The article goes on to say more on the subject of study follow-up and how representative their participant population is, but the answer to the question in the

The article goes on to say more on the subject of study follow-up and how representative their participant population is, but the answer to the question in the headline is extremely simple. It's a measure of risk reduction. It sounds like 94 participants have become ill with COVID. Fewer than 9 of them were in the treatment group. Sample risk of catching COVID: 94/N. Sample risk of catching COVID, knowing you're in the treatment group: 9/N. The risk of catching COVID, knowing you're in the treatment group, compared to the overall risk of catching COVID: (9/N)/(94/N) ≈ 0.096. Decline in risk due to being in the treatment group: 1 − 0.096 = 0.904.