Preventing overdiagnosis: how to stop harming the healthy
BMJ 2012; 344 doi: https://doi.org/10.1136/bmj.e3502 (Published 29 May 2012) Cite this as: BMJ 2012;344:e3502
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Medical Technology; a friend and a foe
Moynihan raised an important issue regarding medicalisation which is very valuable due to the many reasons which may cause over-diagnosis and over-treatment and in other cases the opposite, under-diagnosis and under-treatment. First, medical technology is becoming more and more advanced and sometimes we as health care providers rely on technologies more than the symptoms that patients experience. In addition, we might not do a good history taking and a physical assessment to make a differential diagnosis prior to an advanced diagnostic study. Due to advanced medical technology ( e.g. ECMO), we might keep a patient with stable vital signs. However, it doesn’t mean he or she can recover but it may be a way to give hopes to the family members and allow them to overcome the grief process.
A profound faith in clinical guidelines might take away from individual needs and lead to misuse of medication since the human body is too complicated and responds quite differently under various conditions to use a standardized guideline. The development of a guideline is still valuable because a single problem or a certain condition might be appropriate to use as a guideline and also can be useful for educational purposes.
The health insurance system has a profound impact on the health care delivery system. An inappropriate reimbursement system might be one of the causes of over-treatment or under-treatment. Lastly, a mutidisciplinary team including policy makers, general practitioners, sub-specialty experts, nurse practitioners, nurses, other health care related personnel and patients should be aware of this important issue. One way to increase awareness is to receive constant education since drug treatment and technologies are always on the move and to develop appropriate strategies to reduce unnecessary medicalisation.
Competing interests: No competing interests
Congratulations to the authors on their lucid exploration of the problem. Clearly the ‘diagnosis’ of pseudo disease is always harmful, but they might have pursued the potential hazards of the downsides of the diagnosis of minimal disease opportunistically or by screening. The effectiveness of intervention is virtually impossible to assess. Most published studies are disease specific and rarely use the appropriate outcome measure for public health interventions, overall healthy survival. This approach is likely to bias studies in favour of intervention, which satisfies the author’s prejudices by promoting ‘their’ disease irrespective of any financial interest. This potential bias aside, the intervention is not, as in the treatment of a sick person, the therapeutic process, but screening itself. In treatment of the sick the placebo effect can always be assumed to be positive and hence its magnitude is immaterial. The effect of diagnosis by screening in disillusioning someone about his good health, particularly where he has been cajoled into it by family, his general practice or the government, may itself impinge adversely on healthy survival. If this negative placebo effect is greater then the therapeutic benefit, then harm is done despite apparent success in a comparative trial. The only way this problem can be solved is by comparing unscreened cohort with a screened one in a truly naïve population. This may now be impracticable in sophisticated societies, but the potential problem should be more widely recognised and investigated as far as is possible
Competing interests: No competing interests
I am surprised that Moynihan and colleagues can write a lengthy article on the medicalisation of society without acknowledging the contributions of Illich and Skrabanek in the seventies and eighties as well as others subsequently. Virtually all their 56 references date from 2000, taking care to include their own.
Competing interests: No competing interests
Perhaps the authors ought also to consider the risks of over-vaccination. I note that by 13 months a UK infant is now expected to have received 25 vaccines [1] despite which the Joint Committee on Vaccination and Immunisation contemplate adding Hepatitis B to the schedule as part of a hexavalent vaccine to be administered with Meningitis C vaccine [2] and a rotavirus vaccine [3]. Can anyone put their hand on their heart and say that they know what the total effect of this burden on our children will be, or that we have a generation of children without unexplained neurological impairments or an exceptional proliferation of chronic immune disorders?
Addionally, the JCVI are now proposing that all school children receive a nasal spray flu vaccination annually. Given the poor history of identifying correctly the proliferating flu strain of the following winter and the fact that the vaccine itself renders the children infectious [4] this seems a particularly good way to spread disease.
For whose benefit is this really?
[1] http://www.nhs.uk/Planners/vaccinations/Pages/Vaccinationchecklist.aspx
[2] http://www.ovg.ox.ac.uk/sites/default/files/6COM.pdf
[3] http://www.telegraph.co.uk/health/healthnews/9274847/All-babies-may-be-v...
[4] http://ec.europa.eu/health/documents/community-register/2011/20110127931...
Competing interests: Autistic son
There is an unrealistic societal preference for reducing risks irrespective of costs. The absurd interpretation of health and safety regulations is a manifestation of this societal attitude. For example, Whitehall staff, who are competent enough to run the country, apparently could not be trusted to put up jubilee decorations safely! (1)
When this ‘societal risk aversion’ meets ‘fee for service’ model, over-investigation and over-diagnosis occurs due to commercial interests and defensive medicine. (2) Hence, the new ‘Health and Social Care Bill’ in England has the potential to make the problem of over-diagnosis much more acute. (3)
References
1. Ahmed K. The Queen’s Diamond Jubilee: Whitehall told to remove bunting over H&S fears. Telegraph.co.uk [Internet]. 2012 Jun 1 [cited 2012 Jun 2]; Available from: http://www.telegraph.co.uk/news/uknews/the_queens_diamond_jubilee/930707...
2. Brody H. Medicine’s Ethical Responsibility for Health Care Reform — The Top Five List. New England Journal of Medicine. 2010 Jan 28;362(4):283–5.
3. Ham C. What will the Health and Social Care Bill mean for the NHS in England? BMJ. 2012 Mar 20;344(mar20 2):e2159–e2159.
Competing interests: No competing interests
The easy availability of pathology tests1 allows practitioners to screen well patients by requesting tests, in some cases before the patient is seen in clinic, known as a ‘fishing’ trip. Reference ranges are usually 95% of the normal population, therefore 1 in 20 results will be outside the reference range. A marginally ‘abnormal’ result may cause additional tests to be requested and if the result is again outside the reference range, then the test or test panel maybe routinely monitored every 3, 6 or 12 months.
Most practitioners do not understand that biological and analytical variation contributes to uncertainty of results. Education of practitioners in appropriate requesting has not proved to be rewarding. Appropriate test repeat intervals should be agreed for both inpatients & community patients. Investigation pathways should be established as it may be preferable for the requestor to write the problem on the request form & for the laboratory to do the necessary investigations. Some examples: protocols for monitoring of type 2 diabetes; statin, lithium, thyroxine and carbimazole therapy; subclinical hypo- or hyper-thyroidism repeat test advice; protocols for investigation of secondary amenorrhoea, infertility and hirsutes.
Commissioners could incentivise laboratories to help reduce test requests, and provide the backup to help with these initiatives.
Jeffrey Barron
Competing interests: No competing interests
Thank you for your paper about medicalisation. You described a number of reasons for this trend over the past years but there was one that you have missed. The reason you have missed may not be unique but I think and hope that it is very unusual. The reason I am going to describe to you which results in over-treatment is false teaching and therefore understanding by the vast majority of doctors, of a physiological process. You might think that a failure of many of us to understand properly some aspect of physiology is hardly unusual. The failure to teach correctly may be down to a matter of opinion. This aspect of physiology however is actually quite simple and the false teaching is virtually institutionalised. Equally, common sense tells us that the teaching not correct.
The physiology of transition at birth.
Here is the description in the latest publication of Ganong’s Review of Medical Physiology (23rd Edition edited by Barrett KE and Barman SM page 584).
Because of the patent ductus arteriosus and formen ovale (figure 3-18), the left heart and right heart pump in parallel in the fetus rather than in series as they do in the adult. At birth, the placental circulation is cut off and the peripheral resistance suddenly rises. The pressure in the aorta rises until it exceeds that in the pulmonary artery. Meanwhile, because the placental circulation has been cut off, the infant becomes increasingly asphyxial. Finally, the infant gasps several times, and the lungs expand. The markedly negative intrapleural pressure (-30 to -50 mm Hg) during the gasps contributes to the expansion of the lungs, but other factors are likely also involved. The sucking action of the first breath plus constriction of the umbilical vein squeezes as much as 100mL of blood from the placenta (the "placental transfusion").
I am sure you appreciate as well as I do that this is not a description of physiological changes at birth. The paper in the BJOG by Wiberg N et al DOI: 10.1111/j.1471-0528.2008.01708 clearly showed that there is no cut off of the placental circulation which they showed usually continued for 120 seconds. The Ganong author makes no attempt to explain the mechanism for the “cut off” of the placental circulation although they are quite confident to describe the effect of a sudden rise in the peripheral resistance. (Which has been demonstrated to occur with early cord CLAMPING) They then go on to explain that the infant becomes increasingly asphyxial, although Wiberg et al showed a steady rise in the arterial PO2 and a similar rise in venous PO2 up to 45 seconds after birth.
The statement that the infant gasps because of the asphyxia from the loss of placental circulation may be true sometimes (after cord clamping) but it would not true in a normal physiological birth and as shown by Wiberg et al since breathing is initiated long before there is any loss of the placental circulation, and it is gas exchange within the neonatal lungs which leads to the rise in umbilical artery PO2 .
The sequence of events may not be fully explained in the Ganong description but there is the implication that the sequence is in the same order in which they are described. It is therefore quite confusing, at the very least, to describe “constriction of the umbilical vein squeezes as much as 100mls of blood from the placenta.” when earlier on the author tells us that the placental circulation has been cut off. How has it opened up again? Even if it was open how does constriction of the umbilical vein squeeze blood from the placenta into the baby?
Let me just go back to one part of this description which we know must be incorrect. “At birth, the placental circulation is cut off and the peripheral resistance suddenly rises.” There is no question that at birth the placental circulation continues for several minutes. Not only is this common knowledge it has been demonstrated in many studies and discussion within the profession is not whether the circulation continues but whether it is better to allow it to continue.
This false teaching is common in most other textbooks in the description of physiological transition at birth. I am not going to list these here but you can easily verify it yourself.
How does this lead to over treatment and medicalisation ? The deep seated belief by the majority of obstetricians, midwives and paediatricians that the function and circulation of the placenta ceases the moment the baby is born trivialises the intervention of cord clamping. "It will stop in a minute or so anyway." "It does not serve any function once the baby is born." "Cord clamping is only speeding up physiology a little." The fact that it does not seem to do much harm encourages this view.
To justify the intervention it is often considered the norm and therefore not clamping needs to be proven to be safe or advantageous to the baby.
Perhaps you may be thinking that it is a rather trivial example and what harm can it be doing. Firstly it applies to virtually every birth so there are huge numbers involved. Even a small amount of harm will have enormous consequences. There is evidence of some harm from early cord clamping although this is not overwhelming but it is strong enough for a Cochrane review. Even if significant benefit from early cord clamping had been shown it would not justify distorting the description of physiology.
The description of physiology is confined to healthy term babies so the fact that preterm babies may or may not benefit from cord clamping management is again irrelevant to my case about physiology.
The fact that a number of trials of cord clamping in preterm babies ( such as the APTS in Australia) are underway, and the fact that early cord clamping is slowly being removed from UK routine birth practice, is the result of the evidence base. However as long as the teaching of physiology is distorted there is the chance that practice can revert back.
It would be so simple to provide a proper description of neonatal transition and highlight where there are uncertainties. There can be no uncertainty about the continued placental circulation for several minutes in a healthy term baby after birth.
Competing interests: No competing interests
The authors of this excellent review highlight a problem of enormous significance and list the drivers of overdiagnosis.
The authors mention over-inclusive guidelines as a cause. The three fold increase in the diagnosis of malignant melanoma since 1975 may be due to rising incidence (though the death rate has not risen) or it may be that primary care clinicians are expected to refer patients urgently on the basis of a clinical prediction rule (the “7 point score”) that has a specificity (assessed only twice and in pigmented lesions clinics) of between 30%(1) and 37%(2). General practitioners following this NICE guidance should be urgently referring approximately 2/3rd of all pigmented lesions they assess. Failure to follow this guidance is a frequent cause of litigation (with the NHSLA paying £863m in litigation costs in 2010/11). It is surely a problem that a clinical prediction rule in a national guideline that is effectively mandatory for all general practitioners has only had its performance characteristics checked on 265 patients in specialist clinics nearly a quarter of a century ago (and the results showed very low specificity).
There are numerous other examples of referral guidance that is effectively mandatory for general practitioners that either have no evidence base or have evidence that the clinical features are extremely poor discriminators with very poor predictive values.
Back pain that is not relieved by rest is found in nearly half of patients with back pain, yet it is a “Red Flag” (as is being over the age of 55).(3) Dysphagia is an indication for urgent referral yet questionnaire studies indicate the symptom is common in the background population (4) and in those with endoscopically proven gastro oesophageal reflux.(5) It maybe that the root to this anomaly, and many others, is that a volunteered symptom (“when I swallow, the food sticks”) is far more predictive of disease than an elicited one (“when you swallow, do you feel the food sticks? Yes, a bit.”).
The authors highlight an enormous and growing problem. It maybe that it would be wise that organisations such as NICE, before enshrining referral guidance in what is effectively a mandatory form, check what the evidence base is for the predictive value of symptoms and signs. Sometimes the results are very surprising. Almost all paediatric texts and much international guidance prior to 2011 gave “normal” values of paediatric respiratory rates and heart rates that were very inaccurate, with the “abnormal” ranges incorporating large proportions of healthy children in all age ranges.(6)
If the burden of over diagnosis and over investigation is going to be controlled then the problem of over inclusive referral guidance needs to be addressed.
Reference List
1. HIGGINS EM, HALL P, TODD P, MURTHI R, DU VIVIER AWP. The application of the seven-point check-list in the assessment of benign pigmented lesions. Clinical and Experimental Dermatology 1992;17:313-5.
2. HEALSMITH MF, BOURKE JF, OSBORNE JE, GRAHAM-BROWN RAC. An evaluation of the revised seven-point checklist for the early diagnosis of cutaneous malignant melanoma. British Journal of Dermatology 1994;130:48-50.
3. Deyo RA, Rainville J, Kent DL. What can the history and physical examination tell us about low back pain? JAMA 1992;268:760-5.
4. Lindgren,.et al. Prevalence of swallowing complaints and clinical findings among 50 to 79 year old men and women in an urban population. Dysphagia 1991;6:187-92.
5. Okamoto K,.et al. Clinical symptoms in endoscopic reflux oesophagitis: evaluation of 8031 adult subjects. Digestive Diseases and Sciences 2003;48:2237-40.
6. Flemming S, Thompson M, et al. Normal ranges of heart rate and respiratory rate in children from birth to 18 years of age: a systematic review of observationsl studies. Lancet 2011;377:1011-8.
Competing interests: No competing interests
Over-diagnosis can be due to failure to practice evidence-based medicine properly. The positive result of any screening test should be addressed in the same way as a presenting complaint, by carefully considering its evidence-based differential diagnosis. One should bear in mind that it may turn out to be a self-limiting condition that will cause no harm as well as something dangerous that needs urgent treatment. The trouble is that evidence-based differential diagnoses and evidence-based ‘sufficient’ and ‘necessary’ criteria of diagnoses [1] are not being used. This glaring omission must be addressed if over-diagnosis (or under-diagnosis) is to be reduced.
Over-treatment is often due to the absence of evidence-based treatment indication criteria. The indication for treating patients is often assumed in a theoretical, non-evidence based way to be some artificially ‘positive’ result, for example that is above two standard deviations of the mean of a healthy population. This is how diabetic micro-albuminuria is treated. However there is evidence that normotensive patients with an albumin excretion rate between 20 and 40mcg/min (about 30% of the treated population) do not get less nephropathy after treatment compared to placebo [2].
Screening tests are assessed intelligently in an evidence based way by using appropriate indices such as specificities and likelihood ratios. However, until differential diagnosis, diagnostic criteria and treatment selection criteria are also assessed in a similar, systematic, evidenced-based way, there will continue to be over-diagnosis and over-treatment (and under-diagnosis and under-treatment too).
References
1. Llewelyn H, Ang AH, Lewis K, Abdullah A. The Oxford Handbook of Clinical Diagnosis, 2nd edition. Oxford University Press, Oxford 2009, pp749-772.
2. Llewelyn D E H, Garcia-Puig, J. How different urinary albumin excretion rates can predict progression to nephropathy and the effect of treatment in hypertensive diabetics. JRAAS 2004, 5; 141-5.
Competing interests: No competing interests
Re: Preventing overdiagnosis: how to stop harming the healthy
A particular concern is population screening projects for autism as a potential target group for newly developed psychiatric drugs.
A Time Magazine article last year [1] publicised a pilot screening project for autism, funded by US charity Autism Speaks, that had taken place in the Korean city of Goyang and apparently discovered an incidence among school children of 1 in 38. Lead author Dr Young Shin Kim was apparently not perturbed by the numbers:
“Kim stresses that the results of her study shouldn't alarm parents into thinking that autism has suddenly exploded in schools. "It doesn't mean there is an increase in new cases," she says. "We just didn't know how to find them and diagnose them. Now we know there are kids with social problems who are not being treated, and we know how to help them."”
Time reported:
“The researchers say they would expect to see similarly high rates of autism emerge in the U.S. and elsewhere if the same data collection strategy were used. "The kids picked up in Korea, many had never been recognized in medical records as having autism," says Geraldine Dawson, chief science officer for Autism Speaks. "That's what needs to be done, that kind of broad screening."”
However, the magazine did not report was that 67% of the screened-positive children seem to have pulled out before the study was completed [2]:
"For the 1,214 sampled screen-positive students, 869 parents (72%) consented to participate in the full assessment. Of these, 286 (33%) completed full assessments. Of those who completed the assessment, 201 (70%) were confirmed to have ASDs (autistic disorder, N=101; other ASDs, N=100), yielding a crude prevalence for any ASD of 0.36% (autistic disorder, 0.18%; other ASDs, 0.18%)."
What the report also did not disclose was that appointment that very same week of Robert H Ring to Autism Speaks:
"Robert H. Ring, previously a Pfizer senior director, will join Autism Speaks in Princeton, N.J., on June 1 in the newly created position of vice president of translational research. His focus will be on helping move drug experimentation from laboratories to clinical trials, "with the goal of improving outcomes for individuals with autism spectrum disorders," according to the nonprofit organization." [3]
While still at Pfizer Ring stated:
“This is a real opportunity to really make a difference for a huge unmet need using expertise we’ve acquired over a number of years. In fact we’re working right now to build a pre-competitive consortium amongst our competitors, including Lilly, Roche, Novartis, Janssen, and trying to agree that this is an important population to be developing medicines for…” [4]
This led to an announcement of a partnership between Autism Speaks, Roche and King's College, London earlier this year [5].
[1] Alice Park, 'South Korean Study Suggests Rate of Autism May Be Underestimated' http://healthland.time.com/2011/05/09/korean-study-suggests-rate-of-auti...
[2] Kim YS, Leventhal BL, Koh YJ, Fombonne E, Laska E, Lim EC, Cheon KA, Kim SJ, Kim YK, Lee H, Song DH, Grinker RR., 'Prevalence of autism spectrum disorders in a total population sample.' Am J Psychiatry. 2011 Sep;168(9):904-12. Epub 2011 May 9.
[3] Lee Howard, 'Past Pfizer autism unit chief to join major advocacy group
By Lee Howard' 12 May 2011, http://www.theday.com/article/20110512/BIZ02/305129344/-1/BIZ
[4] Pfizer News, 'Autism research unit at Pfizer to address autism spectrum disorders'
[5] 'Unprecedented Academic-Industry Collaboration Seeks New Drugs and Novel Treatments for Autism', http://www.autismspeaks.org/about-us/press-releases/research-academic-in...
Competing interests: Autistic son