Table 1

Statistical comparison of DKA risk for sotagliflozin versus placebo

Trial no.Sotagliflozin Events/N (incidence rate per 100 patient-years)Placebo Events/N (incidence rate per 100 patient-years)Hazard ratio* (95% CI)Exposure-adjusted Mantel-Haenszel risk difference^ per 100 patient-years (95% CI)Number needed to harm per year (95% CI)
309/31035/1,049 (3.40)1/526 (0.19)17.57 (2.41, 128.20)3.21 (2.04, 4.38)31 (22.8, 49.0)
31221/699 (6.00)4/703 (1.11)5.37 (1.84, 15.64)4.89 (2.17, 7.60)21 (13.2, 46.1)
  • Source: FDA reviewer (16).

  • *A Cox proportional hazards model stratified by trial was used for the 309/310 analysis, and a nonstratified Cox model was used for the 312 analysis, with actual treatment as the only covariate, with the two doses of sotagliflozin combined. Data were truncated 30 days after treatment end date.

  • ^Exposure-adjusted Mantel-Haenszel risk difference, stratified by trial; the 95% CI was calculated using the Sato method.

  • †Number needed to harm is the number of patient-years of exposure to sotagliflozin to observe one additional DKA event.