What Does “A 30% Lower Risk of Disease Progression” Mean?

When a headline says that a treatment “reduces the risk of disease progression by 30%,” it is tempting to imagine thirty more people out of every hundred avoiding progression. To understand the claim, first identify the comparison being made and the outcome the researchers measured.

Start with the comparator, the population and the time period.

A new treatment compared with a placebo answers a different question from a new treatment compared with an established medicine or another combination. People receiving their first treatment and people who have already received several treatments should not be treated as the same population simply because they share a disease name.

A hypothetical example can explain a relative change. Suppose that, at the same specified time point, twenty out of every hundred people in the comparison group have experienced an event, versus fourteen out of every hundred in the new-treatment group. The event proportion has fallen from 20% to 14%. That is an absolute reduction of six percentage points and a relative reduction of 30% from the original 20%. These are two descriptions of the same result, offering different information. This example illustrates arithmetic only. It does not describe a real medicine or estimate anyone’s personal benefit.

In cancer reporting, the 30% may instead come from a hazard ratio.

Studies of time to an event often report a hazard ratio, or HR. An HR of 0.70 is commonly summarized as a “30% relative risk reduction.” A hazard ratio compares event hazards over time, and its interpretation depends on the analysis and its assumptions. It cannot simply be translated into “thirty fewer people out of every hundred will progress” or “people will live 30% longer.” The NCI explanation of hazard ratios describes this time-based comparison.

The earlier 20%-versus-14% example compares event proportions at one defined time point. It is not a conversion of HR 0.70 into those proportions. To understand how many participants were still without progression at one year, look for the study’s reported estimates at that time, or ask a clinician to explain the relevant curve and estimates.

Next, ask how the study outcome relates to what matters to you.

Some cancer trials assess progression-free survival, or PFS. The precise trial definition often treats progression or death as an event. Overall survival, or OS, measures survival time. These outcomes answer different questions. An improvement in PFS, by itself, does not establish the size of an improvement in OS. It also does not fully describe symptoms, quality of life or the burden of treatment. The NCI definitions of PFS and OS help distinguish the concepts.

Alongside a promising headline, look for the uncertainty around the estimate, serious side effects, treatment discontinuation and the duration of follow-up. A striking relative number is one part of a study. The size of the benefit, the strength of the evidence and the risks need to be discussed together.

You do not need statistical vocabulary to have that discussion. Try asking: “Does this 30% refer to event proportions at a particular time, or to a hazard ratio? What was the comparison treatment? Are absolute numbers available? How similar were the participants to me? What happened to side effects and quality of life, as well as disease control?”

Group results can inform treatment discussions, but they cannot predict exactly what will happen to one person. The NCI information on prognosis makes that distinction clear. Turning a headline into a few understandable questions can be more useful than remembering its largest percentage.