BOSTON – A hospital-wide algorithm for diagnosing and managing complicated bacterial pneumonia in children led to a marked cut in unnecessary chest CT examinations and a reduced number of surgical interventions. It also produced better outcomes, with fewer readmissions and no change in average length of stay or vancomycin use.
A key element of the management algorithm, implemented 3 years ago for children with a pleural effusion, empyema, or both complicating community-acquired bacterial pneumonia, was the emphasis on assessing children with ultrasound rather than with CT. This change produced a drop in chest CT examinations in these patients from 60% before the algorithm became hospital policy to 17% after.
The algorithm called for preferentially using ultrasound to assess these cases. During the first 15 months of its use, chest ultrasounds in these patients was performed in 71% of cases, compared with 27% of cases before the algorithm, Dr. Roberta L. DeBiasi said at the meeting.
The preferential use of more ultrasound examinations in children with a pleural-space infection meant that fewer children received the large radiation dose delivered by a CT exam, and the sedation required for CT. While safer, ultrasound also produces better imaging than CT in these patients "to sort out who has a loculated empyema that needs VATS [video-assisted lung surgery] and who has a nonloculated effusion that generally doesn’t need VATS," said Dr. DeBiasi, a pediatric infectious diseases specialist on the staff of Children’s National Medical Center in Washington, D.C.
Creation of a new algorithm for the hospital depended on getting physicians and surgeons from all the divisions and departments involved in managing these children – infectious diseases, surgery, radiology, hospitalists, pulmonology, and emergency department – together to decide on the best management approach and make it hospital policy.
"We asked, don’t you realize there are data that ultrasound is preferable, so why use CT so often? The answer was that an ultrasound technician wasn’t available at night in the emergency room." After seeing the data, the radiologists agreed that having ultrasound available 24/7 was important and so arranged it, she said in an interview. The hospital gets on average one or two patients a week with community-acquired bacterial pneumonia complicated by a pleural space infection.
Although no society guidelines existed in November 2008 when the revised algorithm went into effect, last August the Pediatric Infectious Diseases Society and the Infectious Diseases Society of American issued joint recommendations on the management of community-acquired pneumonia in children and included a recommended approach similar to the Children’s National algorithm, Dr. DeBiasi said (Clin. Infect. Dis. 2011 Aug. 30 [doi: 10.1093/cid/cir531]). The only difference was that her hospital’s guidelines are more specific, and guide the staff through the local protocol step by step. For example, the new society recommendations say that either video-assisted lung surgery or fibrinolytic therapy are appropriate options for managing loculated empyema. Because surgeons at Children’s National Medical Center do not use fibrinolytic therapy on these cases, the algorithm specifies VATS only, she said.
"We asked, don’t you realize there are data that ultrasound is preferable, so why use CT so often?"
To examine the impact of the algorithm, Dr. DeBiasi and her associates analyzed patient management and outcomes during the 15 months before the revised algorithm went into effect and then during the first 15 months after. The review showed that the 83 patients managed before November 2008 were an average of 6 years old, similar to the 87 patients treated during the first 15 months using the algorithm, who were an average of 5 years old.
The reduced number of CT exams and increased ultrasound use led to a reduction of VATS from 45% of cases before the algorithm to 29% after. Patient outcomes were better – with a "nice," statistically significant drop in readmission rates from 8% before the algorithm to none during the period after – but during both periods, vancomycin use and average length of stay remained constant (35% and 8 days, respectively), Dr. DeBiasi noted.
"Our [inference] is that some patients didn’t need VATS, and so the algorithm reduced unnecessary interventions. These patients were just managed medically," Dr. DeBiasi said. "I think it was the ultrasound that led to less VATS, because ultrasound is better than CT to see who needs VATS and who doesn’t."
Dr. DeBiasi said that she did not have any disclosures.