Compare a panel of analytes

A comparison set groups ordinary method comparison studies by test and method pair. Import one wide panel file or the separate files from each method, review their specimen matches, then calculate and bind the saved reports.

Use the same non-patient specimen ID for a specimen measured by each method. Keep one result per specimen, test and method. Leave a missing measurement blank; a zero is a measurement and stays zero. Measurements for a test must use the same unit across methods.

A wide file has a specimen column and one result column for each test and method, for example Glucose | Reference and Glucose | Candidate. Separate method files have a specimen column and one column per test. File names become the proposed method names, which you can edit during review.

Use CSV, TSV or XLSX. The wide panel example and separate Reference and Candidate examples contain example glucose, sodium and potassium measurements. One missing candidate potassium measurement gives that test a different specimen count. These files are examples for learning the import steps. Use the method comparison guide to plan a study.

1. Open Panel comparisons and choose Import panel. Name the set. Its studies go into General studies unless you choose a Project, or type a Project name in a workspace without projects.
2. Select the Analyzer profiles to reuse saved test-code mappings, units, acceptance limits and decision levels. Choose the Panel files, then Review columns.
3. In Review panel columns, select the specimen ID column and the result columns. Confirm each analyte, method and unit. Matching saved import formats also apply their specimen-ID cleanup and ignored columns.
4. Choose the Reference method. The set compares each other method with it. Additional method pairs lets you include other comparisons.
5. Select Include profile reference-interval endpoints as labeled prediction points when those fitted values will help the review. These points describe the fitted relationship. Acceptance comes from the study's limits.
6. Choose Review matched specimens. Open an unmatched count to see which specimen measurements are missing. Only complete pairs are included in each study. Check the copied limits, then select Create n studies.

The retained source tables contain the mapped columns and assigned study codes. A patient-linked specimen column defaults to Replace with study codes; choose These IDs are already coded only for de-identified laboratory IDs. Download the one-time code key after creation and keep it in your laboratory; LabRuns retains no copy. The paired results identify the source import and source row of each measurement.

Choose Calculate set to save each comparison result. The set shows each test's independent specimen count, primary regression, specimen agreement and result. Expand Decision-level bias and predictions to review fitted bias, confidence intervals and labeled reference endpoints. Extrapolated means the point lies outside that comparison's studied reference range.

A study imported without a limit for each specimen's difference reads Not evaluated, and the summary line counts the analytes that need an allowable difference. Choose Set limits for n analytes to see each one with its CLIA / CAP allowable difference, such as Albumin, serum: ±8%, checked. Uncheck any analyte that takes your laboratory's own limit. Add limits and recalculate saves each checked limit in its study's acceptance limits and calculates the set again. An analyte with no usable table limit, such as one graded by a peer-group SD, shows the reason and Open Set up, where you enter your own limit. Any other unavailable decision shows its saved reason beside the result.

Open a test to review or correct its ordinary study. Recalculate after saving changes so the set shows the current saved result. Choose Create binder to review the selected calculations and prepare the set summary with the full reports and evidence archive. Paper approval remains optional.