First author, year, reference | Marker | Study type | Study population | Key findings | Limitations |
---|---|---|---|---|---|
Foltran et al., 2013, [112] | Glucose | Single-center, retrospective case-control study | 1,018 Italian non-intensive care patients with pneumonia | - Plasma glucose levels of mean 86 mg/dl (95 % CI, 61–102 mg/dl) were associated with minimal risk of in-hospital mortality | - Only admission glucose levels |
- Single-center study | |||||
- No information on important confounders (e.g. diabetes) | |||||
- The OR was 1.33 (95 % CI 1.07–1.66) for each 10 mg/dl of increase in plasma glucose in hyperglycemic patients (>86 mg/dl) | |||||
McAlister et al., 2005, [113] | Glucose | Multicenter, observational prospective cohort study (median length of in-hospital stay was 6 days, follow-up until hospital discharge) | 2,471 Canadian adults with CAP | - Patients with an admission glucose level >11 mmol/l had an increased risk of death (13 vs. 9 %, p = 0.03) and in-hospital complications (29 vs. 22 %, p = 0.01) compared to those with glucose levels <11 mmol/l | - Glucose only measured once |
- No examination of antiglycemic treatments | |||||
- Not measured long-term glucose (HbA1c) | |||||
- Patients with admission glucose >11 mmol/l had 73 % higher mortality (95 % CI 12–108 %) and a 52 % higher (12–108 %) risk of in-hospital complications compared to patients with glucose levels ≤6.1 mmol/l | |||||
Lepper et al., 2012, [114] | Glucose | Multicenter, prospective cohort study (6-month follow-up) | 6,891 German patients with CAP (CAPNETZ Study) | - In patients with no pre-existing diabetes an increased serum glucose level was a predictor of death at 28 and 90Â days | - Treatment of pneumonia was left to the discretion of the doctor |
- Mild hyperglycemia (6-10.99 mmol/l) on admission was associated with an increased risk of death at 90 days (HR 1.56, 95 % CI 1.22–2.01, p < 0.001), and this risk increased to 2.37 (95 % CI, 1.62–3.46, p < 0.001) when serum glucose levels were ≥14 mmol/l | - Did not examine changes in serum glucose | ||||
- Glucocorticoid treatments were not recorded | |||||
- No HbA1c tests |