borderline significanceP = .051
The crude model had borderline significance (HR 2.26, P = .051), while propensity score–adjusted (HR 9.92, P = .002) and IPTW models (HR 13.98, P < .001) showed strong effects.
The crude model had borderline significance (HR 2.26, P = .051), while propensity score–adjusted (HR 9.92, P = .002) and IPTW models (HR 13.98, P < .001) showed strong effects.