Evidence base

What AI Dietitians is built on.

Every number the platform produces comes from somewhere. This page says where, so you can trace a figure back to its source, check the version we are using, and decide whether it is the one your service works from.

The licence-required attribution record is Attributions and Data Sources, and it is the authoritative version of everything below.

The data

The reference datasets

Australian Food Composition Database

Food Standards Australia New Zealand (FSANZ)

Release 3, published 23 December 2025, covering 1,588 foods. It is the source of the whole-food figures in the food reference, meal planning, carbohydrate counting and blended tube feeding tools. Energy in kilocalories is derived by us from the kilojoule values FSANZ publishes, because the database publishes kilojoules only, and a value is flagged as imputed where FSANZ records it as calculated, borrowed or estimated rather than analysed.

Licence Creative Commons Attribution-ShareAlike 3.0 Australia

FSANZ AFCD

IDDSI framework

International Dysphagia Diet Standardisation Initiative

The Levels 0 to 7 continuum for drinks and foods, the descriptors and the testing methods, used wherever texture and consistency appear in the platform. Reproduced under the framework licence, with the copyright line carried on every surface that shows it.

Licence Creative Commons Attribution-ShareAlike 4.0 International

iddsi.org/framework

Nutrient Reference Values for Australia and New Zealand

NHMRC and the New Zealand Ministry of Health, 2006, updated 2017

Recommended dietary intakes, adequate intakes, estimated average requirements and upper levels throughout the reference tables and calculators. The Australian Dietary Guidelines and the Australian Guide to Healthy Eating (NHMRC, 2013) supply the recommended daily serves by age and sex.

Licence Used under the Creative Commons Attribution licence NHMRC publishes them under

nrv.gov.au

WHO growth standards

World Health Organization

The WHO Child Growth Standards (2006) and WHO Growth Reference (2007). The LMS parameters behind the paediatric BMI-for-age z-scores and percentiles.

Licence Used in accordance with the WHO permissions terms for that material

who.int

The guidelines

What the calculators cite

Each calculator names its source on the result screen. These are the ones it names most often.

ESPEN European Society for Clinical Nutrition and Metabolism: Protein and energy targets in critical care, surgical and wound nutrition; the glucose infusion rate ceiling in the parenteral nutrition tool; the refeeding guidance on reducing the rate when phosphate falls.
NICE CG32 Nutrition support for adults, 2006, updated 2017: The refeeding syndrome risk criteria and the starting rates that follow from them, including the asymmetric rule of one high-risk criterion or two moderate ones.
GLIM Cederholm T, Jensen GL, Correia MITD, et al. GLIM criteria for the diagnosis of malnutrition, a consensus report from the global clinical nutrition community. Clinical Nutrition 2019;38(1):1-9.: The two-step malnutrition diagnosis, which requires a positive validated screen before the phenotypic and aetiologic criteria are applied, and which feeds the ICD-10-AM coding helper.
KDOQI KDOQI Clinical Practice Guideline for Nutrition in CKD: 2020 Update. American Journal of Kidney Diseases 2020;76(3 Suppl 1):S1-S107.: Protein recommendations by chronic kidney disease stage and on dialysis, in the renal calculator, alongside Dietitians Australia guidance.
Dietitians Australia Professional practice standards and Australian practice guidance: Practice expectations for documentation and communication, and Australian guidance where it differs from the international position.

The instruments

Screening and assessment tools

The platform implements the following published instruments. Each remains the work of its authors and publishers, cited here and again inside the product. Scoring thresholds and interpretation guidance shown alongside them are in our own words.

MST Ferguson M, Capra S, Bauer J, Banks M. Development of a valid and reliable malnutrition screening tool for adult acute hospital patients. Nutrition 1999;15(6):458-464.
NRS-2002 Kondrup J, Rasmussen HH, Hamberg O, Stanga Z. Nutritional risk screening (NRS 2002): a new method based on an analysis of controlled clinical trials. Clinical Nutrition 2003;22(3):321-336.
GLIM Cederholm T, Jensen GL, Correia MITD, et al. Clinical Nutrition 2019;38(1):1-9.

MUST is deliberately absent. It is BAPEN's instrument and reproducing its scoring is what their terms reserve, so we removed it from the product and from this site rather than reimplement it. Our malnutrition screening explainer describes how the tools differ and points at BAPEN's own calculator.

Health service data

Diet codes, coding and medicines

The diet code vocabulary in the oral nutrition support compatibility checks comes from the Therapeutic Diet Specifications for Adult Inpatients published by the NSW Agency for Clinical Innovation. Codes are abbreviated for use in the product; the ACI publication is the authoritative specification.

The indicative complication-uplift figures in the malnutrition coding tools are derived from the National Hospital Cost Data Collection and National Efficient Price determinations published by IHACPA. They are indicative only and are not a billing prediction. The authoritative tariff is the one in effect at discharge, as applied by your health service's coding pipeline.

Drug and nutrient interaction entries are compiled from TGA Product Information, NPS MedicineWise resources and the primary literature, with each entry recording the source it came from. It covers interactions relevant to clinical dietetics and is not a complete medicines reference. It does not replace a current pharmacopoeia, your hospital formulary, or a pharmacist.

What this does not mean

No endorsement. FSANZ, the NHMRC, the WHO, IDDSI, NICE, ESPEN, Dietitians Australia and the authors of the instruments above do not endorse AI Dietitians. None of them has reviewed how their material is applied here.
No validation or accreditation. Using a published guideline is not the same as being validated against it. No body listed on this page has certified, accredited or clinically validated this software.
Composition data is indicative. Nutrient content varies between batches and brands with season, processing, soil and growing conditions. Where precision matters, for research or food labelling, use foodstandards.gov.au directly.
Outputs support your judgement. Calculator results and AI drafts are indicative information, not a diagnosis or a treatment decision. Verify against primary sources before you act. See the Disclaimer.

If something here is wrong, out of date, or missing an attribution it should have, tell us at copyright@aidietitians.com.au. We fix attribution problems quickly.

Australian clinical content

See what it does with all of this.

The full capability list, the free IDDSI and screening references, and how your data is handled.

See every feature How data is handled Start free