Overdiagnosis is an iatrogenic harm. Other iatrogenic harms, such as negligence (for example, failure to investigate or diagnose a serious condition) or medical error (e.g wrong dose of a medication), are widely accepted as negative events for which doctors are held responsible. In contrast, there is controversy over 1) whether overdiagnosis harms constitute real harms; and 2) whether overdiagnosis harms constitute important harms.

In this paper, we investigate why the harms of overdiagnosis are viewed differently from these other iatrogenic cases. First we examine the differential way that harms arising from acts of commission are treated compared with harms arising from acts of omission. Usually in moral theory, harms of commission are considered to be more serious than harms of omission. This distinction is recognised in medicine, where it is invoked in debates about end of life care, such that letting die (e.g. by withdrawal of supportive treatment) is often considered less morally problematic than actively ending a patient’s life with a lethal injection.

However, in the area of harms related to diagnosis, omissions tend to be judged more harshly than commissions. Underdiagnosis results from an omission – failure to perform the relevant examinations and/or investigations – and this is deemed culpable. In contrast, the harms of overdiagnosis arise from an act of commission: the harms ensue as a result of positive intervention such as a screening test or investigation performed by a medical practitioner, but causing these harms is not usually judged as harshly as those arising from underdiagnosis.

We argue that one reason for this differential treatment may arise from the nature of the harms in question. With underdiagnosis, there is usually a concrete harm, such as advanced disease in an identified individual. This tangible harm can be attributed to a lack of action on the part of the treating doctor – they should have made the diagnosis. The harm, advanced disease, is widely accepted as the kind of harm that medicine should strive to avoid. In contrast, the harms of overdiagnosis are both less identifiable and less clearly defined. There is little agreement regarding overdiagnosis harms: in what sense does harm occur; who or what is harmed; who causes the harm; are the harms available to perception; and how are present experience and imagined futures traded off in evaluating harms?

We propose a schema for classifying the harms that arise from overdiagnosis and describe a number of biases, types of uncertainties, moral beliefs, and social processes that may help to explain why the harms of overdiagnosis are rarely judged as harshly as those of underdiagnosis, despite being iatrogenic harms of commission.