On its face, its a frightening projection: Cancer rates worldwide are expected to increase 75 percent by 2030.

That would mean cancer cases nearly doubling over the course of 22 years, according to new data published Thursday in Lancet Oncology. Globally, the number of cancer patients would grow from 12.7 million in 2008 to 22.2 million in 2030.

That sounds a lot like a cancer epidemic. In developing countries, the growth in cancer cases is especially pronounced, potentially growing 90 percent over the same time period. But here in the United States, separate research suggests we could have something entirely different, and potentially more benign on our hands: An epidemic of overdiagnosis.

The British Medical Journal recently combed through American data on cancer deaths and diagnoses. It found that, while American deaths from certain cancers have remained pretty constant since the 1970s, the number of cancer diagnoses has skyrocketed. Here’s what that looks like in chart form (the blue line represents cancer diagnoses and the red line cancer deaths):

For both breast and prostate cancers, as shown above, deaths have actually been trending downward slightly since the mid-1990s. It’s the diagnoses, however, that are significantly higher than they were in 1975.

Some of that likely has to do with improvements in medical technology, as we learn more about how cancer works — and how to treat it. With more testing we’ve caught more harmful cases than we may have previously.

But researchers in the British Medical Journal think it also has to do with overdiagnosis: As screening tests become more sensitive and pervasive, they pick up more cancers or abnormalities that previously would have gone undiagnosed and caused no symptoms or harm.

One review of breast cancer cases finds that about a third of the cases detected likely fall in the the category of “overdiagnosis,” abnormalities that would never have manifested into symptoms or early death. Incidence of prostate cancer spiked in the early 1990s with the introduction of the Prostate-Specific Antigen, or PSA, test in 1986. Mortality rates, however, stayed flat.

Does overdiagnosis matter? There is sometimes a “better safe than sorry” mentality, that suggests erring on the side of caution and going ahead with treatment. Playing it “safe,” however, can come with its own risks. “The downsides of overdiagnosis include the negative effects of unnecessary labelling, the harms of unneeded tests and therapies, and the opportunity cost of wasted resources that could be better used to treat or prevent genuine illness,” write the BMJ researchers.

A rise in cancer diagnosis is certainly alarming, but not necessarily because it suggests we have a greater risk of dying from cancer. The odds of that happening, at least in the United States, are on the decline. What we do see is a growth of cancer diagnosis that could, on the balance, be doing more harm than good.