Table 3.
Summary of derivation and validation datasets.
| Derivation dataset | Validation dataset | |
|---|---|---|
| Final # of patients | 1883 | 286 |
| Final # of BG samples | 12,344 | 4030 |
| Age, mo, mean ± STD | 50 ± 73 | 40 ± 62 |
| CTICU | 60% of patients | 57% of patients |
| PICU* | 41% of patients† | 43% of patients |
| Metabolic acidosis‡ | 14% | 10% |
| Respiratory acidosis‡ | 28% | 22% |
| Metabolic alkalosis‡ | 7% | 17% |
| Respiratory alkalosis‡ | 6% | 10% |
| Mixed‡ | 44% | 41% |
| PARDS‡ | 12% | 9% |
| Before resampling | 11% pH < 7.3 | 7% pH ≤ 7.3 |
| 71% 7.3 ≤ pH < 7.45 | 71% 7.3 ≤ pH < 7.45 | |
| 18% pH ≥ 7.45 | 22% pH ≥ 7.45 | |
| Post resampling and processing | 17% pH < 7.3 | NA§ |
| 54% 7.3 ≤ pH < 7.45 | ||
| 29% pH ≥ 7.45 | ||
| Blood gas type | 90% arterial | |
| 10% capillary | ||
Numbers for post-processed data are shown, except for pH range data. STD standard deviation. *PICU pediatric (multidisciplinary, medical-surgical) ICU. †Patients may have stayed in both ICUs. ‡Definition of PARDS, respiratory and metabolic acidosis and alkalosis are discussed in Supplementary material. §Validation dataset was not resampled.