401 patients were recruited between 2011 and 2013 with a median age of 63 years; 79% were male; 85.5% had type 2 diabetes; 27.5% presented with a recurrent ulcer; and 45.5% had a neuro-ischaemic ulcer, 50.5% neuropathic, and 3.5% ischaemic.
Both swab and tissue reports were available for 395 patients, and at least one pathogen was reported in 70.1% of swab samples and 86.1% of tissue samples. In 58% of patients the two samples resulted in a difference in the reported pathogens, with: 13.2% both reporting different pathogens; 36.7% reporting additional pathogens in the tissue sample; and 8.1% reporting additional pathogens in the swab sample.
In the most prevalent pathogens (those present in >8%), there were significantly higher rates of reporting of the following pathogens from tissue samples than swabs (McNemar’s p-value <0.05): Gram Positive Cocci, Gram Negative Bacilli, Enterobacteriacea, Anaerobes, Streptococcus, Enterococcus, Coagulase-Negative Staphylococcus, Gram Positive Bacilli, Corynebacterium; with differences ranging between 3.3% (Streptococcus) and 13.7% (Gram Positive Cocci). There was no evidence of a difference in reporting for Methicillin-resistant S. Aureus (MRSA) (p=0.22); Staphylococcus Aureus and Pseudomonas (the latter two both had p=1.00). In terms of agreement, the PABAK ranged from 0.58 (Gram Positive Cocci) to 0.97 (MRSA).
Investigation of the influence of baseline factors on agreement of the number of reported pathogens found, despite large centre variation, significant evidence (p=0.02) of an association with ulcer duration, such that older ulcers ( ≥56 days old) had a reduced odds (0.64 95% CI (0.45, 0.95)) of the tissue sample reporting more pathogens than the swab.