26 Defining and measuring overdiagnosis when there is no diagnosis of disease: the experience of the canadian task force on preventive health care with overdiagnosis in the context of fracture risk assessment
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
<h3></h3> ‘Overdiagnosis means making people patients unnecessarily, by identifying problems that were never going to cause harm or by medicalising ordinary life experiences through expanded definitions of diseases’ (Broderson et al). That definition seems intuitive and for certain diseases (like cancers), measurement of the extent of this phenomenon has been described. Measurement relies sometimes on data from randomized trials, e.g. the comparison of the number of incident cases in a screened group versus a non-screened group. In high-quality randomized screening trials with sufficient follow-up time and little or no contamination of the control group, the excess number of diagnosed cases in the screened group represents the degree of overdiagnosis. But, what about when there is no disease, when we are identifying problems defined by an estimated risk of a future event? That was the question faced by members of the Canadian Task Force on Preventive Health Care when developing a protocol for an evidence synthesis to support a guideline on screening to prevent fragility fractures. In that setting, the screened group is given a risk of a future event, not a diagnosis that could become apparent in the non-screened group. In fact, there are no symptoms to diagnose before somebody experiences a fracture, so these individuals would experience the outcome, not the ‘disease’ hereby defined as a risk. For the Task Force, in the setting of screening to prevent fragility fractures, overdiagnosed individuals are those who are deemed to be at excess risk of fracture – either according to a set threshold or based on shared decision-making –but who would have never known they were at risk because, without screening, they would not have experienced a fracture. We will explain the process that lead us to this definition and will give our perspective on how to calculate the degree of overdiagnosis when it is not possible to compare the occurrence of disease in screened and not screened groups. We believe this extension of the definition and the proposed way of calculating overdiagnosis in the setting of risk assessment is a way forward in the conceptualization of the overdiagnosis phenomenon. We will suggest that this could be applied to other chronic diseases, including, for example hypercholesterolemia and the risk of cardiovascular disease, where the value of cholesterol is also used (with other factors) to estimate risk and inform decisions. To our knowledge, it is the first time that overdiagnosis in risk assessment has been defined in this manner. <h3>Objectives</h3> Propose a way to conceptualize overdiagnosis and calculate its extent in the context of risk assessment. <h3>Method</h3> This is the result of a group reflection on the topic that started while trying to define outcomes for a systematic review on the prevention of fragility fractures. <h3>Results</h3> Starting from more common ways of understanding and calculating overdiagnosis we will present how we propose to achieve this in the setting of risk assessment. We will share the logic we followed and some graphical representation of our idea. <h3>Conclusions</h3> Overdiagnosis is not an easy concept to grasp when there is no disease. At times this has been simplified by labeling a risk as a disease (ex: osteoporosis is not a disease in itself; it confers a certain amount of risk of fractures). We will share our thoughts about a way to further understanding of overdiagnosis in the context of risk assessment.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it