The National Institute of Neurological Disorders and Stroke (NINDS), National Institutes of Health (NIH) expects the clinical research it funds to meet the highest standards of scientific rigor, yet appreciates the burden that extensive data collection puts on investigators and study subjects. Furthermore, the Institute leadership recognizes that investigators independently identify data elements and forms for each study, many of which could be common across studies. As part of its effort to facilitate research of the highest quality, while streamlining clinical data collection in neurological studies, NINDS initiated the Common Data Elements (CDEs) project.

The purposes of the CDEs project are to provide a parsimonious data collection model by placing the focus on the most important and useful data, and to consequently limit the collection of data that are merely ancillary to the primary study outcomes. We recognize that the most important variables collected by a study will probably not be CDEs. However, the CDEs project does aim to provide a foundation of variables that are common across studies in order to increase the efficiency and consistency of data collection and to facilitate data sharing. The Institute still encourages creativity and uniqueness by allowing investigators to independently identify and add their own critical variables.

 

CDEs OVERVIEW

The CDEs have been identified through review of the documentation of numerous studies funded by NINDS, review of the literature and regulatory requirements, and review of other Institute’s common data efforts. Other data standards such as those of the Clinical Data Interchange Standards Consortium (CDISC), the Clinical Data Acquisition Standards Harmonization (CDASH) Initiative, ClinicalTrials.gov, the NINDS Genetics Repository, and the NIH Roadmap efforts have also been followed to ensure that the NINDS CDEs are comprehensive and as compatible as possible with those standards.

The CDEs project’s goal of streamlined study data are supported by products and materials that focus on:

o        Data collection – tools to assist study staff as they collect and manage data during a study;

o        Data management – data labels, definitions, formats, etc. that can be used to create study data files, whether implemented during a study or after study completion.

 

CDEs: STREAMLINING DATA COLLECTION

The data collection materials include:

§        Core and Administrative Form Templates – Provide form layouts in Word that can be customized to meet study needs;

§          CDE Reports – Display corresponding data dictionary definitions for all elements found on the form templates. The components (i.e., columns) of the CDEs Reports are:

o        Common Data Group (CDG) – Logical collection of one or more related CDEs (see the CDEs Key Terms for a more complete definition);

o        CDE Name – Variable name for the element and also its XML tag;

o        CDE Definition – Description of the element;

o        Code List – Possible response choices for the element, where applicable;

o        Recommended Data Type – Attribute of the element that describes the kind of data being collected (e.g., date/time, numeric or string);

o        Length – Space allocated for collecting the element;

o        Format – Appearance of the way the element should be collected (e.g., MM/DD/YYYY for a date);

o        Version Number – Record of the release wave;

o        Version Date – Date the current version was uploaded to the Web site.

§          Annotated Forms – Visually show the linkage between the form templates and their associated CDEs; these PDF versions of the form templates have the CDE Names superimposed next to each item;

§         Form Instructions – Provide a Manual of Procedures (MOP) describing how specific data items are collected and suggest related references;

§         Data Range and Logic Checks – Provide range and logic checks for many of the common forms.  These proposed checks would be programmed into an electronic data capture system to help ensure the quality of the collected data.

Development of the CDE data collection materials has identified two types of clinical study data:

o        Core elements – the “core” CDEs and the complementary collection of form templates are not designed to identify the data that address the study hypotheses, but were developed to represent the data necessary to conduct a credible study. 

o        Administrative elements – the “administrative” CDEs and the form templates are designed to facilitate study data management through standard study tracking logs, checklists and other data not related to clinical study outcomes.

The current data collection materials include CDEs organized on the following “core” form templates:

§         Inclusion/Exclusion Criteria – Outlines the format for these forms so that study staff can easily customize them for study specific criteria;

§         Subject Characteristics – Collects study population characteristics for baseline comparisons and to aid in subgroup analysis;

§         Medical History- Review of Systems at Baseline – Identifies co-morbidities across body systems;

§         Surgical History at Baseline – Identifies surgical or otherwise invasive procedures clinically relevant to the study;

§         Behavioral History at Baseline – Collects a history of the subject’s alcohol and tobacco use;

§         Family History at Baseline – Provides a record of known health conditions among the subject’s blood relatives;

§         Pre-existing Medications – Records an individual's medication regime for a protocol specified period prior to study enrollment;

§         Vital Signs, Weight and Height – Collects the subject’s vital signs as well his/her weight and height;

§         Rankin Scale – Assesses the disability of a study subject;

§         Trail Making Score – Assesses the cognitive functions of an individual;

§         Intervention Dosing – Documents the administration of the study intervention;

§         Intervention Compliance – Documents the dispensation of study intervention;

§         Adverse Events – Documents safety issues;  

§         Concomitant Medications – Identifies potential safety issues and avoids problematic drug interactions;

§         Death Report – Captures details of the subject’s decease, should that event occur while the subject is enrolled in the study;

§         Laboratory Tracking – Tracks the transmission of sample specimens to the laboratory and can be used to reconcile laboratory data received electronically from the laboratory;

§         Laboratory Tests – Records the results from commonly collected laboratory tests;

§         Subject Status – Documents subject status within the context of the study.

The data collection materials also include “administrative” form templates that assist with study administration and the tracking of clinical data. These forms do not have data dictionaries associated with them because they are primarily used for data/study management purposes and do not collect actual clinical data. Form templates designated as “administrative” include the following:

§         Visit Schedule –Provides a schematic of the study protocol and details what assessments occur at each subject encounter;

§         Visit Checklist – Enumerates the assessments that are to be completed at a study visit;

§         Screening Log – Describes and documents all individuals screened for study inclusion in a log format;

§         Subject Contact Information – Captures contact information necessary to keep in-touch with a study subject between visits and to help locate an individual that does not return for follow-up. Information is collected not only on the subject but also on alternative points of contact. The information should be kept in a locked location or secure local computer file since it contains personal identifiers;

§         Serious Adverse Events – Collects additional information that describes a serious adverse event (SAE) and its outcome. The SAE report may be sent to the IRB, FDA, or data monitoring committee depending upon the SAE reporting requirements for the study. The SAE data is sometimes kept in a tracking system separate from the study data system since it contains text and information not generally needed for the study (e.g., lot number of drug);

§         Protocol Exceptions – Records instances during the study when the protocol was not followed;

§         Header Information – Specifies the study, site, visit, and subject identification number (ID) so that study forms are attributed to the correct subject, site, and visit (Note: The header elements are already incorporated on all template forms).

The “core” data elements are generally applicable across neurological diseases. Future versions of the CDEs materials will include suggested data elements, definitions, and form templates for studies in specific neurological diseases, such as stroke, epilepsy, Parkinson’s disease, multiple sclerosis, etc. The CDEs materials will expand as elements are vetted in disease-specific therapeutic areas, as shown in Figure 1.

 

Figure 1. CDEs Data Collection - Form Development

[CDE Diagram Text Description]

 

CDEs WEB SITE (http://www.nindscommondataelements.org/)

The NINDS CDEs Web site serves as a repository and dissemination tool for the CDEs project.  All of the data collection materials described in the previous sections of this document are available for Investigators to utilize for trials being planned, and ongoing research. The Web site provides NINDS clinical researchers access to the CDEs, their definitions and instructions, and sample data collection forms. 

 Access to these materials will assist:

§         All researchers prepare their grant application for a clinical study by providing them with a “core” set of data elements, definitions and sample forms;

§         Researchers and coordinators new to clinical studies by identifying the necessary “core” data elements, definitions and suggested forms.  The planned “shopping cart” feature will allow study staff to choose individual elements for their study and design unique data collection forms; and

§         Experienced researchers by giving them the tools to implement the NINDS CDEs and definitions. The availability of range and logic checks for the CDEs will also help investigators improve data quality across studies.

Access to information on the Web site is currently available through two main portals - Common Data Elements and Common Data Forms. Programmers and study statisticians are more likely to use the Common Data Elements portal while Investigators and study coordinators are more likely to access the Common Data Forms portal. It is recommended that users of the CDEs Web site first view the online tutorial,   (http://www.nindscommondataelements.org/Tutorial/CDE Website Tutorial_demo3.htm), as the tutorial walks users through the scope of the CDEs Web site.

In the near future, the CDEs Web site will also have a “shopping cart” feature available to investigators and study staff.  With this tool, users can pick and choose the most desirable CDEs for their study and design and save custom data collection materials through the Web site. 

 

CDEs APPLICATION - Applying Common Data Elements to a Clinical Research Study

Once a study protocol is developed, the data required for the study can be checked against the Common Data Forms and CDEs Data Dictionary to assure that:

§         Forms necessary for a study are included (e.g., inclusion criteria, study completion, adverse events); and

§         Items collected in the study use the NINDS CDEs definitions and data structure wherever possible.

Steps to Implementation

Once a study protocol is developed, the Investigator and site data manager or assigned study staff should become more familiar with the NINDS CDEs. It is recommended that this process begin as soon as possible, such as during the pre-grant award meetings with the Clinical Trials Cluster staff.

Site staff should review the protocol and study forms (if completed already) to determine what data elements match the NINDS CDEs. This process can take place using one of three approaches, depending on the study status:

1. Early Stages of Study Form Development – After a protocol has been finalized, the study staff should work with the investigator to develop the study data collection forms. This Web site provides template study forms that can be downloaded and customized for a particular study. In addition, the Web site provides a data dictionary and MOP for each form. Most studies can, at a minimum, use the “core” set of forms for data collection.    
2. Study Forms Developed – If the protocol, forms and data management system have already been developed for a study, but the study has not begun enrolling, the recommended “core” data collection forms and data elements should be reviewed and incorporated wherever possible.
3. Study is Ongoing – For studies that are currently enrolling subjects, study staff should ensure that that the “core” data elements and forms are included in the data collection process. For data sharing purposes, it may eventually be recommended that the study data elements be “cross walked” or mapped to the NINDS CDEs  before sending a final data set to NINDS .

 

CDEs: FACILITATING DATA SHARING

The data sharing aspects of the CDEs project are still under development.  They concentrate on establishing a metadata framework that will promote data independence from the any particular data collection environment.  Once fully developed, the metadata CDEs can be used by investigators to share their study information, regardless of the system utilized to capture and store their study data.  Ultimately, these standards will facilitate the aggregation of data in a common repository, (e.g. an NINDS Data Warehouse). 

As part of the data sharing activities, data from completed NINDS clinical trials are being utilized to continually test and improve the data sharing framework.  In a data mapping process, data from a completed “legacy” NINDS trial are extracted and transformed to the metadata framework CDEs.  Each data mapping experience provides an iterative feedback mechanism for both the data collection and data sharing activities of the project.  

The current CDE materials that will facilitate data sharing include:

§         Outline of Protocol Elements - Delineates and defines the key elements of a clinical study protocol.  The protocol elements overlap with ClinicalTrials.gov’s data elements and are important for cataloguing and sharing information about NINDS-funded clinical studies in a consistent manner.  It will be just as important to eventually include protocol elements in a common repository as it will be to store the clinical data collected by the protocols.

§         Universal clinical data element descriptions and definitions;

§         CDEs data dictionary with Extensible Markup Language (XML) tags;

§         Clinical data algorithm for uniquely categorizing study variables in the NINDS Data Warehouse.

The following documents provide further information on NIH’s policy for Data Sharing:

§         National Institutes of Health Statement on Sharing Research Data

§         National Institutes of Health Statement on Sharing Research Data National Institutes of Health Grants Policy Statement -- See NIH Grants Policy, Part II Subpart A, Availability of Research Results: Publications, Intellectual Property Rights, and Sharing Biomedical Research Resources for information regarding a detailed description for NIH grantees to share resources, pages 1192-121

 

SUMMARY

The NINDS CDEs project will continue to evolve and expand to meet the needs of clinical investigators conducting studies in neurology. The materials presented represent a refined continuation to be utilized in data collection and storage, as well as in dataset creation.  NINDS hopes that all studies will eventually be able map their data using the standard CDEs to create a common NINDS Data Warehouse.

 CHANGES SINCE VERSION 1.0

There have been significant changes to the content and structure of the CDEs material since the Version 1.0 release in April of 2006.  We have included the minutes from the original working group meeting in the front flap of this binder for your reference.

New Forms:

§ Surgical History at Baseline

§ Intervention Dosing

§ Intervention Compliance

§ Death Report

Modified Forms:

§         Inclusion and Exclusion Criteria

§         Subject Characteristics (formerly Demographics)

§         Medical History - Review of Systems at Baseline (formerly Medical History)

§         Behavioral History at Baseline (Formerly Medical History - Option 1)

§         Family History at Baseline (formerly Medical History - Option 2)

§ Pre-existing Medications (formerly Prior Medications)

§         Vital Signs, Weight and Height (formerly Vital Signs)

§         Rankin Scale (Modified)

§         Trail Making Score

§         Laboratory Tracking

§         Laboratory Tests

§         Adverse Events

§         Concomitant Medications

§         Subject Status (formerly Study Completion)

§         Visit Schedule

§         Visit Checklist

§         Screening Log

§         Subject Contact Information (formerly Participant Contact Information)

§         Serious Adverse Events

§         Protocol Exceptions (formerly Protocol Deviations)

§         Header Information

Removed Forms:

§        Demographics Option 1 – Psychological Data

§        Demographics Option 2 – Psychosocial Data

§        Treatment Compliance (replaced with Intervention Compliance & Intervention Dosing)

§        General Physical Exam

The corresponding data dictionaries and MOPS have been modified in accordance with the form changes. 

Please note that some forms included in the Version 1.0 release are being reviewed as part of our disease-specific element initiative, and have not been deleted from the CDEs. 

Also included in the Version 2.0 materials are range and logic checks for each of the forms.  These checks are designed to be incorporated into the database for a study using the CDEs.  This is another tool NINDS will provide to the investigators to reduce study start-up time.