Demographic data, clinical outcomes, and sputum culture results in study participants grouped according to blood glucose quartiles
| Group 1(n = 69) | Group 2(n = 69) | Group 3(n = 75) | Group 4(n = 71) | p value | |
|---|---|---|---|---|---|
| Logistic regression was used for analysis of univariate relationships between glucose (divided by quartiles into four exposure levels) and categorial variables. Analysis of variance was used to compare age between the four groups and Kruskal-Wallis testing used to compare length of stay (not normally distributed) between the groups. | |||||
| Demographic data | |||||
| Blood glucose (mmol/l) | <6.0 | 6.0–6.9 | 7.0–8.9 | >9.0 | |
| Mean (SD) age (years) | 73.1 (10.7) | 72.9 (11.1) | 76.7 (8.6) | 75.0 (10.5) | 0.087 |
| Sex (M:F) | 43:26 | 39:30 | 43:32 | 42:29 | 0.743 |
| Pre-existing diagnosis of diabetes, n (%) | 1 (1.4%) | 1 (1.4%) | 3 (4%) | 10 (14.1%) | 0.002 |
| Clinical outcomes | |||||
| Composite adverse outcomes, n (%) | 28 (41%) | 37 (54%) | 45 (60%) | 51 (72%) | 0.0001 |
| Mortality, n (%) | 8 (12%) | 11 (16%) | 16 (21%) | 22 (31%) | 0.003 |
| Median (IQR) length of stay (days) | 7 (4–14) | 9 (5–16) | 10 (6–22) | 12 (5–21) | 0.087 |
| Sputum culture results | |||||
| At least one pathogen, n (%) | 14 (56%) | 17 (61%) | 19 (66%) | 22 (73%) | 0.154 |
| Multiple pathogens, n (%) | 3 (12%) | 5 (18%) | 8 (28%) | 10 (33%) | 0.031 |
| Staphylococcus aureus, n (%) | 1 (4%) | 3 (11%) | 8 (28%) | 8 (27%) | 0.011 |
| Yeasts, n (%) | 0 (0%) | 3 (11%) | 2 (7%) | 5 (17%) | 0.065 |