The Risk of Developmental Dysplasia of the Hip in Premature Infants with Breech Presentation at Birth

Abstract Objective  This article evaluates the risk (defined here as incidence per 1,000 infants) of developmental dysplasia of the hip (DDH) in premature infants undergoing hip ultrasound (HUS) for breech presentation only compared with other indications. Study Design  Retrospective study of infants born between January 1, 2009 and December 31, 2018 at <37 weeks' gestation with a HUS obtained in the first year of life, at Nationwide Children's Hospital, the only available facility for pediatric radiology and orthopaedic services in central Ohio. We calculated risk differences and odds ratios of DDH by the indication of HUS. Results  From 2,397 infants reviewed, 89% underwent HUS for breech presentation only. The local incidence of DDH for infants undergoing HUS for breech-only indication was 15 per 1,000 compared with 155 per 1,000 for infants undergoing HUS for other indications. The odds ratio of DDH diagnosis was 12.1 (95% confidence interval: 7.5, 19.6) for infants undergoing HUS for an indication other than breech presentation only relative to infants undergoing HUS for the indication of breech presentation only. Conclusion  The risk of DDH diagnosis in premature infants undergoing HUS for breech presentation is much lower than those undergoing HUS for clinical concerns and other risk factors. Screening these infants with physical examination may be sufficient. Key Points Breech presentation is the most common indication for hip ultrasound in premature infants. The risk of DDH is lower in premature infants with breech presentation compared with premature infants with other clinical concerns. The risk of DDH is higher in infants born at 33 to 36 weeks of gestation compared with those born at an earlier gestation.

Developmental dysplasia of the hip (DDH) is an abnormal hip development during the fetal, neonatal, and postnatal periods that leads to acetabular dysplasia.DDH includes a spectrum of physical and radiological findings.Severitybased broad categories of DDH are subluxation and dislocation.The femoral head moves within the acetabulum in a subluxated hip.However, it moves outside the acetabulum in a dislocated hip. 1 DDH may lead to avascular necrosis of the femoral head, difficulty walking, and significant morbidity.Risk factors for DDH include female sex, family history of DDH, and incorrect swaddling postnatally. 1 A recent study determined that female sex, breech presentation, and family history were independently associated with abnormal hip ultrasound (HUS) findings. 2The risk of dislocation is higher during the final 4 weeks of a term pregnancy because of mechanical factors.The risk increases further with oligohydramnios, breech presentation at birth, and macrosomia. 3tudies in term infants with breech presentation showed increased risk for DDH up to five times the baseline risk. 4reech presentation increases with decreasing gestational age (GA) at birth, occurring in 24.4% of pregnancies at 28 weeks and decreasing to 3.7% at 37 weeks of gestation. 5,6he risk of DDH in premature infants is not well defined.Some studies suggested decreased risk of DDH in premature infants compared with term infants. 7There are conflicting reports on whether breech presentation increases the risk of DDH in premature infants the same way it does for term infants. 8,9National guidelines recommend clinicians to consider HUS for term infants with breech presentation in the third trimester. 1,10However, these guidelines do not address whether clinicians should consider HUS for preterm infants with breech presentation and a normal physical examination. 1,10n this study, we evaluated the risk (defined here as incidence per 1,000 infants) of DDH diagnosis in premature infants who received HUSs.We compared the incidence of DDH diagnosis per 1,000 infants based on the indication for the ultrasound.Specifically, we estimated the incidence difference and odds ratios (ORs) of DDH diagnosis by the indication alone and by the indication stratified by sex.

Study Setting and Design
This is a retrospective cohort investigation based on a chart review performed at the Nationwide Children's Hospital (NCH).NCH and its off-campus facilities are the only available facilities for pediatric radiology and orthopaedic services in central Ohio.Almost all children born in central Ohio receive their imaging, diagnosis, treatment, and follow-up for DDH at NCH facilities.We identified eligible patients and reviewed their electronic medical records (EMRs) for at least 2 years after birth.We manually extracted study variables to an Excel data spreadsheet.The NCH's Institutional Review Board (IRB) approved the study.The IRB granted a waiver of Health Insurance Portability and Accountability Act authorization and informed consent.The study was conducted in accordance with the Declaration of Helsinki.

Patients
We screened the EMRs of all infants born between January 1, 2009 and December 31, 2018 to identify study subjects.The following were the inclusion criteria: (1) GA of < 37 weeks at birth and (2) a HUS is performed during the first year of life.The following were the exclusion criteria: (1) the HUS was performed after the infant was started on DDH treatment or (2) the first HUS was not performed at NCH. Study variables and definitions are provided in ►Table 1.

Statistical Analysis
We stratified the patients based on the indication for their first HUS into: breech only, breech plus, and all other indications.For infants in the breech-only group, the only indication for the HUS was breech presentation.For infants in the breech plus group, breech was at least one of the indications listed.The other indications group contained infants who did not have breech as an indication and included abnormal physical exam only, congenital anomalies associated with HUS such as myelomeningocele, and other rare indications such as transverse lie.Patients' demographics were summarized using count and percent for categorical variables and median (25th and 75th percentile) for continuous variables.The incidence differences per 1,000 using the average marginal effect, and ORs along with 95% confidence intervals (CIs) of DDH diagnosis comparing indications for a HUS were estimated from logistic regression models. 11Due to small number of infants who had breech plus abnormal physical examination, those infants were combined with the all other indications category in the regression models.Evidence suggests that breech position has a different effect by sex, we therefore included indication and sex as an interaction term in some of the models. 12GA and birth weight (BW) might confound the association between indication and risk of DDH. 1 We present four models: (1) indication as the only covariate (breech-only vs. all other indications), (2) indication and infant sex, (3) indication, infant's sex, and the interaction, and (4) the interaction model additionally adjusted for GA at birth and BW.Data processing was done using SAS software (SAS institute, Cary, NC).Analyses were conducted in R and the margins package was used to estimate the incidence differences from the logistic regression models. 11

Study Subjects
We identified 2,446 unique patients who fulfilled the inclusion criteria.We excluded three patients who were treated for DDH based on clinical diagnosis before the first HUS and one patient who started treatment at a different institution before transferring care to our hospital.Out of the 2,442 remaining, 45 had missing BWs and were excluded from the analysis set.Our final sample had 2,397 patients.We provide the details of the number of patients we identified, excluded, and analyzed in ►Fig. 1.

Descriptive Data
In this sample, 2,133 (89%) had breech presentation at birth as the only indication for HUS, 23 (1%) had breech plus another indication, and 241 (10%) did not have breech as an indication for HUS.An abnormal physical exam was the only indication for 8% of the patients, Two percent of the patients had an indication other than breech presentation and/or abnormal physical examination, and these indications included, but are not limited to, congenital anomalies and transverse lie.We provide the baseline characteristics of the study subjects stratified by the indication for the HUS in ►Table 2. ►Fig. 2 provides the number of patients and the incidence per 1,000 infants of DDH for each completed weeks of GA for the study subjects.There is a consistent increase in the risk of DDH from 33 to 36 weeks (►Fig. 2).►Fig. 3provides the DDH diagnosis by sex, GA group, and indication for HUS.

Main Results
The incidence of DDH diagnosis was 155.3 per 1,000 infants for those with other indications while the incidence was 15 per 1,000 for infants with only a breech indication (unadjusted incidence difference: 140.3 [95% CI: 96.3, 184.3],OR: 12.1 [95% CI: 7.5, 19.6], ►Table 3).These estimates were derived from the number of infants undergoing HUS only, not the overall population of all preterm infants.When stratified by sex (by adding the interaction term to the model), females had a higher absolute incidence and higher odds of DDH when other indications for HUS were present, compared with breech-only, than males (adjusted incidence difference: 148.

Discussion
The true incidence of the full spectrum of DDH in the neonatal period is unknown because of the absence of a uniform definition. 1 There is a wide variation in the reported incidence of DDH.The reported incidence is higher when an ultrasound is the screening method rather than a physical examination. 13The American Academy of Pediatrics (AAP) Clinical Practice Guideline: Evaluation and Referral for Developmental Dysplasia of the Hip in Infants estimates clinical hip instability to occur 1 to 2% for term infants.It estimates abnormality detected by imaging to occur in up to 15% of infants.It reports an estimated risk of DDH in 2 to 27% for infants with breech presentation.The American Academy of Orthopedic Surgeons (AAOS) clinical practice guideline about DDH acknowledges that the true incidence of DDH is uncertain.However, it reports the incidence at 5 to 7% of all infants from large ultrasound screening studies. 10Paton et al reported a DDH rate of 37 per 1,000 infants referred for breech presentation. 14Rosendahl et al reported a treatment rate of 34 per 1,000 infants undergoing HUS for universal screening. 15ur results show the risk of DDH is low in premature infants with breech presentation and no other clinical concern.A recent study by Leonard and Kresch found that preterm infants with DDH have the same likelihood of vertex and breech presentation. 9However, this study lacked a uniform screening method for infants born in breech position and infants born in vertex position.Based on the current practice, infants born in breech position are more likely to Fig. 2 The number of patients and the probability of developmental dysplasia of the hip (DDH) for each completed weeks of gestational age for all study subjects across all indications for hip ultrasound (HUS).
Developmental Dysplasia of the Hip in Premature Infants Osman et al. e2393 This document was downloaded for personal use only.Unauthorized distribution is strictly prohibited.
have a HUS and more likely to have an abnormality detected than infants born in vertex position.This makes it difficult to draw a conclusion on the risk of breech presentation in this population. 9Our results are consistent with the study by Lange et al that reported a lower incidence of DDH in premature infants based on universal screening with HUS. 7 Most of the infants in our study underwent HUS for the indication of breech presentation only.This contrasts with the findings of Paton et al for term infants referred for HUS. 16n that study, a breech presentation was the indication for 48% of infants, while clinical concern was the indication for 25%. 16Degnan et al found breech presentation was the indication for HUS for 20.4% of the patients, hip deformity for 27.4%, and abnormal physical examination for 17.7%. 17he probability of breech presentation is higher in premature infants than in term infants. 5This likely makes breech presentation the commonest indication for HUS in preterm infants.
The strengths of this study include the large number of study subjects and the review of medical records for the first 2 years of their lives.Additionally, NCH provides all the pediatric radiological and orthopaedics services in central Ohio, which means the relevant data are available.The limitations include the retrospective observational nature of the study, with inherent limitations including confounding factors such as variations in care between individual  clinicians and over the study period.While almost all pediatric patients in central Ohio receive their radiological and orthopaedics services at NCH, some patients might have received care somewhere else and had a DDH diagnosis that was not recorded in their NCH EMR.The calculated incidence rate of DDH does not reflect a population level risk, because the denominator was the total number of infants undergoing HUS study rather than the total number of preterm infants in the local population.

Conclusion
The goal of screening for DDH is to detect undiagnosed cases and allow earlier less-aggressive interventions to achieve hip reduction.The optimal method to screen for DDH is controversial. 10,18,19The AAP guideline recommends screening all newborns for DDH with a physical examination by a pediatrician.Additionally, the guideline recommends considering a HUS for infants with breech presentation and a normal physical examination. 1The AAOS also recommends imaging before 6 months of age in infants with risk factors including breech presentation. 10A recently published large meta-analysis found a higher rate of early diagnosis and nonoperative treatments with universal HUS screening. 13ate diagnosis and operative treatment were similar in universal HUS compared with screening with physical examination and selective HUS. 13 Another study with a smaller number of patients found no late DDH diagnoses in patients with risk factors and negative HUS. 20 There were reports of delayed diagnosis of DDH during the coronavirus disease 2019 pandemic. 21Considering the relatively low risk of DDH in premature infants with breech presentation, serial physical examination rather than HUS might be an effective screening strategy for this group of infants.This also saves time and resources for the families and health care.

Fig. 3
Fig. 3 Percentage of developmental dysplasia of the hip (DDH) diagnosis by sex, gestational age group, and indication for hip ultrasound.

Table 2
Patient characteristics overall and by the indication for hip ultrasound (n ¼ 2,397) Breech and abnormal physical examination in 20 patients.Breech and other indications including congenital anomalies in 3 patients.
Abbreviations: GA, gestational age; HUS, hip ultrasound.a Median (25th and 75th percentile).b Other race includes mixed race, Nepali, and Asian.c

Table 3
Incidence difference per 1,000 infants, odds ratios, and 95% CI of DDH from logistic regression models comparing breech only as an indication for HUS versus all other indications overall and by sex (n ¼ 2,397) Abbreviations: CI, confidence interval; DDH, developmental dysplasia of the hip; GA, gestational age; HUS, hip ultrasound.a Unadjusted model, indication only.b Indication and infant sex.c Main effects and interaction of indication and infant sex.d Main effects and interaction of indication and infant sex, adjusted for GA and birth weight.