When diagnoses are wrong

THE hallmark of a profession is its ability to self-regulate. This right is grudgingly granted by the government and public to members of the profession. This is because professionals are considered so specialised and highly skilled in their knowledge that it would be difficult for an external agency to regulate without stifling.

Members of the profession in turn understand this concern that the public has and prove their ability to self-regulate by having robust internal mechanisms to keep their profession scientifically updated and professionally ethical. This includes ensuring that specialist qualifications are of the highest standard.

However, a recent editorial co-authored by Yale histopathology professors Dr David Rimm and Dr Robert Homer titled “False-positive pathology: improving reproducibility with the next generation of pathologists” in Laboratory Investigation 2019, points to a situation where this self-regulatory mechanism seems to have failed.

Before I get into that, let me explain who histopathologists are. Briefly, histopathologists (aka surgical pathologists, anatomical pathologists) are medical doctors who work primarily outside of wards/clinics and in laboratories.

Hence, histopathology is a sub-speciality of laboratory medicine (the others being microbiology, chemical pathology, hematopathology and forensic pathology). In the main, histopathologists examine tissue samples from patients sent by surgeons to make a diagnosis.

One of the most important aspects of the work is in cancer diagnostics. Cancers are rarely diagnosed or treated without a written report from a histopathologist. It goes without saying that an accurate diagnosis is essential.

What is the difference between histopathologists and clinic doctors? Clinic doctors see a patient from start to finish, that is, they see the full disease cycle (diagnosis, treatment, outcome). Hence they have a two-pronged feedback mechanism to ensure that they are providing accurate treatment. The first is informal-direct observation of how previous individual patients responded to their treatment, and the second is formal-clinical research whereby data of previous groups of patients are collected and analysed. Together, they provide a check-and-balance mechanism that ensures that future patients are treated based on the best possible medical evidence.

Histopathologists, on the other hand, are only present in one part of the disease cycle (ie, the diagnostic stage). We usually have no access to the informal mechanism – ie, no direct knowledge of how our diagnosis worked out in any individual patient since we do not participate in the complete disease cycle. We are almost fully dependent on the formal mechanism – ie, clinical research – to ensure that we are making the correct diagnosis.

Our “diagnostic criteria” are formulated on the basis of such research. Hence, the check-and-balance mechanism is usually absent. Logic and good ethics would mean that the profession would ensure that histopathologists acquire the “clinical research methods” component to very high standards. Further, if one is not trained in research methods, that would likely mean that one would not be able to even read a complex research paper to keep oneself updated.

So what is that editorial saying? It says that (worldwide) histopathologists specialist degree curricula have not ensured that these high standards have been met. Bluntly put, it seems to imply that we have been conferred degrees with a vital component missing. And this vital component is, usually, not successfully acquired post-qualification via continuing medical education (CME) either.

What is the effect of having incomplete training and ineffective CME? It creates problems like misdiagnosis (calling one type of cancer another type thus misdirecting treatment), over-diagnosis (calling a non-cancer a cancer) or under-diagnosis (calling a cancer a non-cancer).

The editorial focuses on the over-diagnosis of cancers. The problem of over-diagnosis is different from the problem of under-

diagnosis or misdiagnosis.

If a particular cancer was under-

diagnosed or misdiagnosed, it would manifest itself over time and the histopathologist would be alerted to the faulty diagnostic criteria by the surgeon. This is an inefficient method but at least it exists as a check and balance.

In the problem of over-diagnosis, we would take it that the treatment “worked” and hence the cancer has not recurred – not realising that the reason that it has not “recurred” is because it is not a cancer in the first place!

What does this mean for patient care? The editorial gives the example of a “cancer” that was, until very recently, known as “non-invasive encapsulated follicular variant of papillary thyroid carcinoma”. For decades, these were labelled and treated as “cancers” on the basis of incorrect diagnostic criteria formulated by relying on low quality research. This is known as over-diagnosis.

Over-diagnosis can lead to severe consequences like mutilating surgery (including amputation), severe adverse effects of anti-

cancer therapy, and financial burdens among others.

Why wasn’t this competency in clinical research methods made compulsory in histopathologist training? The editorial references a study which shows that the skill-sets required are difficult to acquire using informal methods such as journal clubs or short courses. They require a proper intensive, structured module.

Unfortunately, senior pathologists (including those with PhDs and those from academic centres) themselves never received this training and were not able to successfully acquire it post-qualification. So this means that there isn’t a pool of properly trained histopathologists available to set up an improved curriculum and teach it to trainees and junior pathologists.

Who decides on the curriculum? This can vary from country to country but often involves representatives from regulatory bodies like medical councils, professional organisations like colleges of pathologists, and senior practitioners (especially from academic centres), at the very least.

What many curriculum framers worldwide did was create a sort of pseudo-module by asking trainees to do “research projects” as part of the degree without first teaching them “research methods”!

That is, trainees who were not even trained to read a research paper are actually asked to do and publish research! So this naturally resulted in a flood of low quality research being published. Diag-nostic criteria were formulated on the basis of these low quality papers further worsening the situation.

Is this a new or previously unknown defect in the curriculum? No, it has been known for decades with a lot of agonising and handwringing in medical journals and conferences. It’s just that no one has had the will to act and the public did not object.

Why has this problem escaped media, government or public attention for so long? The editorial is silent on this. However, I suspect that it is because histopathology (and indeed many other laboratory medicine disciplines) are “hidden” from public view.

Also, patients with “false positive” diagnoses don’t realise that they have been wrongly diagnosed. There is no formal mechanism to inform patients if they have been the victims of low quality “diagnostic criteria” and reverse their diagnosis. Hence, no governmental or public pressure has been brought to bear on curriculum framers to seek outside help to improve the histopathology specialist curriculum.

A “false positive” cancer diagnosis can have devastating consequences for a patient, even worse if the patient is a child.

Notwithstanding our right to self-regulation, the editorial by the Yale professors needs to be taken seriously, and the deficiencies in the specialist training programme need to be rectified.


Consultant Histopathologist & Statistician

Department of Pathology

Faculty of Medicine,

Universiti Malaya

opinion , medical , statistics