<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=1843390119021846&amp;ev=PageView&amp;noscript=1">

Managing Your Data Supply Chain: A Holistic Approach to Data Quality

March 22, 2018
By Bruce Leyton
Original Article in Healthcare Analytics News

Everyone talks about the importance of data quality—from software vendors to end users such as hospitals and health systems. But before we talk about why quality matters and how to manage it, shouldn’t we be sure we know what it means? For the purposes of this article, let’s agree that, put simply, quality means the expectation that a product will work as promised, for the use intended, without defect or failure. Inherent in this definition—for all products, not just software—is the idea of meeting the expectations of your customers. When those expectations are not met, consequences can be dire.

Healthcare Data Challenges

As organizations move to align to the new, data-driven paradigm, they are facing challenges within their own data supply chain. The issue has become particularly apparent within healthcare. The ever-increasing complexity of the healthcare industry and healthcare/patient data is the perfect storm for a quality crisis. But more than in any other industry, the cost of not having a strategy to maintain data quality can be severe.

Poor quality data in hospitals can lead to duplicate tests, billing and medical errors, and wasted marketing resources, among other problems. Hospitals operating without a strong data management solution have the very real potential to negatively affect patient outcomes; studies have shown that 4 of every 100 cases involving duplicate records have a negative impact on care, and that more than 100,000 people die annually because of identity or “wrong patient” errors. The financial risk is great as well; a study of one hospital in Texas showed that duplicate records made up 22% (250,000) of its records, and that each of those records cost the hospital more than $96.

The Data “Supply Chain”

To understand how data quality can break down, we can take a look at supply chains in non-digital industries. Let’s look, for example, at the furniture and home goods store IKEA. The retailer has more than 355 locations in 29 countries, and it has a reputation among its customers for good, consistent customer service and quality. Regardless of country or individual store locations, IKEA customers know what to expect. Part of that consistency, which has helped build a loyal customer base, are the delicious and iconic Swedish Meatballs IKEA serves in its restaurant area. All told, the restaurant sales don’t account for a large portion of the company’s profit margins, but they contribute to the customer experience. They stand out as a differentiator that may entice customers to visit an IKEA store rather than its competitors.

With such a strong reputation and loyal fan base, the announcement in 2013 that some of the meatballs being sold in IKEA’s European locations contained horse meat took everyone by surprise. IKEA had not knowingly cut corners or misled its customers; the issue turned out to be a supply chain problem that led to a quality problem, which quickly turned into a public relations problem. IKEA recovered nicely by voluntarily recalling any products that may have been affected, and then went above and beyond to retain customer loyalty by announcing a “farm to fork” initiative, which involved audits along every step of the supply chain journey to ensure quality, as well as DNA testing for each individual batch of meatballs.

IKEA learned a valuable lesson from this experience that is applicable to many other industries, software included. The lesson is, of course, that your products are only as good as the quality of the materials that make up the supply chain. In the software industry, then, this means that the quality of your product relies heavily on the quality of the data that will be ground up and put into it—data, too, need “DNA testing” for each batch. In either case, it comes down to trust and reliability. If you think quality data isn’t your problem, think again.

Airline’s Dreams Are Grounded

Similar examples of poor quality management abound in other industries. In January 2013, a Boeing 787 Dreamliner airplane made an emergency landing in Japan due to issues with the plane’s battery. This eventually led to investigation by the Federal Aviation Administration (FAA), which led to the discovery of battery failures that could result in “the potential for fire in the electrical compartment.” The FAA made the decision to subsequently ground all US Dreamliners until all repairs to existing planes—and those still in production—could be made.

The direct and indirect costs associated with this quality error were staggering. It was estimated that planes simply not being in the air would cost Boeing $125 million per month—not including the cost of repairs. Further, several airlines sought compensation from Boeing to offset their own lost revenue from the crisis. And other countries declined to approve the Dreamliner for flight until the planes in the US were repaired and then approved by the FAA, which resulted in even more lost revenue.

So, where did the break in quality happen? “The 787 was designed to be the most fuel-efficient large commercial jet ever built. That has meant saving weight wherever possible,” reported CBS News at the time. “The almost 1.5 megawatts of power the Dreamliner generates also requires a lot of battery space for storage. In order to save weight, Boeing's designers used lithium-ion batteries that weigh about half as much as the usual nickel-metal hydride batteries…They have to figure out, ‘Do you stick with this technology or take a weight penalty in adopting the older technology?’ That's probably what they are wrestling with now.” Additionally, “Boeing designed the 787 to be built in a modular fashion, with subcontractors supplying completed subassemblies—resulting in a simpler assembly line and faster production…This is the most outsourced plane ever built in the history of aviation," CBS reported.

Perspective is Everything

There are lessons to be learned from both Boeing and IKEA. Boeing learned that cutting costs and leaving supply chain details to “someone else,” and focusing on a benefit in one area (reduced weight) without considering the consequences to other areas, can easily leave a company vulnerable to quality errors. With data management, as well, it’s crucial to remember that quality must be managed from the beginning of the process to the end, and everywhere in between, even if one individual department doesn’t believe it will be impacted.

From the IKEA public relations crisis, we learned that quality can often be a matter of perspective. In some parts of Europe, it’s common to eat horse meat, so what was taboo and a scandal in the US could have been seen as an honest mistake—or even no big deal—in other parts of the world. In fact, the error could have just as easily been a misunderstanding of cultural differences as one of poor supply chain management and logistics.

In a similar way, when one department in an organization like a large health system or payer collects data, staffers may be aware only of their own “cultural perspective,” so to speak. In other words, different departments often collect patient data in different ways, in different systems, for different purposes, and may not be aware of how that information will be interpreted and used elsewhere. If each department is siloed in this way, that missing piece of perspective means each department may not apply the appropriate controls, or “supply chain management,” resulting in a quality breakdown. This is why having a holistic approach is mission-critical in healthcare.

A Holistic Approach to Data Quality

An article from Forbes published in June 2017, titled, “The Importance Of Data Quality -- Good, Bad Or Ugly,” cites several benefits of good data quality, including:

  • Decision making: The better the data quality, the more confidence users have in the outputs they produce, lowering risk in outcomes and increasing efficiency.
  • Productivity: Good-quality data allow staff to be more productive. Instead of spending time validating and fixing data errors, they can focus on their core mission.
  • Compliance: In industries where regulations govern relationships or trade with certain customers, especially in finance, maintaining good-quality data can be the difference between compliance and millions of dollars in fines.

The solution, managing the data supply chain, lies with not just the quality of the data, or the software solution being used, or the hospital’s IT team. The industry is slowly awakening to the idea that data quality is a holistic “supply chain problem.” Much like traditional supply chain management, the data supply chain is a highly sophisticated and interconnected series of interactions with key touchpoints along the way. And it is only with this broader, more strategic perspective that we can begin to manage quality, value, cost, and risk more effectively.

Topics: In the News, Data Quality

Sign up for all the latest from VisionWare