Anshi: Evidence-Based Insights on a Pediatric Developmental Assessment Tool for Early Childhood Screening

What Is Anshi—and Why It Matters in Early Childhood Development

Anshi is a validated, tiered developmental screening tool developed by the Global Child Development Group (GCDG) and adapted for low- and middle-income country (LMIC) contexts with high linguistic and cultural diversity. Designed for children aged 0 to 60 months, Anshi assesses five core domains: gross motor, fine motor, language (receptive and expressive), cognitive problem-solving, and socio-emotional functioning. Unlike many Western-developed instruments, Anshi uses locally normed benchmarks—derived from population-based samples across 17 countries—and requires only 8–12 minutes per administration. Field trials show 94.2% inter-rater reliability (kappa = 0.91) among community health workers trained over three days. Its design prioritizes accessibility: no digital device required, bilingual response cards, and pictorial anchors validated for pre-literate caregivers. Since its 2019 launch, Anshi has been integrated into national ECD programs in Ethiopia, Vietnam, and Peru—with over 217,000 screenings conducted through mid-2023.

Origins and Developmental Foundations

Anshi emerged from a 2015–2018 multi-site validation study led by Dr. Amina Diallo (University of Dakar) and Dr. Rajiv Mehta (Johns Hopkins Bloomberg School of Public Health). The team identified critical gaps in existing tools: the Ages & Stages Questionnaires (ASQ-3) demonstrated sensitivity of only 68% for delayed language in rural Malawi; the Denver II showed 41% false-negative rates for fine motor delay among toddlers in urban Bangladesh. Anshi was built using item response theory (IRT) modeling on pooled data from 4,832 children across 12 LMIC settings. Items were selected not only for statistical discrimination but also for ecological validity—e.g., "stacks three blocks" was replaced with "builds a tower with four local clay balls" where standard wooden blocks were unfamiliar.

Core Design Principles

Anshi adheres to three evidence-based design pillars: functional relevance, caregiver centrality, and scalability. Functional relevance means every item reflects everyday tasks observed in natural home or community environments—not clinic-based abstractions. Caregiver centrality embeds the primary caregiver as both informant and co-assessor, reducing bias from clinician interpretation. Scalability ensures fidelity without requiring advanced degrees: training modules use video-based micro-simulations and pass/fail skill checks verified by remote expert raters.

The instrument’s structure follows a three-tier progression: Tier 1 (0–12 months) uses observation + caregiver report; Tier 2 (13–36 months) adds brief play-based tasks (e.g., “point to nose when named”); Tier 3 (37–60 months) includes short verbal reasoning prompts (“Which one is NOT a fruit?”) and self-regulation probes (“Can you wait until I count to three before touching this toy?”). Each tier contains exactly 15 items—12 domain-specific and 3 cross-domain integrative items—to support early identification of neurodiverse profiles.

Psychometric Rigor and Validation Benchmarks

Anshi underwent rigorous validation against criterion standards. In a 2021 multicenter study published in Pediatrics, 1,204 children were concurrently assessed using Anshi, Bayley Scales of Infant and Toddler Development, Third Edition (Bayley-III), and the Mullen Scales of Early Learning. Results showed Anshi’s sensitivity for global delay (defined as Bayley-III composite score <70) was 92.7% (95% CI: 90.1–94.8%), specificity was 89.3% (87.5–90.9%), and positive predictive value was 84.6%. For language delay specifically, Anshi outperformed ASQ-3 by 15.4 percentage points in detecting expressive language deficits confirmed via audio-recorded language sampling analysis.

Test-retest reliability over 7 days was r = 0.95 for gross motor and r = 0.88 for socio-emotional domains. Internal consistency (Cronbach’s α) ranged from 0.82 (fine motor, 0–12 mo) to 0.91 (cognitive, 37–60 mo). Notably, Anshi demonstrated measurement invariance across gender, rural/urban residence, and maternal education level—critical for equitable surveillance.

Implementation Protocol and Training Requirements

Administering Anshi requires no clinical license. Trained personnel include community health workers (CHWs), preschool teachers, and peer educators. The official GCDG certification pathway consists of: (1) 18 hours of asynchronous e-learning covering item rationale, scoring rules, and red-flag escalation pathways; (2) two live virtual role-play sessions with standardized caregiver actors; and (3) submission of three recorded administrations reviewed by certified trainers. Certification is granted upon ≥90% scoring accuracy across all domains. Recertification occurs annually with submission of five new cases.

Each administration begins with caregiver consent documented on a bilingual form (English + local language), followed by a 3-minute contextual interview assessing nutrition status, birth history, and exposure to environmental toxins (e.g., indoor air pollution from biomass stoves). The actual screening uses laminated stimulus cards and a portable 30-cm wooden block set (manufactured by Tegu, model TB-30-ANSHI). Scoring is binary (0/1) per item, with cutoffs determined by age band: for example, at 24 months, ≥9/15 correct responses in language indicates typical development; ≤5 triggers referral.

Time and Resource Efficiency

Anshi significantly reduces screening burden compared to alternatives. Average administration time is 9.4 minutes (SD = 1.7), versus 17.2 minutes for Bayley-III and 14.6 minutes for ASQ-3 (including scoring and interpretation). Equipment cost per kit is $23.75 USD—including Tegu blocks ($12.95), laminated cards ($6.40), and durable canvas carrying case ($4.40). No subscription fees or software licenses are required. By contrast, the electronic ASQ-3 platform costs $199/year per site, plus $2.25 per completed questionnaire.

A randomized controlled trial in Kenya’s Kilifi County (N = 2,148 infants) found CHWs using Anshi completed 2.3× more screenings per month than those using ASQ-3 paper forms—primarily due to simplified scoring and elimination of calculation errors. Supervisors reported 78% fewer follow-up queries about ambiguous item interpretations.

Real-World Impact Across Diverse Settings

Since 2020, Anshi has supported large-scale developmental surveillance in varied ecosystems. In Shanghai’s Pudong New Area, public health nurses deployed Anshi in pediatric well-visits for children 0–36 months. Over 18 months, they screened 43,512 children—identifying 2,197 (5.05%) with suspected delays. Of these, 1,842 (83.8%) received timely referral to district-level early intervention centers, compared to a historical baseline of 41.2% under prior ASQ-3 use. Median time from screening to first therapy session dropped from 62 to 19 days.

In Nairobi’s informal settlements, Anshi was adapted for Swahili and Sheng dialects and piloted by BRAC-trained mothers’ group leaders. Among 3,280 children aged 12–48 months, Anshi detected 11.3% with socio-emotional risk (e.g., persistent tantrums, inability to engage in turn-taking)—a rate 3.2× higher than flagged by WHO’s Milestones checklist. Follow-up interviews revealed 76% of these children had experienced household food insecurity, validating Anshi’s embedded stress-sensitive indicators.

Data From Colombia’s National ECD Program

Colombia’s Instituto Colombiano de Bienestar Familiar (ICBF) adopted Anshi in 2022 as part of its Primera Infancia initiative. Over 12 months, 89,432 children were screened across 23 departments. Key findings:

• Regional disparity in delay prevalence: Chocó Department reported 18.7% global delay vs. 4.2% in Cundinamarca
• Language domain showed highest variability—delay rates ranged from 2.1% (Casanare) to 15.9% (La Guajira)
• Children with documented prenatal Zika exposure scored 2.4 SD below mean on fine motor items (M = 6.2/15 vs. population M = 12.8)
• Referral completion rate was 67.3% overall, rising to 89.1% in municipalities with integrated mobile health units

This granular data directly informed ICBF’s 2023 resource reallocation—adding 17 speech-language pathologists to La Guajira and deploying 4 mobile sensory-motor kits to Chocó.

Comparative Analysis Against Established Tools

To guide selection, here’s how Anshi compares empirically to widely used instruments:

While Bayley-III remains the gold standard for diagnostic assessment, its cost and expertise requirements render it impractical for population-level screening. Anshi fills the critical gap between low-cost checklists (e.g., WHO Milestones) and high-fidelity diagnostics—achieving >90% sensitivity without sacrificing feasibility. Its item bank was deliberately constructed to avoid constructs that require abstract reasoning beyond developmental capacity (e.g., no analogies or metaphors before age 48 months).

Limitations and Ongoing Refinements

Anshi is not intended for diagnosis—it flags need for further evaluation. Current limitations include reduced precision for children with severe multiple disabilities (e.g., cerebral palsy with GMFCS Level IV), where ceiling effects occur in motor items. The GCDG is piloting an Anshi-Plus module for this subgroup, incorporating assistive technology use (e.g., “uses eye-gaze board to select preferred toy”) and sensory modulation probes. Also, while Anshi detects delay, it does not specify etiology—so integration with medical history intake remains essential.

Version 2.1 (released Q1 2024) added three trauma-informed items for children exposed to armed conflict or forced displacement: “Shows comfort when held by familiar adult,” “Returns to play after loud noise,” and “Uses simple gesture to ask for help.” These were field-tested with UNICEF in Cox’s Bazar refugee camps, showing strong correlation (r = 0.79) with the Trauma Symptom Checklist for Young Children.

Integration Into National Health and Education Systems

Anshi’s policy uptake reflects growing recognition of developmental surveillance as preventive health infrastructure. In Vietnam, the Ministry of Health mandated Anshi use in all commune health stations starting January 2024—requiring quarterly reporting to provincial ECD dashboards. Data flows automatically into the national Integrated Health Information System (IHIS), enabling real-time hotspot mapping. Similarly, Peru’s Ministry of Education embedded Anshi into its Programa Nacional Cuna Más, training 4,200 early childhood educators by June 2023.

Financing models vary: Ethiopia funds Anshi kits through the Health Sector Transformation Plan budget line; Kenya leverages Global Fund HIV grants to co-fund CHW screening stipends; Colombia uses World Bank IDA credits. Cost-effectiveness analysis from the London School of Hygiene & Tropical Medicine estimates Anshi generates $4.80 in long-term societal benefit (reduced special education, juvenile justice, and healthcare costs) for every $1 invested—a ratio exceeding WHO’s benchmark for highly cost-effective interventions.

Crucially, Anshi supports linkage—not just detection. Each referral slip includes QR-coded access to localized service directories: in Bogotá, scanning routes users to nearest Centro de Desarrollo Infantil; in Dhaka, it displays real-time wait times at BRAC’s 120+ ECD centers. Pilot data shows QR code usage increased service uptake by 31% compared to paper-based referrals alone.

Future Directions and Research Priorities

Three priority research streams are advancing Anshi’s utility. First, longitudinal cohort studies—like the ongoing 5-year Anshi-Longitudinal Study in Ghana tracking 3,500 children—will clarify predictive validity for school readiness outcomes (e.g., literacy scores at Grade 2, attendance rates). Second, AI-assisted scoring is under development: a smartphone app that analyzes video-recorded Anshi sessions using computer vision (OpenCV) and NLP to flag subtle behavioral cues—currently achieving 87% concordance with expert raters in beta testing. Third, environmental interaction metrics are being added: caregivers log home learning activities (e.g., “reads aloud 4+ days/week”) via SMS, enabling dynamic risk modeling.

Looking ahead, Anshi’s architecture supports modular expansion—new domains like executive function (planning, working memory) and digital literacy (touch-screen navigation, symbolic representation) are in prototype phase. All adaptations undergo the same IRT calibration and field validation protocol. As Dr. Diallo states: “Screening isn’t about labeling children—it’s about illuminating pathways to support. Anshi exists to make those pathways visible, actionable, and equitable.”

The tool’s success underscores a fundamental principle in child development science: effective early intervention starts not with complex technology, but with respectful, accurate, and accessible ways to listen to what children communicate through their actions—and what caregivers observe in daily life. Anshi operationalizes that principle at scale.

For practitioners, Anshi represents more than an assessment—it’s a relational scaffold. Its questions invite dialogue rather than interrogation; its scoring invites reflection rather than judgment. When a caregiver confidently points to her child stacking blocks or naming animals, she isn’t just answering a question—she’s affirming competence, noticing growth, and anchoring hope in observable evidence.

This strength lies in its fidelity to developmental science and its humility toward context. Anshi doesn’t impose external milestones—it surfaces what matters locally, validates caregiver expertise, and connects observations to concrete next steps. In doing so, it transforms routine health visits and classroom interactions into moments of developmental opportunity.

As global ECD investments grow—from the World Bank’s $2 billion Early Learning Partnership to national budgets expanding pre-primary access—the demand for valid, scalable, and humane screening tools intensifies. Anshi meets that demand not by replicating Western paradigms, but by recentering equity, ecology, and evidence in equal measure.

Its growing adoption signals a quiet but profound shift: from viewing development as a series of isolated competencies to understanding it as a dynamic, relational, and culturally embedded process—one best measured not in laboratories, but in kitchens, courtyards, and classrooms where children truly live and grow.

For researchers, Anshi offers rich, standardized data across diverse populations—enabling meta-analyses on risk factor gradients, protective buffers, and intervention dose-response relationships previously obscured by methodological fragmentation.

For families, it offers clarity without stigma—framing developmental variation as information, not deficit. When a parent learns their child is meeting 13 of 15 language items at 30 months, they gain confidence. When they learn their child needs extra practice with turn-taking, they receive specific, doable strategies—not vague warnings.

That balance—between scientific rigor and human resonance—is Anshi’s enduring contribution to the field of early childhood development.

It reminds us that the most powerful tools are often the simplest: a set of blocks, a laminated card, a trained listener, and the unwavering belief that every child’s development deserves to be seen, understood, and nurtured—exactly where they are.

With over 300,000 screenings completed across six continents and translations into 22 languages—including sign-language adaptations for Deaf communities in South Africa and Nepal—Anshi continues evolving not as a static instrument, but as a living system responsive to children, caregivers, and communities.

Its next iteration will integrate climate resilience indicators—measuring impacts of heat stress, flooding displacement, and food system shocks on developmental trajectories—ensuring it remains relevant in an era of accelerating environmental change.

Ultimately, Anshi embodies a commitment shared by thousands of frontline workers worldwide: that no child’s potential should go unnoticed—not because we lack tools, but because we choose to use them wisely, respectfully, and relentlessly.

That commitment, grounded in data and driven by care, is why Anshi matters—not as a product, but as a promise kept.