In the unofficial design manual for clinical trial software programs, there are numerous unwritten rules but only a single severe commandment. Any new solution must have exactly 52,009 powerful and essential features. Why? So that experts and consultants in the pharmaceutical industry can stand back and exclaim, "Wow! You have precisely the 52,009 features that we require you to have." A San Francisco entrepreneur named Jae Chung is planning to to disregard that commandment with a company called GoBalto.
Chung is no dreamy software architect. A 40-something ex-McKinsey consultant, he founded and sold Celltrion, a Korean biotechnology firm that specializes in contract manufacturing and generics. Chung has sufficient resources to keep GoBalto in the hunt against traditional, established clinical trial management system (CTMS) and document management suppliers that include Oracle, Parexel, SAS, Study Manager, Clinical Force, BioClinica, Intralinks and Mission3.
Chung becomes uneasy when his system is compared to a traditional, legacy CTMS. It's only designed to do one task well. The other 52,008 mandated features are nowhere to be seen. His solution is targeted just to initiating sites and getting them onboard with the proper paperwork.
In looking at his competition, it's fair to say, Chung was not intimidated. He saw a gap in what clinical trial professionals really needed. "Where is the innovation?" he asks. "Where is the tool to help the Joe Schmo study manager? Where are the solutions for these guys? We are here to shake things up. I want this company to be the clinical software company."
Clinical Web 2.0
He started GoBalto in 2008. With 12 employees, the company already offers a sort of biomedical matchmaking or Facebook-like service called "Directory."
And this week GoBalto announced a forthcoming module called "Tracker." Getting up and running on the system is designed to be fast, and typically includes uploading a standard spreadsheet to the GoBalto website. That Excel spreadsheet is the seed from which a large percentage of industry-sponsored studies grow.
Chung says the GoBalto Tracker manages the needed regulatory documents and project milestones; offers metrics and reports on site initiation; and generally expedites the study start process. It could even allow investigators to sign documents electronically, eliminating the need to manage umpteen faxes. "It's a lot more user-friendly and easy to use," Chung says. "A lot of the CTMS systems don't provide core functionality in getting sites up and running."
Chung and his team did 150 interviews to learn what the industry wanted in a study startup program. In that context, he remains puzzled why the internet and technology have not had any dramatic effects on clinical trial productivity. Says Chung: "You look at all these other industries, and how the web has transformed them. I'm scratching my head. Why hasn't this space innovated?"
For inspiration, Chung and his team drew upon Apple and 37signals, a web software firm that has developed potent, beautifully designed web applications. 37signals competes successfully against Microsoft, offering web-based document and project management applications, all with modest monthly fees. The Chicago company raises a trite concept like "ease of use" to an art that even Apple struggles to match, given several sleek Mac programs that can't actually manage the growing numbers of apps, songs, photos and contacts proliferating on our computers.
37signals has made a rhetorical crusade out of removing functionality as it designs its apps, never including some wonderful software feature that a random guy in the accounting department was rumored to have insisted upon. Chung shares that minimalist philosophy, saying he seeks to provide only what is truly essential in a study startup application.
Chung does understand how committee-driven corporations scrupulously compile exhaustive, 51-slide Powerpoint decks describing the essential features for industry software. "Any monkey can go and put up an application with multiple tabs, links, graphs and charts. But when you come to the screen, it's like, 'What the heck am I supposed to do?' We came up with a very simple, pared-down solution that goes to the heart of the matter," says Chung.
Like 37signals, Chung is building his system on Ruby On Rails, an open source development platform that is optimized for speedy code development. (In certain circles, Ruby On Rails has almost as many fans as Apple.) Smaller software firms are the biggest admirers of RoR, but it's powering mega sites that include Groupon and Twitter. So there is no question that Rails is robust enough to handle the modest numbers of users of clinical trial systems.
Chung is puzzled by the industry's complacency about not using more secure systems for so much of the work of clinical trials. He notes that most email floats around the web and is easily intercepted, requiring no decryption. Says Chung: "Tell me about what you do to secure email. It's part of the naiveté of the average user on the internet. They're really not educated on the internet, on security. We're more secure than email."
Chung is betting that small and mid-sized pharma firms will grasp the value of a sleeker, faster way to enroll sites. Biotech outfits just getting started have neither the time, the desire nor the funds for a million-dollar CTMS that will not be ready for a year. Even then, some traditional CTMS strain to provide fresh, actionable data not for technical reasons, but simply because of the sheer size and multi-layered corporate structure that accompany any mid-sized or large life science organization.
When the program debuts for all users in a few months, Chung says he'll be able to offer several tiers of pricing, all based on the complexity and size of a study. "It's going to be significantly cheaper," he says. "I am talking about orders of magnitude."
The reports and metrics generated by his system, Chung believes, will be more intelligible and attractive than anything from Microsoft or a company of that girth. Users will be able to export data into PDF and Microsoft Office formats. Reports will appear within a web browser window, and require no other software to view or interpret. Some of the reports will contain anonymized benchmarking information that will help everyone involved in a trial see where they stand relative to the rest of the user community on the GoBalto site.
Facilitating such insights and comparisons, Chung hopes, will foster a bit of friendly competition to get the work done faster. So GoBalto might display (with no identifying information) the collective averages of real-world performance about, say, how long it took to finalize a contract in India or bring on a clinical site in Berlin. Getting enough users to have a statistically meaningful quantity of data will be a challenge for the firm during its launch period.
Once he has a critical mass of users, Chung is betting that providing visibility into study startup metrics could foster improvement and efficiency gains. If a particular investigative site sees that the global average for a particular task is 28 days, compared to its own completion time of 68 days, there will be a natural opportunity for introspection.
"It's human nature," Chung says. "How can you not ask the question, 'What the hell is going on here?' " But whether all sponsors and contract research organizations (CRO) will want their projects included in such averages or or find a way to opt out is unknown at this stage.
Chung acknowledges the pharmaceutical industry is not known for accelerated adoption of new types of software. Good ideas have withered and died while awaiting the verdicts from big pharma's years-long evaluation and purchasing cycles. But Chung is confident that something called the internet is now sufficiently familiar, and the problems with site initiation sufficiently obvious, that a new approach can gain traction. "Industries can be changed if there are people behind them who believe in positive change," he says.d9A2t49mkex