Edgar L. Mounib
"We must all hang together, or most assuredly we shall all hang separately," said Benjamin Franklin at the signing of America's Declaration of Independence. The same could perhaps be said today of two huge but troubled industries: pharmaceuticals and health care.
The mandate facing each industry is clear: The pharmaceutical industry needs to develop safer and more efficacious treatments, and the healthcare industry needs to consistently deliver evidence-based, high-quality care. The realization is slowly dawning that each cannot attain its goals alone. But whether the two industries will acknowledge, and take, the next logical step-toward collaborating with each other and sharing their data-remains to be seen.
Meanwhile, the patient waits expectantly.
The problems besetting the pharmaceutical and healthcare industries are all too apparent.
On the pharmaceutical side, it takes roughly 13 years to develop a drug, with 46 percent of all drugs failing in Phase III trials. Worse, a full half of all drugs that make it to market are ineffective for the patients for whom they prescribed. Meanwhile, the healthcare industry-constituting 16.2 percent of the U.S. gross domestic product-is staggering under combined impact of forces such as globalization, consumerism, demographic shifts, rising costs to treat diseases and conditions, and the proliferation of new technologies and treatments. For example, chronic diseases now account for 60 percent of all deaths in developed countries and consume up to 75 percent of their healthcare resources.
Caught in a storm, these two industries have good reason to turn to each other.
Devising ways to share information more comprehensively is the critical issue in developing this collaboration. Technology is a major part of the solution. But it also is a big part of the problem.
Consider what happens when a patient visits his or her physician. Data are created. And depending on the physician's location, those data could reside in many different sites, such as physician offices, primary care sites, academic medical centers, hospitals, specialty clinics, ambulatory surgery centers, and pharmacies. But having data reside in many places is not the problem-after all, having broad access to data is desirable. The problem is that the data typically are stored in different media and formats, and thus cannot flow easily, if at all, between those media and formats.
If the data could be securely collected, processed, and disseminated in a timely fashion but without compromising patient privacy, then everyone would benefit. Medical researchers could investigate more complex issues. Pharmaceutical and biotech companies could improve their understanding of disease mechanisms and clinical outcomes, and enjoy improved access to candidates deemed good for clinical trials. Regulators could more accurately assess the drugs they must review. Providers could more consistently deliver more evidence-based, standardized health care. Governments, health insurers, and employers could make more informed decisions about how to deploy limited resources. And in the end, most important, patients would receive better care.
So if both the problem and the desired ends are clear, then what steps should we take to solve the problem and achieve those ends?
As the foundation, we look first for technical solutions. The good news is that the requisite technologies exist. Consider two approaches, for example.
Hub and spoke. One of the more efficient ways to link a wide range of systems and tools is the "hub-and-spoke" structure. The information from each database at the end of the spokes is stripped of sensitive details, replicated to the "hub," formatted using common data standards, amalgamated, and categorized. A service-oriented architecture (SOA) can then be used to deliver an agreed-upon set of "services" for accessing and using the data. SOA supports breaking down applications into common, repeatable business tasks or "services," enabling them to communicate with each other independently of the computing platforms a business and its partners are using, or the locations in which the applications are based. This approach has two main advantages. One is that it enables users to access the data independent of the computing platforms on which the systems are running. The second is that it gives them the flexibility to build the range of services they need around a "single version" of the data.
Trusted virtual domains (TVDs). Although TVDs are in their infancy, this solution offers the potential for a more secure computing environment, which is vital for protecting private data such as patient records. In broad terms, TVDs use "virtualization" and "overlay" techniques to form a "protective wrapper" around each of the computing entities used to perform a service, regardless of the hardware, software, or network configuration involved. This design is meant to prevent accidental or malicious damage from spreading to other computing entities. It also provides a way of identifying different computing entities, authenticating them, and auditing communications among them.
Adopting shared standards is another means to facilitate data sharing between the pharmaceutical and healthcare industries. Standardized data, such as medical vocabularies and dictionaries, can improve data consistency, enable cross-trial analysis, provide an easier way to transfer data with external partners, and help facilitate regulatory reviews. Various initiatives are currently afoot to develop and promote open standards across the two industries. The Clinical Data Interchange Standards Consortium (CDISC), for example, has created standards for exchanging animal, human, and clinical data. The Health Level Seven (HL7) is developing a number of standards for hospital data. And the Biomedical Research Integrated Domain Group (BRIDG) has developed an overarching "semantic" model for harmonizing different standards and bridging the gap between biomedical research and healthcare.
Solution: Stepping Across Industry Boundaries
To date, the pharmaceutical and healthcare industries have focused on data convergence internally (for example, within or between departments). This is a necessary first step. But now the convergence must cross industry lines. Companies must motivate their employees to look not only beyond group or divisional lines, but beyond corporate and industry boundaries to find innovative solutions to the nation's medical challenges. The will to engage in this effort must flow from the top of the organization and permeate throughout. As noted, the technologies to collaborate exist and will only improve, but if they remain unused, they will amount to nothing.
Industry Readiness for Collaboration
It is reasonable to assume that the need for greater collaboration between the pharmaceutical and healthcare industries is largely accepted in both industries. But questions remain: To what extent do pharmaceutical and healthcare companies, deep in their bones, feel the need to collaborate? To what extent does such collaboration already exist? What are the barriers? And what are the palpable benefits?
We sought to answer these questions by interviewing 27 leaders across disciplines in AstraZeneca, the Karolinska Institute, and the Karolinska University Hospital. What we learned is that the scope for convergence is considerable.
First, the vast majority of respondents agreed that there is very little cross-industry collaboration today and that it is essential-to get data faster, to save time and money, and ultimately to do the better research incidental to better patient care. Second, the respondents largely agreed that better technology, and widely accepted technology and data standards, would foster that collaboration. Third, they acknowledged that the current barriers are not only technical (in the sense that the existing technology has not been sufficiently deployed) but also cultural, ethical, and organizational, and that, taken together, these barriers are disabling.
With so much at stake, the pharmaceutical and healthcare industries have every reason to collaborate. So how should they proceed in creating a "patient-centric information chain"?
The current circumstances suggest that both industries must, as a first step, recognize that achieving this collaboration will require huge cultural shifts, that all the interested parties must be committed to it, and that these parties must establish a shared governance structure, with each party having an assigned role and a clear understanding of the benefits it can expect from its cooperation. In this sense, the governance structure would be like a participatory democracy.
As a second step, the interested parties in both industries must collaborate in developing a comprehensive communications plan to secure internal and external support. The third step will be to establish a code of conduct and to fill in any gaps in essential skills. This largely technological step in which the two industries must evaluate the readiness of their technologies, eliminate redundancies or inconsistencies, and safeguard their data-using a trusted third party if necessary.
The final step will be to measure performance-because it is only by tracking their performance that the pharmaceuticals and healthcare industries can achieve their shared goals.
Meanwhile, the patients are waiting.
Heather Fraser is global life sciences and pharmaceutical lead, IBM Institute for Business Value, United Kingdom (firstname.lastname@example.org).
Edgar L. Mounib is healthcare lead, IBM Institute for Business Value, Washington, D.C. (email@example.com).
Sarah Payne is a managing consultant, life sciences and pharmaceutical team, IBM Global Business Services, United Kingdom (firstname.lastname@example.org).
Publication Date: Saturday, December 01, 2007