The aim of this study was to capture 12-month outcomes from a representative multicentre cohort of patients undergoing total ankle arthroplasty (TAA), describe the pattern of patient-reported outcome measures (PROMs) at 12 months, and identify predictors of these outcome measures. Patients listed for a primary TAA at 19 NHS hospitals between February 2016 and October 2017 were eligible. PROMs data were collected preoperatively and at six and 12 months including: Manchester-Oxford Foot and Ankle Questionnaire (MOXFQ (foot and ankle)) and the EuroQol five-dimension five-level questionnaire (EQ-5D-5L). Radiological pre- and postoperative data included Kellgren-Lawrence score and implant position measurement. This was supplemented by data from the National Joint Registry through record linkage to determine: American Society of Anesthesiologists (ASA) grade at index procedure; indication for surgery, index ankle previous fracture; tibial hind foot alignment; additional surgery at the time of TAA; and implant type. Multivariate regression models assessed outcomes, and the relationship between MOXFQ and EQ-5D-5L outcomes, with patient characteristics.Aims
Methods
The responsiveness of the Manchester–Oxford Foot
Questionnaire (MOXFQ) was compared with foot/ankle-specific and
generic outcome measures used to assess all surgery of the foot
and ankle. We recruited 671 consecutive adult patients awaiting
foot or ankle surgery, of whom 427 (63.6%) were female, with a mean
age of 52.8 years (18 to 89). They independently completed the MOXFQ,
Short-Form 36 (SF-36) and EuroQol (EQ-5D) questionnaires pre-operatively
and at a mean of nine months (3.8 to 14.4) post-operatively. Foot/ankle
surgeons assessed American Orthopaedic Foot and Ankle Society (AOFAS)
scores corresponding to four foot/ankle regions. A transition item measured
perceived changes in foot/ankle problems post-surgery. Of 628 eligible
patients proceeding to surgery, 491 (78%) completed questionnaires
and 262 (42%) received clinical assessments both pre- and post-operatively. The
regions receiving surgery were: multiple/whole foot in eight (1.3%),
ankle/hindfoot in 292 (46.5%), mid-foot in 21 (3.3%), hallux in
196 (31.2%), and lesser toes in 111 (17.7%). Foot/ankle-specific
MOXFQ, AOFAS and EQ-5D domains produced larger effect sizes (>
0.8)
than any SF-36 domains, suggesting superior responsiveness. In analyses
that anchored change in scores and effect sizes to patients’ responses
to a transition item about their foot/ankle problems, the MOXFQ
performed well. The SF-36 and EQ-5D performed poorly. Similar analyses,
conducted within foot-region based sub-groups of patients, found
that the responsiveness of the MOXFQ was good compared with the
AOFAS. This evidence supports the MOXFQ’s suitability for assessing
all foot and ankle surgery.
