Every child is seen.
100% of the class moves through calm, non-invasive stations during the school day. Nothing is diagnosed here — a finding is captured for review.
Clinical evidence · for schools & partners
Catching it early matters more than a perfect diagnosis delivered too late.
SKIDS screening is a sequential test. A high-sensitivity hardware net is designed to miss as few children as possible; a pediatrician then reviews every flag and validates the report. This page shows what the device does on its own, and what the device plus a pediatrician do together.
Why sensitivity, then review
In screening, the costly error is the missed child.
For many childhood conditions, when we intervene changes the outcome more than how precisely we label it on day one. Amblyopia is treatable before roughly age seven; hearing caught early protects speech and literacy; anemia and developmental delays respond best to early action. A missed case can close that window; a false positive costs one review visit.
So the hardware is tuned for sensitivity, and the pediatrician over-read restores precision. No child is diagnosed by a device. Screening flags a possibility; a pediatrician decides whether it deserves a closer look.
The sequential workflow
Two layers: a sensitive net, then an expert filter.
Doctors review only the cases the hardware flags — not the negatives. Under the sequential-testing principle, this means sensitivity is set by the first layer, and specificity climbs sharply once the pediatrician rejects false positives.
A high-sensitivity triage net.
Devices and assisted capture are tuned to miss as few children as possible. This casts a wide net, so it will also flag some children who turn out to be well.
A pediatrician reviews every flag.
Tele-pediatricians review the captured evidence, reject false positives, and validate the report before it reaches a parent. This is where accuracy is restored.
How to read the numbers
Three terms make the tables below make sense.
Sensitivity
Of children who truly have a condition, how many the screen catches. A screen is deliberately tuned high, so it does not miss.
Specificity
Of well children, how many are correctly cleared. Lower specificity simply means more children get a second look from a pediatrician.
PPV, in a school
Of children flagged, how many truly have the condition. Because most conditions are uncommon in a general school, a flag means “worth a look,” not a diagnosis.
Standalone figures describe the device or algorithm on its own. Combined (HITL) figures describe the device followed by a pediatrician over-read: the over-read chiefly raises specificity and positive predictive value — it cannot exceed the device’s sensitivity, because a pediatrician reviews only what the device flags. Ranges are indicative of the instrument class and are not a guarantee for any individual child; a definitive diagnosis always requires the gold-standard test named for each domain.
Evidence by clinical domain
Standalone screen, then combined output.
Seven grouped domains cover the whole-child screen. Toggle to the full compendium below for a condition-by-condition view of how the pediatrician over-read changes each finding.
| Clinical domain | Hardware modality | Standalone triage | Combined (device + pediatrician) | Gold standard & grounding |
|---|---|---|---|---|
| Refraction, amblyopia & strabismus Myopia, astigmatism, anisometropia, leukocoria | Binocular infrared photoscreener | Sens 86–88% Spec 75–80% | Sens 85–88% Spec 94–97% | Cycloplegic retinoscopy FDA Class II (510(k)) · CE · CDSCO Class B |
| Auditory pure-tone loss Conductive & sensorineural deficits | Automated Hughson-Westlake audiometer | Sens 91–93% Spec 84–88% | Sens 90–93% Spec 96–98% | Sound-booth diagnostic audiometry IEC 60645-1 · ANSI S3.6 |
| Cardiac murmurs & respiratory sounds Pathologic murmurs, tachycardia, wheeze | Dual-band digital smart stethoscope | Sens 89–92% Spec 82–85% | Sens 89–91% Spec 95–97% | Echocardiography (TTE) CDSCO Class B · CE Mark |
| Tympanic membrane & ear canal Otitis media, perforation, impaction | 3.9 mm rigid video otoscope | Sens 85–88% Spec 80–84% | Sens 85–88% Spec 95–98% | Operative ear microscopy CE Class I · non-invasive |
| Oral cavity & dentition Caries, plaque, gingivitis, tonsils | Macro video wand | Sens 75–82% Spec 78–84% | Sens 75–82% Spec 92–95% | Clinical dental operatory exam Visual screening aid |
| Growth, posture & anthropometry Scoliosis, stunting, goiter triage | Stadiometer + digital HD camera | Sens 88–92% Spec 85–88% | Sens 88–91% Spec 94–97% | Spine radiograph & WHO charts Anthropometric standard |
| Neurodevelopment & behaviour Autism, ADHD, dyslexia, motor delays | Tablet eye-tracking & validated scales | Sens 84–88% Spec 76–81% | Sens 84–87% Spec 93–96% | ADOS-2 & DSM-5 evaluation Validated digital battery |
| Tier | Condition | Hardware modality | Standalone Sens / Spec | Combined Sens / Spec | Pediatrician over-read |
|---|---|---|---|---|---|
| Tier 1 | Pediculosis Capitis | 3.9 mm rigid video wand | 82% / 85% | 82% / 98.5% | Over-read distinguishes active, viable nits from empty nits or pseudo-nits (hair casts, dandruff). |
| Tier 1 | Tinea Capitis | 3.9 mm rigid video wand | 80% / 82% | 80% / 97.0% | Doctor rules out alopecia areata, confirming a fungal kerion or scaling pattern before oral antifungals. |
| Tier 1 | Seborrheic Dermatitis | 3.9 mm rigid video wand | 84% / 79% | 84% / 95.0% | Review differentiates from scalp psoriasis or atopic dermatitis for correct keratolytic guidance. |
| Tier 1 | Microcephaly / Macrocephaly | Calibrated wall tape | 88% / 85% | 88% / 97.5% | OFC correlated against WHO growth charts, ruling out cephalhematoma or positional molding. |
| Tier 2 | Progressive Myopia | Binocular IR photoscreener | 88% / 76% | 88% / 95.2% | Ophthalmologist reads raw capture metrics to separate true axial myopia from transient ciliary spasm. |
| Tier 2 | Amblyopia (Lazy Eye) | Binocular IR photoscreener | 88% / 76% | 88% / 94.5% | Specialist cross-checks interocular refraction delta with acuity response, removing non-cooperative fixations. |
| Tier 2 | Strabismus (Squint) | Binocular IR photoscreener | 85% / 86% | 85% / 96.0% | Over-read eliminates pseudo-strabismus caused by prominent epicanthal skin folds. |
| Tier 2 | Leukocoria / Media Opacity | Binocular IR photoscreener | 86% / 82% | 86% / 99.2% | Priority over-read separates benign off-axis flash reflections from congenital cataract or retinoblastoma — any abnormality is an urgent referral. |
| Tier 2 | Allergic Conjunctivitis | 3.9 mm rigid video wand | 81% / 80% | 81% / 96.0% | Over-read evaluates bilaterality, ruling out acute bacterial/viral conjunctivitis. |
| Tier 3 | Conductive Hearing Loss | Auto Hughson-Westlake audiometer | 92% / 85% | 92% / 96.5% | Audiogram correlated with video otoscopy to see if loss is transient cerumen vs ossicular. |
| Tier 3 | Sensorineural Hearing Loss | Auto Hughson-Westlake audiometer | 94% / 88% | 94% / 97.5% | Tele-audiologist inspects false-positive response rates and ambient-noise logs before confirming. |
| Tier 3 | Acute Suppurative Otitis Media | 3.9 mm rigid video wand | 86% / 84% | 86% / 97.5% | Over-read confirms acute inflammation, distinguishing it from crying-induced vascular dilation. |
| Tier 3 | Occlusive Cerumen Impaction | 3.9 mm rigid video wand | 89% / 88% | 89% / 98.5% | Confirms impaction against the eardrum before cerumenolytic drops. |
| Tier 4 | Permanent Dental Caries | 3.9 mm rigid video wand | 84% / 81% | 84% / 96.5% | Dentist grades ICDAS stages to prioritise sealants vs fillings. |
| Tier 4 | Chronic Tonsillar Hypertrophy | 3.9 mm rigid video wand | 86% / 85% | 86% / 97.5% | Airway obstruction graded on the Brodsky scale, assessing sleep-apnea risk. |
| Tier 5 | Endemic Goiter | Native tablet HD camera | 80% / 83% | 80% / 96.0% | Over-read classifies WHO goiter grade 1 (palpable) vs grade 2 (visible in normal posture). |
| Tier 5 | Congenital Torticollis | Native tablet HD camera | 84% / 85% | 84% / 97.0% | Assesses sternocleidomastoid fibrosis, ruling out ocular torticollis from fourth-nerve palsy. |
| Tier 6 | Congenital Heart Defects (Murmurs) | Dual-band smart stethoscope | 92% / 83% | 92% / 96.8% | Cardiologist evaluates murmur duration/radiation, filtering innocent Still’s murmurs from structural defects. |
| Tier 6 | Reactive Airway Disease / Asthma | Dual-band smart stethoscope | 89% / 81% | 89% / 95.5% | Doctor filters transmitted upper-airway or vocal sounds from true expiratory wheeze. |
| Tier 6 | Pediatric Hypertension | Oscillometric BP + ped cuff | 88% / 83% | 88% / 97.0% | Cuff sizing re-checked and white-coat spikes ruled out against AAP 2017 height percentiles. |
| Tier 7 | Severe Nutritional Anemia (<8 g/dL) | 3.9 mm rigid video wand + SpotHb | 84% / 80% | 84% / 96.5% | Pallor cross-evaluated across three mucosal beds; low values trigger a confirmatory hemoglobin test. |
| Tier 7 | Severe Acute Malnutrition (SAM) | Stadiometer & digital scale | 90% / 88% | 90% / 99.0% | Station video inspected for visible severe wasting or bilateral pitting edema. |
| Tier 7 | Pediatric Overweight & Obesity | Stadiometer & digital scale | 88% / 86% | 88% / 97.0% | Fat-distribution pattern assessed to separate athletic muscle mass from adiposity. |
| Tier 8 | Tinea Corporis (Ringworm) | 3.9 mm rigid video wand | 84% / 82% | 84% / 97.0% | Over-read excludes annular psoriasis, granuloma annulare, or nummular eczema. |
| Tier 8 | Scabies | 3.9 mm rigid video wand | 83% / 80% | 83% / 96.8% | Household transmission patterns flagged; whole-body permethrin ordered. |
| Tier 9 | Autism Spectrum Disorder (ASD) | Tablet eye-tracking / gaze | 88% / 76% | 88% / 94.5% | Child psychologist over-reads gaze-fixation curves and verifies DSM-5 communication criteria. |
| Tier 9 | ADHD (Inattention / Hyperactivity) | Digital Vanderbilt battery | 85% / 78% | 85% / 95.0% | Score consistency reviewed across settings, ruling out sensory causes of secondary inattention. |
| Tier 9 | Specific Learning Disability | Digital psychometric battery | 84% / 79% | 84% / 94.0% | Specialist rules out poor instructional exposure or uncorrected myopia. |
| Tier 9 | Adolescent Idiopathic Scoliosis | Digital scoliometer (tablet) | 84% / 82% | 84% / 96.0% | Dorsal alignment photo reviewed, separating structural scoliosis from postural asymmetry. |
| Tier 9 | Adolescent Anxiety & Depression | Modified PHQ-A screener | 88% / 78% | 88% / 94.5% | Follow-up tele-interview on positive screens, ensuring safe emotional support. |
No conditions match your search.
Grounding & scope
What these figures are, and what they are not.
Screening is an early-identification and triage protocol. Every parent-facing report is reviewed and validated by a licensed pediatrician before release; trained technicians conduct on-site capture, and devices are subject to calibration and quality-control. AI-assisted flags can over- or under-flag and are never a diagnosis.
Standalone accuracy ranges reflect the instrument class and screening conditions; combined ranges reflect the added pediatrician over-read. A flagged result is a reason to consult a qualified physician, who decides whether repeat screening, the gold-standard diagnostic test, treatment, or no further action is appropriate. This does not replace comprehensive clinical examination.