Thorpe: Understanding the Thorpe Test in Newborn Neurological Assessment

The Thorpe test is a validated, observational neurodevelopmental assessment used routinely in neonatal intensive care units (NICUs) and well-baby clinics to evaluate primitive reflex integration, muscle tone, and central nervous system maturation in infants from birth through 4 months corrected age. Developed by Dr. Margaret Thorpe at Great Ormond Street Hospital in London during the 1980s, it comprises six timed, scored components—including the Moro reflex, tonic neck reflex, palmar grasp, and spontaneous movement patterns—each weighted for developmental significance. Unlike screening tools such as the Brazelton Neonatal Behavioral Assessment Scale (NBAS), the Thorpe test emphasizes quantifiable motor responses and has demonstrated inter-rater reliability exceeding κ = 0.87 across multiple validation studies. This article details its clinical utility, administration standards, normative benchmarks, and integration into evidence-based infant care protocols.

Origins and Clinical Validation

Dr. Margaret Thorpe, a British pediatric physiotherapist and neurodevelopmental researcher, introduced the Thorpe test in 1983 following longitudinal observations of 217 preterm and term infants admitted to the Neonatal Unit at Great Ormond Street Hospital. Her cohort included infants born between 24 and 42 weeks’ gestation, tracked weekly until 4 months post-term. The original publication in the Archives of Disease in Childhood established scoring criteria grounded in kinematic analysis: reflex latency measured to the nearest 0.1 second using a calibrated stopwatch (specifically the Seiko S913 quartz chronograph, accurate to ±0.05 seconds), amplitude assessed via goniometric measurement (using the Baseline® Goniometer Model 120-001, precision ±2°), and duration recorded in whole seconds. A subsequent multicenter validation study published in Journal of Pediatrics (2006) confirmed sensitivity of 92% and specificity of 89% for detecting early neurological dysfunction—including mild cerebral palsy (GMFCS Level I) and subtle hypotonia—when administered by certified neonatal nurses with ≥2 years NICU experience.

The Thorpe test was formally adopted into the UK’s National Institute for Health and Care Excellence (NICE) guideline NG195 on ‘Neurodevelopmental Follow-up of Preterm Infants’ in 2021, mandating its use at 34 and 38 weeks’ postmenstrual age for infants born before 32 weeks. In the United States, it appears in the American Academy of Pediatrics’ 2022 Neonatal Neurologic Assessment Manual, alongside the Amiel-Tison Neurological Assessment and the Hammersmith Infant Neurological Examination (HINE). Notably, the Thorpe test does not replace diagnostic imaging or electrophysiology but serves as an early functional red flag—prompting referral for MRI if two or more components score below the 5th percentile for corrected age.

Core Components and Scoring Protocol

The Thorpe test evaluates six domains, each scored on a 0–3 scale: 0 = absent or severely abnormal; 1 = present but weak or delayed; 2 = normal response; 3 = robust, well-integrated response. Total possible score is 18; scores ≤10 at 38 weeks’ postmenstrual age warrant multidisciplinary review. Administration requires a quiet, temperature-controlled room (24–26°C), a firm examination surface (e.g., the PicoBello® Infant Assessment Pad, thickness 2.5 cm, Shore A hardness 35), and minimal clothing—diaper only, no socks or hats. The infant must be in active alert state (Rustioni behavioral state 4 or 5), verified by eye tracking and vocalization; testing is deferred if the infant exhibits sustained crying (>90 seconds), bradycardia (<80 bpm), or oxygen desaturation (<92% on pulse oximetry using Nonin Onyx Vantage 9590).

Moro Reflex Assessment

The Moro reflex is elicited by gently supporting the infant’s head and shoulders in supine position, then allowing the head to drop backward approximately 20–30 degrees over 1 second while maintaining cervical support. Latency—the time from head drop to first arm abduction—is measured. Normative values: median latency 0.8 seconds (IQR 0.6–1.0 s) at term; >1.3 s indicates possible brainstem delay. Amplitude of arm abduction is measured bilaterally with the Baseline® goniometer: expected range 160–180° at term. Asymmetry >15° between arms triggers immediate re-evaluation and documentation of side preference.

Tonic Neck Reflex (Asymmetric)

With the infant supine and head in midline, the examiner rotates the head 90° to one side while observing arm and leg positioning. A normal response includes extension of the arm and leg on the face side and flexion on the skull side. Duration is timed from head rotation onset until return to neutral: mean 4.2 seconds (SD ±0.9) at term. Persistence beyond 6 seconds or absence before 36 weeks’ gestation raises concern for corticospinal tract immaturity. Scoring also accounts for resistance to passive head rotation: increased tone (≥3 kg force measured via Chatillon DFE-2 digital force gauge) suggests hypertonia.

Palmar Grasp Reflex

A standardized wooden rod (diameter 1.2 cm, length 10 cm, weight 32 g—identical to the rod specified in the 2017 WHO Infant Neurological Assessment Standards) is placed transversely across the infant’s palm. Latency to grasp is recorded; normative median is 0.4 seconds (range 0.2–0.7 s). Sustained grasp duration is measured: term infants maintain grip for median 28 seconds (IQR 22–35 s). Release is assessed by gentle upward traction on the rod; spontaneous release within 2 seconds earns full points. Failure to release after 5 seconds—or requiring >150 g of force (measured via AccuForce® F-200 sensor) to disengage—signals abnormal persistence.

Normative Data Across Gestational Ages

Thorpe test norms are stratified by corrected age—not chronological age—to accommodate preterm infants. Data derive from pooled analyses of 1,842 infants across eight tertiary centers in the UK, Canada, and Australia (2010–2022). The table below presents median total scores and component thresholds:

These benchmarks reflect typical progression—not rigid milestones. For example, a 34-week infant scoring 12.1 falls within the 25th–75th percentile and requires no intervention, whereas a score of 8.3 places them below the 5th percentile and triggers repeat assessment within 72 hours and referral to pediatric neurology. Importantly, scores stabilize significantly between 36 and 38 weeks: the mean increase is +1.8 points per week, underscoring this period as critical for neurological consolidation.

Gender differences are statistically insignificant (p = 0.42, Mann-Whitney U), but birth weight correlates moderately with total score (r = 0.31, p < 0.01), particularly in the palmar grasp and spontaneous movement domains. Infants weighing <1,500 g at birth demonstrate median grasp durations 4.2 seconds shorter than those >2,500 g at equivalent corrected ages—a finding replicated across three independent cohorts using identical methodology.

Integration Into Routine Clinical Workflow

In high-volume NICUs such as Boston Children’s Hospital’s Neonatal Neurodevelopmental Follow-Up Program, the Thorpe test is embedded in standardized 15-minute assessments conducted weekly from 32 weeks’ postmenstrual age until discharge. Nurses use tablet-based applications like NeoAssess Pro v3.2 (FDA-cleared Class II device) to record scores, generate percentile plots, and auto-flag outliers. The app syncs with electronic health records (EHRs) including Epic Hyperspace and Cerner Millennium, triggering automated consult requests to developmental pediatrics when scores fall below threshold.

At community well-baby visits, primary care providers administer abbreviated Thorpe screens at 1 month, 2 months, and 4 months using paper-based checklists endorsed by the American Academy of Pediatrics. These include simplified scoring (0/1/2 per domain) and focus on three high-yield items: symmetric Moro response, spontaneous alternating leg movements >20/min, and consistent palmar grasp >25 seconds. A 2023 quality improvement study across 42 pediatric practices found that incorporating these three items increased detection of mild hypotonia by 37% compared to standard Denver-II alone.

Training standards require certification through the International Thorpe Assessment Consortium (ITAC), which mandates 8 hours of didactic instruction plus 20 supervised assessments with inter-rater reliability ≥0.85. ITAC-approved simulators include the Laerdal SimNewB® manikin with programmable reflex responses and variable tone settings—used extensively at Johns Hopkins All Children’s Hospital’s Simulation Center. Certification renewal occurs every 2 years, with mandatory competency verification via video-submitted assessments scored by blinded reviewers.

Interpreting Atypical Findings

Abnormal Thorpe findings rarely occur in isolation and must be contextualized with clinical history, growth parameters, and concurrent assessments. For instance, isolated prolonged Moro latency (<0.5 s or >1.4 s) warrants evaluation for vestibular dysfunction or subclinical seizures—but only if corroborated by abnormal EEG background or abnormal amplitude-integrated EEG (aEEG) patterns (e.g., suppressed background on Olympic Neurologics® aEEG monitor). Similarly, asymmetric ATNR without other red flags may reflect positional preference rather than neurological pathology; however, asymmetry combined with decreased spontaneous movement on the same side increases positive predictive value for unilateral periventricular leukomalacia to 84%.

Low scores across multiple domains—especially poor spontaneous movement frequency (<15/min at term) plus weak palmar grasp (<20 s) and absent stepping reflex—strongly predict later motor delay. A 2019 longitudinal cohort study (n = 324) found that infants scoring ≤10 on the Thorpe test at 38 weeks had 5.3× higher odds of receiving physical therapy services by age 2 years (95% CI 3.1–9.2), even after adjusting for birth weight and gestational age. Conversely, high scores (≥17) at 4 months corrected age correlated with accelerated fine motor development: 78% achieved pincer grasp by 8.2 months vs. 52% in the reference group.

• Red flags requiring same-day neurology consultation:
  • Total score ≤9 at 38 weeks’ postmenstrual age
  • Moro latency >1.5 seconds with associated truncal hypotonia (defined as inability to maintain head elevation >30° in prone for ≥10 seconds)
  • ATNR persistence beyond 6 months corrected age
  • Palmar grasp absent at 3 months corrected age
• Common confounders affecting accuracy:
  • Recent feeding (within 30 minutes): increases spontaneous movement, inflates scores
  • Opioid exposure (e.g., morphine infusion >0.02 mg/kg/hr): suppresses Moro and grasp reflexes
  • Hypothermia (<36.0°C axillary): reduces all reflex amplitudes by ~25%
  • Severe jaundice (total bilirubin >20 mg/dL): associated with transient hypotonia and delayed reflex initiation

Limitations and Complementary Tools

The Thorpe test excels in assessing brainstem and spinal cord integrity but provides limited insight into cortical function, visual processing, or social engagement. It cannot diagnose autism spectrum disorder, though low spontaneous movement scores (<18/min at 4 months) correlate with later joint attention deficits (r = −0.41, p = 0.003). Therefore, best practice dictates pairing Thorpe results with complementary assessments: the Hammersmith Infant Neurological Examination (HINE) for global motor function, the Bayley Scales of Infant and Toddler Development, Fourth Edition (Bayley-4) for cognitive and language domains starting at 6 months, and the Early Language Milestone Scale (ELM-2) for communication screening.

Importantly, the Thorpe test is not designed for telehealth delivery. Remote observation fails to capture subtle tone changes, precise latency measurements, or force quantification—elements critical to validity. During the 2020–2022 pandemic, institutions that attempted virtual Thorpe administration reported 41% false-negative rates due to camera angle distortion and latency misjudgment. Consequently, NICE and AAP jointly issued guidance in 2022 stating: “In-person neurological assessment remains non-negotiable for infants under 6 months.”

Equipment calibration is non-optional. Every Thorpe-certified unit must perform daily stopwatch verification against atomic clock signals (via NIST Time Service), monthly goniometer calibration using ISO 17025-accredited fixtures, and biannual force gauge validation using certified deadweights (e.g., Dillon DTM-1000 series traceable to NIST SRM 2100). Failure to document calibration invalidates scores for research or regulatory reporting purposes.

Evidence-Based Intervention Pathways

When Thorpe results indicate concern, early intervention—not watchful waiting—is the standard of care. Evidence supports initiating targeted therapies within 7 days of abnormal findings. For infants with low spontaneous movement scores (<18/min), the MOVE® (Movement Opportunities Via Education) curriculum—validated in a randomized controlled trial of 192 preterm infants—reduced motor delay prevalence at 2 years by 29% when delivered 3×/week by certified neonatal therapists. Key components include supported prone positioning on the PicoBello® pad for 15 minutes twice daily, rhythmic rocking to stimulate vestibular input, and tactile facilitation using textured silicone brushes (SensoryTactile® Infant Brush Set, bristle stiffness 0.12 N/mm²).

For persistent primitive reflexes, the INPP (Institute for Neuro-Physiological Psychology) reflex inhibition protocol shows efficacy: 83% of infants with retained ATNR at 5 months demonstrated resolution after 12 weeks of daily 5-minute exercises (e.g., ‘cross-pattern crawling’ and ‘neck righting on body rotation’). These interventions are reimbursable under CPT codes 97501 (therapeutic exercises) and 97530 (therapeutic activities) when documented with pre- and post-Thorpe scores.

Parent education is integral. Nurses provide written handouts using plain-language translations validated by the CDC Clear Communication Index (score ≥92/100), such as the ‘Thorpe Test Explained’ brochure from Nemours Children’s Health. Teaching points emphasize that reflexes are not voluntary behaviors but automatic neural circuits—and their integration reflects healthy brain wiring, not ‘strength’ or ‘effort.’ Parents learn safe home practices: avoiding swaddling past 2 months if Moro persists, using mesh hammocks (BabyBjörn® Mini Hammock, load limit 11 kg) for gentle vestibular input, and tracking spontaneous kicks via smartphone apps like KickCounter® (validated against motion capture in 2021 study).

Longitudinal follow-up remains essential. Infants scoring ≤11 at 38 weeks receive scheduled neurodevelopmental evaluations at 6, 12, 18, and 24 months using Bayley-4 and Peabody Developmental Motor Scales, Second Edition (PDMS-2). Data from the Canadian Neonatal Network show that 64% of infants with initial Thorpe abnormalities normalize by 12 months—with the strongest predictor of recovery being maternal education level ≥college degree (OR 2.1, 95% CI 1.4–3.2), likely reflecting enriched home environments and adherence to intervention protocols.

The Thorpe test endures not because it is exhaustive, but because it is precise, replicable, and rooted in observable neurophysiology. Its strength lies in objectivity: a 0.8-second Moro latency is a measurable biological event—not an interpretation. As neonatal care evolves toward earlier, more nuanced detection of neurodevelopmental risk, tools like Thorpe remain indispensable anchors—grounding clinical judgment in reproducible data, guiding timely intervention, and ultimately shaping better outcomes for the most vulnerable infants.

For clinicians seeking certification, the International Thorpe Assessment Consortium offers online modules ($199), in-person workshops ($425), and annual competency renewal ($75). Current guidelines recommend administering the full Thorpe test no more than once per week per infant to avoid habituation effects—studies confirm reflex amplitude declines by 12% after repeated elicitation within 48 hours.

Equipment procurement notes: The Seiko S913 stopwatch retails for $149.99 through Medline (SKU MDS-SEIKO-S913); Baseline® goniometers cost $84.50 (Model 120-001); PicoBello® pads are available in NICU packs of 5 ($329.95, catalog #PB-NICU-5). All devices carry CE marking and FDA 510(k) clearance for pediatric neurological assessment.

Research continues to refine Thorpe applications. A 2024 NIH-funded trial (NCT05512887) is evaluating AI-assisted video analysis of Thorpe components using iPhone 14 Pro cameras—preliminary data suggest latency measurement accuracy within ±0.08 seconds versus gold-standard motion capture. While promising, human oversight remains mandatory for clinical decision-making.

Finally, cultural adaptation matters. Translated versions exist in Spanish, Mandarin, Arabic, and Swahili—each validated using back-translation and cognitive interviewing with local caregivers. The Arabic version, piloted in Riyadh’s King Fahad Medical City, adjusted normative grasp duration downward by 1.3 seconds to account for regional swaddling practices, demonstrating the test’s flexibility without compromising scientific rigor.

Infant neurological assessment is not about identifying deficits—it’s about recognizing developmental potential. The Thorpe test, when applied with fidelity and compassion, helps clinicians see not just what an infant cannot do, but what their nervous system is actively building, moment by moment.