**3. Overdiagnosis and Overtreatment**

It is probable that the single most important cause of the rising DTC incidence over the last few decades has been the increased detection of asymptomatic thyroid nodularity revealed by the liberal use of thyroid ultrasound. This has unveiled a huge reservoir of mainly benign but sometimes malignant disease. The junction at which diagnosis becomes overdiagnosis is the point at which the identification of disease does not lead to overall population benefit. Overtreatment is an almost inevitable product of overdiagnosis and is observed when a disease is diagnosed and optimally treated but the net result is an unfavourable balance between patient benefit and the overall adverse effects of care. This risk was recognised almost 30 years ago [24] as increasing access to ever-improving imaging techniques replaced medical examination.

The widespread use of neck ultrasound has led to either a preoperative increased detection of early-stage tumours and/or indirectly to thyroid surgery where incidental thyroid cancers, mainly papillary thyroid carcinomas under 10 mm (papillary thyroid microcarcinomas (PTMC)), are found. From 1975 to 2009, the proportion of incidental PTMC has increased from 25% to 39% [25]. In some European countries and the US, 45 to 70% of thyroid malignancies are considered "over diagnosed", based on studies comparing the expected and observed prevalence of thyroid cancer [26]. This conclusion is

also inferred by the rise of early-stage thyroid cancer and the incongruity of there being more cancer with no change in mortality over decades [27]. The alternative explanation, that early diagnosis coupled with excellent treatment has compensated for a real thyroid cancer epidemic, appears less substantiated.

Overdiagnosis and overtreatment are clear generic population-based concepts but are more difficult to define at an individual level since it has not been possible to reliably predict the natural history of an individual PTMC in a specific patient. The early diagnosis of a PTMC will have saved some patients from a late diagnosis and despite the excellent overall prognosis some PTMC may become larger and metastasise or manifest an aggressive clinical behaviour even without enlargement. However, it remains the case that at autopsy after deaths unrelated to thyroid cancer 6.7–16.1% of thyroids present one or more foci of PTMC [12] so that the vast majority of newly diagnosed PTMC is implicitly of no clinical significance. The diagnosis and treatment of PTMC offers no benefit to the patient in the vast majority of cases. Apart from some difficult to identify patients who will benefit, most are exposed to potential morbidity without gain. In addition to the personal risk of surgery, one must consider the psychological effects of a cancer diagnosis that is cancer in name but infrequently in behaviour. A frequently overlooked additional consequence of the surge in DTC diagnoses is the drain on healthcare resources.

The overdetection and treatment of PTMC comes with an economic cost irrespective of whether it is treated surgically or subjected to a surveillance programme. These costs must be balanced against the consequences of a delayed diagnosis in the minority of patients that may come to harm if a timely cancer diagnosis is not made. This argument effectively overlaps with the health economic debate that applies to cancer screening in general. Does thyroid cancer screening fit the cancer screening principle that a presumptive diagnosis of subclinical disease and an early diagnosis improves outcome? [28]. For screening to be effective, the time interval between a disease becoming detectable by the screening tool and the presentation of clinically detectable disease must be shown to be detrimental to the patient [29]. For a screening program to be considered the disease must be common, have an identifiable risk group and the screening tool must be both sensitive and specific. There is no evidence that this applies to thyroid cancer. A Polish group reviewed 4701 patients surgically treated for thyroid cancer [27] with patients divided according to whether the diagnosis was made with a clinical presentation or without symptoms or risk factors. The asymptomatic group predictably presented a lower TNM stage, a lower rate of multifocality, and no characteristics of aggressive clinical behaviour. The use of screening results in the diagnosis of indolent cases and may lead to overdiagnosis and overtreatment. Very few countries have implemented a thyroid ultrasound screening program for thyroid cancer and several governments have now acknowledged the possible detrimental effects of unwarranted neck ultrasound use in asymptomatic patients.

In an attempt to reduce unstructured neck ultrasound screening in the UK, only a thyroid specialist should request a thyroid ultrasound [30]. The American Preventive Service Task Force (USPSTF) recently released its guidelines, in which it strongly recommends against using neck ultrasound for thyroid cancer screening in asymptomatic patients [31]. The impact of changing guidelines was seen in South Korea where screening with neck ultrasound for thyroid cancer became part of a National Cancer Control Program in 1999 [32], creating an epidemic of low-risk PTC. After recognising the morbidity of unnecessary thyroid surgery due to the thyroid screening this practice was discouraged from 2014 with a corresponding decrease in the incidence of thyroid cancer and the number of thyroid operations decreased significantly [33]. However, if patients do not undergo surgery for PTMC, the management dilemmas and costs associated with an active surveillance (AS) programme with repeated clinical review and ultrasounds is also problematic since a reliable predictor of progression is still not available [34,35].
