Association between persistent hyperlactataemia and mortality*
| Setting | Population | Overall mortality (%) | Thresholds for lactate reduction | Relative risk of death | Sensitivity | Specificity | PPV | NPV | Mortality timeframe |
| Hospital admissions (East Africa)11 | 60 days to 12 years with severe febrile illness (n=1906) | 5.98 | Failure to fall by >40% and/or lactate >2.5 mmol/L at 8 h | 2.0 | 64% | 54.3% | 8.2% | 96% | 72 h |
| Hospital admissions (Uganda)10 | <5 years with a diagnosis of pneumonia (n=75) | 17.33 | Failure to fall by >40% by 8 h | 2.8 | 53.8% | 75.8% | 31.8% | 88.7% | In-hospital death or survival |
| PICU admissions (India)12 | 1 month to 12 years with septic shock (n=148) | 63.5 | Failure to fall >10% from baseline at 24 h | 2.5 | 78.7% (68.8% to 86.2%) | 72.2% (58.1% to 83.1%) | 83.1% (73.4% to 89.9%) | 66.1% (52.5% to 77.5%) | In-hospital death or survival |
| PICU admissions (India)15 | 1 month to 17 years with septic shock (n=112) | 31.3 | Failure to fall >20% from baseline at 24 h | 12.1 | 92% | 63% | 84.5% | 78.6% | 60 days |
*95% CIs provided (in brackets) where available in text or via author correspondence.
NPV, negative predictive value; PPV, positive predictive value.