Data Integrity (Rpl)

A printed, optical, or electronic document designed to record all of the protocol-required information to be reported to the sponsor on each trial subject.

All information in original records and certified copies of original records of clinical findings, observations, or other activities in a clinical trial necessary for the reconstruction and evaluation of the trial.

Source data are contained in source documents (original records or certified copies).

Source documents are documents in which data collected for a clinical trial is first recorded (i.e. Original documents, data, and records).

Such data usually entered later in the CRF or eCRF.

Examples include: hospital records and GP medical summaries, subjects' diaries or evaluation checklists e.g. VAS scores, recorded data from automated instruments e.g. ECG MUSE recordings, copies or transcriptions certified after verification as being accurate and complete, x-rays, subject (CRF) files, and records kept at the pharmacy, at the laboratories, and at medico-technical departments involved in the clinical trial.

PSQs.

VDB print outs.

PSQ:

Only used to look at potential suitability for a trial.

(Essentially part of pre-screening).

Gives a preliminary indication of suitability and what may need to be clarified.

Intended that it is filled out without guidance.

Volunteers may not understand terminology and may make errors.

VDB:

Contains information from last encounter with volunteer/last time they screened, which may not be up-to-date when they come back for another study’s screening e.g. weight may have changed, smoking status may have changed. The latest information may not yet have been entered.

The operational techniques and activities undertaken within the quality assurance system to verify that the requirements for quality of the trial related activities have been fulfilled.

CSP, SOM or (for company procedures) SOPs.

CSP, SOM or (for company procedures) SOPs.

Authorisation by the PI that the staff member can perform specific tasks on a specific trial.

You ensure you are on the delegation log for that task and study by ensuring you:

GCP: Have read GCP/done MCQs – if not read/acknowledged then will not be on any delegation.

Study: Have read/acknowledged all relevant study documents e.g. CSP and SOM any updates to these.

Task: Have read/acknowledged/passed MCQs of relevant sections of relevant SOPs and for certain skills, have passed the exam in that skill (where relevant) and any study-specific training.

Any of the below not read/acknowledged:

GCP expired (this takes you off general AND study/procedure specific delegation).

Protocol amendment (CSP update).

SOM updated.

SOP updated.

MCQ or SOP read status has expired.

Specific skills exam expired (e.g. ICF, dosing).

Impact of performing tasks when not on delegation:

Means they have done procedure when not authorised to do so which is a violation of GCP. It implies you are not operating within the limits of their competence which is a GMC-reportable issue.

Results in an Incident Report being raised.

Can jeopardise integrity of the data entered.

Potential impact on business – e.g. future studies with the client.

Paper CRF:

Black ink pen should be used.

Writing must be legible.

Only pages with the correct VDB label should be completed.

Correct date and time format must be used.

All corrections should be made in a standard way, i.e. crossed out with one single line and signed & dated.

No free text is allowed with exception of free text boxes and comments sections.

Paper and eCRF:

Do not leave anything blank. If unsure about something, write unknown as ‘NK’ or ‘UNK’. If something is not applicable ‘NA’ should be written.

Do not use any abbreviation that is not listed in the SOP/CSP/SOM.

Corrections that are in any way suspicious or ambiguous must be explained in Investigator’s comments; i.e. where there is a potential for misunderstanding or to clarify amendments to data.

Investigators comments should be easily understandable, relevant and have supporting evidence when applicable.

PSQ

Medical history

HCF/medical records

Other patient letters (e.g. from specialists) or clinical results

How to ensure consistency: Check HCF/medical record against the PSQ and medical/surgical history page and ensure consistency.

Hierarchy of sources of data: PSQ is the lowest, then HCF/medical record, then doctor’s evaluation (the RP’s assessment of the medical history trumps the other sources).

If there is a discrepancy between the volunteer’s account and the HCF/medical record this should be clarified with the GP and/or volunteer as appropriate and this should be documented in the Investigators comment pages.

Comments regarding the medical history or concomitant medication should be entered in the Investigators comment page rather than the HCF i.e. only CS/NCS, initial can be written on the printout.

Standard reference ranges (e.g. TDL) or reference ranges set by CSP.

If none provided by either, RPL’s own ranges in the study SOM (or SOP).

Clinical tests repeated three times – gives consistent, stable readings. Validates the data/observed results by ensuring no experimental bias and random errors; also checks that the technique used gives the same results each time on the same sample.

Used in ECGs and BMI/weight, vital signs (usually only if 1st reading is abnormal or borderline).

Clinical tests repeated three times – gives consistent, stable readings. Validates the data/observed results by ensuring no experimental bias and random errors; also checks that the technique used gives the same results each time on the same sample.

Used in ECGs and BMI/weight, vital signs (usually only if 1st reading is abnormal or borderline).

Consistent HR over all three – within 2 beats difference.

Minimum of 90 seconds between first and last, maximum of 4.5mins to avoid hysteresis. Minimum of 30 secs between each ECGs.

No interference in any of the three readings.

Abnormal morphology – can do a repeat triplicate to confirm, or holter for better assessment.

Parameters e.g. QRS, QTcF assess as normal/abnormal as per usual ranges. CSP gives definite criteria about what can/can’t be included.

Abnormalities of potential significance should be investigated with a holter.

If several ECGs recorded, it is always the last 3 that are used (as once a valid triplicate is obtained ECG recording can stop).

Usually maximum of 10 ECGs allowed.

Clinical tests repeated three times – gives consistent, stable readings. Validates the data/observed results by ensuring no experimental bias and random errors; also checks that the technique used gives the same results each time on the same sample.

Used in ECGs and BMI/weight, vital signs (usually only if 1st reading is abnormal or borderline).

Consistent HR over all three – within 2 beats difference.

Minimum of 90 seconds between first and last, maximum of 4.5mins to avoid hysteresis. Minimum of 30 secs between each ECGs.

No interference in any of the three readings.

Abnormal morphology – can do a repeat triplicate to confirm, or holter for better assessment.

Parameters e.g. QRS, QTcF assess as normal/abnormal as per usual ranges. CSP gives definite criteria about what can/can’t be included.

Abnormalities of potential significance should be investigated with a holter.

If several ECGs recorded, it is always the last 3 that are used (as once a valid triplicate is obtained ECG recording can stop).

Usually maximum of 10 ECGs allowed.

Body weight can be variable and therefore BMI can change, and we need to repeat borderline values to ensure BMI criteria is met prior to inclusion. For each gender-specific BMI range, there are borderline values within 5% either side of the upper of lower border.

Borderline ‘in’ range exists to ensure those volunteers only just in the acceptable range are truly eligible (and not in by chance). Borderline ‘out’ range ensures volunteers are not unnecessarily excluded.

Values within the borderline ranges (either borderline in or out) need to be repeated on a separate visit.

Borderline values get repeated to get two consistent readings ‘(borderline) in’ the study’s accepted range to be eligible.

The third triplicate is only needed if the 2 first readings are inconsistent – i.e. one in, one out. If a second result is clearly in (when the first was borderline out), a third measurement is needed as there is too greater variation in the results which suggests measurement error.

1. Clinically significant medical history e.g. current symptoms or treatment.

2. Any history that may leads to their exclusion from the trial (or inclusion, for patients on a trial requiring a specific medical diagnosis).

3. Non-clinically significant but relevant medical history - e.g. for medicolegal reasons, to document that you have seen/considered it; potential for relapse during the trial; potential for scheduled/non-urgent treatment; or if sponsor requests that all medical history regardless of significance is recorded – will be stated in CSP.

Quality control is performed by trained members of staff who have passed the QC exam e.g. CTAs, SMs, some members of VR, who have worked in that particular setting.

To verify that the rights and wellbeing of human subjects are protected

To ensure the reported trial data is accurate, complete and verifiable from source documents

To ensure the trial is conducted in compliance with the Clinical Study Protocol, ICH GCP guidelines and applicable regulatory requirements.

Resolving queries in the data/source documentation raised by QC staff by amending/clarifying the original data recorded in the source document/CRF.

Staff must ensure the accuracy of the data they are recording and check their own work afterwards.

QC staff checking the CRFs that day who identify queries raise these directly to the relevant member of staff in person.

Queries should be resolved immediately.

These queries can be completed directly i.e. missing data can be filled in if available and no further documentation required.

Queries that cannot be resolved on the same day will be raised on an eMDF (electronic Monitoring Discrepancy Form) which is filed at the front of the CRF/screening file.

Once an internal query is raised a notification will be sent out.

Once a query is resolved the eMDF should be signed and dated. This will be checked by QC staff/EM. If the query was answered appropriately the QC staff/EM will close it.

For external queries: eMDFs generated by the external monitor are printed and filed by the internal QC staff. An appointment will be booked through the doctor’s calendar by the CQM for the resolution of these queries.

Queries must be resolved within 72hours.

Only the staff member who made the original entry can resolve the query. If this is not possible then the query can be resolved by another member of staff who is able to verify the resolution. If neither option is possible then a senior member of staff should resolve the query by documenting that the information is unknown.

Only the staff member who made the original entry can resolve the query. If this is not possible then the query can be resolved by another member of staff who is able to verify the resolution. If neither option is possible then a senior member of staff should resolve the query by documenting that the information is unknown.

An eCRF is:

It mirrors the paper CRF into which we enter trial data recorded in the clinic.

Not every trial uses an eCRF.

The benefit of it is that the sponsor (and their monitors) can remotely access the case report forms.

Paper CRF must match this to ensure correct data is being captured.

Impact on query resolution:

The impact on query resolution is that there are two documents to update.

After RPL staff enter data in the paper CRF, data entry staff input it into the eCRF usually within 72 hours.

Once this is done, the paper CRF page gets stamped by DM to indicate data on that page of the paper CRF must not be amended without DM being informed via email, to ensure that the eCRF is also updated. The paper and eCRF must be updated within a close timespan.

Data from the eCRF is exported into the database containing all study data. This is done by a single DM person.

Data from paper CRFs (where no eCRF is being used) is directly entered into the database twice, then compared for inconsistencies.

An eCRF therefore has less checks and greater potential to miss updates to the source from query resolution.

First process:

Queries are resolved in person with a data manager where doctor and data manager update both paper CRF and eCRF at the same time in a meeting. No further documentation is then required.

Second process:

Queries that are not resolved in person are exported from the eCRF into an eMDF (electronic Monitoring Discrepancy Form) and emailed/taken in person to the doctors by data management.

Once a query is resolved in the paper CRF the eMDF should be signed and dated to indicate it has been resolved. DM staff then take the paper CRF and amend the eCRF

First process:

Queries are resolved in person with a data manager where doctor and data manager update both paper CRF and eCRF at the same time in a meeting. No further documentation is then required.

Second process:

Queries that are not resolved in person are exported from the eCRF into an eMDF (electronic Monitoring Discrepancy Form) and emailed/taken in person to the doctors by data management.

Once a query is resolved in the paper CRF the eMDF should be signed and dated to indicate it has been resolved. DM staff then take the paper CRF and amend the eCRF

Queries in the data that has been entered in the database after the PI has signed off the paper and/or eCRF. If a paper CRF is used, this is indicated by a red page present (usually at the front of the paper CRF) and means the file has been signed off by the PI must not be amended.

Raised by DM.

Resolved using DCFs

First process:

Queries are resolved in person with a data manager where doctor and data manager update both paper CRF and eCRF at the same time in a meeting. No further documentation is then required.

Second process:

Queries that are not resolved in person are exported from the eCRF into an eMDF (electronic Monitoring Discrepancy Form) and emailed/taken in person to the doctors by data management.

Once a query is resolved in the paper CRF the eMDF should be signed and dated to indicate it has been resolved. DM staff then take the paper CRF and amend the eCRF

Queries in the data that has been entered in the database after the PI has signed off the paper and/or eCRF. If a paper CRF is used, this is indicated by a red page present (usually at the front of the paper CRF) and means the file has been signed off by the PI must not be amended.

Raised by DM.

Resolved using DCFs

If there is any error or discrepancy found, a Data Clarification Form (DCF) will be raised by DM.

Any queries that need to be resolved by the doctors will be forwarded to them by DM.

When completing a DCF, the query should be confirmed by checking the relevant CRF/source page however the correction can only be made on the DCF – the paper CRF cannot be amended.

If additional information is required e.g. evaluation of a blood result, it should be attached to the DCF.