“There’s not enough evidence to weigh the benefits and harms of performing routine visual examinations to screen for skin cancer.”
This is the conclusion from the U.S. Preventive Services Task Force in a recently concluded review of screening.
But not everyone agrees and there have been formal statements from dermatology groups opposing the conclusions. What is clear is that screening the entire population comes with significant costs and with significant over treatment, including biopsies.
Over treatment is a feature of medicine and not a sign of poor care but of a complex problem and while not pleasant for those involved, it’s significantly better than metastatic disease from a spreading melanoma. The associated costs though, can cripple already struggling healthcare systems who have to prioritise other medical interventions.
All of this debate is based on two facts.
Firstly, you have to see a lot of people to find melanoma with lifetime risk less than 2% even in countries of high melanoma rates like Australia and New Zealand.
Secondly, of all the lesions sent for biopsy to rule out melanoma, only roughly 4-5% of them turn out to be melanoma. The procedure is safe, relatively cheap and the consequences of missing melanoma are devastating, so clinicians don’t take any chances.
Combining the two means there is a lot of cost to find a melanoma. Which is where we think the technology we’re developing has such utility. We’re using Artificial Intelligence (AI) to decide whether a lesion needs to be biopsied and right now the performance suggests we can accurately reduce the number of biopsies to find a melanoma. Significantly.
If you combine the fact that we can provide assessments for a low cost to anybody who is concerned, we can change the economics of population based screening.
We like to think that means we can make a skin cancer screening service available to everyone and if we can do that, we can make sure that more people survive the disease.