Introduction: Our aim was to question the usefulness of a three-phase bone scan in the evaluation of pain in the knee region after TKR. Our hypothesis was that an abnormal investigation had a poor association with the presence of infection or loosening, and did not provide any additional diagnostic information above that already available through other standard investigations.
Methods: A retrospective study over a 24-month period was performed comprising 118 patients investigated with a TPBS. Investigations were summarised and analysed, and were classified as entirely normal, possibly abnormal, and definitely abnormal.
Results: Thirty-three per cent (39/118) of TPBSs were reported as being entirely normal, 59% (69/118) as possibly abnormal, and 8% (10/118) as definitely abnormal. During the 24-month study period, 131 revision TKR procedures were performed at our institution; 9% (12/131) were investigated with TPBS and 91% (119/131) were not. No patient with an entirely normal pre-operative TPBS underwent revision TKR surgery. Eighty-five per cent (67/79) with an abnormal TPBS were managed conservatively. In our series, a TPBS had a positive predictive value of 2.53%, a negative predictive value of 100%, with an overall accuracy of 34.75% with 100% sensitivity (97.5% one-sided confidence interval 0-24.71%), and 33.62% specificity (95% confidence interval 53.29-72.37%), in the diagnosis of infection, or loosening with concurrent infection in determining the indication for revision surgery.
Conclusion: A TPBS should only be considered following clinical evaluation, serological investigation, diagnostic imaging, and microbiological analysis of fluid obtained from arthrocentesis by a specialist revision arthroplasty surgeon. A TPBS may be useful in the situation where abnormal serology is present, but where repeated joint aspirations samples are inconclusive.
Keywords: Cemented total knee replacement; Pain in the knee region; Revision arthroplasty; Three-phase Technetium-99m bone scan.