Table 6.
ROC analysis results of the biochemical data investigated in the study as predictors of GLIM-based malnutrition.
| ROC Results | Albumin | TLC | CRP | Total Protein |
|---|---|---|---|---|
| AUC (CI 95%) | 0.672 (0.58–0.77) | 0.431 (0.32–0.54) | 0.701 (0.61–0.79) | 0.582 (0.48–0.69) |
| SE | 0.049 | 0.056 | 0.047 | 0.054 |
| z | 3.472 | −1.233 | 4.292 | 1.507 |
| p | 0.0005 | 0.218 | 0.0000 | 0.132 |
| Sensitivity (95% CI) |
67% (59% to 73%) |
NA | 52% (45% to 60%) |
NA |
| Specificity (95% CI) |
62% (44% to 78%) |
NA | 82% (65% to 93%) |
NA |
| Positive Predictive Value (95% CI) |
90% (85% to 93%) |
NA | 94% (88% to 97%) |
NA |
| Negative Predictive Value (95% CI) |
26% (20% to 33%) |
NA | 25% (21% to 29%) |
NA |
| Accuracy (95% CI) |
66% (59% to 72%) |
NA | 57% (50% to 64%) |
NA |
NA—not analyzed, ROC—receiver operating characteristic, AUC—area under curve, CI—confidence interval, SE—standard error, TLC—total lymphocyte count, CRP—C-reactive protein.