What Should a Baby Tracker Measure?
By Lybadora Editorial Team4 min read
Not yet reviewed by a clinician.
Three things, and a tracker should be judged on them. It should record events a caregiver can observe without equipment, each with a timestamp, and it should only produce a comparison where a published reference exists or where the baby's own earlier days provide one. That rules more out than in: the American Academy of Sleep Medicine publishes total sleep per 24 hours but nothing about wake windows, the CDC states its milestone checklists are not a screening or diagnostic tool, and the WHO growth standards describe a reference population rather than a target for one baby. Anything outside that has to be presented as an observation, never as an assessment.
Explanation
Most baby trackers answer the question of what to measure by measuring everything, then presenting the result as insight. A more useful question is what a record can support, and that has a short answer: observable events, timestamped, compared only against something real.
Supporting evidence
What health authorities publish, and what that leaves a tracker able to say:
| Published | What a tracker can do with it |
|---|---|
| Total sleep per 24 hours by age, naps included (AASM) | compare a 24 hour total, if every sleep including naps is recorded |
| Wake windows by age | nothing, no health authority publishes them |
| Feeding frequency and diaper counts in the newborn weeks (AAP) | compare counts, not amounts |
| Growth standards from a reference population (WHO) | place a measurement beside a range, never turn it into a target |
| Developmental milestone checklists (CDC) | prompt a caregiver to look, and nothing more |
The CDC states its milestone checklists are not a screening or diagnostic tool, which means a tracker that scores development against them has stepped past its evidence.
CDC: Key points about the developmental milestone checklists
The WHO growth standards describe how a reference population of healthy children grew. A percentile is a position in that population, not an instruction.
And the input itself has known error. Parent-reported infant sleep compared against objective measurement of the same nights did not agree closely, which limits what any conclusion drawn from a log can claim, however the log was collected.
Rise & SHINE: parent-reported and objectively measured infant sleep at 6 months
What the numbers mean
The three-part test falls out of the table.
Observable without equipment. A caregiver can see that a feed started, that a nap ended, that a diaper was wet. A caregiver cannot see sleep stages or millilitres taken from the breast, so a tracker that reports them is reporting a guess.
Timestamped. Without a time, an entry cannot produce an interval, a total or an order of events, which is most of what published guidance is written in.
Comparable to something real. Either a published figure, or this baby's own recorded days. A comparison against other babies of the same age is the one thing an app has the data to do and the least reason to do, because a range built from many babies says nothing about whether this one has changed.
What parents can observe
- Whether the app shows you the entries behind a statement. A claim with no entries under it cannot be checked.
- Whether it distinguishes a total from a recommendation. 13 hours is a number. "Not enough sleep" is a verdict.
- Whether it admits missing data or quietly averages over the gap.
- Whether it names a source for any age-based figure it shows.
- Whether it compares this baby to this baby, or to a population.
What may explain a pattern
When an app's number surprises a caregiver, the ordinary causes are usually in the record rather than the baby: a nap nobody logged, a feed counted twice across a handover, a day with a three hour gap in it, a sleep that crossed midnight and landed in the wrong day, or an interval measured start to start in one entry and end to start in the next.
That is why the entries matter more than the number. A figure without its entries cannot be corrected, only believed or doubted.
What parents can do
Ask of any statement an app makes which entries it rests on. Treat anything unsourced and age-based as someone's estimate. And judge a tracker on whether it stops at what happened, because everything past that is a conversation for a paediatrician, not for an app.
What Lybadora does with this record
Lybadora records observable events with timestamps, and its trackers hold the fields a published figure is written in: start and end for sleep, start and side for feeds, counts for diapers, measurements for growth. Nothing is inferred that a caregiver did not see.
Where it looks for patterns, it compares this baby against this baby's own recent days, it says which entries the comparison rests on, and it says when there are not enough days to say anything yet. Published ranges may stand beside a baby's own number, never in place of one. Lybadora does not call a number normal, behind or on track, because the record can say what changed and only a paediatrician can say what it means.
Safety considerations
No tracker, and no article about trackers, screens or diagnoses anything. The CDC is explicit that its developmental milestone checklists are not a screening or diagnostic instrument, and a growth chart percentile is not a verdict on a baby. Concerns about feeding, breathing, weight, alertness, development or behaviour belong with a paediatrician, and an app that appears reassuring is not a reason to delay that conversation. Where a tracker offers an interpretation, it is worth knowing what evidence it rests on before acting on it.
Sources
- American Academy of Sleep Medicine -- Recommended Amount of Sleep for Pediatric Populations: A Consensus Statement
- Centers for Disease Control and Prevention -- Key Points about CDC's Developmental Milestone Checklists
- World Health Organization -- WHO Child Growth Standards
- Rise & SHINE Study (PMC) -- Associations between parent-reported and objectively measured sleep duration and timing in infants at age 6 months