It’s easy to give a sharp salute to "quality." What modern research organization would oppose it?
Quality refrigerators, quality cars, quality overnight packages can be defined easily. Quality in a clinical trial is a bit murkier. By definition, every trial is one of a kind, with clinical, operational, regulatory and technological aspects that are unique. Having said that, quality research is a beguiling topic in an age of defect-riddled banks and oil wells.
At this year’s annual Drug Information Association (DIA) meeting, Thomas Grundstrom of Quintiles made a persuasive argument for quality as a core practice, not some bit of marketing verbiage. Grundstrom is VP of global data solutions at the large contract research organization (CRO), and he spoke with a level of candor and directness that can be rare at jargon-filled conferences where old terminological wine is decanted into new bottles.
Grundstrom said all sponsor and contract research organization (CRO) communities are seeking hard data to be used in operational decisions. “Most of us want more than anecdotal stories," he said. “You want to be able to measure it systematically along the way.” He advanced the idea of not just reacting to the trial's progress, but monitoring and anticipating events of interest to sponsors and regulators.
In theory, he said, a CRO should be uniquely positioned not only to say how a trial is proceeding, but to determine what should happen in the future. Key performance indicators (KPI), if chosen judiciously, should be actionable. Said Grundstrom: “A CRO is in a relatively unique position to be able to look at these KPI and not simply look at bad things that have happened, but be proactive and look at when they might happen. You can build those signals in.”
Which is where quality comes in. Contacted after the DIA event, Grundstrom related that for his firm, quality is centered around three elements: patient safety; integrity with regulations; and customer expectations. To some degree, he related, all serious firms sensitive to the complexities of clinical research share such values. Quintiles isn't claiming it invented them.
The company does believe, however, that in exploring the interplay of its processes, technologies and people, it may be able to deliver quality in ways that set it apart. "We have embarked on a journey that takes quality to the next level and becomes a real differentiator," he told ClinPage.
Silos or compartmentalized knowledge in CROs and sponsors, Grundstom says, impair the ability to ensure quality. Data managers don't have the ability to comprehend what's happening with central labs; the pharmacologists are uncomfortable reaching out to biostatisticians; senior managers await reports that never arrive; sites are not told why an edit check is triggered repeatedly.
All of those individuals may be doing their jobs, but not seeing trends that affect patients. "They may be totally different organizations, totally not talking to each other. That creates problems," Grundstrom says. So like many competitors, Quintiles is working on ways to connect the silos.
For now, it's also advocating the usage of KPIs such as biomarkers or safety signals to support quality. The idea is to not just respond to scientific fire alarms, but to undertake fire prevention. “Most of what we find here is very reactionary,” he said of the current clinical trial environment. “We have become such specialists that nobody sees the complete patient any more.”
At DIA, Grundstrom suggested building informational dashboards so that a sharp change in the incidence of queries is immediately noticed, with new up- or down-ticks becoming triggers for corrective action. All too often, he said, clinical trials unfold in a rear-view mirror.
In a similar vein, Grundstrom's presentation explored the concept of signal detection not in a pharmacovigilance or drug safety sense, but as it might apply to someone designing airline cockpits, allowing them to change direction in real time. Signal detection of that sort in a clinical trial, he suggested, could help establish that an edit check in an electronic data capture (EDC) system was working; that additional training of staff was needed; that cultural biases are leading to misunderstandings of the protocol.
If applied to future projects, such lessons could lead to better case report form design and other factors that prevent the problems from cropping up in the first place. Just setting the edit check, in an environment obsessed with quality, might not be enough. Knowing that the edit check is being triggered half as much as expected or 500 percent more than expected could lead to a reassessment of it, he suggested.
In some cases, KPIs could even lead to stopping the trial. In the case of Hy's Law and drug-induced liver injury (DILI), Grundstrom said, even a few prognostic incidents are likely to doom the drug’s chances in the market. “You see 2-3 [of these],” Grundstrom said gravely, “you may as well stop the compound. Even when you’re as far downstream as a CRO, sometimes you’re in a unique position to do this kind of signal detection.”
A few of his remarks hinted that Quintiles is incorporating KPIs into its projects as a standard practice. “The industry is a bit tired of saying, we’re going to pay you for every hour you work,” he said. “I don’t blame them.” He provided few details on that front, but impllied that some life science organizations could be exploring more comprehensive contracts that connect quality research and the CRO's compensation.
Editor's note: Here's another story about the software that Quintiles is developing to support quality.d9A2t49mkex