also suggest that while distressing symptoms should be treated without delay, clinicians might consider delaying the use of the PCOS label for women at high risk of overdiagnosis (e.g. adolescents, young women with mild symptoms and women with non-hyperandrogenic phenotypes). If there is any change in symptoms, or additional symptoms are noted at future visits, the diagnosis of PCOS can be confirmed and detailed counselling offered.

    We agree that under-recognition of PCOS can cause distress for affected women, and that research is urgently needed to improve the diagnostic criteria and patient experience. However, we are concerned that current diagnostic criteria likely include women with symptoms that are mild. A few studies in different community based populations examining the prevalence of PCOS by age found that prevalence fell considerably after the age of 25, suggesting that the condition can improve with time (6-8), whereas studies recruiting clinic based subjects show improvement in symptoms usually by the fourth decade of life (9,10). Although high quality longitudinal research in community samples is needed to investigate this further, this suggests that a milder form of the syndrome may exist, however the current diagnostic criteria classify women with both mild and transitory symptoms as having the condition. The benefits and harms of the diagnosis are unclear for this group.

    An increasing body of research in various medical conditions, such as hypertension, cancer and common paediatric conditions, such as reflux and conjunctivitis, suggests that there are psychological, social and behavioural harms associated with labelling people unnecessarily (11-13). Similarly in women with PCOS with mild symptoms, who are likely at very low risk of future illness (3,14), a permanent label may provide more harm than benefit, inducing unnecessary fear and anxiety about future fertility and long-term health. Further research to clarify the benefits and harms of the diagnosis for women particularly at the mild end of the spectrum is urgently needed, and we strongly advocate for this.

    We share the same goal with Gibson-Helm and colleagues – we want the best possible outcomes for women who are experiencing distressing symptoms. Unfortunately, some patients have had unfavourable experiences with healthcare providers related to delays in diagnosis and inadequate or inaccurate information about PCOS (15). Identification of accurate and reliable diagnostic cut offs for each of the PCOS criteria will maximise the benefits that come with appropriate and timely diagnosis of PCOS for women who are affected by significant symptoms, while minimising the harms that come with unnecessarily labelling of healthy women who are not likely to benefit. Given the spectrum of this syndrome, all women should receive personally relevant information that addresses their individual concerns. We are currently conducting research studies to determine the benefits and harms of diagnosis across the different PCOS phenotypes and spectrum of severity, including women with milder forms of PCOS who have been significantly underrepresented in research to date (4,16). Until there is better evidence regarding the relationship between the milder forms and future health risks, women need to be informed of the uncertainties involved with the current diagnostic criteria.

    References

    1. Gibson-Helm M, McAllister V, Norman R. J, et al. BMJ rapid response 13th September 2017. http://www.bmj.com/content/358/bmj.j3694/rr
    2. Copp T, Jansen J, Doust J, et al. Are expanding disease definitions unnecessarily labelling women with polycystic ovary syndrome? Bmj 2017;358:j3694.
    3. Lizneva D, Suturina L, Walker W, et al. Criteria, prevalence, and phenotypes of polycystic ovary syndrome. Fertility and sterility 2016.
    4. Azziz R, Carmina E, Chen Z, et al. Polycystic ovary syndrome. Nat Rev Dis Primers 2016;2:16057.
    5. Bozdag G, Mumusoglu S, Zengin D, et al. The prevalence and phenotypic features of polycystic ovary syndrome: a systematic review and meta-analysis. Human reproduction 2016;31(12):2841-55.
    6. Zhuang J, Liu Y, Xu L, et al. Prevalence of the polycystic ovary syndrome in female residents of Chengdu, China. Gynecologic and obstetric investigation 2014;77(4):217-23.
    7. Lauritsen MP, Bentzen JG, Pinborg A, et al. The prevalence of polycystic ovary syndrome in a normal population according to the Rotterdam criteria versus revised criteria including anti-Mullerian hormone. Human reproduction 2014;29(4):791-801.
    8. Tehrani FR, Rashidi H, Khomami MB, et al. The prevalence of metabolic disorders in various phenotypes of polycystic ovary syndrome: a community based study in Southwest of Iran. Reproductive biology and endocrinology : RB&E 2014;12:89.
    9. Carmina E, Campagna AM, Lobo RA. A 20-year follow-up of young women with polycystic ovary syndrome. Obstetrics and gynecology 2012;119(2 Pt 1):263-9.
    10. Pinola P, Piltonen TT, Puurunen J, et al. Androgen Profile Through Life in Women With Polycystic Ovary Syndrome: A Nordic Multicenter Collaboration Study. The Journal of clinical endocrinology and metabolism 2015;100(9):3400-7.
    11. Ogedegbe G. Labeling and hypertension: it is time to intervene on its negative consequences. Hypertension 2010;56(3):344-5.
    12. Pickering TG. Now we are sick: labeling and hypertension. Journal of clinical hypertension 2006;8(1):57-60.
    13. Nickel B, Barratt A, Copp T, et al. Words do matter: a systematic review on how different terminology for the same condition influences management preferences. BMJ open 2017;7(7):e014129.
    14. Daan NM, Louwers YV, Koster MP, et al. Cardiovascular and metabolic profiles amongst different polycystic ovary syndrome phenotypes: who is really at risk? Fertility and sterility 2014;102(5):1444-51 e3.
    15. Gibson-Helm M, Teede H, Dunaif A, et al. Delayed diagnosis and a lack of information associated with dissatisfaction in women with polycystic ovary syndrome. The Journal of clinical endocrinology and metabolism 2017:jc20162963.
    16. Lizneva D, Kirubakaran R, Mykhalchenko K, et al. Phenotypes and body mass in women with polycystic ovary syndrome identified in referral versus unselected populations: systematic review and meta-analysis. Fertility and sterility 2016.

  2. Anonymous says:

    Who are they kidding? Overdiagnosis? Not in my city, I’m sure. And THAT’S the problem. So please forge ahead with this very important work. There is, as stated, a sliding scale as to the severity eg having one or two of the symptoms, or all three, and of course there is a need to address all patients presenting with any number. But over diagnosis? I really think not. As for causing stress, would you not tell a patient with say, an auto immune disease, that they shouldn’t know. This is about lifting the stigma and treating patients asap, as the effects down the line will cause much more distress if unknown and therefore left untreated. Knowledge is power, even when it is not the most welcome knowledge…

  3. Julie Duffy Dillon says:

    Thank you for addressing this horrendous BMJ commentary. I found it to be harmful and full of misogyny. I’m glad to know most of the PCOS community is standing together on this. We wrote a reaction to the BMJ analysis here http://www.juliedillonrd.com/overdiagnosed

  4. Anonymous says:

    I think that it is important for the parents of adolescents to be able to get information from health care providers. It is unhelpful for parents to be told: you are not the patient, your daughter is, she doesn’t want to know about this now. As a parent it can be difficult getting satisfactory information to help deal with an adolescent with PCOS.

Leave a Reply

Your email address will not be published. Required fields are marked *