We report the largest multicentre series analysing the use of bone scans investigating painful post-operative Total Knee Replacements (TKR). We questioned the usefulness of reported scintigraphic abnormalities, and how often this changed subsequent management. 127 three-phase bone-scans were performed during a two-year period. Early and late flow phases were objectively classified. Reported incidences of infection and loosening were determined. Reports were subjectively summarised and objectively analysed to establish the usefulness of this investigation. Eight cases were excluded.

Scans were classified as: 33% (39) normal, 53% (63) as possibly abnormal, 6% (7) probably abnormal, and 8% (10) as definitely abnormal. Thirteen patients (11%) underwent revision TKR surgery. Intra-operative analysis revealed loosening of one femoral component, and massive metallosis of the patella in another. Cultures were negative in all cases. The sensitivity and specificity of a definitely abnormal investigation in predicting need for revision surgery was 23% and 82% (respectively). High instances of ambiguously reported abnormalities were observed.

This investigation has no role to play in the routine investigation of a painful TKR. It is unnecessary in investigation of periprosthetic infection and should not be used in a routine assessment of a painful TKR. If used it should be limited until an experienced revision surgeon has made a full assessment.