Low Hanging Fruit in the Fight Against Inefficiency


Applied Clinical Trials

Applied Clinical TrialsApplied Clinical Trials-03-01-2011
Volume 20
Issue 3

Direction from regulatory agencies would help eradicate wasteful 100 percent source data verification.

The vast majority (84%) of sponsor companies report that they are checking 100 percent of their source data against case report form data. Ask sponsors why they support this practice and they'll tell you that regulatory pressures require it. Ask regulatory agencies if this is true and they insist that such behavior is overkill and unnecessary. In the absence of unequivocal direction, sponsors are reluctant to use any of a variety of cost- and resource-saving alternatives to 100 percent source data verification (SDV).

Now facing a more intense and challenging operating environment, sponsors are clearly and more actively questioning whether the practice of 100 percent SDV is practical and essential. Many argue that the practice diverts attention and resources from far more mission critical clinical trial activities. Study monitors spend approximately half of their time verifying source data. For example, one-quarter of CRA time is spent checking GCP compliance and drug accountability, and only 5 percent of their time is interacting with study staff. Sponsors also contend that there is no evidence to support whether 100 percent SDV improves data quality. A recent Society for Clinical Data Management report noted that SDV is an intensely detailed task that is prone to human error and inaccuracies.

Kenneth A. Getz

The potential savings from a more selective and targeted source data verification approach are substantial. These savings, if realized, would go a long way toward providing much-needed additional support to promising candidates in company portfolios and to resource needs. A recent study by Funning and colleagues found that SDV costs companies on average one-third of the entire Phase III pivotal trial budget. Another recent study found that using a risk-based approach to reduce the number of monitoring visits and data points verified would save pharmaceutical and biotechnology companies $3- $5 billion each year.

Necessity or overkill?

Sponsors have taken an extremely risk-averse stance. No sponsor wants to be caught in a situation where the Food and Drug Administration (FDA) and the European Medicines Agency have found clinical study data errors and inconsistencies. As such, study monitoring and source data verification practices are based on the most conservative interpretation of GCP-ICH regulations. Although sponsors have suspected for years that 100 percent SDV is overkill, most sponsors have dug in their heels and refused to alter historical practices believing that 100 percent SDV remains the best way to ensure the validity and integrity of clinical data.

Regulations, on their part, neither specify how monitoring must be conducted nor require specific data review methodologies. GCP guidance developed by the International Conference on Harmonization (ICH E6) does not mandate 100 percent SDV. Technically, regulatory agencies don't want study monitors to check every source data point at every clinical site. The FDA guideline on GCP, for example, explicitly calls for the review of a "representative" number of study volunteer records, not all records.

Is partial and alternative source data verification feasible? Absolutely. One need not look any farther than government-funded research. Study monitors checking studies sponsored by the National Institutes of Health (NIH) have been doing partial SDV for decades providing insight into ways that CRF data can be verified with less frequent and more targeted monitoring visits. According to a report released by the Clinical Trials Transformation Initiative (CTTI) in 2010, onsite monitoring visits are performed on less than one-third (31%) of studies sponsored by the NIH. Of those studies in which there is an onsite monitoring visit, only half involve CRAs verifying CRF data to the source data.

Improving the SDV process

A number of alternatives to 100 percent SDV hold promise in reducing monitoring visits and promoting higher levels of monitoring efficiency while improving data quality. Some sponsors are considering a risk-based monitoring approach in which clinical teams identify the highest risk areas for the data and then tailor their source data verification strategy to these risks. Both the study parameters—including the size, phase, and complexity of the protocol—and the experience of the site staff are considered when assessing the study risk.

Figure 1. SDV costs companies about one-third of the total Phase III study budget.

Targeted SDV uses more sophisticated statistical methods to select a subset of specific data that will be verified during the on-site monitoring visit. Under a fixed field targeted SDV approach, only case report form data for the most critical variables (e.g., primary endpoints and adverse events) are 100 percent verified and other data variables are not scrutinized as closely. Random fields targeted SDV uses random sampling techniques to select specific case report form fields that will be verified during each monitoring visit.

Another approach gradually reduces or increases the volume of source data verified based on past onsite monitoring visit experience. The initial monitoring visits might start with 100 percent SDV. If no significant data quality issues are identified at the site, the volume of SDV decreases on subsequent monitoring visits. Or the initial monitoring visit might start with 50 percent SDV and on subsequent visits the verification of source data is increased to higher levels as needed.

Major electronic data capture systems were largely developed to support 100 percent SDV. In order for alternative SDV approaches to be used, technology solutions are being adapted to accommodate risk-based and targeted SDV approaches. Market leaders Medidata Solutions and Oracle have already introduced solutions to support these alternative approaches.

Awaiting agency guidance

Most drug development sponsors are aware of the variety of alternative approaches and they anticipate more widespread use when operating conditions (e.g., infrastructure, processes, controls, and staff experience) and regulatory direction converge. The overarching sentiment at this time, however, is that of caution.

Sponsors appear fearful that broad use of a novel alternative monitoring approach—without regulatory acceptance—may invite additional agency scrutiny. Many sponsors are leery about adopting novel SDV approaches in the absence of real case examples of companies that have successfully pioneered alternatives to support new drug applications. Another major obstacle to implementing risk-based and targeted SDV approaches is the aversion and reluctance to change legacy processes and systems.

New guidance from regulatory agencies is expected to break this logjam of caution. At the present time, the FDA is working on a draft guidance to provide clarity of source data verification guidelines and to promote alternative approaches to 100 percent SDV. According to agency statements, the draft guidance will describe FDA's current thinking about risk-based and targeted SDV monitoring approaches. Recognizing the high cost, logistical complexities, and potential inefficiencies with onsite monitoring visits, regulatory authorities appear very open to evaluating new SDV approaches. It is unclear when the draft guidance will be released.

More can be learned and communicated about data quality and the validity of alternative SDV approaches based on real company experiences. Certainly more can be learned from the unique SDV practices found in NIH funded studies as well. Not only is the incidence of onsite monitoring visits supporting NIH-funded studies low (only 31% of all studies) according to CTTI, but also certain data elements are rarely verified (e.g., non-serious adverse event reports and secondary endpoint reports). These practices can and will inform FDA and industry sponsors about alternative approaches most conducive to reducing overkill while maintaining a sufficiently robust approach to verifying source data.

There is much riding today on initiatives that can generate substantial drug development cost savings, study conduct efficiency, and cycle time improvement. One hundred percent SDV represents low hanging fruit—if and when regulatory agencies say so.

Kenneth A. Getz MBA, is a Senor Research Fellow at the Tufts CSDD and Chairman of CISCRP, both in Bost, MA, e-mail: Kenneth.getz@tufts.edu

© 2024 MJH Life Sciences

All rights reserved.