Analytically Speaking: MDS data accuracy

Share this content:
Since the late 1990's, the Minimum Data Set has been a primary driver in the long-term care industry, influencing all areas of skilled nursing facility operations.

The clinical data that are collected in this assessment determines Medicare, and in many states Medicaid, reimbursement. In addition, the MDS greatly influences survey/certification outcomes and public perception of the facility through CMS' Five-Star system.

Most importantly, the MDS is an assessment and screening tool and the first step in a comprehensive assessment of your residents. Through the MDS, you are able to tell each resident's “story”: identify their individual strengths, weaknesses and areas of risk in order to develop the most appropriate plan of care.

With so much depending on this one assessment tool, the accuracy of the MDS is vital. Since the implementation of MDS 3.0 in October 2010, facilities have had to adjust to a new item set with different coding rules.

In addition, there have been three major revisions to the MDS coding and scheduling rules since implementation. October 2011's represented the most significant. Because of the relative newness of the MDS and these changes, the likelihood of errors has increased. My company applied 320 data integrity tests to approximately 3 million MDS 3.0 assessments and found that the volume of errors has increased from version 2.0 to 3.0.

Eighty-nine percent of MDS assessments had data integrity issues, and if an assessment had an issue, on average it had 2.71 per assessment. MDS data that went through major MDS software's “data scrubber” module demonstrated similar error rates. These errors resulted in inaccurate representation of resident acuity, missed care planning requirements and quality improvement opportunities. Most errors were seen in section G (functional status), I (active diagnosis), N (medications) and M (skin conditions).

Do these results suggest that the MDS 3.0 is an inferior assessment tool compared to 2.0?

Absolutely not. It will take several more years and additional clarifications from CMS before we see the desired data integrity. The industry will need to continue to adjust to the significant increase in volume created by the October 2011 implementation of the “three-day rule” that requires an End of Therapy Other Medicare Required Assessment (EOT OMRA) be completed, as well as the new Change of Therapy (COT) OMRA. With a higher number of required assessments and increase in length come an increased workload and more opportunity for error.

So what can your facility do to make sure that your residents' “stories” are accurately told? First, be sure that you have access to the most recent revision of the MDS 3.0 Manual and that your facility's interdisciplinary team is updated on any changes to the coding guidelines. Second, review your documentation capture tools to align them with MDS nomenclature. Finally, implement a system to audit the completed MDS before it is submitted to CMS. There are automated tools, which go well beyond what your MDS software has, available to help you with this process.

The key is ensuring that your MDS — which drives pretty much everything from care to reimbursement to survey — is accurate and can hold up to external scrutiny.

Steven Littlehale is EVP and chief clinical officer at PointRight Inc. He is a gerontological clinical nurse specialist and former university instructor.

Guest Columns

Guest columns are written by long-term care industry experts, ranging from academics and thought leaders to administrators and CEOs.