ICD-10 Move Could Lead to Inaccurate Patient Safety Data, Study Finds
September 8, 2014 in News
Transitioning to ICD-10 coding could negatively affect patient safety reporting and perceptions of hospital quality because of inaccurate comparisons between the new codes and those used under the ICD-9 system, according to a study published in the Journal of the American Medical Informatics Association, Health IT Analytics reports (Bresnick, Health IT Analytics, 9/5).
U.S. health care organizations are working to transition from ICD-9 to ICD-10 code sets to accommodate codes for new diseases and procedures. The switch means that health care providers and insurers will have to change out about 14,000 codes for about 69,000 codes.
In April, President Obama signed into law legislation (HR 4302) that pushed back the ICD-10 compliance date until at least October 2015.
Study Details, Findings
For the study, researchers from the University of Illinois at Chicago examined 23 different types of patient safety indicators (Hall, FierceHealthIT, 9/4). They used the General Equivalent Mappings to determine the accuracy of PSIs when comparing those used in ICD-9 and ICD-10 code sets (Goth, Health Data Management, 9/8).
The study found that of the 23 PSIs examined:
- Three PSIs under ICD-9 had straightforward equivalents in ICD-10;
- 15 PSIs under ICD-9 resulted in convoluted mapping to ICD-10; and
- Five PSIs under ICD-9 had no equivalents in ICD-10.
According to the study, the transition between the systems might inadvertently increase the number of PSIs and make it impossible to compare the two as ICD-10 is implemented. The researchers wrote that such an issue could increase the risks of:
- Under-reported safety incidents;
- Unwarranted inflation of PSIs because of ICD-10′s more-specified codes; and
- Increased variability of calculations (Health IT Analytics, 9/5).
The researchers wrote that the “transition could incent unethical ‘numbers-focused’ translations to improve adverse event statistics, while observed events may be unchanged or worsened.” They added that PSI data “published on public reporting sites, such as Hospital Compare, could be non-representative of a safe hospital environment, which is a burden to patients” (Health Data Management, 9/8).
The study authors said that better coding guidelines are needed to help providers use ICD-10 (Health IT Analytics, 9/5).
Further, they said CMS should publicly list improved and deleted PSIs to help “inform the public about the changes” (Health Data Management, 9/8).