Table 2. Estimated effects of intervention compared with control on primary and secondary outcomes, with binary outcomes.
| Raw data at EOS, n (%) | Effects based on primary model* | ||||
|---|---|---|---|---|---|
| Control | Intervention | OR (95% CI) | RR (95% CI) | RD (%, 95% CI) | |
| HbA1c <7.0% (primary outcome) | 2,171 (37.5) | 5,120 (43.6) | 1.335 (1.180, 1.510) | 1.186 (1.105, 1.267) | 7.0 (4.0, 10.0) |
| FBG <7.0 mmol/L | 2,357 (40.9) | 5,160 (44.1) | 1.115 (0.993, 1.252) | 1.065 (0.996, 1.135) | 2.7 (−0.2, 5.5) |
| BP <140/80 mm Hg † | 2,513 (43.4) | 5,282 (45.0) | 1.055 (0.952, 1.169) | 1.030 (0.972, 1.089) | 1.3 (−1.2, 3.9) |
| LDL-C <2.6 mmol/L | 2,615 (45.8) | 5,235 (45.0) | 0.984 (0.865, 1.120) | 0.991 (0.922, 1.062) | −0.4 (−3.6, 2.8) |
| Composite ABC control ‡ | 509 (8.9) | 1,258 (10.9) | 1.245 (1.068, 1.451) | 1.218 (1.062, 1.395) | 1.9 (0.5, 3.5) |
Notes: Composite ABC control is defined as HbA1c level <7.0%, BP <140/80 mm Hg, and LDL-C <2.6 mmol/L.
* Primary model: logistic regression with GEE accounting for clustering and with adjustment of the baseline value of the analyzed outcome. RR are indirectly derived from OR using the following formula: RR = OR ÷ [1 − p0 in control × (1 − OR)].
† Only baseline SBP and clustering were adjusted in the primary model for BP control.
‡ No baseline variable was adjusted in the primary model for the composite ABC control.
BP, blood pressure; CI, confidence interval; EOS, end of study; FBG, fasting blood glucose; GEE, generalized estimating equation; LDL-C, low-density lipoprotein cholesterol; OR, odds ratio; RD, risk difference; RR, relative risk; SBP, systolic blood pressure.