# Weighted predictive modelling estimation of patient acceptable symptom state for forgotten joint score, Oxford hip score, and EuroQol health index 3 and 12 months after total hip arthroplasty in a United Kingdom cohort

## TL;DR

This study defines new thresholds for acceptable patient outcomes after hip replacement surgery using a novel method and data from over 2000 Scottish patients.

## Contribution

A new weighted predictive modelling method is introduced to estimate patient acceptable symptom state (PASS) thresholds for three PROMs after hip arthroplasty.

## Key findings

- PASS thresholds for FJS, OHS, and EQ-5D-5L increased between 3 and 12 months post-surgery, reflecting ongoing recovery.
- Male patients required higher scores to consider their outcomes satisfactory compared to female patients.
- The study provides specific score thresholds for acceptable outcomes at 3 and 12 months for three widely used PROMs.

## Abstract

Patient acceptable symptom state (PASS) enhances interpretation of patient-reported outcome measures (PROMs). However, very few studies have defined PASS values for widely used PROMs and are based on methods sensitive to distribution of PROM scores or weak correlation. This study utilises a new anchor-based method to estimate PASS thresholds for Forgotten Joint Score (FJS), Oxford Hip Score (OHS) and EuroQol health index (EQ-5D-5L) at 3 and 12 months after primary total hip arthroplasty (PTHA).

This retrospective cohort study used data for PTHAs performed at a high-volume arthroplasty centre in Scotland between April 2021 and March 2023. PROMs were FJS, OHS and EQ-5D-5L. A new weighted predictive modelling method was used to define PASS values. Anchor questions used were surgery-specific satisfaction for the FJS and OHS, and EQ-Visual Analogue Scale for the EQ-5D-5L.

A total of 2793 PTHAs were performed, of which 65 to 73% had complete PROMs data. Respective median age and BMI were 69 years and 29.1 kg/m² and 57% were female. The Spearman correlations between anchors and PROMs ranged between 0.35 and 0.54. PASS thresholds (95% CI) at 3 and 12 months respectively were 31.5 (29.5–33.9) and 38.3 (36.5–42.4) for FJS, 31.5 (29.1–32.8) and 36.2 (35.1–36.8) for OHS, 0.814 (0.795–0.822) and 0.867 (0.845–0.875) for EQ-5D-5L.

We report new PASS thresholds for FJS, OHS and EQ-5D-5L 3 and 12 months following PTHA. These thresholds reflect the symptom (health) state at which an average Scottish patient considers their outcome acceptable.

Patient-reported outcome measures (PROMs) are increasingly being used to capture patients’ perception of the impact of total hip replacement (THR) on their health-related quality of life. However, interpreting these scores in a clinically meaningful way is still a challenge. Patient acceptable symptom state (PASS) is a crucial way to interpret PROMs data. It specifies a threshold or cut-off value in a PROM, such that patients whose PROM scores are above the threshold are considered to have an acceptable outcome. Despite its usefulness, PASS values at early to mid-follow up for commonly used PROMs such as Oxford Hip Score (OHS), Forgotten Joint Score (FJS) and EuroQol Health Index (EQ-5D-5L) are scarce with existing studies relying on methods that are biased to distribution of PROM data or correlation strength between anchor variable and PROM score. In this study, we determined PASS values for the OHS, FJS and EQ-5D-5L at 3 and 12 months after THR using a new weighted predictive modelling approach and data from over 2000 patients in Scotland. Our results indicate that 3 and 12 months after THR respectively, an average patient in Scotland requires scores of at least 31.5 and 38.3 for the FJS, 32 and 36 for the OHS, and 0.814 and 0.867 for the EQ-5D-5L to perceive their outcome as acceptable. PASS thresholds increased between 3 and 12 months reflecting ongoing recovery process. Male patients require higher scores to consider their outcome satisfactory.