TB or not TB? A new clinician decision support tool for tuberculosis

Infection Prevention and Control (IPC) programs are tasked with implementing the appropriate isolation of patients with suspected or confirmed communicable diseases to reduce the risk of health care-associated infections to patients and mitigate occupational risks to health care personnel.

However, patient isolation can negatively impact clinical workflows, access to care, and hospital capacity, highlighting the critical importance of safe and efficient de-isolation protocols. System-based solutions for isolation and deisolation, including clinical decision support tools, are increasingly needed to:

• Support patient and health care personnel safety
• Reduce cognitive and administrative burden on clinicians
• Enhance clinician well-being

A team of infection prevention experts, infectious disease physicians and researchers at Massachusetts General Hospital (MGH) developed a risk scoring tool by reviewing the records of past patients who had been diagnosed with TB and those who had been deemed at risk but turned out not to have tuberculosis (TB). They created a validated model to assess the likelihood of TB infection before discontinuing isolation precautions.

Then this same team joined with developers in clinical informatics and digital health to program the scoring model into the electronic health record in the form of a tool called “TBorNot TB.”

If a patient is identified as needing evaluation for TB, they are labeled as “TB-Risk,” prompting the correct isolation in the specialized hospital room and use of personal protective equipment. Once the health care provider believes, given the patient’s symptoms, test results and/or risk factors, that TB is no longer a possibility, they can use the TB or Not TB tool to determine if it is safe to discontinue isolation precautions.

What are the clinical implications and next steps?

This tool will positively impact patient care and clinician workflows. In a rigorous analysis of TB evaluations across the MGB system over six years, the tool correctly identified all instances in which TB infection was present and ensured that all those patients would remain safely isolated during their work up.

At the same time, the tool was able to identify about a quarter of cases where TB was highly unlikely. This tool allows those patients to come out of isolation safely and efficiently, which may increase bed capacity, alleviate workload, reduce the time spent on infection control consultations, and improve the patient experience.