Medical Information Sources, Implementation Methods, and Boundaries

Last reviewed: August 10, 2026

CuddleKeep is a record-keeping, organization, and reminder tool for parents and caregivers. It is not a medical device and does not provide diagnosis, treatment, prescriptions, dosage, or advice on the next time medication should be taken. This document describes the sources for the current core health content and how that content is simplified in the App, to support review, testing, and future updates.

Neither this project nor its developers have provided verifiable medical-institution credentials, medical-practitioner credentials, regulatory approval, or clinical validation in the repository. Product copy and store materials therefore must not imply that any such qualification exists.

For the item-by-item source comparison, active-path corrections, closed gates, and professional sign-off requirements in this round, see the [Evidence-Based Engineering Audit of Medical Content (2026-08-08)](MEDICAL_CONTENT_AUDIT.md). That record is an evidence-reviewed engineering audit, not clinical reviewed. Without verifiable sign-off by a licensed professional, the engineering-verification status must not be rewritten as “medically reviewed and approved.”

Verified Core Content

Content	Primary sources	Implementation in the App	Important limitations
Growth percentiles	WHO Child Growth Standards, WHO Q&A	The local table stores only five anchor points: P3, P15, P50, P85, and P97. Values are rounded to 0.1, and results between whole months are linearly interpolated.	This is an approximate display, not the official WHO calculator or a complete LMS/Z-score calculation. A single position cannot be used to diagnose a growth problem; continuous trends and professional assessment should be considered.
Neonatal jaundice	NICE CG98 Recommendations	Saves a transcutaneous bilirubin value or a caregiver-described visual range; it only flags seeking medical care for onset within 24 hours of birth, persistence beyond 2 weeks in a full-term infant or 3 weeks in a premature infant, and warning signs such as white/gray stool or dark urine.	It does not determine severity from a value or visual range and does not provide phototherapy/exchange-transfusion thresholds or treatment advice. Visual observation cannot replace bilirubin measurement. Follow-up testing is arranged by a medical institution.
Children’s temperature	NHS: Fever in children	Neutrally stores temperature, measurement site, and time; for high or low temperatures in young infants and unsuitable measurement methods, it displays conservative age-related prompts to seek medical care.	It does not automatically diagnose the cause of fever from a single value or provide treatment advice. The specific threshold, measurement site and method, and the baby’s overall condition all affect interpretation and still require review by pediatric/emergency professionals for the target region. Other symptoms or caregiver concern may also require medical care.
Medication records	FDA: Got a Sick Kid? Don’t Guess. Read the Label	Displays only the medication name, free-text dose, time, and number of records in 24 hours as entered by the caregiver.	It does not infer active ingredients from a product name or check duplicate ingredients in combination products, interactions, minimum intervals, daily limits, or whether another dose can be given. Users should follow the medicine label and consult a doctor or pharmacist.
Breast-milk storage	CDC: Breast Milk Storage and Preparation	Displays household-storage references for freshly expressed milk: up to 4 hours at room temperature ≤25°C, up to 4 days refrigerated, best quality within about 6 months frozen and acceptable within 12 months, and use within 24 hours after complete thawing in the refrigerator; warmed milk or milk left after feeding is referenced for handling within 2 hours. Prompts are generated from the time recorded by the user, and a conservative “do not feed; discard” action appears once the applicable limit corresponding to the record is reached.	This is household-handling guidance based on a user-entered time, not microbiological testing or individualized “safety certification” of a particular bag of milk. For prematurity, illness, temperature fluctuation, uncertain hygiene conditions, or separate medical-institution requirements, follow individualized professional advice.
Starting complementary foods, texture, and general safety information	NHC, “婴幼儿营养喂养评估服务指南（试行）” (Guidelines for Infant and Young Child Nutrition and Feeding Assessment—Trial), “婴幼儿喂养健康教育核心信息” (Core Health-Education Messages for Infant and Young Child Feeding), CDC: When, What, and How to Introduce Solid Foods	“Around 6 months” and texture stages are general planning examples only. The interface also tells caregivers to consider the baby’s developmental readiness and professional feeding advice; food texture should match current chewing and swallowing ability; ingredients should be handled hygienically; food requiring cooking should be heated thoroughly; and an adult should supervise continuously during feeding. The complementary-food observation window only organizes caregiver records.	The App does not decide from age alone whether complementary foods should begin, does not claim that an ingredient is “safe” or “has passed” based on observation days, does not automatically determine whether the next food can be introduced, and does not provide individualized portion, cooking-time, or temperature advice through recipes. Severe eczema, a known food allergy, swallowing difficulty, prematurity, or other special circumstances require individualized guidance from a doctor or qualified professional.
Local menu, inventory, and shopping organization	No medical or nutritional threshold is used. The ingredient library contains only a “reference starting age” label; all other processing concerns menus, inventory, and shopping data entered manually by the user.	Stores household menus and inventory locally, organizes missing items by comparing the text entries in the two, and generates a shopping TXT that the user may actively export. When generating a one-week record plan, it excludes from the bundled ingredient library only items whose reference starting age has not yet been reached and skips items that the user placed on an avoid list; the remainder is ordered by category and user preference.	The exclusion only keeps clearly unsuitable items out of the list and is not a suitability determination for a particular baby. It does not claim nutritional balance, allergy safety, or that an ingredient “can be introduced”; it does not make individualized recommendations based on medical history or records; and it does not claim that an ingredient has passed merely because enough observation days have elapsed. The caregiver decides whether to feed a particular item to the baby, and special circumstances require professional guidance.
Mainland China vaccination-schedule reference and manual completion	National Disease Control and Prevention Administration, “国家免疫规划疫苗儿童免疫程序及说明（2026 年版）” PDF (National Immunization Program Vaccine Schedule for Children and Explanatory Notes—2026 Edition)	The offline page displays the CN region, source version, source review date, reference age, and not-clinically-signed-off status; the user may independently mark an item as completed.	The content pack remains `draft` and does not generate calculated vaccine or well-child-care notifications. Page dates are not appointment dates, deadlines, or catch-up conclusions. Vaccine formulation and route, contraindications, special health conditions, combination-vaccine substitution, local supply, free-coverage scope, catch-up vaccination, and appointments are all subject to confirmation by the local disease-control authority and vaccination provider.
Age-based themes and parent-child activity ideas	NHC, “3 岁以下婴幼儿健康养育照护指南（试行）” (Guidelines for Nurturing Care for Children Under Age 3—Trial), plus sources stored item by item in the current static content	Displays themes, parent-child activities, and toy ideas only according to profile age; the caregiver chooses whether to try them.	It does not display a milestone-completion rate, score, determine normality/delay, or generate automatic training, care plans, or notifications. Prematurity, developmental concerns, and other individual issues require assessment by well-child-care or developmental-behavioral professionals.
51 offline reference knowledge articles	Each article carries its own organization, title, URL, and source review date. Sources are primarily official pages from NHC, WHO, CDC, AAP / HealthyChildren, NHS, NICE, and similar organizations.	All articles remain at `review.state: draft` without clinical sign-off. The release build displays only the 51 articles that match both an explicit article ID and the current content hash, and each article displays its sources and disclaimer.	This is a reference library that users may browse actively, not “clinically reviewed and approved” content. It does not use individual records to push, diagnose, score, or generate medical/developmental conclusions. A change to the body invalidates the old hash allowlist; the article can be restored to display only after renewed review.
Postpartum recovery, follow-up, and lochia records	WHO: Raising the importance of postnatal care, WHO recommendations on maternal and newborn care for a positive postnatal experience (2022), NHS: Your body after the birth	The timeline only helps users record physical condition, emotions, and matters to discuss at follow-up, and expressly states that every person’s recovery process differs. The lochia and wound modules save the user’s selections and display a safety prompt to contact a medical institution for persistent heavy bleeding, a clear increase, or other concerning signs.	It does not determine from the number of postpartum days whether recovery is “normal,” infer the cause of bleeding, pain, wound, or emotional change, or replace local postpartum follow-up and emergency arrangements. WHO and NHS sources do not automatically resolve differences in Chinese service pathways; in an emergency, contact a local medical institution.
Supplement records	No dosage source or uniform supplementation regimen is built in; actual use follows advice the user has already received from a doctor, pharmacist, or health-care institution.	Vitamin D, AD, calcium, DHA, probiotics, iron, and similar entries only record whether the user checked an item at a certain time and any note entered by the user.	The App does not determine from age, weight, feeding method, or existing records whether supplementation is needed and does not provide dosage, frequency, course length, maximum, or next-use time. A category name and check-in do not constitute a recommendation for use.
Seven-day emotional care	The questions and options are original to CuddleKeep; urgent safety wording references WHO Suicide Q&A	Six questions reviewing daily life. It does not calculate or display a score, risk level, or diagnostic conclusion. Only when the user directly selects that thoughts of self-harm occurred or that safety cannot be assured does it display safety guidance to contact local emergency/crisis services, not remain alone, and tell a trusted person.	This is not a clinical scale and has not been validated for screening efficacy. It cannot be used to determine whether depression or another mental-health condition is present. Regardless of the answers, the user may proactively seek professional help.
Child health-management time points (automatic planning currently disabled)	NHC interpretation of “0～6岁儿童健康管理服务规范” (Standards for Health Management Services for Children Aged 0–6), National Basic Public Health Service Standards (Third Edition) PDF	Build 3 does not generate a well-child calendar or automatic visit-preparation reminder. Users may create their own to-dos and actively generate a well-child summary from their own records. This source is used only to constrain future content review.	Local service arrangements may change and are subject to notices from local primary medical and health institutions. Before any automatic time point or notification is enabled in the future, it must be separately reviewed, signed off, and its metadata re-examined.
Infant and young child nutrition and feeding assessment	NHC, “婴幼儿营养喂养评估服务指南（试行）” (Guidelines for Infant and Young Child Nutrition and Feeding Assessment—Trial)	Used to understand that feeding circumstances, growth records, and professional services should be considered together, rather than drawing a conclusion from a single App metric.	The current App’s five-anchor approximation table, feeding records, and trend summaries are not equivalent to the complete assessment described in the guideline and performed by primary medical/health or maternal-and-child health institutions.

Unified Product Boundaries

“Record” means only that user input is saved; it does not mean that a doctor has verified the content.
A “red flag” identifies a circumstance in which prompt contact with a medical institution should be considered; it does not mean that the App has diagnosed a disease.
The App does not make treatment or medication decisions from a jaundice value, temperature, or medication name.
Stool color, projectile vomiting at any age, and green/yellow-green or bloody vomit trigger conservative medical-care or emergency red flags, but the interface does not infer a specific disease or surgical timing.
The current release includes an offline reference page for the Mainland China vaccination schedule and user-entered completion records, but does not generate calculated vaccine / well-child notifications or provide conclusions on vaccination eligibility, formulation and route, contraindications, catch-up vaccination, or appointments.
The local Kitchen only organizes menus, inventory, missing items, and shopping TXT files; age-based pages provide only themes and activity ideas; the knowledge library displays only reference articles locked by ID + content-hash allowlists. None of the three generates an individual safety determination, developmental score, diagnosis, or automatic plan / notification.
All source links, version numbers, thresholds, and interface wording should be included in pre-release regression testing.

Release-Level Review Not Yet Completed

The table above records a comparison of the current implementation’s wording/values against official sources. It is not a comprehensive clinical review completed by a qualified professional, nor is it a conclusion of medical-device, regulatory, or Apple approval.

The current release candidate allows users to actively browse the Mainland China vaccine reference and 51 knowledge drafts without clinical sign-off, and enables unscored age-based activity ideas and local Kitchen organization. The final interface, sources, disclaimers, hash invalidation, and the boundaries of no notifications / scoring / determinations / automatic planning still require release-level regression on the final candidate. “Available for reference reading” must not be rewritten as “clinically reviewed.”
Currently visible red-flag prompts concerning temperature, jaundice, vomiting, bleeding/lochia, wounds, emotional safety, and similar subjects still require independent professional review to confirm that they do not omit important information, misleadingly delay medical care, or use emergency-service wording unsuitable for users in China.
Complementary-food sources have been compared for start timing, developmental readiness, texture, and general safety boundaries. Kitchen can organize missing items and shopping text from menus and inventory entered by the user. Its one-week record plan filters entries using the bundled ingredient library’s “reference starting age” and skips ingredients placed by the user on an avoid list. This is exclusion at the list level and does not constitute portion advice, an individualized menu, or an ingredient-safety conclusion. Special foods and allergy-related circumstances still require review by a registered dietitian, pediatrician, or allergy professional.
Postpartum sources have been used to limit the feature to recording, communication, and prompts to seek medical care, but they do not establish that the recovery process, follow-up arrangements, or combinations of red flags have completed obstetric clinical review.

“Seven-day emotional care” has been changed to an original, non-diagnostic review of daily life and does not use the name, questions, total score, or score thresholds of a clinical scale. Records from older versions are still retained under the internal type='epds' for compatibility, but the interface and summaries no longer display their old score or level. This internal key exists only for data compatibility and does not mean that the product still provides the corresponding scale.

The sources, implementation, and principal limitations used in this round have been recorded for breast-milk storage, complementary foods, vaccines, the knowledge library, and postpartum content. Supplements are expressly limited to user records without dosage advice. These completed items should still be reviewed by an appropriately qualified professional before release; the vaccines and 51 knowledge articles remain explicitly marked as draft / not clinically signed-off reference content. Until a verifiable review record has been completed and saved, store copy must not claim “medically reviewed,” “clinically validated,” “authoritative diagnosis,” or a similar conclusion.

Content Maintenance Process

Each time a release is prepared:

Check whether an official source has been updated, withdrawn, or moved to a different link;
Compare source code and the interface to confirm that numbers, age boundaries, units, and disclaimers are consistent;
Run tests to confirm that health modules have not reintroduced diagnostic, treatment, dosage, medication-interval, or absolute-correctness wording;
Record the review date, source version, change summary, and reviewer;
Arrange independent review of high-risk content by a person with appropriate professional qualifications and retain verifiable records of the review scope and date. Do not publicly claim that review has been completed when it has not actually been completed.

Apple Review Reference

App Review Guidelines: pay particular attention to Section 1.4.1 on the accuracy of medical information and Section 1.4.2 on drug-dosage calculations.
A medical disclaimer does not replace content accuracy, evidence, and compliance review; the product description should accurately state that this is a care record-keeping tool.

Current official sources for this region

General reference only. This App has not clinically reviewed this content and does not create automatic schedules, judgments, scores or reminders from these sources. Open the official source and consult a local qualified professional.

postpartum