117 / 225
DOS Kongressen 2017 ·
117
Incidence of hip and knee replacements in rheumatoid
arthritis patients following introduction of biological
DMARDs: an interrupted time series analysis using na-
tionwide Danish health care registers
René Cordtz, Samuel Hawley, Daniel Prieto-Alhambra, Kristian Zobbe, Pil Højgaard, Lars
Erik Kristensen, Søren Overgaard, Anders Odgaard, Lene Dreyer
Department of Rheumatology, Gentofte Hospital, Center for Rheumatology and
Spine Diseases, Rigshospitalet; Musculoskeletal Pharmaco- and Device Epidemiology,
NDORMS, University of Oxford; Musculoskeletal Pharmaco- and Device Epidemiology,
NDORMS, University of Oxford; Department of Rheumatology, Gentofte Hospital, Cen-
ter for Rheumatology and Spine Diseases, Rigshospitalet; The Parker Institute, Bispebjerg
and Frederiksberg Hospital; The Parker Institute, Bispebjerg and Frederiksberg Hospi-
tal; Department of Orthopaedic Surgery and Traumatology, Odense University Hospital;
Department of Orthopaedic Surgery, Copenhagen University Hospital Herlev-Gentofte;
Department of Rheumatology, Gentofte Hospital, Center for Rheumatology and Spine
Diseases, Rigshospitalet
Background:
Previous data have been conflicting regarding an impact of biological Dis-
ease Modifying Anti-Rheumatic Drugs (bDMARDs) on the incidence rate (IR) of total
hip/total knee replacement (THR/TKR) in rheumatoid arthritis (RA) patients.
Purpose / Aim of Study:
To investigate the impact of bDMARD introduction for the
treatment of RA on the IR of THR and TKR compared with general population compara-
tors (GPC).
Materials and Methods:
Interrupted time-series analysis using the National Patient
Register. Each incident RA patient diagnosed at a rheumatology department from 1996-
2011 was matched with 10 GPC. We calculated 5-year age- and sex-standardised IR
of THR and TKR for RA patients and GPC diagnosed/matched in each 6-month period
from 1996-2011. Trends in the pre-bDMARD era (1996-2001) were compared with
trends in the bDMARD era (2003-16) using segmented linear regression and a 1-year
lag period (2002-03) at the time of bDMARD implementation.
Findings / Results:
We identified 30 868 incident RA patients (mean age 58 years,
70% women) and 301 527 GPC. THR: For GPC, the IR increased throughout the entire
study period (1996 IR: 2.9/1000 PY; year 1996-2001: +0.11/1000 PY; year 2003-
16: +0.02/1000 PY). For RA patients, the IR decreased from 1996 to 2016 (1996 IR:
8.7/1000 PY; year -0.36/1000 PY). TKR: The IR increased among RA patients from
1996-2001 (1996 IR: 5.9/1000 PY; year +0.19/1000 PY), but immediately started
decreasing from 2003 (year -0.20/1000 PY). The IR increased in GPC throughout the
entire study period (1996 IR: 0.4/1000 PY; year 1996-2001: +0.21/1000 PY; year
2003-16: +0.08/1000 PY).
Conclusions:
In 1996, the IR of THR and TKR was 3 and 15-fold higher among RA
patients compared with GPC. In RA patients, bDMARD introduction was associated with
decreasing IR of TKR, but not THR. IR of THR and TKR increased for GPC throughout the
entire study period.
No conflicts of interest reported
69.