In 1987, biotech company Genentech launched a better clot buster, Activase (tPa). Clot busters are used after a patient sufferers a stroke or heart attack to break up problematic blood clots and restore blood flow. The standard of therapy before Activase launched was streptokinase, which was substantially cheaper: $320 vs. $2,300. This posed a problem for Genentech because potential customers needed evidence that the newer drug was good enough to justify its higher price. In response, Genentech ran the so-called GUSTO trial involving 1,081 hospitals in 15 countries with 41,021 patients who had suffered a myocardial infarction, colloquially referred to as “heart attack.”
In 1993, the GUSTO trial provided highly statistically significant results that showed that 30-day mortality among Activase-treated patients was 6.3% versus 7.3% for streptokinase and the 24-hour mortality rates were 2.4% and 2.9%, respectively.
People focused on the first result—6.3% vs 7.3%—and I can’t tell you how many otherwise smart and educated people said that there was “just a 1% difference between the two therapies.” A difference of one percentage point! Not a difference of one percent.
Based on these results, patients treated with Activase were 13.7% less likely to die within 30 days of therapy. The “just a 1% difference” people were implying that patients treated with Activase were 1% less likely to die within 30 days of therapy.
Their terminology made it sound like 100 people would die on the old drug while only 99 people would die on the new drug—a 1% difference. When, in fact, 100 people would die on the old drug and only 86 would die on the Genentech drug.
The 24-hour mortality results showed a 17.2% benefit for Activase, effectively supporting the 13.7% benefit seen over 30 days.
Was the additional $2,000 worth the better patient outcomes? For me, the answer is clear. Giving Activase instead of streptokinase to 1,000 patients would, based on these results, cost an extra $2 million and save ten patients (1,000 x 0.073 – 1,000 x 0.063 = 10), for an effective cost per life saved of $200,000 (1993 dollars). In today’s dollars, the amount would be $448,000, or about 5% of the value the Environmental Protection Agency uses to value an American life.
Some people dismissed the GUSTO results by saying there was just a 1% difference between the two therapies. Had they paid more attention, they would have celebrated Activase’s 13.7% improvement over streptokinase and noted its economic value.
Good analysis, Charley.
I do wonder, though, about your value of life analysis. You're right about the EPA's number and it's well backed by analysis of risk premiums for risky work.
The data that you discuss on the drugs, though, show the difference in the reduction in mortality over just 30 days. You would need to know the difference in the reduction of mortality over at least a few years.