DATA MANAGEMENT

            Data from clinical trials must be collected, recorded, and reported in a systematic and
            accurate manner. There are many steps in the data management of clinical trials:
                      Developing forms for data collection.
                      Collecting the data.
                      Accurately recording the data.
                      Entering data into the database and assuring the quality of the data.
                      Analyzing the data.
                      Reporting the data.
            The Clinical Research Coordinator (CRC) is key in assuring that the cycle of data
            collection and reporting runs efficiently.

            Data in clinical trials may be collected by a variety of means. Most studies collect data
           prospectively using data flow sheets or Case Report Forms (CRFs). Data flow sheets are
            designed to record selected data over time in various columns. The data may be
            summarized on a separate summary page, or all data may be computerized for analysis.
            CRFs are generally used by pharmaceutical companies and are often very detailed and
                 Most studies gather data in the following general areas:
                      Demographics (sex, birth date, age, weight, height, race, marital status).
                      Patient medical history.
                      Diagnosis (primary and secondary, coexisting medical disorders).
                      Investigational agent administration (dates, time, identification, and code

© 2002 by CRC Press LLC
            138      Clinical Research Coordinator Handbook

                      Other drugs and treatments (previous, concomitant, adjuvant).
                      Physical examinations.
                      Vital signs (temperature, blood pressure).
                      Clinical laboratory (blood chemistry, hematology, urinalysis, other specific tests).
                      Adverse reactions.
                      Efficacy parameters.
                      Study completion.
            The CRF should reflect the protocol and be designed to record and document all of the
           required data. The criteria for CRF development are as follows:
                      The CRF relates directly to the parameters of the protocol.
                      The CRF is easy to complete, clear and concise, and does not duplicate data.
                      The CRF is easy to review.
                      The data are compatible with the computer database.
                      The CRF does not collect extraneous information not related to the objective of
                      the protocol.

            The second step in data management is collecting the data. Data should be collected as
            CONSISTENTLY as possible. How is this achieved?
                      All persons making assessments on patients should agree to the same guidelines.
                      Ideally, the same person should assess the same subject at each visit.
                      Use the same clinical laboratory for testing.
                      Use study aids (see Table 7.1).
                      Develop a rapport with subjects so that they understand their obligations in the
                      study and are willing to return as scheduled.
                      Record data directly on the medical record; transcribe to data flow sheet or CRF
                      as so011 as possible.
                      Advise ancillary personnel (e.g., lab technicians) of the study requirements.

© 2002 by CRC Press LLC
                                                                                          Data Management         139


                      Use the protocol schedule of study visits and evaluations (see Appendix D) or study schema.
                      Design subject worksheets by visit date and required parameters.
                      Design a calendar for each subject with study visits outlined.
                      Refer to relevant CRF pages.

                      Design visit-specific pages listing required parameters for each visit to attach to the subject's
                      chart during the visit.

                      Prepare all paperwork, e.g., clinical lab requisitions, ahead of time.
                      Establish a query system or prompt for subject return appointments.
                      Keep a calendar of all subject return appointments.

            Accurate recording of data in a clinical trial is a very critical step. Accuracy, legibility,
            and consistency are key. Recording data on flow sheets or completion of CRFs is a very
            time-consuming and tedious process. Data entries on CRFs should be completed on a
           regular basis (ideally, immediately following the subject visit) to avoid a backlog of
            entries and to ensure that accurate data are captured. Table 7.2 lists guidelines for recording
            data and completion of CRFs.
                 Data on CRFs will be reviewed by the monitor and then reviewed again by the data
            managementldata entry group at the sponsor's offices. Table 7.3 presents common areas
            where data are verified and errors may be detected.
                 Ultimately, the CRF is a part of the New Drug Application (NDA) and the FDA
            (U.S. Food and Drug Administration) has access to each CRF for all studies.

© 2002 by CRC Press LLC
            140      Clinical Research Coordinator Handbook

                      CASE REPORT FORMS

                      Follow directions for completing the CRF and obtaining variables as written in the protocol, study
                      procedures manual, and CRF.
                      Make all entries in black ink to maximize legibility and facilitate copying.
                      Print legibly and neatly.
                      Use acceptable medical terminology.
                      All comments should be brief, concise, clear, and confined to "comment" sections of the CRF.
                      Write comments only if they apply to the study and the topic.
                      Correct errors by drawing a single line through the incorrect entry, reenter the correct data nearby,
                      initial, and date the correction. DO NOT write over data, erase data, or use correction fluid. If data
                      are to be rewritten onto another CRF page, make a note on the original page, line-through the
                      page, initial, and date. If the CRF was signed by the investigator, heishe also must initial the new
                      page verifying agreement with the data.
                      Complete all blanks. If data are missing, use one of the following conventions (verify use with
                          NA       Not Applicable.
                          ND       Not Done.
                          UNK      Unknown.
                      Do not draw lines to indicate missing data.

                      Verify that items entered in "Other" categories do not fit in a coded field.
                      Verify UNITS and TIME are recorded as specified on CRF.
                      (Appendix E contains algorithms for conversions.)

                      Data must be complete for the day, month, and year during the study.
                      When using study-specific rating scales or codes, use only the options provided.
                      Use only CRFs designed for the particular study.
                      Check that all errors and crossouts are initialed.
                      The investigator must review and sign each CRF after completion.

            Data are recorded from a variety of source documents: clinic notes, medical records, nurse's
            notes, radiology reports, and lab reports. Even the appointment book and hospital census
            are considered source documents. The investigator is required to maintain source documents
            as well as provide specific data for the analysis of the study. Sponsor representatives and
            FDA inspectors need to have access to subject records. (Permission from the subject should
            be obtained as part of the informed consent process.) The monitor will verify data entries
            against source documents for accuracy and content during periodic site visits.

© 2002 by CRC Press LLC
                                                                                       Data Management           14 1


                      lnclusion/exclusion criteria: Were subjects entered appropriately? Verify per source documents
                      (medical record, lab results).
                      Verify entry date to date informed consent signed.
                      Verify all dates to source documents.
                      Especially be aware of transcription errors commonly occurring for patient ID number, initials,
                      dates on each page.
                      Verify all visits according to protocol.
                      Verify birthdate (especially the year).
                      Verify all data (clinical labs, PE, etc.) to source documents.
                      Check adverse events. Review the event, relationship to drug, severity. Do any of the adverse
                      events qualify as serious adverse events requiring an Investigative New Drug (IND) Safety Report
                      to the FDA?
                      Has the Principal Investigator or a registered subinvestigator signed the CRF?
                      Are all boxes completed?
                      There are no missing or incomplete data.
                      Review clinical lab results for indication of adverse events.
                      Investigational agent administration is ACCURATELY recorded, agrees with protocol dosing
                      guidelines, and matches the drug accountability record.
                      If changes are made, assure that all related data are updated accordingly.
                      All corrections are clearly written, initialed, and dated.

                 Every subject in the trial should have a separate clinic chartlmedical record.
            Minimally, visit dates and clinician's comments should be recorded. Lab reports should
            be included in the record as well. All investigative and/or clinic staff examining subjects
            should be very diligent and precise when writing clinic notes.
                  The following specific items should be recorded:
                      Data not reported elsewhere, e.g., temperature, height, weight, blood pressure,
                      physical exam findings.
                      Subject's comments suggestive of an adverse experience.
                      Any comments suggestive of noncompliance.
                      Any irregularities, e.g., missed doses or taking concomitant medication.
                      Phone contact with subject.
                      Attempts at contacting subjects "lost to follow-up."

© 2002 by CRC Press LLC
            142      Clinical Research Coordinator Handbook

                 Be aware that any discrepancies corrected in the source document might require a
            correction on the CRF. Similarly, corrected information on the CRF should be supported
            by source documents.
                 Healthcare professionals receive specific training in recording information in medical
            documents. Table 7.4 contains guidelines for medical record documentation.
                 Source documents must be maintained as long as all other study materials: for a
           period of two years following the date of NDA approval for marketing for the indication
            studied, or if no application is filed or one is not approved, until two years after the
            notification to the FDA that the investigation (IND application) is discontinued (2 1 CFR
            3 12.62). For studies conducted to meet ICH GCP Guidelines, records must be retained
            for two years after approval in the last country submitted. Verify that the institution's
           policy for the retention of medical records meets these requirements. Also, determine
            the institutional policy for disposition of medical records of deceased subjects. Where
            are the archives? Are the files readily accessible?

            Electronic Data Transfer
            Central labs used for clinical trials or individual labs often arrange for transfer of data
            through computerized methods or by electronic tape. These data transfers are less likely
            to produce transcription errors (and relieve the CRC of a lot of data entries). However,
            a hard copy of the lab results must be maintained with the subject's permanent
            record and CRF.
                Additionally, hospitals and clinics are becoming more sophisticated in the collection
            of patient data by specific software programs. It may he possible to transfer (by disk or
            modem) selected pieces of patient data to the sponsor electronically, eliminating
            transcription of data to CRFs. However, as above, it is necessary to maintain hard copies
            (printouts) of the data with CRFs, and there must also be a way to verify data (source
            document verification). The investigator also will have to review the data and sign
            appropriate forms.

            Remote Data Entry
            One method of collecting clinical trial data is by remote data entry. A computer terminal
            is set up at the site and connected to the sponsor's data bank by modem. The purpose is
            to decrease the time involved in data collection and entry by directly entering the data.

© 2002 by CRC Press LLC
                                                                                             Data Management              143


                      Write dateitime at each entry (military time is preferred or use A.M.or P.M.).

                      Write legibly, clearly, neatly.

                      Use black ink (reproduces best in case of legal action).

                      Cross out errors with a single line, write "errornandcorrect entry, then initial and date the correction.
                      If there is not enough room, use the next line to make the change.


                      Make a line through empty space remaining after a note, then sign your name and title.

                      If a page portion is skipped (blank), cross a single line through it and begin your note at the next
                      page at the top or at the next available space.


                      Use proper units of measurement for all entries.

                      Never use correction fluid. The old entry should always be visible.

                      Never "estimate" data points that are quantitative. If you forgot a data point, write "omitted," or
                      circle and leave blank. Consult with the study sponsor about handling missed data.

                      Be sure to receive proper permission before removing parts of the patient chart or copying it.

                      Remember that all patient information is strictly confidentialand constitutes a legal document.

© 2002 by CRC Press LLC
            144      Clinical Research Coordinator Handbook

            The data will be reviewed by the monitor either at the site or electronically, and
            clarifications may be discussed with or e-mailed to the CRC. Remote data entry systems
            must have security mechanisms to assure the integrity of the data (21 CFR 11). When
            using remote data entry, it is imperative to maintain a hard copy of the data at the site.

            MaintaininglStoring CRFs
            The original copy of the CRF generally goes to the sponsor and is kept at the sponsor's
            facilities. Some studies will require that copies of the CRF be submitted to the sponsor
            by fax to facilitate timeliness of data collection. Accurate copies (photocopy or NCR
           paper) of the CRFs must be kept by the investigator with the study files for the following
            time periods:
                      Two years following the approval of the NDA
                      Two years following the discontinuation of the IND application.
                 Note that this is not two years after your site completes the study. Actually, this can
            be quite a long time! If your study was an early Phase I1 trial, it may be five to seven
            years before enough data are gathered through the completion of the Phase I11 studies,
            the writing and submission of an NDA, and review at the FDA.
                 Also note that electronically submitted clinical lab data and data submitted by remote
            data entry must be backed up with hard copies in the study file.

            Data from CRFs or flow sheets are generally entered into a computer database. A
            statistician analyzes the data according to the statistical analysis plan stated in the protocol.
            Queries may be generated during this process, and the monitor will discuss them with
            the CRC and investigator. The data are summarized into tables and statistical figures for
            inclusion in the final study report and submission to the FDA in an NDA.

            Data from clinical trials are reported in final study reports. Each investigator must
            summarize the events of the trial at that particular site with hidher patient population
            and submit the report to the Institutional Review Board (IRB) and the sponsor of the
            trial. This report should include the following:

© 2002 by CRC Press LLC
                                                                             Data Management        145

                      Dates of the beginning and end of the study.
                      Number of subjects enrolled.
                      Number of subjects completing the study.
                      Number of dropouts and reasons.
                      Summary of adverse events.
                      Notation of patient deaths, if any.
                      Overall opinion of the effect of the investigational agent.
                      Comparison with other effective treatments.
                      Summary and interpretation of clinical laboratory values outside of the
                      normal range.
                      Listing of protocol violations and explanation.
                      Discussion of any specific subject experiences, if warranted.
            Additionally, the investigator may wish to publish the results with the approval of the
                 In the bigger picture, the sponsor will summarize all of the data from all trial sites
            and prepare a final study report or final medical report. This report is a detailed summary
            of the statistical and clinical results of the trial. Emphasis is placed on the safety and
            efficacy of the investigational agent. Generally, a discussion is included that positions
            the investigational agent as an effective treatment in the disease being studied. The report
            may be part of an NDA, a line extension, an OTC conversion, marketing, or for publication
            in medical journals.

            2 1 CFR 3 12.62.
            21 CFR 11.

                 Recordkeeping in Clinical Investigations (10195).
                 Compliance Program Guidance Manual (Clinical Investigators) (8118/94).
                 Guide for Detecting Fraud in Bioresearch Monitoring Inspections (4193).
            See also references for site inspections.

© 2002 by CRC Press LLC
            146       Clinical Research Coordinator Handbook

            A Multidisciplinary Approach to Data Standards for Clinical Development. Rebecca
            Kush, Wayne Kubick, Kaye Fendt, Dave Christiansen, and Judith Sromovsky, Applied
            Clinical Trials, Vol. 9 (6), p. 76, 2000.

            A Review of the Source Document Verification Process in Clinical Trials. M. Schuyl,
            and T. Engel, Drug Iizfornzution Jou7-izal,Vol. 29 (4), pp. 129 1-1 299, 1995.

            Auditing Clinical Data. H. Ruth Pyle, Applied Clinical Trials, Vol. 9 (3,65, 2000.

            Clinical Study Conduct/Procedures. R. Guarino, Drug Iizfornzution J o ~ ~ r i ~ u l , (2),
           pp. 481488, 1994.

            CRF Design and Planned Data Capture. Wendy Bohaychuk, Graham Ball, and Katy
            Sotirov, Applied Clinical Trials, Vol. 8 (6), p. 64, 1999.

                                  in        Triuls. B. Spilker, Raven Press, New York, 199 1.
           Data Collectioiz Fo7-MIS Clii~icul

            Designing Case Record Forms for the World Wide Web. Paul Bleicher, Richard Dab,
            Patricia Giencke, and Jeffrey Klofft, Applied Clinical Trials, Vol. 7 (1 l), p. 36, 1998.

           Drug Ii7fornzutioiz Jou7-izal,Vol. 28 (2), 1994. Contains multiple articles on the following

                      Clinical Laboratory Data: Practical Approaches to Using Control Laboratories
                      in Error.
                      Clinical Safety Data Management in Japan.

            Electronic Exchange of Laboratory Data: Recommendations for Improving Quality and
            Reducing the Hidden Costs. PMA Task Force, Drug Iizfornzutioiz Jou7-izal,Vol. 27 (3),

            Identify Yourself! Computer Security and Authentication. Paul Bleicher, Applied Clinical
            Trials, Vol. 8 (lo), p. 40, 1999.

            Sign Here Please, Mr. Bond. Paul Bleicher, Applied Clinical Trials, Vol. 9 (8), p. 28,

© 2002 by CRC Press LLC

Shared By: