A Case Study: Improving the Availability of Clinical History


David Larson, MD, MBA, radiology department chair at Stanford and a known quality improvement expert in radiology, shared a series of clinical histories that commonly cause heartburn in every radiology department and imaging center in the nation: pain, cough, football injury, abdominal pain.

In a presentation at the RSNA meeting on Monday, Nov. 28, one of three comprising “Mission Critical: How to Increase Your Value by Mastering the Intersection of Quality Improvement and Informatics,” Larson shared a case study that elucidated everything involved in a recent quality improvement project he spearheaded with C. Matt Hawkins, MD, to improve the availability of clinical history at Cincinnati Children’s hospital, Larson’s previous post.

“This is what is often seen in radiology as the impossible thing,” he noted, acknowledging the role of informatics, but focusing on the change management aspects of the project. “Yes, [informatics] is critical, it’s necessary, but it is not nearly sufficient.”

To illustrate the problem, he flashed a radiograph of a foot on the screen and asked the audience to identify the problem based on the clinical history provided: pain. When there were no takers, he provided the actual clinical history: “kicked bedpost two hours ago, pain in the pinky toe,” and pointed out a small fracture at the base of the fifth middle phalanx.

“If we have an adequate history it gives us a better shot at making the right diagnosis,” he began, outlining a case study in organizational change in eight steps.

No. 1: Define the problem. The team at Cincy Children’s Hospital began by defining the problem: An adequate patient history—defined as what, when, and where the symptom or accident occurred—improves patient care by maximizing the chance of an accurate diagnosis, and they weren’t always getting that. They measured the extent of the problem and found that they were getting an adequate clinical history 38 percent of the time.

No. 2: Develop interventions. “We decided we would like the technologist to augment the history,” Larson explained.  “At Cincy, we have the luxury of having parents, so almost always there is a caregiver in the room so let’s take advantage of that.”  The technologist would ask for their comments and description of problem, and type it into the medical record.

No. 3: Anticipate and prepare. “When we change something, we mess with someone’s life, and in this case, it was the technologist,” Larson noted. “They are going to have to acquire more history than we already have. How will they react?”  You need to anticipate problems and develop a strategy to connect with those who will be on the frontlines of change. The team at Cincy felt there was no more powerful path than to connect around the shared and noble mission of providing the best possible care for their pediatric patients.

No. 4: Get support. Larson recommends getting buy-in not only from those who supervise those most directly affected by the change, but the next level above as well. “We know someone is going to go in and complain, and you have a few milliseconds to influence this. If that person’s supervisor says, ‘I know it’s tough, but we have to do it,’ then you’ve won.”

Beyond the standard organizational hierarchy, pay attention also to the informal organizational networks described in this article that appeared in the Harvard Business Review, describing the various agents who can make or break your project, notably endorsers, fencesitters, resisters, and influencers.

“We use both formal and informal hierarchies,” Larson said. “I know it may sound Machiavellian, but we are trying to get the organization going in the right direction, to do the right thing. We focus on the fencesitters, and these are the people you want to win over, because once they tip, then the whole organization tips. We enlist the influencers to help, by strategically getting them on board early to help.”

No. 5: Make the case. You prepare to launch your project by making your case. “Have respectful and sincere dialogue with technologists as to why getting a clinical history is important,” Larson suggested, by explaining that they set the radiologists up for success, making them true partners in the care provided.

“If you are sincere, then that is what people need to hear,” Larson said. “Do it in person, go out and talk with them, connect with the heart, and express appreciation. Listen to feedback and change your strategy based on that.”

No. 6: Clarify expectations. The implementers must know exactly what is expected of them.  The team at Cincy explained the “what, where, and when” needed for each exam, making liberal use of examples that showed what the clinician provided, and what was needed from technologist to augment the history.

This was a time of support, answering questions, using encouragement and patience, and thanking people for their effort. Once the needle started to move toward adoption, the project was officially launched.

No. 7: General feedback. General feedback was provided by sharing performance data at each site, providing and some tricks and tips. This is training, as differentiated from individual coaching. “They are all making the same mistakes at first and then by the end they are all making different mistakes,” Larson said. “Then you have to provide individualized feedback.”

No. 8: Provide specific feedback. At Cincy, cooperation took a turn for the worst after implementation. “We had a revolt,” Larson noted. “People said, ‘We aren’t going to do this.’” The Cincy team moved into coaching mode by providing individualized feedback to each technologist, and Larson cautioned: “This is going to hurt. Tell them why and what you are doing. You have to prepare people.”

No. 9: Make it the new norm. It is important to continue to monitor the project and institute accountability to maintain your results. At Cincy, individual performance data was used in performance reviews at the request of the technologists, who successfully raised the percentage of orders with an adequate clinical history from 38% to 92% a year later.

“It’s above 99% now,” Larson added. “It has become a part of the culture.”