Physical examination before diagnostic ultrasonography: Alive and kicking or a relic of the past?
April 7, 2026Overdiagnosis of Papillary Thyroid Cancer
| Objectives | To estimate population-level rates of papillary thyroid cancer overdiagnosis in USA.
To identify the implications of reducing thyroid ultrasonography use for thyroid cancer incidence and overall mortality. |
| Authors |
David O. Francis; Louise Davies; Yichi Zhang; Natalia Arroyo; Sara Fernandes-Taylor; Peter Nordby; Ben A. Y. Cher; Manasa Venkatesh; Erin J. Aiello Bowles; Oguzhan Alagoz. |
| Journal |
JAMA Network Open. 2026;9(2):e2559852. doi:10.1001/jamanetworkopen.2025.59852 |
| Methods |
This study used a validated simulation model (Papillary Thyroid Carcinoma Microsimulation Model, PATCAM) of contemporary PTC incidence in the adult US population (aged 18 years or older), between 1991 and 2019 to study thyroid cancer incidence and referrals for ultrasonography of nonpalpable thyroid nodules. Data analysis was conducted from June 2024 to August 2025. The model represents the probability that an individual is referred for ultrasonography through an age- and sex-specific function of tumour size. PATCAM assumes that all PTC cases prior to 1990 were detected through the palpable pathway. As Ultrasonography referral rates in the US have substantially increased over time, PATCAM incorporates ultrasonography utilization data from Kaiser Permanente Washington starting in 1997 to estimate the yearly probability of referral to ultrasonography in the nonpalpable pathway The outcomes are:
|
| Results |
The model estimated that 72% to 94% of PTC cases were overdiagnosed. While the proportion of overdiagnosis was slightly higher for women compared with men (75%-95%vs 63%-90%), the absolute rate of overdiagnosis was higher among women compared with men (13-17 per 100 000 individuals vs 3-5 per 100 000 individuals). This rate translated into 443 212 to 573 705 women and 107 804 to 154 504 men who were overdiagnosed with PTC over 28 years. Reducing use of thyroid ultrasonography for nonpalpable nodules by 33% and 67% would have decreased PTC incidence by 17% (18 to 15 per 100 000 individuals) and 41% (18 to 11 per 100 000 individuals) in 2019, respectively, regardless of sex, with the greatest decrease in absolute rates occurring among women. These reductions would have resulted in a less than 0.1% change in overall mortality between 1991 and 2019. |
| Conclusions |
Between 72% and 94% of PTC cases diagnosed in the US between 1991 and 2019 were overdiagnosed, even after accounting for an underlying increased risk in the development of thyroid cancer. The detection of these cancers and their treatment yielded no benefit to population mortality. Although some degree of overdiagnosis is unavoidable in clinical practice, this suggests an opportunity to reduce unnecessary thyroid ultrasonography referrals, particularly for nonpalpable nodules. Doing so could lower the number of cancer diagnoses and associated psychological and treatment burdens without compromising population health or increasing mortality. |
| Key points | Thyroid cancer overdiagnosis in the US remains substantial, even after accounting for possible increases in true incidence.
Reducing unnecessary thyroid ultrasonography referrals, particularly for nonpalpable nodules, would reduce unnecessary diagnoses and also unnecessary treatments with their associated harms and costs without increasing mortality. Doing so we could lower the number of cancer diagnoses, the thyroid cancer incidence, and the associated psychological and treatment burdens without compromising population health or increasing mortality. |
| Link (DOI) | https://dx.doi.org/10.1001/jamanetworkopen.2025.59852 |
| Ultrasound speciality | Thyroid Ultrasound |
Short-Review by:
Prof. Dr. Jose Luis del Cura
Department of Radiology
Donostia University Hospital
Spain
Strengths:
Used a simulation model that was developed and validated using high-quality data that had high accuracy in estimating rates of tumours. Thyroid overdiagnosis is the elephant in the room, with a great impact in the amount of US-exams performed daily. Both topics, thyroid US and overdiagnosis are trendy, and have been extensively discussed in the last years. The study covers a long period of time.
Weaknesses:
Mainly theoretical. The results are not generalizable. The model is based on the specific characteristics of the population and medical system in USA, and the use and indications for ultrasound there. The study does not take into account the changes in the standards of diagnosis and treatment of thyroid cancer throughout the years when drawing its conclusions. Uses data from USA, but to study prevalence of thyroid disorders on US screening, uses a source from the working population of Germany.
Personally thinking:
Although mainly theoretical biased and probably not generalizable to other healthcare settings or populations, the study confirms something that have been repeatedly highlighted: systematic US screening for thyroid does more harm than good. US exams of thyroid have no benefit over life-expectancy and should not been performed except in case a pathology is clinically suspected.
