Shared Best Practice Data: The Key To Error Reduction

By Dr. Viet Nguyen, Chief Medical Officer at Lockheed Martin Information Systems & Global Solutions

A recent Johns Hopkins University study citing medical errors as the third-leading cause of death in the United States has been met with significant controversy in the press. While the mixed reception is understandable, the study has initiated an important dialog. Whether or not medical errors fall immediately behind heart disease and cancer as a cause of death, there are a significant number of medical errors that can — and should — be prevented.

According to the study, medical errors are defined as unintended acts, failures to complete planned actions or the use of improper care plans. At their core, the majority of medical errors are the result of deviation from medical best practices. As a result, combatting this issue will require us to share best practices nationally — in the same way clinicians share research and innovation about illnesses — and to leverage them at the point of clinical decision-making.

Why Medical Errors Occur
Most episodes of care start with an event, typically an accident or an illness. Clinicians gather information about the patient, analyze the information, and make an informed decision about an intervention. They might choose a diagnostic intervention to gather more information, or select a treatment involving a medication, therapy or surgery. The process of information gathering, analysis, and intervention continues throughout the care journey, as the patient’s status changes.

Errors can occur in the information-gathering stage if important clinical facts about the patient aren’t collected. They can also occur during the analysis stage if clinical information is overlooked or misinterpreted. Moreover, errors can occur during diagnostic intervention — as with the ordering of unnecessary radiology tests — or during a treatment if important steps are omitted or completed improperly or on an untimely basis. Preventing these errors will require a three-pronged approach that hinges on clinical decision support (CDS).

3 Steps Toward A More Error-Free Health System
The following steps will help enable healthcare organizations to reduce medical errors throughout the industry:

• Share uniform best practices: Healthcare organizations should aim to promote uniform best practices in all care settings. Leading provider organizations often prepare best-practice guidelines for their clinicians and deploy them in paper form or through an automated CDS system. Yet, while these efforts are helpful, knowledge still tends to remain siloed within care settings. If one hospital’s guidelines are less complete than another’s, for example, a knowledge gap exists between the two facilities. One way to close the gaps is through a shared, national database of industry best practices, continuously updated with new information.

• Create a best practice checklist: Providers can leverage these shared best practices to create clinical checklists that promote adherence to guidelines. Renowned American surgeon, writer and public health researcher Atul Gawande makes the case for a simple to-do list approach in The Checklist Manifesto. Gawande notes when The Johns Hopkins Hospital ICU implemented a five-point checklist in 2001, it prevented an estimated 43 central line infections and eight deaths over 27 months. When this same checklist was tested in multiple Michigan ICUs, it reduced infections by 66 percent within three months and likely prevented more than 1,500 deaths in a year and a half.

• Implement an industry-wide standard: Presenting uniform best-practice guidance automatically to clinicians within normal workflows, in conjunction with EHR documentation, is the final step in reducing medical errors. CDS systems can present useful information to clinicians at the appropriate point in time and verify checklist completion, beginning at the start of each care episode. Although this step represents an IT challenge, emerging technologies hold promise for its accomplishment.

Delivering The Solution Through IT
The big health informatics challenge in all of this boils down to one question: How do we make decision support universal? One group of provider organizations, The Health Services Platform Consortium, is attempting to answer the question by developing a technical framework for knowledge representation in computable form that can be implemented consistently across all EHRs. Its efforts focus on Fast Healthcare Interoperability Resources (FHIR®), which is picking up momentum as a replacement for traditional HL7 EHR interfaces.

FHIR holds the potential for eliminating the need to customize CDS algorithms for each EHR vendor to address differences in data models, coding and logic. The FHIR data model supports uniformity of terms through a common computer language, allowing hospitals to share newly developed best practices as computable knowledge.

Some CDS vendors are hard at work improving their own solutions along these lines, and their efforts are important as well. As with all health IT breakthroughs, it’s difficult to predict which approach will eventually lead the market — and provide true results. The important thing is that we escalate the promotion of uniform, evidence-based best practices on a national scale to reduce medical errors, improve patient safety and achieve a more reliable healthcare system as a whole.

About The Author
Viet Nguyen, MD, is a board certified clinical informaticist and Chief Medical Officer for Lockheed Martin Information Systems & Global Solutions, working toward improving continuity of care and patient safety through integrated technology. He works as an informaticist with federal agencies and standards development organizations, supporting terminology and interoperability efforts. He was trained in internal medicine and pediatrics at the University of Cincinnati and Cincinnati Children’s Medical Center, and at the University of Utah as a medical informatics fellow.