| Metric | Value | Copy |
|---|---|---|
| {{ row.label }} | {{ row.value }} | |
| No metrics yet. | ||
| {{ header }} | Copy |
|---|---|
| {{ cell }} | |
| No detail rows yet. |
| Priority | Recommendation | Rationale | Timeline |
|---|---|---|---|
| {{ row.priority }} | {{ row.recommendation }} | {{ row.rationale }} | {{ row.timeline }} |
| No guidance recommendations available yet. | |||
| Level | Signal | Count | Action lane |
|---|---|---|---|
| {{ row.level }} | {{ row.signal }} | {{ row.count }} | {{ row.action }} |
| No ladder signals available yet. | |||
Growth screening is about noticing when a sequence of measurements starts to look less reassuring than the latest number alone. A child can still appear to be gaining height and weight while drifting into a lower percentile channel, or can stay in a familiar percentile band despite a single noisy visit. This checker turns serial measurements into a structured screening view of that pattern.
The package reads visit rows as age in months, height in centimeters, and weight in kilograms, calculates height, weight, and body mass index percentiles for each visit, and then looks for the specific signals this slug ships: low linear-growth percentile, weight extremes, persistent high body mass index percentile after age 24 months, downward height band crossing, large recent weight-percentile shift, and low recent height velocity. Instead of leaving those signals scattered across a manual chart review, it collects them into summary badges, a flag table, a guidance tab, an alert ladder, a trend chart, and a JSON export.
That makes it useful when the real question is triage rather than pure charting. A pediatric follow-up, nutrition review, school-health conversation, or parent update often needs a quick answer to, "Does this look routine, should it be watched closely, or does it deserve faster review?" The tool keeps that evidence visible instead of hiding it behind one headline label.
The caution is equally important. Its WHO-style, CDC-style, and hybrid choices are internal reference modes layered onto bundled anchor curves, not live official chart lookups. The alert rules are configurable screening thresholds, not diagnostic criteria, and a concerning result still depends on measurement quality, visit spacing, age context, and the broader clinical picture.
The safest way to use this result is as an organized screening summary for serial growth data. It can sharpen follow-up planning, but it does not replace standardized pediatric growth assessment or professional diagnosis.
Start with clean visit rows and the default threshold posture. Enter one visit per line in Visit data rows as age_months,height_cm,weight_kg. The parser ignores blank lines, sorts the kept visits by age, and keeps the latest row when the same age appears twice. That is a sensible baseline because it stops duplicate-month edits from double-counting a visit.
For many first passes, the default settings already answer the main question. You get a neutral Balanced risk posture, a default follow-up cadence, a standard body mass index alert threshold, and a single-interval height-velocity check. Run that once before tightening anything.
The advanced controls change the meaning in concrete ways. Growth reference set changes which internal reference adjustment is used. Band crossing alert and Percentile-shift alert determine how much recent movement counts as notable. Height velocity floor and Velocity persistence decide when slow linear growth becomes a critical signal. BMI alert percentile and BMI persistence requirement decide how sticky a high body mass index trend must be before it is flagged. The guidance controls then shape how assertively the tool words follow-up recommendations once the flags already exist.
Critical flags, Latest height percentile, and the detail row for recent height velocity before anything else.The best use case is a short, believable series measured with the same method over time. This slug is less helpful when the input history is sparse, mixed in quality, or entered after a unit or technique change that the package cannot detect on its own.
The checker is fully browser-side for this slug. It does not ship a helper endpoint, so the visit data, calculations, guidance rows, chart image exports, and JSON payload all stay in the current browser session unless you choose to copy or download them.
Each visit row is normalized into age, height, weight, and body mass index. The code then calculates height, weight, and body mass index percentile and z-score values from built-in sex-specific median and spread curves, with modest adjustments for the chosen reference mode. Those outputs are therefore package-specific screening estimates. They are inspired by common growth-chart interpretation practice, but they are not a direct rendering of official CDC or World Health Organization software.
The flag logic is intentionally specific rather than generic. Some signals come from the latest visit only, some depend on the recent interval, and some require persistence across multiple visits. That mix matters because the tool is trying to separate one surprising point from a short pattern that looks harder to dismiss.
| Signal | What the package checks | Why it matters |
|---|---|---|
Low linear growth percentile |
Latest height percentile below the 3rd percentile | Promotes a low-height pattern into a critical review signal |
Low weight percentile and High weight percentile |
Latest weight percentile below the 3rd or above the 97th percentile | Captures weight patterns that may deserve context even when height is less concerning |
Elevated BMI percentile |
Consecutive visits at 24 months or older at or above the configured body mass index threshold | Adds persistence so one isolated high point does not automatically dominate the screen |
Downward height band crossing |
Latest height percentile drops across enough major bands between the last two visits | Highlights a channel change that can be easy to miss in raw centimeter gain |
Large weight percentile shift |
Latest weight percentile jump crosses the band threshold and exceeds the configured point shift | Separates ordinary fluctuation from more notable short-interval movement |
Low recent height velocity |
Annualized height gain between recent visits falls below the configured floor, optionally across two intervals | Pushes slow linear growth into the triage logic even when the absolute percentile is not yet extreme |
The major percentile bands used for crossing checks are fixed at 3, 15, 50, 85, and 97. The chart tab plots height, weight, and body mass index percentile lines against dashed 3rd and 97th threshold lines, while the Alert Ladder tab compresses the outcome into critical, watch, and composite triage lanes.
The guidance layer reads the computed counts back through Risk tolerance, Guidance follow-up interval, Intervention urgency, and Catch-up priority to generate recommendation wording and timing. Exports are broad for a screening slug: the metrics and details tables can copy or download CSV and export DOCX, the guidance tab can copy or download CSV, the chart can be saved as PNG, WebP, JPEG, or CSV, and the full payload can be copied or downloaded as JSON.
The core limitation is that this slug is deliberately a screening engine. It organizes serial measurements, but it cannot tell whether a tape measure changed, whether shoes were removed, whether a visit was acute-illness affected, or whether a concerning pattern fits the full clinical story.
Sex and paste the visit history into Visit data rows as age_months,height_cm,weight_kg. A valid series needs at least two visits.Screening Flags next. That table tells you exactly which signals fired, what evidence line triggered them, and what action cue the package attaches to each one.Guidance when you need wording for follow-up framing. The tone line and recommendation priorities change when you adjust risk tolerance, urgency, or catch-up emphasis.Alert Ladder when you need the fastest triage view for handoff. It separates immediate-critical items, watch items, and the composite score.Screening Flag Map if you want to see how the percentile lines behave across time and export that chart as an image or CSV.JSON for a machine-readable record of the inputs, summary, tables, and chart payload.If the red error banner appears, fix the rows before interpreting the output. The package will not produce a trustworthy screen from an incomplete or non-numeric visit history.
A reassuring run usually looks like zero critical flags, zero or few watch flags, and a latest visit that stays inside a familiar percentile neighborhood. A more concerning run usually combines low linear-growth percentile, a clear recent drop in height percentile bands, or low recent height velocity.
The watch category deserves nuance. A high weight percentile or persistent high body mass index percentile does not mean the same thing as downward linear-growth crossing, and the tool does not pretend that they are interchangeable. The details table is where that distinction becomes useful: it tells you whether the current question is about size, trend, persistence, or recent growth speed.
The follow-up language also has two layers. The metrics summary uses a straightforward baseline of 0.5 months when any critical flag is present, 1 month when watch flags are present without critical flags, and 3 months when no flags are active. The guidance tab can then sharpen or soften that baseline through the selected risk and urgency posture.
Treat the result as a signal to review, not as proof of cause. The package can show that the latest interval looks slower, that the latest visit fell below a threshold, or that a body mass index pattern persisted across visits. It cannot by itself explain why that happened.
Example 1: A routine series with no active flags. Enter visits such as 6,65.2,7.2, 12,73.8,8.9, 24,84.6,11.3, and 36,92.4,13.2. A pattern like this usually keeps height, weight, and body mass index in a familiar range and leaves the summary in the routine lane.
Example 2: A recent low height-velocity concern. Suppose the later visits are 24,84.6,11.3, 30,86.1,12.1, and 36,87.3,12.9 with the default velocity floor. The child still gains height, but the annualized gain between the recent visits is slow enough that the checker can raise a critical low-height-velocity flag. That is exactly the kind of pattern raw centimeters can hide if you do not annualize the interval.
Example 3: A persistent high body mass index trend after age 24 months. Imagine a series where visits at 24 months and later remain above the configured body mass index alert percentile. With the default persistence requirement of one visit, the tool can raise the body mass index flag immediately; with a higher persistence setting, it waits for repeated visits to stay above the threshold.
No. The slug uses built-in growth anchors plus internal WHO-style, CDC-style, or hybrid adjustments. It is better understood as a screening-oriented estimate engine than as an official chart renderer.
Because the shipped flag logic only counts body mass index persistence once visits are 24 months or older. That matches the general practice of using body mass index-for-age as a later-infancy and childhood screening measure rather than applying it the same way in very young infants.
The checker uses fixed major percentile marks at 3, 15, 50, 85, and 97. A crossing alert depends on how many of those bands are crossed between the recent visits and what threshold you set in the advanced controls.
No helper is shipped for this slug. The calculations, tables, chart, and exports are built in the browser for the current session.
No. A single result can trigger a screen, but diagnosis still depends on careful measurement technique, serial context, and professional clinical interpretation.