To compare pre-referral microbiology and previous bone excision in long bone osteomyelitis with intra-operative microbiology from a specialist centre.
A prospective observational cohort study of patients referred to a single tertiary centre who met the following criteria: (i) aged ≥18 years, (ii) received surgery for long bone osteomyelitis and (iii) met diagnostic criteria for long bone osteomyelitis. Patient demographics, referral microbiology and previous surgical history were collected at the time of initial clinic appointment. During surgery, a minimum of 5 intra-operative deep tissue samples were sent for microbiology. Antimicrobial options were classified from the results of susceptibility testing using the BACH classification of long bone osteomyelitis as either Ax (unknown or culture negative), A1 (good options available) or A2 (limited options available). The cultures and susceptibility of pre-referral microbiology were compared to the new intra-operative sampling results. In addition, an association between previous osteomyelitis excision and antimicrobial options were investigated.
79 patients met inclusion criteria during the study period. From these, 39 (49.4%) patients had information available at referral regarding microbiology obtained from either sinus swab (n=16), bone biopsy (n=11), previous osteomyelitis excision sampling (n=7), aspiration (n=4) or blood culture (n=1). From these 39 patients, microbiology information at referral fully matched microbiology samples taken at operation in 8 cases (20.5%). Fifteen of the 39 patients (38.5%) had a different species isolated at surgery compared to referral microbiology. The remaining 16 patients (41.0%) had a culture-negative osteomyelitis on surgical sampling. Based on the microbiology obtained in our centre, 35 patients were classified as A1 (44.3%), 15 as A2 (18.9%) and 29 as culture negative, Ax (36.7%). Patients who had received previous excision of osteomyelitis before referral (n=32, 40.5%) had an increased odds ratio (OR) of having microbiology with limited antimicrobial options compared to those undergoing primary osteomyelitis excision (OR: 3.8, 95% CI 1.2 – 11.2, P=0.023, Fisher's exact test).
Patients are frequently referred with limited microbiological information. Pre-referral microbiology in long bone osteomyelitis correlated with intra-operative samples taken at our centre in less than one quarter of cases. Pre-referral microbiology data should be used with caution for planning treatment in osteomyelitis. Previous surgery for osteomyelitis was associated with microbiology culture with limited antimicrobial treatment options.