Sonali Desai

One of a provider’s worst fears is missing a diagnosis, especially of a disease like cancer where early detection can offer far better options and outcomes for a patient. But even if a clinician flags a concern, such as an incidental lung nodule found on a chest CT scan or a few non-cancerous polyps detected during a colonoscopy, a patient can still fall through the cracks due to lack of follow up after their abnormal test result.

“One of the reasons physicians experience burnout is the intense cognitive load from the volume of test results coupled with the rapid pace of care delivery — it’s the sense that you’ve missed something. It’s a feeling that often keeps you up at night,” said Sonali Desai, MD, MPH, director of Ambulatory Patient Safety for the Brigham and director of Quality for the Department of Medicine. “We wanted to find a way to offset that — a system that helps all of us sleep better at night knowing that our patients received appropriate follow up with the care they need.”

Desai and colleagues in the departments of Medicine, Radiology and Surgery have piloted an initiative known as Ambulatory Safety Nets to help catch potentially important findings that could lead to cancer diagnoses. The program provides a way to ease the cognitive burden of frontline clinicians by establishing a central support team — the ambulatory patient safety team — that proactively identifies, contacts and tracks patients who need follow-up tests for colon or lung cancer.

Track and Catch

The Ambulatory Safety Nets program currently includes ongoing efforts in two areas — colon cancer and lung cancer — with plan to expand to other areas, such as abnormal pap smears which can lead to cervical cancer, abnormal prostate cancer related blood tests, and breast imaging. Desai and colleagues are working at the Partners level to replicate their model across the system. In addition to Desai, the Safety Nets team includes a project manager and coordinator who collaborate with primary care physicians (PCPs) and specialists.

For the Lung Cancer Safety Net, which began in 2017, the team identified patients with incidental lung nodules through natural language processing and chart review. The team then emailed PCPs about patients that may require further chest CT imaging or referrals and continued to track completion of imaging and outcomes for the patients.

In collaboration with radiology, the team has deployed a more automated lung nodule tracking system known as Radiology Result Alert and Acknowledge for Development of Automated Resolution (RADAR), which allows radiologists and ordering clinicians to collaboratively develop follow-up plans and track patients. This system builds upon an existing Alert Notification of Critical Test Results (ANCR) web-based platform that the Brigham has used for many years to notify clinicians of critical or unexpected radiology findings. The Lung Cancer Safety Net program has expanded to the inpatient setting with plans to spread to the Emergency Department in the coming months.

The team used a similar approach for the Colon Cancer Safety Net, using electronic registries to identify patients who had a prior colonoscopy with pathology, iron deficient anemia rectal bleeding and might be at risk for delayed diagnosis of colon cancer. In close collaboration with specialists in gastroenterology, surgery, endoscopy and population health management, as well as PCPs, the team created an infrastructure for the safety net. In July of 2018, they hired a patient navigator to conduct active patient outreach, scheduling and navigation, including motivational interviewing, to increase the rate of colonoscopy completion.

“This multi-disciplinary effort has been invaluable in helping us identify patients who are due for surveillance colonoscopy because of their medical history or family history,” said Kunal Jajoo, MD, the clinical director of the Division of Gastroenterology. “It truly represents a safety net in that a portion of these patients may not have otherwise been identified and scheduled for these necessary procedures.  It’s been an excellent collaborative effort.”

Net Results

To date, the colon cancer safety net has identified 974 patients needing follow up over a two-year period, with 341 colonoscopies scheduled and 302 completed. Of these colonoscopies, 53 percent found polyps including two high-risk lesions. The lung cancer safety net has detected 120 incidental lung nodules requiring follow-up in primary care annually with 94 percent follow-up completed.

Desai and colleagues published results from both the colon cancer and lung cancer pilot projects in The Joint Commission Journal on Quality and Patient Safety. They report that of the 252 patients identified for the colon cancer safety net, 111 (44 percent) either scheduled or completed a colonoscopy. Of the 123 patients identified for the lung cancer safety net, 70 (57 percent) received appropriate follow-up as a result of the safety net’s outreach to the patient’s PCP. The team also conducted interviews with physicians and administrators about the Safety Net programs. Participants used the terms “thoughtful” and “collaborative” to describe the overall adoption of the safety nets and said they agreed that the safety net fit the mission of the practice and department.

Since the time of the paper’s publication, Desai and colleagues have done additional follow-up and testing for patients. They have also tracked down reasons why patients have not completed colonoscopies — these include changes in insurance, moving seeking care elsewhere and inability to reach a patient after three to five phone call attempts. For the Lung Cancer Safety Net, they have adapted their approach since the paper’s publication to centralize radiology scheduling and make the role of the radiologist and PCP more defined. 

Balancing on a Diagnosis Tight Rope

When it comes to diagnosis, providers must walk a thin line between overdiagnosis and underdiagnosis. Desai notes that one of the ways the pilot projects have evolved over time is by moving away from artificial intelligence tools that flagged more incidental findings from chest CT scans than were needed. Instead, the system they’ve developed places an onus on the radiologist to make a clear recommendation about whether follow-up evaluation was needed.

Desai and colleagues have found that defining clear responsibilities for PCPs and specialists is also key to the success of the safety nets. In colon cancer, for example, the team decided that the gastroenterologist performing the procedure should be responsible for updating the chart.

Desai offered a patient example to illustrate the power of the safety net. A 28-year old patient with inflammatory bowel disorder who had previously had a colonoscopy was identified and contacted for a repeat colonoscopy, which detected precancerous lesions that required treatment with surgery. 

“The whole idea behind working on safety nets is to reduce diagnostic error,” said Desai. “It’s a hard problem to fix. There’s a lot of research on why it happens but less on implementation of concrete solutions. This type of program can be very rewarding because you can see tangible results when patients gets the care they need.”

Home | Safety Matters