Consider patient risk factors for melanoma when prescribing antihypertensives

Drucker and colleagues provide a great paper1 that raises questions about antihypertensive medications and adverse effects. The adjusted hazard ratio for melanoma with thiazide exposure is quite high at 1.34 (95% confidence interval 1.01–1.78).1

In comparison, the baseline incidence rate for melanoma among men 40–44 years old is 23.50 per 100 000.2 Therefore, a very rough translation of hazard ratios (with apologies to all biomedical statisticians) shows that prescribing thiazides increases the risk of melanoma by about 34%, increasing the incidence rate from 23.5 to 31.49 per 100 000 people.

That means if 100 000 people were prescribed hydrochlorothiazide, then about 8 more people (plus or minus the confidence interval) would get melanoma — a number needed to harm of about 12 500 people per year. If the patient had a tenfold greater risk of melanoma, based on skin colour, family history and so on, the number needed to harm becomes 1250 people. For some patients, that may cross the decision threshold of whether or not to take thiazide.

I would appreciate someone checking my “back of the envelope” numbers, but if they are correct, this paper shows that providers should be considering melanoma risk in our review of drug options for patients with hypertension.