Table 4.
Multinomial and ordinal regression models for individually fitted baseline factors
| OR (95% CI) | AIC§ | Reduction in
−2 Log L |
df | P value | ||
| Multinomial summary of isolates | Both swab and tissue report the same pathogens. | |||||
| Null model | 941.29 | |||||
| Ulcer type* | 945.72 | 1.570 | 3 | 0.666 | ||
| Any ischaemic versus neuropathic only | Swab>pathogens compared with the tissue | 1.03 (0.48 to 2.20) | ||||
| Any ischaemic versus neuropathic only | Tissue>pathogens compared with the swab | 0.86 (0.53 to 1.40) | ||||
| Any ischaemic versus neuropathic only | Swab and tissue report totally different pathogens. | 0.68 (0.35 to 1.31) | ||||
| Ulcer grade | 949.16 | 4.125 | 6 | 0.660 | ||
| Grade 2 versus grade 1 | Swab>pathogens compared with the tissue | 0.68 (0.26 to 1.78) | ||||
| Grade 2 versus grade 1 | Tissue>pathogens compared with the swab | 1.08 (0.60 to 1.93) | ||||
| Grade 2 versus grade 1 | Swab and tissue report totally different pathogens. | 1.14 (0.51 to 2.54) | ||||
| Grade 3/4/5 versus grade 1 | Swab>pathogens compared with the tissue | 1.28 (0.52 to 3.11) | ||||
| Grade 3/4/5 versus grade 1 | Tissue>pathogens compared with the swab | 1.60 (0.87 to 2.95) | ||||
| Grade 3/4/5 versus grade 1 | Swab and tissue report totally different pathogens. | 1.55 (0.69 to 3.45) | ||||
| Previous antibiotic therapy* | 946.28 | 1.005 | 3 | 0.800 | ||
| Yes versus no | Swab>pathogens compared with the tissue | 0.80 (0.36 to 1.80) | ||||
| Yes versus no | Tissue>pathogens compared with the swab | 1.14 (0.69 to 1.89) | ||||
| Yes versus no | Swab and tissue report totally different pathogens. | 1.10 (0.56 to 2.16) | ||||
| Antimicrobial dressing* | 943.44 | 3.850 | 3 | 0.278 | ||
| Yes versus no | Swab>pathogens compared with the tissue | 1.13 (0.51 to 2.51) | ||||
| Yes versus no | Tissue>pathogens compared with the swab | 0.69 (0.40 to 1.19) | ||||
| Yes versus no | Swab and tissue report totally different pathogens. | 1.38 (0.66 to 2.89) | ||||
| Wound duration (median split)* | 941.48 | 5.802 | 3 | 0.121 | ||
| <56 days vs ≥56 days | Swab>pathogens compared with the tissue | 0.94 (0.43 to 2.04) | ||||
| <56 days vs ≥56 days | Tissue>pathogens compared with the swab | 1.75 (1.08 to 2.86)† | ||||
| <56 days vs ≥56 days | Swab and tissue report totally different pathogens. | 1.14 (0.59 to 2.17) | ||||
| Log wound duration (continuous)* | 944.97 | 2.318 | 3 | 0.509 | ||
| Swab>pathogens compared with the tissue | 0.95 (0.72 to 1.25) | |||||
| Tissue>pathogens compared with the swab | 0.88 (0.74 to 1.04) | |||||
| Swab and tissue report totally different pathogens. | 0.93 (0.74 to 1.18) | |||||
| Ordinal summary of isolates | ||||||
| Null model | 917.72 | |||||
| Ulcer type*: any ischaemic versus neuropathic only | 0.90 (0.61 to 1.33) | 919.45 | 0.271 | 1 | 0.603 | |
| Ulcer grade | 920.16 | 1.559 | 2 | 0.459 | ||
| Grade 2 versus grade 1 | 1.33 (0.82 to 2.15) | |||||
| Grade 3/4/5 versus grade 1 | 1.27 (0.78 to 2.07) | |||||
| Previous antibiotic therapy*: yes versus no | 1.25 (0.81 to 1.91) | 918.56 | 1.154 | 1 | 0.283 | |
| Antimicrobial dressing*: yes versus no | 0.76 (0.49 to 1.18) | 918.16 | 1.553 | 1 | 0.213 | |
| Wound duration (median split)*: <56 days vs ≥56 days | 1.56 (1.05 to 2.33) | 914.62 | 5.097 | 1 | 0.024‡ | |
| Log wound duration (continuous)* | 0.92 (0.80 to 1.05) | 918.15 | 1.571 | 1 | 0.210 |
Based on the evaluable population n=395.
*Factors with missing data from the 28 (7.1%) patients with at least one missing data item.
‡Significant at the 5% level.
§Smaller is better.
AIC, Akaike information criterion.