A comparison of postoperative microbiological and histological test results with the indication for revision as listed at the time of surgery in the National Joint Registry: a single-centre cohort study of metal-on-metal total hip replacement and hip resurfacing revision procedures.

Introduction: Surgeons allocate an indication(s) for revision surgery on NJR Minimum Data Set forms in the immediate perioperative period, when results from intraoperative histology and microbiology samples are not available. We evaluated to what extent the postoperative diagnostics obtained for patients undergoing metal-on-metal hip arthroplasty (MoMHA) revision surgery align with the indication for revision listed in the NJR.

Methods: NJR data for MoMHA revision patients (2004-2015) at a single centre were linked to hospital records of microbiology and histology testing for prosthetic joint infection (PJI) and adverse reaction to metal debris (ARMD). The Musculoskeletal Infection Society and Mirra classifications were used as "gold standards' to classify PJI and ARMD respectively and to assess the diagnostic accuracy of the clinician selected revision indication recorded on MDS forms.

Results: Of 301 cases, 5.6% (n = 17) and 3.7% (n = 11) were revised for PJI and ARMD respectively. In a further 6.6% (n = 20, PJI) and 15.6% (n = 47, ARMD) the indication for revision selected at the time of surgery did not corroborate with the test results from intraoperative samples. Diagnostic accuracy analysis revealed sensitivity and specificity of indication for surgery at the time of revision for PJI of 35.5% (95% CI, 19.2-54.6) and 97.8% (CI, 54.6-95.2) respectively, and for ARMD of 7.41% (CI, 2.06-17.9) and 97.2% (CI, 94.3-98.9) respectively.

Conclusions: We observed that the results of histology and microbiology analyses of intraoperative samples may provide evidence to support alternative diagnoses in 20% of cases. This indicates a need for a change in the work-up of these cases to allow a more accurate prediction of the indication for revision. Alternatively, if there was evidence of a difference in outcome for patients where the indication may be changed when such data becomes available, the feasibility of a facility to add additional information from samples taken intraoperatively when it becomes available should be explored.