Logo

Laboratory Reference Interval (LabRI) (v5.5.0)

Tool for Estimating and Verifying Reference Intervals

  • Configuration
  • Report

Name of the responsible specialist

NOTE 1: Provide the name of the person responsible for the process or analysis performed.


Define the Dataset

NOTE 2: Select the data file to be analyzed. The file must be in .csv, .xls, or .xlsx format.

NOTE 3: Select the column name from the dropdown list containing the data to be analyzed.

NOTE 4: Select 'Yes' if the measurand name is the same as the column name in the dataset, or 'No' otherwise. If 'No' is selected, an additional field will appear to specify a custom measurand name, which will be used in the HTML report generated by the tool.

NOTA 5: Provide the data source, specifying where the dataset was obtained from.


Set the Type of Reference Interval and Confidence Level

NOTE 6: Select the desired type of reference interval. The default setting is 'Double-sided,' which calculates both the lower and upper reference limits. 'Right-sided' considers only the upper reference limit.

NOTE 7: Select the desired confidence level for estimating the reference limits. The available options are 90%, 95%, and 99%. If no selection is made, the tool will apply a default confidence level of 90%


Information for Study Traceability

NOTE 8: Describe the measurement procedure and the analytical method used. This method is the practical process that applies the analytical principle to obtain the result.

NOTE 9: Describe the unit of measurement for the measurand (e.g., mg/dL, g/dL, mmol/L etc.).

NOTE 10: Describe the sample type (e.g., whole blood, serum, plasma, 24-hour urine, random urine etc.).

NOTE 11: Describe the age range concisely to ensure readability in the HTML report. Suggested formats for reporting the age range include simple, direct expressions like '7 to 16 years', 'under 2 years', '7 - 16 years', '< 2 years' etc.

NOTE 12: Describe the sex concisely to ensure readability in the HTML report. Suggested formats include simple, direct expressions like 'Male', 'Female', 'M,' 'F', 'Male and Female', 'M and F' etc.

NOTE 13: Defining exclusion criteria is essential to ensure that reference intervals accurately represent the healthy population. In direct methods, these criteria enable the careful selection of reference individuals by eliminating factors that could introduce unwanted variability, such as recent illnesses or medication use.

In the indirect sampling method, filtering the dataset is essential. Exclusion criteria should be applied, such as:

  • Exclude potentially pathological results: Eliminate values that may skew the reference distribution, ensuring it represents healthy individuals. To increase dataset selectivity, abnormal results from co-requested laboratory tests can be used to help identify subclinical conditions. Additionally, methods like Latent Abnormal Values Exclusion (LAVE) or similar approaches allow for iterative exclusion of latent abnormal values, refining the dataset more precisely.
  • Remove data from specific departments (e.g., oncology, Intensive Care Unit, Home Care etc): These departments typically handle patients in critical conditions that can significantly alter results.
  • Limit the frequency of results per patient within the study period: Including, for example, only the first result for each patient prevents data overload from the same individual, which could bias the sample.

Provide the Comparative Reference Interval (if applicable)

NOTE 14: To perform reference interval verification, a comparative reference interval must be provided. This comparative reference interval can be selected from various sources, such as:

  • Reagent Kit Package Inserts: Specified reference ranges for laboratory tests based on manufacturer's validation studies.
  • Scientific publications and guidelines from specialized organizations: Intervals derived from multicenter studies and reviews that reflect widely accepted practices in the field.
  • Academic Books (Textbooks): Consolidated and widely accepted reference intervals based on extensive scientific literature and available in academic books.
  • Local or multicentric studies: Results obtained for a specific population or harmonized across multiple regions.
  • Indirect methods: involve previous studies that estimated reference intervals using patient data stored in laboratory information systems (LIS), applying algorithms to identify and exclude potentially pathological values.

NOTE 15: If a Comparative Reference Interval is not specified, the reference interval estimated by the LabRI method itself will be used as the comparative reference. This is because the verification module algorithms of the LabRI method require a comparative reference to perform the verification.


Define the subsampling approach used in the study (if applicable)

NOTE 16: If no maximum subsample size is provided, a default value of 10,000 observations will be applied. In the LabRI method study, this threshold was used to balance computational efficiency with the retention of a dataset size considered adequately representative of the population served by the laboratory.


Main Commands


Report an Issue

Did you encounter any issues with the LabRI Tool? We appreciate your feedback! To report a problem, click the link below to access our issues system on GitHub: Report an Issue in GitHub GitHub

NOTE 17: To submit an issue report, you need to have a GitHub account and be logged in.

Developed by Alan Carvalho Dias | Powered by LabR Group | License: GPL-3.0 | Code available at: GitHub Lab R Group Website