An analysis of data reported in the American College of Surgeons National Surgical Quality Improvement Program found the most commonly reported data is often unreliable, according to a study in JAMA Surgery.
Researchers assessed the reliability on a scale of zero to one, with zero being completely unreliable and one being perfectly reliable. They examined the reliability of morbidity, severe morbidity and mortality for six common procedures. They found that the biggest reason for unreliable data had to do with significant differences in sample size.
For example, reliability for overall morbidity for abdominal aortic aneurysm repair was 0.29, with an 18.3 percent occurrence rate and 25 cases per year. However, for colon resection, the reliability for overall morbidity was 0.61, with 26.8 percent occurrence rate and 114 cases per year.
Researchers suggest sampling minimizes clinical data reliability, and advanced modeling strategies are needed to increase reliability.
More Articles on Data Reporting:
Patient Safety Tool: Guidelines for Intervention Description and Replication Reporting
5 Tips for Measuring Patient Safety Data
CMS: Quality Reporting Deadline Again Delayed Due to Weather