Standard Deviation Index Calculator

Background.

A laboratory that measures glucose on a control sample and gets 5.4 mmol/L, when other laboratories using the same analyser and reagents average 5.2, is reading high. Whether that gap is trivial or a problem depends on how much those other laboratories disagree among themselves. If their results scatter widely, 0.2 is lost in the noise; if they agree closely, it stands out.

The standard deviation index captures exactly that. It divides the difference between a result and its target by the standard deviation of the comparison group, turning a gap in mmol/L, ng/mL or seconds into a count of standard deviations. Because the units cancel, the same scale applies to a cholesterol assay and a clotting time, and results across a whole menu of tests can be read side by side.

In clinical laboratories the index is the backbone of peer comparison and external quality assessment. Each month a laboratory sends its control results to a programme that pools them with peers on the same instrument and method; the report returns the peer mean, the peer standard deviation and the laboratory's SDI for each test. A positive value means the laboratory reads high relative to the group, a negative one low. A value that stays on the same side month after month suggests a calibration bias, even when it is small.

The arithmetic is the same as an ordinary z-score. What it means depends on the target and spread you supply, and acceptance limits are set by each programme or by the laboratory's own quality plan, not by the formula. This page only accepts positive results and targets, so it does not suit quantities that can be zero or negative.

What is standard deviation index calculator?

The standard deviation index (SDI) is the signed difference between a result and a target value, divided by a standard deviation that describes the expected spread of comparable results. In laboratory quality control the target is usually a peer-group mean or assigned value and the standard deviation is the peer group's, but the calculation is the same for any setting. The result has no units: it is a count of standard deviations.

How to use this calculator.

1. 01Enter the result being assessed — a laboratory's mean for a control material over the month, or a single proficiency-testing result.
2. 02Enter the comparison target in the same unit: the peer-group mean or the value assigned by the scheme.
3. 03Enter the comparison standard deviation in that unit too, taken from the same report as the target. Use the peer group's spread or the scheme's stated standard deviation, not your own laboratory's imprecision.
4. 04Read the sign as well as the size: positive means above target, negative below. Compare it with the acceptance limits of your scheme or quality plan.

The formula.

SDI = (x − T) ÷ s, where x is the result, T the comparison target and s the comparison standard deviation, all in the same unit. The numerator is the bias — how far, and in which direction, the result lies from the target. Dividing by s rescales that bias into standard deviations, so the index is unit-free. The denominator shapes the answer as much as the difference does: the same 5-unit bias is an SDI of 2.5 when the peer standard deviation is 2 but only 0.5 when it is 10. A tight peer group therefore makes small biases visible, while a scattered one hides them. The page treats all three inputs as positive magnitudes and refuses zero or negative values.

A worked example.

Frequently asked questions.

References& sources.

1. [1]ISO 13528:2022, Statistical methods for use in proficiency testing by interlaboratory comparison (BIBLIOGRAPHIC/PAYWALL).
2. [2]NIST/SEMATECH e-Handbook, standard scores.
3. [3]BIPM, The International System of Units (SI Brochure), 9th ed., version 3.01, coherent derived units and quantity equations.