There is a wide variety of criteria to be considered. Price matters, but so does the performance of the product or service under consideration. When it comes to sourcing in healthcare, the latter can be particularly challenging.

Sourcing for the human body is more complicated than in industries such as automotive or electronics, where products must meet pre-determined specifications. Everyone’s anatomy is different, and factors such as genetics, lifestyle, and other medical conditions play a role as to which products work best for which patients.

That’s the thinking behind the federal government’s Precision Medicine Initiative, which seeks to match treatments based on a patient’s genome, environment and lifestyle. Much of the work to date has been in the field of pharmagenomics – understanding how a person’s genes impact the safety and effectiveness of different drugs. Pharmaceutical researchers work under regulations and commercial requirements that demand documentation in electronic health records (EHRs), and claims forms of standard identifiers for medicines administered to patients.

The same opportunity is now becoming a reality for medical devices. In 2013, the U.S. Food and Drug Administration (FDA) began requiring manufacturers to assign and label products with unique device identifiers (UDIs). Corresponding federal regulations from other government agencies now require EHRs to hold a list of a patient’s implantable devices using the UDIs, and for hospitals to share that list with other providers caring for the same patient.

There’s still a lot of work to be done before all medical devices bear UDIs. Providers are routinely capturing the information, but the existence of UDIs lays the groundwork to better understand how products are performing in routine clinical practice.

Generating real world evidence on medical devices is particularly important, given differences in how the FDA evaluates and approves medical devices for sale in the U.S. compared with pharmaceuticals.

First, clinical trials for implantable medical devices typically involve far fewer patients than those for pharmaceuticals. Drug trials are conducted in phases, with thousands of patients in the later stages. For implantable medical devices, it is neither practical nor appropriate to have large numbers of patients undergo surgical implantation as part of the trials. In recent years, between 100 and 200 patients have been deemed sufficient for many of the products evaluated and approved under the FDA’s most stringent pathway, the Pre-Market Approval (PMA).

Second, only one percent of medical devices on the market are evaluated using the PMA. That’s because most new medical devices are iterations of older devices and are approved under supplements to the original PMAs. Between 1979 and 2012, more than 5,800 supplements were submitted for approval based on just 77 original PMAs.

Third, the FDA Modernization Act of 1997 stipulates that the FDA must use the least burdensome pathway for manufacturers to provide supporting evidence for documentation. Only one of six different supplement pathways requires clinical data based on human testing. In some cases, the FDA will require additional testing in “real-world” clinical practice, but that is still the exception and not the rule.

The underlying rationale is to minimize regulatory barriers to bringing new medical device innovation to market. Still, there remains an underlying need to understand how these products are performing in the real world, and increasingly for which kinds of patients. That’s why the FDA sees the capture of UDIs in EHRs, registries and claims as key to more effective post-market surveillance. This can help more quickly to identify any serious problems that did not come to light in the evaluation process, and which could lead to product recalls. In other cases, real-world data about the use of medical devices can support their approval for new uses. That’s because physicians in the real world will sometimes use a medical device for an indication beyond what the FDA has approved.

As more data is collected on the use of specific devices in EHRs and registries, researchers will have a robust enough database to understand the performance of medical devices by type of patient – e.g., those of a particular gender, age group, or with additional medical conditions such as diabetes. This is particularly important given the wide range of devices and price ranges in a particular category.

Real-world data can help patients and their physicians make better decisions on whether the additional cost for one product versus another is warranted based on its performance for specific patient types. In some cases, the decisions are relatively easy to make; an 85-year-old woman likely does not need the high-performance knee or hip implant intended for a world-class athlete. But other decisions may be harder. For example, some hernia patients are at greater risk for complications and may warrant use of biologic mesh, which is considerably more expensive than synthetic mesh. With real-world performance data, these kinds of sourcing decisions can be made based on what’s best for a specific patient.

There has been much written about waste in healthcare, in the form of unnecessary procedures or unwarranted use of resources. Use of standards and commonplace supply-chain practices such as barcode scanning can go a long way toward minimizing that waste, by making sure we are matching the right supply or service to meet the specific patient’s needs. That’s what delivers value in healthcare.

Karen Conway is vice president of healthcare value for GHX.