UKLS

UKLS approached 247,354 people, of whom 30.7 percent responded positively, and randomised 4,055 at high predicted risk to a single low-dose CT scan or to usual care. Of 1,994 scanned, 42 (2.1 percent) were diagnosed with lung cancer; 36 of those 42 (85.7 percent) were stage 1 or 2 and 35 (83.3 percent) had surgical resection as their primary treatment. Lung cancer was commoner in the lowest socioeconomic group, and the psychosocial harms of screening were reported as modest and temporary.

At a median 7.3 years of follow-up there were 30 lung cancer deaths in the screened arm against 46 in the control arm, a relative rate of 0.65 (95 percent confidence interval 0.41 to 1.02, p = 0.062), too small a trial to reach significance alone. Pooled with eight other randomised trials in a random-effects meta-analysis the relative rate was 0.84 (0.76 to 0.92), which the authors called unequivocal support for screening in identified risk groups. The trial also established that a risk model, rather than age and pack-years alone, is the practical way to choose who to invite, which is how the NHS programme now works.