Using Patient-Reported Outcomes to Predict Revision Arthroplasty Following Femoral Neck Fracture: Enhancing the Value of Clinical Registries through Data Linkage

Christina L. Ekegren, Richard de Steiger, Elton Edwards, Richard S. Page, Raphael Hau, Susan Liew, Andrew Oppy, Belinda J. GabbeInternational Journal of Environmental Research and Public Health 2019 · DOI

TL;DR

This study demonstrates that patient-reported outcomes (PROMs), specifically pain and discomfort six months after a hip fracture, are strong predictors of future revision arthroplasty. By linking two large clinical registries, the research highlights how PROMs can significantly enhance the predictive value of existing healthcare data for better patient management.

What Problem Does This Paper Address?

Hip fractures are a major health concern, particularly in aging populations, leading to reduced quality of life and increased mortality. While initial surgical fixation aims to restore function, a significant number of patients require further surgery, such as revision arthroplasty, due to complications or implant failure. Predicting which patients are at higher risk for these subsequent procedures is crucial for optimizing surgical decision-making and resource allocation. Traditionally, risk assessment has relied on demographic data, pre-fracture health status, and radiological findings. However, these factors may not fully capture a patient's post-injury experience, which can significantly impact their need for further intervention. This paper addresses the gap in understanding how a patient's subjective experience of pain and functional limitations, as captured by PROMs, can serve as a powerful predictor of future arthroplasty needs following a femoral neck fracture.

Methods

The researchers employed a retrospective cohort study design, linking data from two large Australian clinical registries: the Victorian Orthopaedic Trauma Outcomes Registry (VOTOR) and the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR).

  • Data Linkage: The core methodological innovation was the secure linkage of these two datasets using unique patient identifiers. This allowed for the integration of detailed information on initial hip fracture management (from VOTOR) with subsequent joint replacement procedures (from AOANJRR).
  • Patient Selection: The study focused on patients aged 55 years and older who sustained a femoral neck fracture from a low fall between March 2007 and June 2015. Patients with pathological fractures or those requiring revision/conversion within six months of the initial surgery were excluded.
  • Outcome Measurement: The primary outcome was the time to revision arthroplasty or conversion to arthroplasty.
  • Predictor Variables: Six-month post-fracture PROMs, including measures of residence, mobility, personal care, usual activities, pain and discomfort, and anxiety/depression (primarily using the EQ-5D-3L), were collected via telephone interviews. Other factors included age, gender, pre-injury disability, Charlson Comorbidity Index (CCI), and the type of initial surgery.
  • Statistical Analysis: A multivariable Cox proportional hazards model was used to identify factors associated with future arthroplasty. This model allowed for the assessment of the time-to-event data while adjusting for potential confounding variables. A backward stepwise approach was used to refine the model.

Datasets

  • Victorian Orthopaedic Trauma Outcomes Registry (VOTOR): This registry collects data on hip fracture patients admitted to four major trauma hospitals in Victoria, Australia, including demographic information, injury details, surgical procedures, and six-month PROMs.
  • Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR): This registry captures data on almost all arthroplasty procedures performed in Australia, including details of the surgery, implant types, and reasons for revision.

The linkage of these two high-quality, population-level registries is a key strength, providing a comprehensive dataset for analysis.

Key Insights and Ideas

  • PROMs as Powerful Predictors: The most striking insight is the strong predictive power of patient-reported pain and discomfort six months post-fracture for future arthroplasty. This underscores the importance of subjective patient experience in clinical decision-making, moving beyond purely objective clinical or radiological markers.
  • Data Linkage Enhances Registry Value: The study demonstrates a highly effective method for enhancing the value of existing clinical registries. By linking trauma and joint replacement registries, researchers can gain a more complete picture of patient pathways and outcomes.
  • Pain is Paramount: While other PROMs (mobility, self-care, etc.) were assessed, only pain and discomfort showed a significant association with future arthroplasty. This suggests that pain is a critical indicator of underlying issues that may necessitate further surgical intervention.
  • Younger Age and No Pre-Injury Disability: These factors were also independently associated with a higher risk of future arthroplasty, suggesting that patients with greater functional demands and fewer pre-existing limitations may be more likely to seek or require revision surgery.

Key Results

  • Pain and Arthroplasty Risk: Patients reporting pain and discomfort six months post-fracture had a 9.5-fold increase in the risk of future arthroplasty (95% CI: 3.81, 23.67) compared to those reporting no pain.
  • Initial Surgery Type: Internal fixation for the initial hip fracture surgery was associated with a higher risk of future arthroplasty compared to cemented hemiarthroplasty.
  • Demographic Factors: Younger age and the absence of pre-injury disability were also significant predictors of future arthroplasty.
  • Other PROMs: Notably, PROMs related to mobility, self-care, usual activities, and anxiety/depression did not show a significant association with future arthroplasty in this study.

Limitations and Future Work

  • Small Number of Internal Fixation Patients: The study acknowledges a limitation due to the relatively small number of patients who underwent internal fixation as their initial procedure. This means that the analysis of revisions specifically for internal fixation failure might be less robust.
  • Observational Study: As an observational study, it can only demonstrate associations, not causality. Unexplained confounding factors may still be present.
  • ICD-10-AM Coding: The reliance on ICD-10-AM coding meant that information on fracture displacement, a known risk factor for surgical failure, was not available.
  • Future Work: The authors suggest that their findings reinforce the importance of routinely collecting and acting upon pain levels as a critical outcome measure following femoral neck fracture. The successful linkage of registries also opens avenues for future research into the long-term outcomes and predictive capabilities of PROMs in orthopaedic care.

Summary updated

Abstract

The aim of this study was to determine the association between patient-reported outcome measures (PROMs) six months following femoral neck fracture after a low fall and future arthroplasty, and the factors associated with this. Six-month post-fracture PROMs were collected from the Victorian Orthopaedic Trauma Outcomes Registry (VOTOR) for patients aged >55 years who were admitted for a femoral neck fracture after a low fall between March 2007 and June 2015. These cases were linked with those registered by Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) up to October 2016. Multivariable analysis was performed using a Cox proportional hazards model to determine factors associated with future arthroplasty, including six-month PROMs. Of the 7077 hip fracture patients registered by VOTOR during the study period, 2325 met the inclusion criteria. Internal fixation being used for the initial hip fracture surgery, being younger and having no pre-injury disability were all independently associated with future revision or conversion to arthroplasty. Out of all PROMs, reporting pain and discomfort six months post-fracture was associated with a 9.5-fold increase in the risk of future arthroplasty (95% CI: 3.81, 23.67). The value of clinical registries can be enhanced via data linkage, in this case by using PROMs to predict arthroplasty following femoral neck fracture.
BibTeX
@article{ekegren2019using,
  author = {Christina L. Ekegren and Richard de Steiger and Elton Edwards and Richard S. Page and Raphael Hau and Susan Liew and Andrew Oppy and Belinda J. Gabbe},
  title = {{Using Patient-Reported Outcomes to Predict Revision Arthroplasty Following Femoral Neck Fracture: Enhancing the Value of Clinical Registries through Data Linkage}},
  journal = {International Journal of Environmental Research and Public Health},
  year = {2019},
  doi = {10.3390/ijerph16081325},
}

Loading related content...

Figure