A volcano plot displays the significance and relative risk of adverse event incidence by treatment exposure.
Bridging the communication gap
JMP® Clinical has doctors and statisticians speaking the same language
|Challenge||To help pharmaceutical companies quickly characterize emerging safety issues found in clinical trial data.|
|Solution||JMP Clinical, combining the industry-leading power of SAS® Analytics with the graphical flexibility of JMP.|
|Results||Applied Clinical Intelligence (ACI), a leader in data safety monitoring and regulatory submission services, has found a tool to support efficient, cost-effective drug development processes. JMP Clinical helps ACI bridge the "clinical-statistical divide."|
Mark Williams is a co-founder of a research firm that, among other services, helps pharmaceutical companies gain approval from the US Food and Drug Administration (FDA) for new drugs and devices, and he once worked for NASA. So you can take it to the bank when he says that bringing a drug to market is more complicated than rocket science.
“That old cliché should actually be, ‘It ain’t drug development,’” says Williams, Vice President and CIO of Applied Clinical Intelligence, a Pennsylvania-based contract research organization serving the pharmaceutical, biotechnology, medical device and health care industries.
It’s a very complex, time-consuming process – and, in the pharmaceutical industry, time is certainly money. Which is why Williams champions JMP Clinical.
“From my perspective, JMP Clinical has become absolutely essential to an efficient, cost-effective drug-development process,” he says.
Bridging the divide
Simply put, this visual discovery software from SAS allows for better conversations between statisticians and clinicians.
“JMP Clinical doesn’t replace the statistician or the SAS programs,” Williams emphasizes. Rather, it “bridges the clinical-statistical divide.”
By dynamically linking advanced SAS Analytics and flexible JMP graphics, JMP Clinical eases the migration from a static, document-centric world to an interactive, data-driven environment.
Its controlled set of configurable workflows provides clinicians and biostatisticians with a common platform on which to quickly reveal safety issues that must be characterized both clinically and statistically.
Williams affirms that JMP Clinical enhances the ability of clinicians to identify and explore issues that lead to better analysis questions.
Because most doctors aren’t trained as statisticians, during clinical trials they traditionally have depended on statisticians to program, analyze the data and return the results – a process that typically took up to three weeks.
On the other hand, statisticians could report what happened, but they couldn’t attach clinical significance to the event.
"Statisticians can tell you with absolute certainty that this happened, and that happened, but they don’t have the clinical training to tell you which is more important," Williams explains. "They don’t have the clinical training to know what it means when the bilirubin of a participant in a trial is skying way above normal.
"Now, with JMP Clinical, the clinician can answer many of those questions in seconds. To say that’s a really big improvement would be somewhat of an understatement."
Meanwhile, advanced analysts can access the statistics behind the graphics – a benefit not usually found in similar software tools.
Williams believes that asking the tough questions before a submission is filed reduces the risk of unpleasant surprises during regulatory reviews. JMP Clinical enables better collaboration for characterizing potential product safety issues during the pre-submission phase.
A 'leap ahead'
The FDA and pharmaceutical industry have relied on JMP for analysis since long before the advent of JMP Clinical.
"The beauty of JMP was that if you were good at manipulating data but didn’t care to learn the actual SAS programming language, that was OK," Williams says. "You could still use it to good effect.
"But from the perspective of a longtime JMP user, doing the work that I do, the leap from JMP to JMP Clinical really is very dramatic."
Right out of the box, JMP Clinical does all the data manipulation and analysis with practically no programming.
JMP Clinical is designed to make use of standards developed by the Clinical Data Interchange Standards Consortium (CDISC), the globally recognized standard for clinical analysis and reporting. It uses Study Data Tabulation Model (SDTM) data – the most mature of CDISC’s data standards – and is among the first tools to use the new Analysis Data Model (ADaM) data standard for analysis and reporting.
ACI is a CDISC-registered service provider, so JMP Clinical aligns well with the company’s paradigm. For instance, the JMP Clinical Starter menu helps users load CDISC data and follow FDA reporting guidelines in a logical sequence.
When Williams began working in drug development, he was shocked to find that there were no standards.
“Something as simple as the question, ‘What is the sex of the patient?’ could be asked in so many different ways,” Williams recalls.
"Now, based on CDISC's SDTM data, a doctor doesn’t have to know all aspects of the underlying data. As long as the data is compliant to CDISC’s standard, they’re just point-and-clicking to get their analysis results."
A critical role in patient safety
ACI uses JMP Clinical to provide graphical analyses of clinical trial safety data to members of independent data safety monitoring boards (DSMBs) charged with protecting patients in clinical trials. DSMBs analyze the accumulating safety data collected in the trials. When a safety issue emerges, the board must make recommendations to either halt the study or take other actions to protect the safety of the participants.
With its fluid movement between results for the entire trial and for specific subjects via multifaceted drill-down and swim-up capabilities, JMP Clinical allows for better and faster characterization of emerging safety issues.
A user can choose an individual subject from any graph to generate a patient profile that provides an integrated visualization of an adverse event, as well as enough information to potentially establish causality. This gives DSMBs higher confidence in the decisions and recommendations they must make.
"Any good SAS programmer can program a patient profile. But it’s all text: 'Female, 55 years old, history of cardiovascular disease, entered trial on day two …' It’s a narrative, and unless you're really good at reviewing text, nothing stands out," Williams explains.
JMP Clinical allows clinicians to quickly assess whether there are other patients with similar profiles who haven’t yet reached the threshold of concern but might be headed there. They can compare information about multiple patients by overlaying their profiles on a single timeline.
The software also uses multiple testing methods to reduce false-discovery rates, mitigating the risk of over-reporting adverse events.
Before submission, ACI uses JMP Clinical to help clients stress-test their clinical safety data and to reproduce many of the standard safety analyses that likely will be performed by FDA medical review officers.
So often, Williams says, the clinician can’t see an emerging safety event soon enough to prevent it.
"Without an interactive visualization tool, the only alternative was to flip back and forth between hundreds of pages of tables and listings to figure out what was going on," he says. "What JMP Clinical allows us to do is rapidly identify emerging safety events and characterize them."
And that can help prevent pharmaceutical companies from spending hundreds of millions of dollars for a drug that the FDA won't approve. "We want to kill off the bad ones as early as possible and move on to the high-potential drugs," Williams explains.
Using the same tools
Bottom line, JMP Clinical allows clinicians to ask better questions and identify potential problems earlier than before.
"JMP Clinical doesn’t necessarily take you where you need to go," Williams offers, "but it takes you where you think you want to go very quickly. It can save weeks of going back and forth in search of the answers you were looking for.
"But the reason I’m so excited about the product is that it acts as a bridge between clinicians – clinical scientists and doctors – and statisticians.
"Before, the clinicians had all the questions; the statisticians had all the answers. Now you have a mechanism for doctors to work with statisticians for greater clarity."
Early on, Williams says, some statisticians at the FDA were a little skeptical about placing a data visualization and analysis tool in the hands of a physician. "But when they saw that the clinicians were finding things that they weren't, they took note."
And Williams points out that an FDA medical review officer is likely using JMP Clinical to evaluate clinical trials data. He estimates that 40 percent of FDA reviewers use JMP.
"Don't you think it's a good idea to have the same tools to look at your data as the FDA is using?" he asks. "That just sort of makes sense."
Which is to say, it ain't rocket science.
JMP Clinical doesn't replace the statistician or the SAS programs. Rather, it bridges the clinical-statistical divide.
Vice President and CIO
Applied Clinical Intelligence