Clinical research data management tasks and … research data management tasks and definitions...
Transcript of Clinical research data management tasks and … research data management tasks and definitions...
Clinical research data management tasks and definitionsMeredith Nahm, MS,1,2 Constance Johnson, PhD,3 Anita Walden,1,2 Todd Johnson, PhD,2 Jiajie Zhang, PhD2
1Duke Translational Medicine Institute, Durham, NC; 2School of Health Information Sciences, Houston, TX;3Duke University School of Nursing, Durham, NC
Abstract
The tasks associated with clinical research data management have not been systematically identified and defined. We surveyed existing sources to create a draft list and vetted the tasks, their relationships, and their definitions through focus groups, followed by peer debriefing, followed by 2 rounds of broad distribution to the membership of the Society for Clinical Data Management (SCDM). A total of 75 tasks were identified and defined. We present the current version of the tasks.
Introduction
The first formal list of clinical trial data management tasks was published in Data Basics in 1999.1 This task list helped scope the Good Clinical Data Management Practices (GCDMP) document, and provided a foundation for the certification exam. Periodically revisiting foundations like the growing evidence base and the task list is important to our field. We started work in the fall of 2007 to re-survey those tasks associated just with the collection and management of data in clinical research. Limited to the tasks that directly operate on data, this work is considerably narrower than the original task list; for example, project, vendor, and people management tasks are not included. However, this work went a step deeper in searching for authoritative definitions for each task and defining those tasks for which no definition could be found. Additionally, this work went beyond tasks common to industry-sponsored clinical trials to include the broader field of clinical research as defined by the National Institutes of Health. We posit that although we use different methods and different terms to describe them, at some not too abstract level, we all do similar things and can benefit from the richness of each other’s knowledge. Further, data collection and management across the different areas of clinical research share the same literature base. A primary goal of this work is to provide the tasks and definitions to the Medical Subject Headings (MeSH) curation team, to improve the indexing of our literature base and ultimately decrease the fragmentation that exists.
References
1. Bornstein S. Clinical data management task list. Data Basics 1999;5:8-10.
2. Society for Clinical Data Management Certification Handbook. Available at: www.scdm.org. Accessed July 20, 2008.
3. Deitzer JR, Payne PRO, Starren JB. Coverage of clinical trial tasks in existing ontologies. AMIA Proceedings 2006:903.
Funding
The project described was supported by grant number UL1RR024128 to Duke University and grant number UL1RR024148 to University of Texas Health Science Center Houston, from the National Center for Research Resources (NCRR), a compo-nent of the National Institutes of Health (NIH), and its contents are solely the re-sponsibility of the authors and do not necessarily represent the official view of NCRR or NIH.
Background
The initial SCDM task list consisted of 67 tasks grouped into 9 categories but contained no definitions. Later, in 2004, SCDM released their National Certified Clinical Data Man-ager (CCDM) certification exam and associ-ated core competencies. The current core competency set for the exam consists of 112 tasks grouped into 26 categories called core competencies.2
A different approach to identification of clinical research related tasks was taken by Deitzer et al, who abstracted tasks from 20 cancer clinical trial protocols.3 This list, however, is significantly broader than operations performed on clinical trial data.
Methods
Focus groups were held at the 2008 SCDM fall conference and followed by peer de-briefing through the SCDM newsletter. The full task set with definitions was provided to attendees at the 2009 fall conference and later released to the membership. This work was approved by the Duke University insti-tutional review board at the corresponding author’s institution.
Results
A total of 75 tasks were identified. Authori-tative definitions were found for 55 of the tasks. The remainder were defined by the authors.
Conclusion
Task commonalities exist across industry and academically oriented studies. Thus, it is likely that a common vocabulary can be achieved. Maintaining such a task list is important to a discipline because it scopes and defines competencies required for practice in the discipline, and provides a link to the evidence base on which practice recommendations should be made.
Manage Clinical Research Data
Identify data
005
Develop database
244–250
Process data
Design form
representa-tion
012–024
Collect data
Track data
Define data
149, 164, 168–170
Store data, metadata
222–226
Query database
097
Define document
data handling process
029–034, 050–051, 148, 155, 165, 173–177
Support system users
Report status
047–049, 150
Secure data
Measure data
quality
188–208
Source document verify data
Electronic data
acquisition
Transform data
166–167
Data surveillance
Impute data
Integrate data
138–147
Export data
Clean data
Reconcile data
Incremental batch
import
Cumulative batch import
Messaging
Specify measure-
ment method
Annotate form
Specify metadata
Measure/ observe
Abstract data
Record data
Record data
changes
Record provenance
Record state
043–046, 082–091
Comparison Error rate estimation
Choose applicable standards
116, 133
Specify data input structure
Restrict access
Define user roles
Grant access
Revoke access
Key entry
156–163
Interactive voice
response
Optically scan
Import
053–058
Calculate (derive)
data
Code data
110–126, 127–137
Map data
Transpose data
Notify source
072–075
Identify discrepancy
071
Receive resolution
Resolve discrepancy
077–081
Manual review
061–064
Manual Auto Program-matic ID
059, 060, 065, 070, 151, 152, 092–096, 104
Double CumulativeSingle Incremental Add data
Confirm data (as is)
Update data
099–103
Delete data
Manage account
Archive data
Lock database
179–187
Develop load scripts
Develop screens &
tables
036, 037, 039
Program data
validation checks
038, 040
Program reports
041, 042
Validate program-
ming
252–253
Small numbers under activities document mapping to the Society for Clinical Data Management Core Competencies, www.scdm.org
type of
type of type of option
type of part of
part of
part of part ofpart of
part ofpart of
part of
type of
type of
type of
type oftype of
type of
nahm_p_22feb10.indd 1 2/22/10 1:34 PM