2.1

A questionnaire-based PROM, with a 5-point Likert scale (0−4) was chosen .

2.2

In the first stage of the initial validation, potentially relevant items from existing questionnaires were identified . Given the low relevance of items related to sports activities for some diagnostic groups, it was decided at this point to develop two separate scores: a general item score and a sports-specific score . In total, 31 general items and 7 sports-specific items were taken forward into the second phase of the project .

2.3

Through a process of forward and backward translation performed by bilingual translators, the original English pool of 38 items was translated into German, French and Swedish . These four language versions were then used for the Stage 2 data collection . Participants were recruited from orthopaedic foot and ankle surgery departments . Inclusion criteria for participants were clinical and imaging indications for foot and ankle surgery and age ≥ 18 years . No exclusion criteria were used other than an inability to complete a written questionnaire . Data collection was performed in France, Germany, Sweden and Ireland . In addition to providing an answer to each item on a 5-point scale, all participants also rated the relevance of the item to their situation on a 5-point scale .

Following data collection, the following analytic steps were taken to reduce the item pool into one general PROM and one sports PROM .

• 1.

  Items with a ceiling effect, low perceived relevance and a high proportion of missing values were noted and shortlisted for exclusion in subsequent steps .

• 2.

  A principal component analysis (PCA) was performed . At the end of this step, the remaining items in their respective principal components would provide optimal scale reliability according to classic test theory .

• 3.

  An item-response theory (IRT) analysis was performed for each of the identified scales (i.e., principal components) to further reduce the number of items and optimize scale unidimensional l .

2.4

Data collection for this final stage of the initial validation took place in the four original language versions, as well as Dutch, Italian and Polish .

2.5

Data collection stage of the validation was performed in Norway. Inclusion criteria for participants were being scheduled for foot and ankle surgery and age ≥ 18 years. No exclusion criteria were used other than an inability to complete a written questionnaire. Data were collected pre-operatively and at postoperative follow-up. A minimum postoperative follow-up of 3 months and mean follow-up of 6 months were planned, with a target of at least 100 completed score sheets. To confirm the internal consistency for each language version, Cronbach’s Alpha of the EFAS Score was computed for each language version separately . To establish the responsiveness of the EFAS Scores, both distribution-based and criterion-based analyses were used . Distribution-based measures of responsiveness included the effect size (ES) and minimal important difference (MID) . The criterion-based measure of responsiveness used was the linear association (Spearman correlation) between improvement on the EFAS Score and a 5-point Likert scale anchor question: did the surgery improve the foot and/or ankle problem? (0 = no, not at all; 4 = yes, very much) .

The ES was calculated as the difference between the baseline and three to six-month follow-up mean EFAS Score, divided by the standard deviation of the baseline EFAS Score .

The MID was considered to be equal to the standard error of measurement (SEM) of the baseline EFAS Score. The SEM was calculated as :

SD = standard deviation of the EFAS Score baseline score

r = value of Cronbach’s Alpha for the EFAS Score at baseline.

To assess the responsiveness of the EFAS Score using the MID, the percentage of participants with an improvement in their EFAS Score between baseline and follow-up exceeding the MID was identified .

Statistical analyses were performed in SPSS (IBM SPSS Statistics 28.0.1, IBM, Armonk, NY, USA). The IRT modelling was performed in XCalibre 4 (Assessment Systems, Stillwater, MN, USA).