Part 1: Individual Data Entry

Step 1: Set Parameters

• Click the “Individual” tab
• Cutoff Selection: Choose 1, 1.5, and/or 2 SD below the normative mean to define cognitive impairment.
  • Tip: Remove any others that you are not interested in using.
• Default Score Scale: Choose from:
  • Standard score, scaled score, T-score, z-score, or percentile.
  • Note: This is only the default — you can change it per test.

Step 2: Enter Cognitive Test Data

Cognitive domains include:

• Language
  • Subdomains: Naming, Fluency
  • Example Tests: Boston Naming Test, Category Fluency, etc.
  • Enter the test name, score, and scale on which the test score is provided for this and all other domains.
  • Note: The test names provided are commonly used measures. If you do not use these measures, you may leave them blank or type over them to include the names of the tests you have available.
  • Note: The specific test and test version used does not matter. The calculator will work for any test that you enter into the Measure field.
• Memory
  • Subdomains: Word List Recall, Story Recall, Word Pair Recall, Design Recall
  • Example Tests: CVLT, Logical Memory, BVMT, etc.
  • Tip: Make sure the score scale matches the data (e.g., T-score, z-score, scale score).
• Executive Function
  • Subdomains: Set-Shifting, Problem-Solving, Response Inhibition
  • Example Tests: Trail Making Test Part B, WCST
  • Tip: When possible, use scores from different measures within the same domain (e.g., Trail Making Test – Part B, WCST Perseverative Errors) rather than two scores from a single measure (e.g., WCST perseverative errors and conceptual level responses).
• Attention/Processing Speed
  • Subdomains: Attention, Processing Speed
  • Example Tests: Digit Span, Trail Making Test Part A, Coding
  • Tip: Make sure the score scale matches your source (e.g., scaled scores from WAIS subtests, T-scores from Trail Making Test).
• Visuospatial
  • Subdomains: Visuoconstruction, Visuoperception
  • Example Tests: Block Design, Judgment of Line Orientation
  • Tip: Adjust the scale type and include the relevant scores.
• MOOD AND BEHAVIOR (Optional)
  • Subdomains: Depression, Anxiety, Behavior
  • Example Tests: BDI (Depression), BAI (Anxiety), CBCL (Behavior)
  • Note: The mood and behavior inventories are not used in phenotype classification, but may be included as potential modifiers.

General Tip: It is recommended that users include tests of different types within a domain whenever possible (e.g., naming and fluency rather than two fluency tasks in the Language domain).

General Note: A small warning button may appear to the right of a test score to indicate that a score may have been entered incorrectly or the incorrect scale type selected. This is simply a prompt to check for errors; the calculator will still work if this warning button is present.

Step 3: Submit

• Scroll back up and review all entries.
• Make sure at least two tests per domain are included in at least four domains.
• Click "Submit Entries".
• The IC Code system will display:
  • Cutoff(s) used
  • Overall phenotype (e.g., Generalized, Bi-Domain Impairment)
  • Domains impaired
  • Any selected modifiers
  • Tests included in the phenotype calculation
• You may download results or reset to enter a new patient/participant.

Individual Data Entry – Workflow Summary

Part 2: Group Data Entry

Step 1: Set Parameters

• Select the “Group” tab
• Cutoff Selection: Choose 1, 1.5, and/or 2 SD below the normative mean to define cognitive impairment.
  • Tip: Remove any others that you are not interested in using.
• Default Score Scale: Choose from:
  • Standard score, T-score, z-score, scaled score, or percentile.
  • Note: This is only the default — you can change it per test.

Step 2: Create Group Template

• Under the Group tab, click Create Template.
• As described above under the Individual Data instructions, for each cognitive domain:
• Select or enter all tests in your battery.
• Set the correct score scale (T-score, z-score, scaled score).
• Place a check in the “Include” box for all tests you want included in the cognitive phenotype generation.
• If your test is not listed, type it in.
• Include at least two tests per domain in at least four domains.
• Enter Filters (optional).
• Note: You can use any of the provided filters or enter other filters of your own (simply type over the existing filter names). Be sure to check the “Include” box for any filters you want included in your dataset.
• Once complete, click Download Group Table Template.

General Tip: It is recommended that users include tests of different types within a domain whenever possible (e.g., naming and fluency tasks rather than two fluency tasks in the Language domain).

Step 4: Upload and Generate Results

• Return to IC Code Portal → Group tab.
• Click “Upload Data and Generate Results.”
• Use Browse to select your completed template file.
• Results will include:
• Display results presented as pie chart(s) showing IC Code phenotype distribution(s) — if you hover over the single domain or bi-domain slices, you can see a breakdown of impaired domains.
• Option to visualize pie charts by desired filters (e.g., sex, side of seizures).
• Option to download the results in a spreadsheet format.

BEST PRACTICES

Group Data Entry – Workflow Summary

The International Classification of Cognitive Disorders in Epilepsy (IC-CoDE)

The IC-CoDE represents a consensus-based, empirically-driven approach to diagnosing cognitive disorders in adults with epilepsy. It was developed in 2020 through a memorandum of understanding (MOU) between the International League Against Epilepsy (ILAE) Neuropsychology Task force and the International Neuropsychological Society (INS). The main goal of the IC-CoDE is to accelerate global research in the neuropsychology of epilepsy by providing an internationally-applicable framework for cognitive diagnostics in epilepsy with clear operational criteria and established impairment cut-offs. The IC-CoDE was developed as a guide for harmonizing multi-site cognitive research in epilepsy. It has not been validated as a diagnostic tool for individual patients in clinical settings.

The IC-CoDE framework was originally tested in a large, multicenter cohort of 2,485 adults with temporal lobe epilepsy and subsequently in a multicenter cohort of 455 with frontal lobe epilepsy. The IC-CoDE is now currently being tested in youth with epilepsy and in adults at international sites. IC-CoDE has not yet been applied to other epilepsy syndromes (e.g., juvenile myoclonic epilepsy, absence seizures). IC-CoDE has also been modified and applied to several other disorders outside of epilepsy (e.g., multiple sclerosis, COVID-19). A list of published studies that have used IC-CoDE is provided under the Publications tab above.

Initial Considerations for IC-CoDE Users

The goal of the IC-CoDE is to apply the cognitive model (Figure 1) using the operational definitions provided in order to arrive at a diagnostic cognitive phenotype (Figure 2). Step-by-step instructions for using the IC-CoDE calculator are provided below. For more information refer to Norman et al., 2021, Epilepsia, McDonald et al., 2023, Neuropsychology, and Hermann et al., 2024, Epileptic Disorders.

Funding

Funding for development of the IC-CoDE Portal was provided by an American Epilepsy Society Infrastructure Award (Award ID 1153665).

Privacy policy

We value your privacy. The full policy is available here .

Frequently Asked Questions

How to cite the IC-CODE Portal

When publishing manuscripts that make use of the IC-CoDE portal to generate phenotype data, we ask that you please include an acknowledgement. We recommend the following boilerplate language: IC-CoDE phenotype data were generated using the IC-CoDE Portal (IC-CoDE-Portal.ccf.org), which was funded by an American Epilepsy Society Infrastructure Grant (Award ID 1153665).

Can the IC-CoDE calculator be used for clinical purposes?

At present, IC-CoDE is recommended only for use in the context of research. While the longer-term goal is to make IC-CoDE applicable for clinical use, more research is needed before IC-CoDE can be used in this manner. For users interested in early clinical application, IC-CoDE should only be used in the context of contemporary clinical standards and practices.

What normative data should be used to generate IC-CoDE phenotypes?

The IC-CoDE does not require use of specific normative data. We encourage each center to carefully select and apply normative data based on best practices (e.g., examination of psychometric properties, clinical characteristics that align with the patient sample, etc.). Research is underway to better understand several important issues, i.e., how varied selection of normative data across centers may impact IC-CoDE phenotypes and how the number of tests per domain as well as variable base rate sensitivities of specific tests may impact diagnostic outcomes.

Can the IC-CoDE calculator be used if patients in your sample completed a different version of the same test (e.g., CVLT, CVLT-2)?

The IC-CoDE is not a test-specific taxonomy; thus, different versions of the same test can be used interchangeably. However, research is ongoing to better understand the influence different tests and test versions may have on IC-CoDE phenotypes.

Can the IC-CoDE calculator be used in children with epilepsy?

Yes! The IC-CoDE calculator can easily be applied to children by simply entering data for pediatric measures. The default measures provided in the calculator are only examples. The user can replace these default measures with any measure they would like. For example, under naming, they could replace the Boston Naming Test with the Expressive One Word Vocabulary Test if they have that measure of naming for a child they assessed. Similarly, under Attention, they could replace the WAIS Digit Span score with the WISC or WPPSI Digit Span score or any other measure of attention in their pediatric battery.
The IC-CoDE has recently been applied to children with new and recent onset idiopathic epilepsies as well as to those with pharmacoresistant focal epilepsies. These findings were recently published, and the references can be found under the “Publications” tab.

Can the IC-CoDE calculator be applied to patient populations outside of epilepsy?

The IC-CoDE taxonomy has been applied to several populations outside of epilepsy, including adults with multiple sclerosis and adults with COVID-19. Findings from these studies have been published in scientific journals, and references are provided under the “Publications” tab above.

How many cognitive domains are required to generate IC-CoDE phenotypes?

Ideally, at least two cognitive measures from all 5 cognitive domains (i.e., language, memory, executive, visuospatial, attention/speed) should be used when generating IC-CoDE phenotypes. However, in instances where there were not enough measures administered within each domain, it may be possible to generate an IC-CoDE phenotype with only 4 cognitive domains. However, careful consideration should be given to the patient population of interest to ensure that an “essential” cognitive domain is not omitted (i.e., domains known to have a high base rate of impairment in that population). For example, memory and language would be considered essential domains in temporal lobe epilepsy and visuospatial would be considered an essential domain for parietal lobe epilepsy. We do not recommend generating IC-CoDE phenotypes in any patient with fewer than 4 cognitive domains.

Can the IC-CoDE calculator be applied to patient populations outside of the United States?

Yes! The IC-CoDE calculator can be applied to patient populations in other regions and to patients who were assessed in other languages. The default measures provided in the calculator are only examples. The user can replace these default measures with any measure they would like. For example, under naming, they could replace the Boston Naming Test with whatever measure they used to assess naming in whatever region or language the patient was evaluated. Similarly, under Attention, they could replace the WAIS Digit Span score with whatever measure they used to assess attention in their neuropsychological battery.
To date, the IC-CoDE has been tested with temporal lobe epilepsy patients in Mumbai, India where there is considerable language and cultural diversity. The findings were published in Shah et al., 2024. Epilepsia. We are currently working with other research groups in South Africa and Japan, who are validating the IC-CoDE in their local samples.

Can the IC-CoDE calculator be applied to bilingual/multilingual patient populations?

Yes! The IC-CoDE taxonomy was validated in a multilingual sample from Mumbai, India that included individuals who were bilingual and multilingual. Notably, we recommend that researchers interpret the calculator findings, particularly verbally-mediated cognitive domains in the context of the procedures used for testing (e.g., language of testing, tests and norms used) and the degree of bilingualism/multilingualism of the patient or sample.

Can the IC-CoDE calculator be applied to non-English monolingual speakers of other languages?

The IC-CoDE was validated in a sample of Spanish-speaking patients with temporal lobe epilepsy. The findings were published in Epilepsia, Reyes et al., 2023. Although researchers can use any test, as the IC-CoDE is not a test-specific taxonomy, we recommend that the battery used for non-English monolingual speakers follows the same or similar cognitive domains outlined in the IC-CoDE Taxonomy. We recognize that there is a dearth of neuropsychological tests available in other languages, and we hope that the flexibility of the taxonomy allows for validation of the IC-CoDE across other languages and cultural groups.

How do you use the IC-CoDE modifiers?

We have examined the relationship between depression and anxiety and IC-CoDE phenotypes, and results were published in Epilepsia, Bingaman et al., 2023. While we have included the opportunity to code mood in your dataset, the influence of mood will not automatically be explored in this IC-CoDE version. Thus, interpretation and further exploration is up to individual investigators.

Is IC-CoDE technical assistance available?

Technical assistance is not currently available, but detailed instructions are provided under the “Instructions for Use” tab above to aid the user in navigating the IC-CoDE portal and calculator.