Hammersmith Infant Neurological Examination (HINE)

The Hammersmith Infant Neurological Examination is a structured clinical assessment designed to evaluate the neurological development of infants between two and twenty-four months of age. It scores 26 individual items across five domains and produces a global score that helps clinicians identify early signs of cerebral palsy and other neurodevelopmental conditions, often months before a formal diagnosis would otherwise occur. The exam has become one of the most widely recommended tools in international guidelines for early detection of cerebral palsy, and its reach extends beyond motor disorders to predicting cognitive outcomes as well.

What the Exam Actually Looks At

The HINE evaluates an infant across five categories: cranial nerve function, posture, movements, tone, and reflexes and reactions.¹ Each of the 26 items is scored individually, and the item scores are summed to produce a global score. The maximum possible score varies by age because some items are not scorable in very young infants, but by around three months most of the exam can be completed.

The cranial nerve section checks visual tracking, facial movements, and sucking and swallowing. The posture section looks at how the baby holds their head, trunk, and limbs when placed in different positions. The movement section assesses the quantity and quality of spontaneous movements. Tone is evaluated through resistance to passive movement in the arms and legs. And the reflexes and reactions section tests protective responses, the ability to reach and grasp, and age-appropriate reflexes that should appear or disappear on schedule.

A separate asymmetry score is also calculated. This tracks whether the infant shows consistent differences between the left and right sides of the body, which can be an early flag for hemiplegic cerebral palsy, a form that affects one side more than the other.

It is worth noting that the HINE is distinct from the Hammersmith Neonatal Neurological Examination, sometimes called the HNNE, which is a 34-item tool used in the first weeks of life and covers a somewhat different set of categories including abnormal signs and behaviors.² The two exams are related but designed for different developmental windows.

How Well It Predicts Cerebral Palsy

The primary clinical use of the HINE is to identify infants at high risk for cerebral palsy before the condition becomes clinically obvious. A systematic review and meta-analysis found that a global HINE score below 58 at three months of age predicted cerebral palsy with a pooled sensitivity of about 80% and a specificity of roughly 89%.³ In practical terms, that means the exam catches about four out of five infants who will later receive a cerebral palsy diagnosis, while correctly reassuring nearly nine out of ten infants who will not.

These numbers shift depending on the infant’s age at assessment. One study looking at high-risk infants reported that the HINE at three months had 88% sensitivity and 62% specificity for detecting cerebral palsy.⁴ The specificity there was notably lower than in the meta-analysis, which reflects the reality that very young infants are harder to read. Immature nervous systems produce findings that can look concerning at three months but resolve by six. The exam tends to get more accurate the older the infant is at the time of testing.

A separate study testing the HINE against another motor assessment tool found that a total score below 57 yielded 77% sensitivity and 91% specificity for predicting cerebral palsy, with strong overall discriminative ability.⁵ The HINE outperformed the Test of Infant Motor Performance in that comparison. What emerges across studies is a consistent pattern: the HINE is a strong predictor, though no single cutoff works perfectly for every population and every age.

Predicting Not Just Cerebral Palsy but Its Severity

One of the HINE’s most useful features is that the global score does not just flag whether an infant is likely to develop cerebral palsy. It also correlates with how severe that cerebral palsy is likely to be. A study tracking infants through their first year found a strong negative correlation between HINE scores and later functional classification on the Gross Motor Function Classification System, or GMFCS, which rates motor ability on a five-level scale.⁶ Lower HINE scores consistently predicted more severe motor limitations.

This matters because early interventions differ depending on the expected trajectory. An infant whose HINE scores suggest mild motor involvement might benefit most from targeted physiotherapy, while one with scores pointing to severe cerebral palsy may need a much broader package of support, including feeding assistance, orthotics, and early planning for adaptive equipment. The HINE gives families and clinicians a head start on making those decisions rather than waiting years for a clearer clinical picture.

Research on combinations of tools has reinforced this point. When structural brain imaging and the HINE are used together, the pairing achieves particularly high specificity for predicting moderate to severe forms of cerebral palsy (GMFCS levels II through V), even though sensitivity for milder cases remains a challenge.⁷

The Asymmetry Score and Hemiplegic Cerebral Palsy

Hemiplegic cerebral palsy, where one side of the body is affected more than the other, can be particularly tricky to spot early. Infants do not use their hands with clear laterality until several months of age, so a subtle difference between the left and right side is easy to miss in a standard exam. The HINE addresses this through its asymmetry subscale, which tallies left-right differences across the scored items.

Research has shown that combining a low total HINE score (below 63) with a high asymmetry score (above 5) distinguishes hemiplegic cerebral palsy from typical development with about 92% sensitivity and 100% specificity.⁸ That 100% specificity is striking: in the study population, no typically developing infant was incorrectly flagged.

More recent work examined the predictive value of individual HINE subscores and found that the reflexes and reactions subscore had the single best discriminative power for cerebral palsy overall, with the cranial nerve function and asymmetry subscores close behind.⁹ This suggests that clinicians should pay attention to more than just the global score. Looking at how the score breaks down across domains adds diagnostic nuance that a single number alone cannot capture.

How the HINE Compares to Other Screening Tools

The HINE does not exist in isolation. Two other tools are commonly used alongside it: General Movements Assessment (GMA) and structural brain imaging (usually MRI). Each evaluates something different. GMA is a video-based method that analyzes the quality of an infant’s spontaneous movements, particularly the presence or absence of “fidgety” movements around three to five months. MRI provides a direct look at brain structure. The HINE, by contrast, is a hands-on neurological examination.

A study of very preterm infants found that the HINE flagged abnormal motor findings at a much higher rate than GMA, and the two tools showed low agreement with each other. The authors concluded that the HINE and GMA appear to measure different things and should be used alongside MRI rather than swapped for each other.¹⁰ This is an important point for clinicians who might assume one tool can substitute for another. Each brings a different lens.

In terms of raw sensitivity, GMA sometimes outperforms the HINE for detecting cerebral palsy at specific ages. One study found that absent fidgety movements had 95% sensitivity and 97% specificity for cerebral palsy, compared with 88% sensitivity and 62% specificity for the three-month HINE.¹¹ But the HINE has practical advantages: it can be performed at any clinic visit without specialized video equipment, and it produces a quantitative score that can be tracked over time. GMA, by contrast, requires trained raters to analyze video and is generally most informative in a narrow developmental window.

Where things get most powerful is in combination. A study from a developing nation found that using a single GMA or a single HINE assessment predicted cerebral palsy with sensitivity under 95%, but combining writhing GMA, fidgety GMA, and HINE raised sensitivity to 100%.¹² Another retrospective study found that combining GMA and HINE achieved 100% specificity for cerebral palsy, though sensitivity varied depending on severity.¹³ The emerging consensus is that no single assessment tool is enough on its own for a confident early diagnosis. The HINE is strongest when it is one part of a broader evaluation.

Beyond Cerebral Palsy

While the HINE was developed with cerebral palsy in mind, it turns out to be informative for other outcomes too. Research in preterm infants has found that HINE scores, particularly after three months of age, have good sensitivity and specificity for detecting significantly delayed cognitive performance, even in infants who do not go on to develop cerebral palsy.¹⁴

A study that followed very preterm children to school age found that HINE global scores were associated with IQ, motor coordination, and results on a later neurological examination, even after excluding children with cerebral palsy.¹⁵ Each one-point increase in the HINE global score was associated with about a 1.2-point increase in full-scale IQ, which adds up meaningfully across the scoring range. This association held even after adjusting for other factors, suggesting the HINE is tapping into something about early brain integrity that extends well beyond motor pathways.

The combined use of GMA and HINE across multiple time points also improves detection of non-cerebral-palsy developmental delays. Single assessments caught less than 65% of motor, cognitive, and social-emotional delays, but the full series of assessments raised that to above 83% for all domains.¹⁶ For clinicians, this reinforces the value of serial assessments rather than relying on a single visit.

Reliability Across Clinicians and Settings

A tool is only useful if different people get similar results when they use it. The HINE performs well on this front. A Turkish validation study reported high inter-rater reliability for the total score, meaning that two different examiners assessing the same infant tended to arrive at very similar scores. Test-retest reliability was also strong, with no significant difference when the same examiner scored the same infant at two time points.¹⁷ A Brazilian Portuguese translation study similarly found strong internal consistency and high inter-rater reliability for the total score and all five categories.¹⁸

The exam has also been tested through telehealth, which matters increasingly in the era of remote follow-up. A prospective study comparing telehealth and in-person HINE assessments at 24 months found moderate reliability for total scores, with minimal average difference between the two approaches. All differences fell within a prespecified clinical acceptability threshold, and clinical classification, meaning whether a child’s score crossed a concerning cutoff, was preserved perfectly across both methods.¹⁹ Telehealth HINE is not as precise as a hands-on exam, particularly for the tone items that require the clinician to physically handle the infant, but for triaging follow-up or maintaining surveillance when families live far from a specialist center, the evidence suggests it works.

Training and Who Can Perform It

One practical strength of the HINE is that it does not require a pediatric neurologist to administer. Implementation research in a high-risk infant follow-up clinic showed that a structured training workshop enabled a diverse team, including advance practice nurses, general pediatricians, developmental-behavioral pediatricians, neonatologists, and fellows, to learn, perform, and document the exam. Completion of HINE domains improved after training.²⁰ This is important because pediatric neurologists are in short supply almost everywhere, and a tool that only specialists can use has limited population-level impact.

A study from Ethiopia demonstrated that the HINE can be implemented effectively in low-resource settings. With a structured knowledge-translation framework, clinicians at a teaching hospital in Addis Ababa were able to integrate the HINE into routine clinical practice. Of 98 children with risk factors for cerebral palsy who were assessed, about 36% received a diagnosis of cerebral palsy or high probability of cerebral palsy. The mean age at detection was roughly 10.6 months, which aligns with international guidelines recommending diagnosis before 12 months. That represents a dramatic improvement from the previously reported average diagnosis age in Ethiopia of five years and eight months.²¹ Early diagnosis triggered immediate referral to rehabilitation, compressing the gap between suspicion and intervention from years to weeks.

Longitudinal Monitoring, Not Just One-Time Screening

The HINE is not just a one-and-done screening tool. Because it produces a quantitative score at each visit, clinicians can track an infant’s neurological trajectory over time. A study in a Chinese clinical setting confirmed that HINE findings at multiple ages during the first year all related to neuromotor outcomes at 12 months corrected age, supporting its use for ongoing longitudinal monitoring.²²

This serial use is what makes the HINE especially valuable in high-risk infant follow-up clinics, where babies born very preterm or with complications like birth asphyxia are seen repeatedly during the first year or two. A single low score at three months might reflect transient neurological immaturity. A persistently low or declining score across several visits is a much stronger signal that something is wrong. Conversely, an infant whose scores improve over time provides more reassurance than a single normal assessment can.

The Brief-HINE

Recognizing that the full 26-item HINE takes time to administer, researchers have developed a shortened version called the Brief-HINE. This tool retains the items most predictive of cerebral palsy and produces a streamlined score that can be obtained faster, potentially making it more practical in busy clinic settings or as a first-pass screening.

The Brief-HINE has shown good diagnostic accuracy across the first year. At three months, a score below 22 was associated with cerebral palsy with 88% sensitivity and 92% specificity. At six months, the cutoff shifted to below 25, with 93% sensitivity and 87% specificity. At nine and twelve months, a score below 27 achieved 95% and 100% sensitivity respectively, with specificity in the 81-86% range.²³ That 100% sensitivity at 12 months is a notable finding: every infant who later received a cerebral palsy diagnosis scored below the threshold. The tradeoff with the Brief-HINE is that you lose some of the nuanced domain-level information that the full exam provides, particularly the asymmetry data and the subscore profiles that can distinguish different patterns of cerebral palsy. For screening purposes, though, the abbreviated version appears to be a reasonable option when time or resources are limited.

What the HINE Cannot Do

For all its strengths, the HINE has limitations that are worth understanding. It is a clinical assessment, not a diagnostic test. A low score is a red flag, not a confirmed diagnosis. The exam’s specificity at younger ages is sometimes modest, meaning some infants will be flagged who turn out to develop normally. This can create anxiety for parents who are told their baby’s score is concerning. Managing that communication, framing the HINE as part of an ongoing monitoring process rather than a definitive verdict, is a skill in itself.

The HINE is also less sensitive for milder forms of cerebral palsy and for non-motor neurodevelopmental conditions like autism or isolated language delay. While there is evidence that scores correlate with later cognitive outcomes, the exam was designed to detect motor abnormalities, and conditions that primarily affect cognition, behavior, or communication may not show up clearly in the HINE’s scoring framework. High-risk infant follow-up programs typically pair the HINE with standardized developmental assessments such as the Bayley Scales to get a fuller picture of how a child is progressing.

Finally, the exam relies on a cooperative infant, which is not always available. A baby who is fussy, hungry, or drowsy during the assessment may produce scores that do not reflect their true neurological status. Experienced examiners learn to work around this by timing the exam around feeds and sleep, but it introduces a source of noise that no scoring system can fully eliminate.