Using medical claims patterns, analysts have observed an increase in sepsis incidence in recent years. However, clinical data may more accurately reflect these trends, in part because claims information is affected by changing diagnosis and coding practices over time, according to a study published in JAMA.
The researchers — led by Chanu Rhee, MD, a population medicine researcher at Harvard Medical School and Harvard Pilgrim Health Care Institute, both in Boston — compared national sepsis incidence estimates using EHR data to those based on medical claims data between 2009 and 2014. To estimate sepsis incidence with EHRs, they collected clinical data from 2,901,019 patients across 409 hospitals.
Here are three of their findings.
1. The researchers identified 173,690 sepsis cases from the cohort of 2,901,019 patients, representing a 6 percent incidence.
2. Using clinical data, researchers determined sepsis incidence remained relatively stable between 2009 and 2014. However, using claims data, incidence increased by 10.3 percent during the timeframe.
3. Clinical data indicated a decline in hospital deaths and no significant change in patients discharged to hospice. Claims data indicated a significant decrease in both hospital deaths and discharge to hospice.
The researchers concluded detailed clinical data from EHRs were more sensitive to identifying sepsis trends than claims-based information. "The findings also suggest that EHR-based clinical data provide more objective estimates than claims-based data for sepsis surveillance," the study authors concluded.