Predicting developmental dysplasia of the hip in at-risk newborns

Andreas Roposch,Olivia Malaga-Shaw,Deborah Ridout,Paul Humphries,Yael Gelfer,John H Wedge,Evangelia Protopapa

doi:10.1186/s12891-020-03454-4

Abstract

BackgroundThe development of developmental dysplasia of the hip can be attributed to several risk factors and often in combination with each other. When predicting the likelihood of developing this condition, clinicians tend to over and underestimate its likelihood of occurring.Therefore, the study aim is to determine among at-risk newborns how to best predict developmental dysplasia of the hip (DDH) within 8 weeks post-partum.MethodsProspective cohort study in secondary care. Patient population included newborns at-risk for DDH – we assessed 13,276 consecutive newborns for the presence of DDH risk factors. Only newborns with at least one of the predefined risk factors and those showing an abnormal examination of the hip were enrolled (n = 2191). For the development of a risk prediction model we considered 9 candidate predictors and other variables readily available at childbirth.The main outcome measure was ultrasonography at a median age of 8 weeks using consensus diagnostic criteria; outcome assessors were blinded.ResultsThe risk model includes four predictors: female sex (OR = 5.6; 95% CI: 2.9–10.9; P < 0.001); first degree family history of DDH (OR = 4.5; 95% CI: 2.3–9.0; P < 0.001), birthweight > 4000 g (OR = 1.6; 95% CI: 0.6–4.2; P = 0.34), and abnormal examination of hip (OR = 58.8; 95% CI: 31.9, 108.5; P < 0.001). This model demonstrated excellent discrimination (C statistic = 0.9) and calibration of observed and predicted risk (P = 0.35). A model without the variable ‘hip examination’ demonstrated similar performance.ConclusionThe risk model quantifies absolute risk of DDH within 8 weeks postpartum in at-risk newborns. Based on clinical variables readily available at the point of childbirth, the model will enhance parental counselling and could serve as the basis for real time decisions prior to discharge from maternity wards.

Highlights

The development of developmental dysplasia of the hip can be attributed to several risk factors and often in combination with each other
We retained 4 in the risk prediction model: female sex (OR = 5.6, 95% CI, 2.9–10.9; P < 0.001); family history of developmental dysplasia of the hip (DDH) (OR = 4.5, 95% CI, 2.3– 9.0; P < 0.001); birthweight > 4000 g (OR = 1.6, 95% CI, 0.6–4.2; P = 0.34); abnormal examination of hip (OR = 58.8, 95% CI, 31.9–108.5; P < 0.001) (Table 3)
This model discriminated between cases with and without DDH (C statistic 0.7, 95% CI, 0.7–0.8; goodness-of-fit P = 0.76)

Summary

Introduction

The development of developmental dysplasia of the hip can be attributed to several risk factors and often in combination with each other. When predicting the likelihood of developing this condition, clinicians tend to over and underestimate its likelihood of occurring. Less widely accepted factors include foot deformities, [2, 3] high birth weight, [4] and torticollis [5]. With use of these factors clinicians on maternity wards estimate the likelihood of DDH to be present or to develop; counsel parents of affected newborns; and make triage decisions. A risk model that could help clinicians making predictions about DDH for individual at-risk newborns would be a useful tool for clinicians

Objectives

Methods

Results

Discussion

Conclusion