# Standard Deviation Index Calculator (SDI) URL: https://speedycalculator.com/calculators/statistics/sdi/ Last updated: 2026-10-30 > The standard deviation index calculator compares a laboratory's mean with the peer-group (consensus) mean in units of the group's standard deviation – the SDI used in proficiency testing and quality control – and, with the lab's own SD, also gives the coefficient of variation ratio. ## Inputs - Laboratory mean - Consensus group mean - Consensus group standard deviation - Also compare precision? - Laboratory standard deviation ## Results - Standard deviation index (SDI) - Performance - Direction - Bias (lab − consensus) - Bias as % of consensus mean - Laboratory CV - Consensus CV - Coefficient of variation ratio (CVR) ## Formula ``` SDI = (laboratory mean − consensus mean) ÷ consensus SD · CVR = lab CV ÷ consensus CV · bias % = (lab mean − consensus mean) ÷ consensus mean × 100 ``` ## Explanation Clinical and testing laboratories regularly analyse the same control material as many other labs and compare results. The standard deviation index (SDI) condenses that comparison into one number: how many peer-group standard deviations your lab's mean lies away from the consensus mean. It is essentially a z-score for your laboratory. An SDI of 0 means your results match the group exactly. A positive SDI means you read high, a negative one that you read low. Values within ±1 indicate good agreement, up to about ±1.25 is generally acceptable, values between 1.25 and 2 call for a closer look at calibration and trends, and an SDI of ±2 or more is usually considered unacceptable and needs corrective action. Watching SDI over several months reveals a developing bias long before a single result fails. SDI looks at accuracy (bias). Precision is judged by the coefficient of variation ratio (CVR, sometimes CVI): your lab's CV divided by the group's CV. A CVR above 1 means your imprecision is larger than that of your peers. ### How to use the Standard Deviation Index Calculator - **Laboratory mean** - **Consensus group mean** - **Consensus group standard deviation** - **Also compare precision?** - **Laboratory standard deviation** Results update as you type, and everything is calculated in your browser. ### How it works Enter your laboratory's mean for the control, the consensus (peer-group) mean and the consensus standard deviation. Optionally add your lab's standard deviation to get the CV ratio. SDI = (lab mean − group mean) ÷ group SD. ### Worked example A laboratory's monthly mean for a glucose control is 5.32 mmol/L, while the peer group reports 5.10 mmol/L with SD 0.16 mmol/L. SDI = (5.32 − 5.10) ÷ 0.16 = 1.375 – marginal, the lab reads high by 0.22 mmol/L (4.3%). The lab's own SD is 0.14, so its CV is 2.63% against the group's 3.14%, and CVR = 0.84: precision is fine, the issue is calibration bias. ### SDI limits |SDI|InterpretationTypical action0 – 1.0GoodNone1.0 – 1.25AcceptableMonitor1.25 – 2.0MarginalReview calibration, reagents, trends≥ 2.0UnacceptableInvestigate and correct ### Inputs explained FieldWhat to enterLaboratory meanA numberConsensus group meanA numberConsensus group standard deviationA numberAlso compare precision?Choose one: No, Yes – enter the laboratory SDLaboratory standard deviationA number ### What each result means - **Standard deviation index (SDI)** = ([Laboratory mean] - [Consensus group mean])÷[Consensus group standard deviation] - **Bias (lab − consensus)** = [Laboratory mean] - [Consensus group mean] - **Bias as % of consensus mean** = ([Laboratory mean] - [Consensus group mean])÷[Consensus group mean]×100 - **Laboratory CV** = [Laboratory standard deviation]÷[Laboratory mean]×100 - **Consensus CV** = [Consensus group standard deviation]÷[Consensus group mean]×100 - **Coefficient of variation ratio (CVR)** = [Laboratory CV]÷[Consensus CV] ## Worked example Inputs — Laboratory mean: 5.32; Consensus group mean: 5.1; Consensus group standard deviation: 0.16; Also compare precision?: Yes – enter the laboratory SD; Laboratory standard deviation: 0.14. Results — **Standard deviation index (SDI) = 1.375**; Performance = Marginal – check calibration and trends (1.25 – 2); Direction = Laboratory reads high; Bias (lab − consensus) = 0.22; Bias as % of consensus mean = 4.314 %; Laboratory CV = 2.632 %; Consensus CV = 3.137 %; Coefficient of variation ratio (CVR) = 0.839. ## Key terms - **Consensus (peer-group) mean**: The average result of all laboratories using the same method and instrument. - **SDI**: The distance between the lab mean and the consensus mean in consensus standard deviations. - **CVR**: The laboratory's coefficient of variation divided by the peer group's. ## FAQ **What is a good SDI?** 0 is ideal. |SDI| ≤ 1 shows good agreement with peers, up to 1.25 is acceptable, 1.25–2 warrants investigation and 2 or more is unacceptable. **How do I calculate SDI?** Subtract the consensus mean from your lab mean and divide by the consensus SD. A lab mean of 9 against a group mean of 8 with SD 2 gives SDI = 0.5. **What does a negative SDI mean?** Your laboratory reads lower than the peer group. The size of the problem is the absolute value. **What is the difference between SDI and CVR?** SDI measures bias (accuracy) relative to peers; CVR compares your imprecision with theirs. CVR below 1 means you are more precise than the group. **What CV ratio is acceptable?** A CVR up to about 1.0 is good, 1.0–1.5 acceptable, 1.5–2.0 needs investigation and above 2.0 requires action. **Is SDI the same as a z-score?** Yes in form – it is a z-score of your lab's mean within the distribution of peer means. Proficiency-testing schemes often report z-scores with the same ±2 and ±3 limits. **How often should SDI be reviewed?** Monthly for each analyte and control level is common. Look for trends: three consecutive months moving the same way can signal a shift before limits are breached. ## Sources - [CLSI C24 – Statistical quality control for quantitative measurement procedures](https://clsi.org/) - [NIST/SEMATECH e-Handbook of Statistical Methods](https://www.itl.nist.gov/div898/handbook/) - [OpenStax – Introductory Statistics 2e](https://openstax.org/details/books/introductory-statistics-2e)