Revisions of knee arthroplasties due to pain; survivorship, use of analgesics, patient-reported outcomes and validation of the indication

Background and aimUp to 20% of patients experience persistent knee pain after insertion of a primary knee arthroplasty. Some patients are revised because of pain without any other obvious knee pathology present. It is unknown if these patients benefit from revision. Therefore, this thesis aimed to investigate ”pain without loosening” as indication for revision knee arthroplasties.

MethodsWe identified 4,456 procedures of first time knee arthroplasty revisions for the indications “pain without loosening” and “aseptic loosening” in Denmark in 1997-2020 from the Danish Knee Arthroplasty Register (DKR). 1,825 revisions were performed for the indication “pain without loosening” and 2,631 revisions were performed for the indication “aseptic loosening”.  We conducted four studies based on data from the DKR. Study 1 validated the indication “pain without loosening” from medical records, radiographs, and computerized tomography (CT) scans. Study 2 investigated the re revision rate of the pain revisions compared to the better established indication “aseptic loosening”. Study 3 investigated use of analgesics one year before and after revision for the indications “pain without loosening” and “aseptic loosening”. Data from the Danish National Patient Registry were used in study 2 and 3 and study 3 further required data from the Danish National Prescription Registry. Study 4 investigated patient-reported outcomes (PROMs) 1-3 years after revision comparing “pain without loosening” versus “aseptic loosening”.
ResultsThe indication “pain without loosening” covered knee arthroplasties revised because of pain in 99% of the investigated cases. We found hidden indications in 42% of these, with stiffness and prosthesis malposition occurring most frequently. We found similar re-revisions rates of about 23% (CI 20-25) and 19% (CI 18-21) for “pain without loosening” versus “aseptic loosening” with a 20-year follow-up. The analgesic consumption did not change considerably after revision for any of the indications. 9% and 8% of the revised patients for each indication respectively became new long-term users of opioids after revision. Patients revised for “pain without loosening” scored significantly worse on PROMs than patients revised for “aseptic loosening”, and a larger proportion of pain patients were unsatisfied with the result of the revision.
Conclusion and perspectivesThe indication “pain without loosening” in the DKR identifies pain revisions, but a broad variety of other underlying indications were present as well. Stiffness and malposition of components lack as indication options in the DKR, and implementation of these indications would strengthen the register. Further, the register data would improve if pre- and postoperative PROMs were captured routinely.Revision for “pain without loosening” performed similar to revisions for “aseptic loosening” regarding prosthesis survival and use of analgesics. A large proportion of long-term opioid users were generated after revision for both indications, but the pain revision patients scored worse on PROMs and were less satisfied.Therefore, revising for the indication “pain without loosening” should be carefully considered, and in most cases avoided, when no obvious knee pathology is present.