Table 4.
Cox regression analyses of discontinuation of basal insulin
| Baseline variables | Unadjusted hazard ratio (95% CI) | P‐value | Adjusted hazard ratio (95% CI) a | P‐value |
|---|---|---|---|---|
| Tertiary hospital vs Secondary hospital | 1.136 (1.072‐1.204) | <.0001 | 1.268 (1.192‐1.349) | <.0001 |
| Inpatient vs outpatient clinic | 1.474 (1.388‐1.565) | <.0001 | 1.317 (1.229‐1.411) | <.0001 |
| Age (y) | 0.995 (0.992‐0.998) | .0004 | 0.996 (0.993‐1.000) | .0243 |
| Education (vs Primary school or lower) | ||||
| Junior high school | 0.822 (0.764‐0.884) | <.0001 | 0.843 (0.780‐0.912) | <.0001 |
| Senior high school | 0.735 (0.679‐0.796) | <.0001 | 0.801 (0.732‐0.877) | <.0001 |
| Junior college or higher | 0.643 (0.586‐0.705) | <.0001 | 0.713 (0.639‐0.796) | <.0001 |
| Urban residence vs rural residence | 1.227 (1.155‐1.303) | <.0001 | 0.925 (0.854‐1.002) | .0556 |
| Out‐of‐pocket ratio (%) | 1.005 (1.004‐1.006） | <.0001 | 1.004 (1.002‐1.005） | <.0001 |
| BMI group (kg/m2), (vs <24) | ||||
| 24‐27 | 0.969 (0.910‐1.033) | .3331 | 1.062 (0.996‐1.134) | .0671 |
| ≥28 | 1.099 (1.011‐1.195) | .0263 | 1.154 (1.058‐1.258) | .0012 |
| Diabetes duration (y) | 0.973 (0.967‐0.979) | <.0001 | 0.972 (0.966‐0.978) | <.0001 |
| Times of SMBG per month before initiating BI (vs 0) | ||||
| 1‐5 | 0.868 (0.814‐0.927) | <.0001 | 0.975 (0.911‐1.042) | .4506 |
| ≥6 | 0.726 (0.671‐0.786) | <.0001 | 0.910 (0.839‐0.988) | .0243 |
| Number of OAD types concomitant with BI (vs 0) | ||||
| 1 | 0.571 (0.532‐0.614) | <.0001 | 0.847 (0.730‐0.982) | .0275 |
| ≥2 | 0.547 (0.508‐0.589) | <.0001 | 0.767 (0.586‐1.003) | .0526 |
| BI type (vs Glargine) | ||||
| Detemir | 0.958 (0.873‐1.052) | .3679 | 0.918 (0.835‐1.009) | .0763 |
| NPH | 1.742 (1.627‐1.864) | <.0001 | 1.558 (1.443‐1.681) | <.0001 |
| Basal‐bolus insulin used at baseline (vs BI only) | 2.319 (2.185‐2.461) | <.0001 | 1.857 (1.706‐2.020) | <.0001 |
a
Adjusted for all listed covariates.