BSA calculator: Mosteller, Du Bois, Haycock and Gehan-George

Every formula above takes weight in kilograms and height in centimetres. A site that records pounds and inches must convert first: 1 lb = 0.45359237 kg and 1 in = 2.54 cm. A participant who weighs 154.3 lb and is 68.9 in tall is 69.99 kg and 175.0 cm, and the Mosteller BSA is 1.84 m². The common errors are using metres for height in a formula written for centimetres (the Mosteller result is then off by a factor of 10), swapping height and weight in the Du Bois exponents, and recording weight in stone. If you capture both unit systems at different sites, record the unit with the value and convert once, in one place. The lab unit converter shows the same discipline for laboratory units, and the vital signs eCRF template is where height and weight are collected.

Rounding matters when a quantity is calculated from the BSA. 1.8447 and 1.8481 differ by 0.2%, which usually disappears into rounding, but a protocol should say whether BSA is rounded to two decimals before it is used, whether the unrounded value is used, and what happens at a boundary such as a banding table. Different systems (the site pharmacy tool, the EDC, the IWRS) that each calculate BSA with a different rounding rule will disagree on some participants, which is a reconciliation problem that has nothing to do with the participant. Pick one source of truth, write the rule down, and use it everywhere.

BSA is a derived estimate, not a measurement. The formulas were fitted to relatively small samples, and none is exact for an individual, least of all at extreme sizes. The spread in the table is small for a typical adult and larger at the ends: for 3.5 kg and 50 cm the formulas run from 0.21 to 0.23 m², roughly 10 to 12% apart, and for 120 kg and 165 cm from 2.23 to 2.41, about 8%. For adults of ordinary size, which formula you choose rarely matters; for children, infants and people with obesity it can. Pediatric protocols often name Haycock or Gehan-George, adult oncology protocols often name Mosteller, and the safest rule is to use the formula the protocol and the reference you are following name.

Some protocols cap BSA at a stated maximum before it is used in a calculation. Mechanically, a participant whose BSA is 2.24 m² and who is capped at 2.00 m² has a BSA input about 11% lower than the calculated value. Whether to cap, at what value, and what to do with the actual weight are clinical and protocol decisions on which guidelines and institutions differ; the capped value and the uncapped value should both be stored, so a reviewer can see what was done. Nothing on this page is a recommendation about how to dose, and the investigator and the pharmacist should apply the protocol’s own rule.

BSA is also used outside dosing: normalising cardiac output (cardiac index), glomerular filtration (eGFR is reported per 1.73 m², which the eGFR calculator explains) and some physiological measures in pediatrics. In those uses the choice of formula should match the reference ranges you compare against, otherwise the normalised value carries a hidden bias.

The inputs deserve as much attention as the formula. Weight and height should be measured, not recalled, on calibrated equipment under a standard procedure (the same clothing rule, no shoes, the same units), and the protocol should say whether the participant’s BSA is calculated once at baseline or recalculated at later visits, and if recalculated, how large a weight change triggers it. Self-reported height and weight carry known errors that feed straight into the BSA. A change in the BSA between visits that is bigger than the formula’s own uncertainty usually means a data entry or unit error rather than a physiological change, which is a good reason to add a visit-to-visit consistency check on the inputs. Where a study reports BSA as a baseline characteristic, state the formula in the methods so the table can be reproduced.

Entering BSA by hand invites transcription errors and rounding disagreements. In Capture, BSA by Mosteller and by Du Bois are built-in calculated fields: the site enters height and weight on the vital signs or anthropometry form, and the BSA appears read-only during entry. The same formula is therefore applied to every participant at every site, and the value cannot be edited by hand. Range high and range low edit checks on height and weight raise an auto-query when a value is implausible, such as a height of 17.5 because a decimal point slipped, which is exactly the error that would otherwise carry into BSA. Query management follows the query to resolution, and the field-level audit trail records the old value, new value and reason for any later correction, so the BSA that was used can be reconstructed. See eCRF calculated fields and skip logic for how these fields are set up.

The built-in list names Mosteller and Du Bois. If your protocol specifies Haycock or Gehan-George, ask about it in a demo rather than assuming it is available. Pediatric studies can use the same height and weight fields with age-specific ranges (see EDC for pediatric clinical trials), and oncology protocols that use BSA to calculate cycle quantities fit the structure described in EDC for oncology clinical trials. The physical exam eCRF and the vital signs eCRF templates are the usual starting points, and a blank eCRF PDF export lets the ethics committee see exactly which fields feed the calculation.