Nationwide and international benchmarks for quality, safety and patient outcomes must be created to generate consistency for imaging specialties and set a standard of care, according to a column from American Journal of Roentgenology (AJR) Editor-in-Chief Thomas H. Berquist, MD, published in the March issue of AJR.

While there are multiple registries that provide benchmarks for practice patterns and quality documentation, such as the American College of Radiology’s National Radiology Data Registry, he said that developing consistent research approaches and uniform data utilization has been "difficult."

Consistency in quality improvement research is hampered by variables in practice and patient populations, according to Berquist. These variables include size of population studied, patient sex and age, severity of disease, socioeconomic factors and variations in approaches to quality research or measurement.

“Despite these challenges, we must develop common evaluation frameworks that are transferable at local, regional, and national levels,” wrote Berquist. A starting point for successful quality improvement requires leadership support and a definition of the customer. He explained that the customer base extends beyond patients and their families.

“Although the patient is the most important, we must also deal with referring physicians and other providers, CEOs, the public, the press, regulatory agencies, third-party payors, industry, vendors, academic institutions and medical schools.”

Berquist went on to summarize suggested steps for process consistency, including:

• Identify the problem;
• Collect and analyze data related to the issue;
• Determine contributing factors to the problem;
• Develop solutions and metrics for each contributing process issue;
• Perform the evaluation with the new solutions and metrics; and
• Evaluate the data and standardize the new process or modify and repeat.

By using common approaches and data systems, Berquist said specific projects can be translated into national databases or benchmarks. “The challenges are many, and developing consistent and optimal research approaches will take time. However, we will and must strive to use the massive data available in a way that continuously improves patient care, with documentation that improvements have affected patient outcomes.”