SDV by ramchandavolu

VIEWS: 392 PAGES: 7

									Indian Journal of Pharmacology 2000; 32: 180-186

EDUCATIONAL FORUM
EFFICIENT SOURCE DATA VERIFICATION

EFFICIENT SOURCE DATA VERIFICATION
R. KHOSLA, D.D. VERMA*, A. KAPUR**, S. KHOSLA***
Sarabhai Piramal Pharmaceuticals Ltd., Dr. Vikram Sarabhai Marg, Baroda, Gujarat, India. *Philipps-Universität Marburg, Institute für Pharmazeutische Technologie und Biopharmazie Ketzerbach 63, D-35037 Marburg, Germany. **Medical Affairs & Clinical Research, Ranbaxy Research Laboratories, 20, Sector 18, Gurgaon-122 001, Haryana, India. ***Faculty of Dental Sciences, King George’s Medical College, University of Lucknow, Lucknow, India. Manuscript Received: 5.1.2000
SUMMARY

Revised: 1.2.2000

Accepted: 21.2.2000

The ICH harmonised tripartite guidelines for good clinical practice, the WHO guidelines for good clinical practice for trials on pharmaceutical products and the FDA’s Code of Federal Regulations require that source data verification must be undertaken for all clinical trials in phases I-IV. SDV, which is an evaluation of the conformity of the data presented in case report form with source data, is conducted to ensure that the data collected are reliable and allow reconstruction and evaluation of the study. The responsibilities of the principal investigator, sub-investigator, study coordinator, monitor, quality assurance auditor and the clinical trial manager in SDV must be made clear at the outset of the clinical trial and adequate training should be provided to the personnel involved. Special emphasis should be placed on direct access to data and confidentiality, so that there are no misunderstandings and errors when SDV is undertaken. Meticulous recording of what was done and found, including an evaluation of the findings, must be made to arrive at an indication whether the errors are random or systematic errors and are arising due to carelessness or deliberate actions. All personnel involved must realize that SDV adds to the scientific and ethical integrity of the clinical trial. This review highlights the various aspects governing the conduct, extent, difficulties and stream lining of the process of SDV and gives some suggestions for fast and effective SDV avoiding the common pitfalls observed by the clinical trial monitors. Source data verification good clinical practice

KEY WORDS

Introduction Good Clinical Practice [GCP] guidelines provide an international ethical and scientific quality standard for the design, conduct, performance, monitoring, auditing, recording, analysis and reporting of clinical trials that provides assurance that the data and reported results are credible and accurate, and that the rights, integrity, and confidentiality of the trial subjects are protected1,2. GCP guidelines for performance of high quality clinical research have been developed nationally and internationally. The ICH Harmonized Tripartite guidelines for GCP3, the WHO guidelines for GCP for trials on pharmaceutical products4, and Code of Federal Regulations Good Clinical Practice Guidelines, Parts 50, 54, 56, 312 and 3145 are three major guidelines which are being followed widely in a large number of countries. GCP
Correspondence: D.D. Verma e-mail: ddverma@hotmail.com

requires Source Data Verification [SDV] to be undertaken for all studies in phases I-IV. Source data All information in original records and certified copies of original records of clinical findings, observations or other activities in a clinical trial are necessary for the reconstruction and evaluation of the trial3. Source data are contained in source documents (original records or certified copies) and can be said to be the first place where information is recorded/ captured. Source documents Original documents, data, and records e.g. hospital records, clinical and office charts, laboratory notes, memoranda, subjects’ diaries or evaluation

181

R. KHOSLA et al.

checklists, pharmacy dispensing records, recorded data from automated instruments, copies or transcriptions certified after verification as being accurate copies, microfiches, photographic negatives, microfilm or magnetic media, X-rays, subject files, and records kept at the pharmacy, at the laboratories and at medico-technical departments involved in the clinical trial 3. The source documents are one of the “Essential Document” that permit evaluation of the study and the quality of the data and serve to demonstrate sponsor, investigator and monitor compliance with GCP and regulatory requirements. Timely filing of source documents greatly assists in the management of a clinical trial and they are of prime importance to the Regulatory Authorities in inspecting the trial and to the quality assurance auditor of the company in auditing the study. Source data verification Source Data Verification [SDV] is an evaluation of the conformity of the data presented in case report forms [CRFs] with source data. In this process, information reported by an investigator is compared with the original records to ensure that it is complete, accurate and valid. The aim of SDV is to ensure that the data collected are reliable and allow reconstruction and evaluation of the study. Strictly speaking, every item of data that appears in a CRF should be documented somewhere else to allow verification, audit and reconstruction. The main objective of SDV is to confirm that the information collected during a clinical study is complete, accurate, reliable and verifiable so as to give confidence to both the sponsor company as well as the regulatory authorities in the data being used to support a marketing application. SDV is also required to provide confidence in any data reported, for example, in published manuscripts and at scientific conferences. Without SDV, no scientist can have confidence in the data presented and the conclusions derived. Source data verification vs data monitoring In data monitoring [DM], the clinical trail monitor checks completed CRF pages to ensure that the data have been correctly entered and that all corrections

have been dated and initialled. Data monitoring is essentially a review of the CRFs by the sponsor company to check for errors and inconsistencies. On the other hand, in SDV, the data collected in CRF is checked against source documents. Confidentiality Continuous adherence to the maintenance of privacy of the trial subjects, including their confidential personal identity and all medical information, is of vital importance. The designated investigator and the clinical trial monitor must make every effort to maintain the confidentiality of information about subjects’ identity when checking the source documents. ICH GCP requires that any persons e.g., monitors, auditors, sponsors and domestic and foreign regulatory authorities, with direct access to clinical trial documents should take all reasonable precautions within the constraints of the applicable regulatory requirements to maintain the confidentiality of subjects’ identities and sponsors’ proprietary information. Informed consent of the subject/ patient should be obtained and documented prior to the study, and assurance given that confidentiality will be maintained. Custody and archiving The monitor should ensure that source data are retained at the study site to enable verification of information recorded in the CRF. The investigator/institution should maintain the trial documents as specified in the essential documents for the conduct of a clinical trial and as required by the applicable regulatory requirement(s). The investigator/institution should take measures to prevent accidental or premature destruction of these documents3. The essential documents should be retained until at least 2 years after the last approval of a marketing application in an ICH region and until there are no pending or contemplated marketing applications in an ICH region or at least 2 years have elapsed since the formal discontinuation of clinical development of the investigational product. These documents should be retained for a longer period however if required by the applicable regulatory requirements or by an agreement with the sponsor. It is the responsibility of the sponsor to inform the investigator institution as to when these documents no longer need to be retained3.

EFFICIENT SOURCE DATA VERIFICATION

182

Original/copy Source Data may be the original or a certified copy as agreed to by the investigator with the sponsor. Certain types of source data e.g., fax on a magnetic paper which fades away with time should always be photocopied, signed, dated and noted to be a true exact copy of the original. Key data in SDV The following data are considered key data in SDV and any gross errors in these might be detrimental to the scientific and ethical quality of the clinical trial: 1) Primary efficacy data, 2) Inclusion / Exclusion criteria, 3) Medical and medication history, 4) Physical examination/vital signs, 5) Visit dates, 6) Adverse events, 7) Concomitant medication and 8) A record that the patient has entered a clinical study and the date of informed consent. Methods of SDV There are two basic methods of SDV, viz. back to back and direct inspection. In the back to back method, the investigator holds all the source documents and answers specific questions asked by the monitor regarding the data without letting the monitor look at the documents. In the direct method, the monitor has direct access to source documents. The ICH GCP guidelines3 state that the sponsor should ensure that it is specified in the protocol or other written agreement that the investigator(s)/institution(s) will permit trial-related monitoring, audits, IRB/IEC review, and regulatory inspection(s), providing direct access to source data/documents. ICH GCP defines direct access as: “permission to examine, analyze, verify, and reproduce any records and reports that are important to evaluation of a clinical trial” 3. FDA has also recognized and acknowledged the importance of reviewing the actual source data. It has identified problems during its reviews of photocopies of data, e.g., instances where names of patients on X-rays and EKGs were obliterated and changed to match study patients6. FDA must be able to make a copy of any records or reports made by the investigational team as part of the trial 5. FDA regulations state that the investigator is not required to divulge subject names unless a more detailed study of the records for a particular individual is required. In these circumstances, it is not necessary for

patient identifiers to be visible on the documents provided for FDA review, as long as there is some linkage to the case report form, such as patient number and initials5. Conduct and extent of SDV Source data verification should be undertaken as the study progresses and as often as possible and every company should have standard operating procedure (SOP) giving precise instructions of how the monitor should undertake SDV. SDV should preferably be undertaken before data are retrieved from the investigator(s) site. In this way, only the data that have been confirmed to be accurate and valid are forwarded to the data management team. This can be done on an ongoing basis or at the conclusion of the study, depending on the duration of the study and the number of patients and the volume of data generated/ collected. However, it is better to conduct some SDV in the early stages of the clinical trial so that any problems elicited can be resolved before they have a significant impact on the integrity of the data generated. Although SDV is a requirement of GCP and is mandatory, however in the ICH3, the WHO4 and the FDA5 good clinical practice Guidelines, it has not been specified how much percentage is required. It is more of a company’s decision to decide whether to do 100% SDV for all patients or to do 100% SDV of key data and a sample of the rest. It is preferable that thorough SDV should be done for first few patients at the center, then quick SDV for each of the remaining patients in a study. The monitor should check some item in each CRF for every investigator, at every center. British standard specifications for SDV are as follows:
Patients Data Specification Monitor points 100 100 CRF 65 data points Random selection Result Action

<3 errors Data accepted

of 65 data points from several > 3 errors Batch patients rejected and data checked in greater detail

183

R. KHOSLA et al.

Although sampling may be acceptable method of SDV, there are occasions when 100% SDV is more appropriate. For example, when the number of subjects at a trial site is very less, it is preferable to check all data during a monitoring visit. Another way is to do 100% SDV until confidence is reached in the quality of the data. The extent of source data verification depends on a number of factors viz. phase of the clinical trial, company policy, regulatory requirements, quantity of data, man power available to the company, time availability, Investigator research experience, investigator staff changes etc. A common approach in SDV is to identify different categories of data, for example “critical” and “noncritical” data. Critical data are those that are focal to the aims and objectives of the study, and which must be correct. The non-critical data should, also be correct, but if an item of such data is in error, it is not critical to the outcome of the study. 100% of the critical data should be verified against the source documents for every patient in the trial. Examples of critical data that should be checked for each subject include the following: 1. 2. 3. 4. 5. 6. Informed consent to participate in the study. Physician/ dental notes, preferably on the hospital/ clinic stationary Primary efficacy end points Secondary efficacy end points Recording and reporting of serious adverse events A confirmation in medication history or elsewhere that the study drug was prescribed and in the specified dosage Conformance to the patient inclusion/ exclusion criteria Visit dates as per the window period specified in the protocol of the study.

high numbers of write-overs, poorly completed, very neat, same ink over a long period of entries, presence of serious adverse events, lack of adverse events when they would be expected, rapid recruitment, withdrawals, and so forth. It is of importance that each type of data is checked in the sample selected for SDV of non-critical data. Some of the common problems encountered during SDV are: Data entered directly on to the CRF, short/ brief medical history, scant clinical notes, several volumes of source documents, source document/CRF mismatch, maximum acceptable error rate, illegible handwriting of investigator/subinvestigator. Management support for the SDV process can sometimes be lacking due to time or budgetary constraints that may be placed on the study monitoring activities. SDV process must be meticulously recorded, so that retrospectively, the following points are verifiable: 1. 2. 3. 4. 5. 6. 7. What source documents were examined Which CRFs were checked and why they were selected The critical data items checked on each CRF The non-critical data items checked on each CRF The frequency of errors The nature of errors Whether any ameliorative action was undertaken at the site

7. 8.

Non-critical data can be checked for a proportion of CRFs. The percentage varies from trial to trial and company to company. Generally it is 15 to 25%. This proportion is generally chosen randomly as the trial progresses, rather than being predetermined at the start. This method allows the sample to include those CRFs that meet specific criteria for checking, e.g.,

The company should initiate, introduce and sustain the SDV process with emphasis on understanding the process, educating the monitors, specifying requirements (SDV plan), documentation of the data checked, understanding and use of sampling techniques, action to be taken in case of questionable / unreliable data [ e.g. the monitors should not panic ; should collect evidence and document facts and take appropriate action after discussing with the study director]. The instructions for carrying out SDV must be clearly documented in an SOP or a working practice document (WPD). The SOP/ WPD should address such elements such as: 1. Prior agreement with the investigator for direct access to source data during monitoring, auditing and regulatory inspections. Procedures specifying how to conduct SDV

2.

EFFICIENT SOURCE DATA VERIFICATION

184

3. 4. 5.

Procedures specifying when to conduct SDV Procedures specifying the determination of sample size / % of SDV to be undertaken Procedures specifying how to record their observations e.g. in standard forms created for the purpose Procedures specifying how to evaluate the findings and what to do and whom to contact if they encounter questionable/ unsubstantiated data.

Table 1. Source Document Verification Checklist. Monitored • SDV •

Patient Initials..................... Randomization No............... Assessment Eligibility Pt Demographics Randomization # Consent Medical History Diagnosis Phyical. Exam Drug History Vital Signs Blood Concomittant Meds Invest.Global Ass Compliance Adverse Events Diary Coll. Lab Test COMMENTS :

Trial Number..................... CRF Collected................... ..../..../.... Final Visit ..../.../.... W/Draw

..../..../.... Baseline • • • • • • • • • • • • • • • • • •

6.

The results of SDV should be critically evaluated to determine whether the standard of data and documentation is appropriate. An acceptable standard for data accuracy should be determined in the SDV SOP/ WPD or the clinical trial protocol. An appropriate level of accuracy may be zero errors for critical data items and <0.5% for non-critical data. Reasons for any identified inconsistencies should be ascertained. For example, are they due to lack of source data, or due to conflicting information in different documents, or due to incorrect methodology, or inappropriate training or carelessness. If the monitor has any suspicions about data being fabricated, the matter should be reported immediately to the clinical trial management and they will decide whether to conduct additional SDV by monitor, or to have an independent audit by the clinical quality assurance auditor to confirm the findings. Subsequently, on the basis of the information obtained, a decision would be made whether to stop the trial at the particular site and to exclude the center and the data generated from further participation in the study and to inform the regulatory authority(ies). A sample Source Data Verification Checklist is provided in Table 1 which might vary from company to company and trial to trial. Implication for not performing SDV There could be serious implications of not performing SDV. There is always the danger of collecting inaccurate or spurious data from which false or inappropriate conclusions can be drawn. In clinical trials, a relatively small amount of information is collected from a small number of patients that can subsequently result in massive patient exposure if the drug is marketed. Also, the rejection of clinical trial data by the regulatory authorities, if they are not satisfied with the integrity of the data, can lead to a refusal to grant a marketing license and this can lead to severe financial and time loss to the sponsor company.

• • • • • • • • •

• • • • • • • • •

• • • • • • • •

• • • • • • • •

•

•

DRUG ACCOUNTABILITY CHECKED • NAME OF MONITOR

SIGNATURE OF MONITOR

Suggestions for efficient SDV * The monitor should always discuss SDV with investigator at the site assessment. Monitor should be honest, frank and assertive. The monitor should have good interpersonal and negotiation skills and the new monitor should be accompanied by a mentor or a senior monitor to get practical exposure to the problems encountered in real life settings. Direct SDV should be undertaken as per ICH GCP guidelines. In the pre-site visit letter to investigator, ask for the CRFs you require to be available at the monitoring visit, for your inspection prior to the meeting with the investigator. Plenty of time should be spared to undertake SDV, particularly at early visits when you may be unfamiliar with the patient files. It is best that the investigator is not present when you undertake SDV. Undertake SDV in peace and quiet, without interruption.

*

*

*

185

R. KHOSLA et al.

*

Company SOP/ WPD should instruct on what data to verify and how to document it. If the SOP/ WPD is vague, prepare a plan of action before you visit, identifying which items you will check, in which CRFs. Stick of your plan of action. All the different facilities involved in the clinical trial should be considered, for example, pharmacy, laboratory, X-ray room, etc., as they may generate data that need to be examined. The existence of other associated documentation related to the study, such as subject diaries, appointment logs, etc., may contain valuable information about the patients’ participation in the trial, and an effort should be made to check these out. A good way to record which data have been verified is to write a “V” [meaning verified] against the item(s) verified, on the top copy of the CRF using a coloured ink (eg. green): This makes it clear to you on future visits what has already been verified and easily identifies items checked during an audit or inspection (it is also faster than making a report of every item checked!) Concentrate on items that will affect the outcome of the study (e.g. patient inclusion criteria, key measures of efficacy and safety). An important and often overlooked part of SDV is that of determining the actual process employed in generating the data. The correct methodology needs to be used, by appropriately trained personnel, otherwise the data may be invalid (even if it may be complete and accurate!). Make notes and a separate report about the errors or inconsistencies that you discover. Ensure your report is put in the trial master file. Never tell your investigator which data you will verify. Never try to carry out data monitoring and SDV at the same time. Keep these tasks separate even if it means going through the CRFs twice (which is never a bad thing). Be honest - only report that you have undertaken SDV on certain books / items if you have in fact done so. Cutting corners does not help anyone. Reassure the investigator that he /she is not *

*

being singled out and the SDV is performed at every site in every study. SDV by monitor will be more efficient if right from the beginning of the trial, the investigator is given the message that the monitor is not a policeman, but rather, the investigators’ own QC manager, working to ensure that whatever goes out with the investigators’ signatures is not questioned later. * The monitor must be assertive. The bottom line in any discussion must be “we have to do SDV”. Simply do not use an investigator who does not allow this. You may, of course, also decide to exclude investigators who do not allow direct SDV. Direct SDV is definitely the best way of fulfilling this important GCP requirement. Current experience suggests that direct SDV is becoming more and more acceptable in areas with special requirements regarding confidentiality ( e.g., AIDS, infertility). Back to back/ indirect methods may be unacceptable and monitors should therefore research the best available methods of SDV with the investigator. Some general aspects should always be considered during SDV, for example, is documentation available sufficient to confirm that the patients existed, had the disease for which they were treated, and were available for the entire duration of the study? Also check the recruitment process of the patients into the trial and who collected the trial data and was this appropriate? See if photocopies of the documents are present, then where are the originals? It is worth emphasizing that there is a need to educate hospital administration, medical record keepers, nurses, pharmacists, lab staff and others about their role in ensuring the credibility of research done in their institution. The institutions’ policy and procedures need to be in tune with the principles of GCP; they cannot be molded easily for an occasional research project.

*

*

*

*

*

*

*

* *

Future trends FDA has issued its requirements for electronic records and signatures in 21 CFR Part 11 in 1997 7. The purpose is to provide the criteria under which FDA will consider electronic records as being equivalent to paper records, and electronic signatures

*

*

EFFICIENT SOURCE DATA VERIFICATION

186

equivalent to handwritten signatures. The principles of conducting SDV on electronic records are the same as for paper records. Electronic records, like paper records, must be accurate, original, legible, attributable and contemporaneous. The computerized system should be validated. In view of the many advantages of electronic recording, like increased speed, less errors and easier storage and transmittal, the use of computerized systems for gathering, analyzing, and reporting clinical trial data is becoming increasingly popular. Conclusions Source Data Verification is required for all clinical trials phase I - IV and is an evaluation of the conformance of the data presented in CRFs with source data. Sponsor should have detailed SOPs which should state precisely the extent and amount of SDV to be undertaken. The sponsor company should also ensure that it is specified in the total or other written agreement that the investigator(s)/ institution(s) will permit trial related monitoring, audits, IRB/IEC review, and regulatory inspection(s), providing direct access to source data/ document. The clinical trial monitor should discuss SDV with the investigator at an early visit, follow his SOPs carefully and document care-

fully all SDV undertaken and the results in the trial master file. It is essential that the data reported by the investigator to the sponsor are accurate and valid, to allow a proper assessment of the efficacy and tolerability of any new drug. REFERENCES
1. Khosla R. GCP guidelines: differences between the ICH and WHO guidelines. Good Clin Pract J 1998;3:24-30. Verma D.D. , Sharma G., Arora H. ; GCP guidelines : FDA Vs WHO. Good Clin Pract J 1999;6:18-22. ICH Harmonized Tripartite Guidelines for Good Clinical Practice [E6: good clinical practice:consolidated guideline], 1997. WHO guidelines for Good Clinical Practice for Trials on Pharmaceutical Products, 1994. Code of Federal Regulations (CFR) Good Clinical Practice Guidelines Parts 50, 54, 56, 312 & 314, April 1999. Bruckheimer M. FDA’s Inspections of Clinical Investigators. Drug Inf J 1993;27:213-6. US Department of Health and Human Services, US Food and Drug Administration, 21 CFR Part 11, Electronic Records; Electronic Signatures: Final Rule, 1997.

2.

3.

4.

5.

6.

7.


								
To top