When and Why Is an Infant Hip Ultrasound Needed?

Infant hip ultrasound is a painless imaging scan used to check whether a baby’s hip joint is developing properly, and it is the most reliable tool available for catching a condition called developmental dysplasia of the hip (DDH) in the first months of life. DDH covers a spectrum of problems, from a slightly shallow hip socket to a fully dislocated joint, and early detection makes an enormous difference in outcomes. The scan works because a newborn’s hip is still mostly cartilage, which ultrasound can image clearly, unlike an X-ray, which cannot see unossified cartilage well until a baby is several months old.

Why Ultrasound Rather Than a Physical Exam Alone

Newborns are routinely checked for hip instability using hands-on maneuvers in the delivery room and at early pediatric visits. These physical tests look for a “clunk” that suggests the hip can be pushed out of or back into the socket. While a positive finding on these exams is a strong signal that something is wrong, the exams miss a lot. A systematic review of over 44,000 hips found these maneuvers had a sensitivity of only about 46%, meaning they caught fewer than half of the unstable hips, even though they were highly specific, rarely flagging a normal hip as abnormal.¹ The sensitivity also varied by the experience of the examiner, with senior clinicians performing somewhat better than trainees.²

Ultrasound fills this gap. It can visualize the actual shape and depth of the hip socket, the position of the femoral head inside it, and how stable the joint is under gentle pressure. This makes it far more sensitive for detecting milder forms of dysplasia that a physical exam would miss entirely. The combination of clinical examination and ultrasound is now part of screening in every European country that has a formal DDH program, though the details of who gets scanned and when vary widely.³

When the Scan Should Happen

Timing matters more than many parents realize. A baby’s hip socket ossifies rapidly during the first three months after birth, forming a cup-like cavity under the pressure of the femoral head sitting inside it.⁴ Scanning too early can be misleading because many newborn hips are still physiologically immature and look mildly abnormal on ultrasound even though they will develop normally without any intervention. Research comparing early ultrasound results to radiographic findings at one year found that scans done before the fourth week of life had significantly lower accuracy, while scans performed from the fourth week onward were accurate about 79% of the time or better.⁵

Most guidelines therefore recommend performing the screening ultrasound somewhere between four and six weeks of age. This window strikes a balance: the hip has matured enough that normal physiological immaturity has largely resolved, but it is still early enough to start treatment during the period when bracing is most effective. If a baby has clear clinical signs of instability at birth, an earlier scan or immediate referral to an orthopedic specialist is appropriate, but routine screening ultrasound at one or two weeks old leads to a high rate of findings that would have resolved on their own.

Who Should Get One

Not every country scans every baby. Some run universal programs where all newborns receive a hip ultrasound, while others use a selective or targeted approach that reserves the scan for babies with known risk factors. The risk factors themselves are well established. A meta-analysis found that breech presentation raised the odds of DDH roughly six-fold, a family history of DDH raised it nearly five-fold, and female sex raised it about four-fold.⁶ Clicking hips on clinical exam carried the strongest association of all. A second meta-analysis using relative risk found a similar hierarchy: breech presentation, female sex, left hip involvement, firstborn status, and family history all contributed independently.⁷

The left hip is affected more often than the right, likely because the most common fetal position places the left hip against the mother’s spine, restricting its movement. Firstborns face slightly higher risk, possibly because the uterus has not yet been stretched by a prior pregnancy, creating a tighter space. These risk factors are additive: a firstborn girl born in breech presentation with a family history of DDH has a much higher probability of dysplasia than any single factor alone would suggest.

How the Scan Works

The most widely used technique is the Graf method. A sonographer places a high-frequency probe on the side of the baby’s hip while the infant lies on their side. The image captures a coronal section of the hip, showing the bony and cartilaginous rim of the acetabulum and the femoral head sitting within it. Two angles are then measured from the image. The alpha angle reflects how much bony coverage the socket provides over the femoral head; a larger alpha angle means better coverage. The beta angle reflects the cartilaginous roof.⁸

Of the two measurements, the alpha angle does the heavy lifting. A recent study found it had far greater diagnostic value than the beta angle, with each one-degree increase in the alpha angle roughly halving the odds that the hip would be classified as abnormal.⁹ Beta angle and the baby’s sex did not add independent predictive value beyond the alpha angle in that analysis. Other ultrasound techniques exist besides the Graf method, including the Harcke dynamic technique and the Morin method, which emphasize different aspects of hip stability and coverage. A systematic review found the Morin technique had the highest sensitivity for detecting dysplasia, while the Harcke technique was highly specific but had low sensitivity.¹⁰ In practice, though, the Graf method dominates worldwide, and most published classification schemes and treatment guidelines reference it.

What the Results Mean

Based on the alpha and beta angles, the Graf classification sorts hips into several types. A Type I hip, with an alpha angle of 60 degrees or greater, is normal. Type IIa is labeled “physiologically immature,” meaning the bony socket is a bit shallow but is expected to mature on its own within the first few months. Type IIb indicates an immature hip that has not caught up by the expected age. Types IIc through IV represent progressively more serious dysplasia, subluxation, and dislocation.

For most parents, the important question is: does a Type IIa result mean my baby has a problem? Usually not. The rate of spontaneous normalization for Type IIa hips runs around 95%.¹¹ One study tracking 285 Type IIa hips found that about 79% had normalized by six to seven weeks, with boys normalizing faster than girls.¹² Essentially all of the slightly better-off “IIa-plus” hips matured into normal hips without treatment. The standard approach for a Type IIa finding is watchful waiting with a follow-up scan a few weeks later, rather than immediate bracing. This is worth understanding because the IIa label can sound alarming, and some parents leave the appointment convinced their baby has a serious hip disorder when in reality the vast majority of these hips will be completely fine.

The Overtreatment Dilemma

The high rate of spontaneous resolution in mild findings leads to a genuine dilemma: screen more babies, and you catch the rare serious case earlier, but you also identify a large number of immature hips that would have been fine anyway. Some of those babies end up in harnesses or braces for weeks, introducing both physical treatment and parental stress for a condition that was never going to cause a problem. A Cochrane review comparing universal ultrasound screening to targeted screening (only babies with risk factors) found no significant reduction in late-diagnosed DDH or in the rate of surgery when every baby was scanned compared to when only at-risk babies were scanned.¹³ There was, however, heterogeneity in treatment rates, suggesting that universal screening may lead to more babies being treated.

This does not mean universal screening is pointless. Countries that use it tend to report lower rates of late-presenting dislocations, which require surgery rather than simple bracing. But the trade-off is real. A recent cost-effectiveness analysis modeling English and Welsh birth data found that universal ultrasound screening was more clinically effective but came at significantly higher cost, and that the current targeted approach was the most cost-effective strategy in the overwhelming majority of simulated scenarios.¹⁴ Earlier economic analyses found that the total program costs of universal, selective, and clinical-only screening were often within about 5% of each other, though the distribution of those costs shifted dramatically, with universal screening spending less on treating late cases but more on the screening itself.¹⁵

What Happens When Treatment Is Needed

If ultrasound reveals a hip classified as Type IIc or worse, the first-line treatment for babies under about six months of age is usually the Pavlik harness. This soft brace holds the baby’s legs in a flexed and outward position, encouraging the femoral head to sit deeply in the socket so the acetabulum can remodel around it. Progress is monitored with repeat ultrasounds, often at intervals of a few weeks. If there is no improvement within roughly three weeks, clinicians typically discontinue the harness rather than continue indefinitely.¹⁶

Success rates with the Pavlik harness are generally high when treatment starts early, but the treatment is not without risks. A recognized complication is avascular necrosis (AVN), a condition where the blood supply to the top of the femoral head is disrupted, damaging the bone. One study found a strong correlation between the length of harness treatment and the appearance of AVN, even in the healthy contralateral hip.¹⁷ Failed Pavlik treatment carries its own downstream risk: babies who tried and failed harness treatment before moving to surgical management had substantially higher odds of developing AVN than those who went directly to surgery without a prior harness trial.¹⁸ This is why clinicians pull the plug on the harness quickly if it is not working: prolonging an ineffective treatment does not just waste time, it actively raises the risk of complications.

How Reliable Is the Scan Itself

One concern parents sometimes hear is that ultrasound is “operator-dependent,” meaning results can vary depending on who performs and reads the scan. That concern is not baseless in many areas of ultrasound, but for infant hip imaging specifically, the evidence is reassuring. A study of trained examiners using the Graf method found that both within-observer and between-observer variability were low, with agreement rates above 91% and classification shifts occurring in a statistically insignificant number of cases.¹⁹ A separate comparison of Graf’s alpha-angle method against an alternative measurement approach found that the alpha angle had better repeatability on intra-observer testing, meaning a single examiner was more consistent when using the Graf system.²⁰ The key qualifier here is “when the examination is correctly executed.” Proper probe positioning and image acquisition require training, and less experienced operators can produce misleading images. This is part of why some regions have explored expanding who can perform the scan.

Swaddling and Hip-Safe Positioning

One of the most practical things parents can learn from the DDH story is that how you wrap and carry a baby affects hip development. Traditional straight-leg swaddling, where the baby’s legs are bound together in extension, has been linked to higher rates of DDH in populations that practice it. A randomized controlled trial in Mongolia compared swaddled and non-swaddled infants and found strikingly different outcomes: 40% of swaddled babies had an abnormal hip finding at some point during follow-up compared to 7.5% of non-swaddled babies, and all eight cases of frank DDH occurred in the swaddling group.²¹ The number of hours per day spent swaddled correlated strongly with the likelihood of an abnormal finding. Animal studies have confirmed the mechanism: straight-leg swaddling in neonatal rats increased the prevalence of hip dysplasia, especially when the swaddling started early or lasted longer.²²

The solution is not to avoid swaddling entirely but to swaddle in a hip-safe way. A review of the evidence concluded that swaddling that allows the baby’s hips to flex and spread naturally, rather than pinning the legs together, is compatible with healthy hip development.²³ Baby-wearing devices that hold the infant in a spread-squat position with hips flexed and knees below the buttocks are considered optimal. The pattern to avoid is anything that forces the legs together and straight, whether that is a tight swaddle, a narrow-based carrier, or a cultural practice of binding the lower body.

The Long View If DDH Is Missed

The reason everyone cares about catching DDH early is that untreated dysplasia does not just go away on its own past the window of cartilage remodeling. If a shallow or unstable hip is not corrected in infancy, it can lead to abnormal gait, chronic pain, and early osteoarthritis in adulthood.²⁴ A large Norwegian registry study following over two million newborns found that babies identified with hip instability at birth had roughly 2.6 times the risk of needing a total hip replacement as young adults compared to those without hip pathology. The absolute risk was still quite low, however. And strikingly, most of the young adults who needed a hip replacement for dysplasia-related arthritis had not been flagged with hip instability at birth at all, suggesting that the clinical exams at birth missed many cases that later became symptomatic.²⁵

Midwife-Performed and Point-of-Care Scanning

One barrier to broader ultrasound screening is access. In many healthcare systems, the scan requires a referral to a specialist or a radiology department, adding appointments, wait times, and logistical hurdles for families. Point-of-care ultrasound performed by trained midwives offers a potential way around this. A diagnostic accuracy study found that midwife-performed point-of-care hip ultrasound had significantly higher accuracy in classifying abnormal hips compared to standard selective screening, increasing the detection rate of dysplastic hips by 60% while actually reducing referral rates.²⁶ Fewer unnecessary referrals and more true positives is an appealing combination, and it fits well with the push in many health systems to move routine screening closer to the point of first contact with families.

Artificial Intelligence and the Future of Hip Ultrasound

One of the more promising developments is the use of AI tools to assist with image interpretation. Measuring the alpha and beta angles requires identifying specific anatomical landmarks on the ultrasound image, and the precision of those measurements depends on the operator’s training. AI algorithms trained to segment the hip structures and measure the angles automatically could reduce variability and support less experienced clinicians. A meta-analysis of 34 studies found that AI-assisted angle measurement showed strong agreement with manual measurement, with pooled reliability scores above 0.86 for both alpha and beta angles.²⁷ A separate evaluation of one commercially available AI tool confirmed that it achieved accuracy and reliability comparable to manual assessment while improving measurement speed and reducing differences between operators.²⁸

These tools are not replacing sonographers or orthopedic specialists. What they could do is make quality screening more accessible in settings where trained experts are scarce, essentially serving as a second set of eyes that ensures the anatomical landmarks are correctly identified and the angles are measured consistently. Some newer algorithms are moving toward fully automated classification, segmenting the image, measuring the angles, and assigning a Graf type in one step.²⁹ Combined with the expansion of point-of-care scanning by midwives, this technology could eventually make it practical to screen far more babies without proportionally increasing the need for specialist radiologists.

What Parents Often Worry About Unnecessarily

Getting called back for a hip ultrasound or receiving a result that says “immature” or “Type IIa” is genuinely anxiety-provoking. A trial looking at the psychosocial effects of ultrasound-managed DDH found that mothers whose babies were splinted early showed increased anxiety and more persistent worries about their child’s hips, with elevated worry scores still present at one year.³⁰ Some of this worry is well-placed, but much of it stems from the gap between what a screening result means medically and what it feels like emotionally. A Type IIa finding in a four-week-old baby is the ultrasound equivalent of being told your child is in the 20th percentile for height at one checkup: it is a data point worth watching, not a diagnosis. The vast majority of these hips are normal by the next scan. Understanding that the classification system has a built-in category for “normal but not finished developing yet” can help parents calibrate their response. The time to worry is when a scan shows Type IIc or worse, or when a previously immature hip has not improved on follow-up. Those are the situations that need treatment, and even then, early intervention with a harness resolves most cases without surgery.