16 March 2021 : Original article
Use of Imputation and Decision Modeling to Improve Diagnosis and Management of Patients at Risk for New-Onset Diabetes After Transplantation
Vidit N. Munshi1ACDEF*, Soroush Saghafian2AEG, Curtiss B. Cook3BE, Sumhith Veda Aradhyula2C, Harini A. Chakkera4BEDOI: 10.12659/AOT.928624
Ann Transplant 2021; 26:e928624
Table 2 Observed and imputed HbA1c, FBG, and insulin use at 1, 4, and 12 months after transplant for liver transplantation patients who did not have pre-transplant DM.
Variable | Observed data | Imputed/perfect collection | ||||
---|---|---|---|---|---|---|
n | Missing values | Mean* | n | Missing values | Mean* | |
1-month HbA1c | 33 | 204 | 5.5/37 | 237 | 0 | 5.4/36 |
4-month HbA1c | 161 | 76 | 5.5/37 | 237 | 0 | 5.5/37 |
12-month HbA1c | 142 | 95 | 5.7/39 | 237 | 0 | 5.6/38 |
1-month FBG | 226 | 11 | 100/5.6 | 237 | 0 | 100/5.6 |
4-month FBG | 223 | 14 | 105/5.8 | 237 | 0 | 105/5.8 |
12-month FBG | 208 | 29 | 111/6.2 | 237 | 0 | 111/6.2 |
On insulin at 1 month | 237 | 0 | 51.1% | 237 | 0 | 51.1% |
On insulin at 4 months | 237 | 0 | 33.8% | 237 | 0 | 33.8% |
On insulin at 12 months | 237 | 0 | 21.5% | 237 | 0 | 21.5% |
* Displayed as (%/mmol per mol) for HbA1c and (mg per dl/mmol per l) for FBG. 95% confidence intervals not displayed, but were the same between observed and imputed variables |