A new study finds that using amyloid positron emission tomography scans can slightly reduce hospitalizations in people with mild cognitive impairment or dementia.

“In the context of the increased morbidity and mortality associated with hospitalization in patients with dementia, even a small reduction in hospitalization rates may have important implications for public health,” the authors wrote in the study. The report was published Monday in JAMA Neurology.

The scan can detect amyloid plaques, which are a neuropathological feature of Alzheimer’s disease. Having amyloid PET scan results has been shown to change how physicians manage care for some people with dementia or MCI. The researchers wanted to know whether it could reduce hospitalizations and emergency department visits in those populations over the course of a year. 

Though the researchers saw a 4.5% relative reduction in 12-month hospitalizations, that percentage was less than the 10% needed to show a significant decline. (But they say it can have an effect on overall health.) 

No statistically significant differences were seen in rates of 12-month hospitalizations and visits to the emergency department.

The researchers evaluated 12,684 Medicare beneficiaries aged more than 65 years who had MCI or dementia that had been diagnosed within the past two years and met criteria to have a PET scan done. Each person was matched to a control who didn’t have an amyloid PET scan. Of the people studied, 51% were women, and the median age was 77.

Overall, 21.4% of people with positive PET scans were hospitalized within a 12-month span compared with 25.7% of people who had negative amyloid PET scans. Between people who had and didn’t have the scan and later visited the emergency department, the results were almost similar. 

People with a positive amyloid PET scan result were at a lower risk of hospitalization than participants with a negative result, the authors said.

Overall, the findings suggest that referral for an amyloid PET scan had a possible association (which was below the pre-specified threshold) with 12-month hospitalization rates but no association with 12-month emergency department visit rates.