Unfortunately, the quality of data is driven less by opportunity and more by incentives for those creating the data.
Prior to the implementation of ICD-10, the key selling point of the new coding set was that it provided the opportunity for more detailed data about the nature of each patient’s condition. In theory, this improved data would allow us to analyze more specific patterns of illness in populations and understand the risk severity and complexity of health conditions at a much more granular and meaningful level.
Most would agree that ICD-10 gives us that opportunity to get much better data, but the operative word is “opportunity.” Unfortunately, the quality of data is driven less by opportunity and more by incentives for those creating the data.
As I have gone around the country speaking to clinicians about code selection in their electronic health records (EHRs), I ask them about the main reason they select one code over another. The answer is always the same: “the one that is easiest to find and gets paid.” Rather than coding focused on representing as accurately and completely as possible the nature of the patient health condition, the primary driver is reimbursement. Codes tend to be used at a higher and less specific level because they are easier to find and greater detail does not seem to affect payment.
While we have the opportunity for greater detail about the nature of each patient’s condition, including risk, severity, and comorbidities, the opportunity for better detail is not sufficient to ensure that the data is actually more accurate and complete. The clinician that makes the diagnosis ultimately must see value in being more specific and complete in documentation and coding. A simple exhortation to achieve better data capture just won’t do it. While there are some incentives that would drive better and more accurate data, such as quality measures, risk adjustments, or improved DRG payments, the clinician does not always value these incentives as significant in the traditional fee-for-service world. Payment rules may drive towards the use of one code over another, but this is not necessarily the code that best represents the most accurate and detailed description of the patient condition.
The impact on population data analysis:
- Analysis can only be done at a high categorical level since there can be no assurance that more detailed coding is used consistently.
- Differing payment incentives may result in different patterns of coding, independent of the patient’s specific condition.
- Clinicians that work in an organization that pushes for more complete and accurate coding and documentation will document different disease patterns with differing levels of severity, as opposed to clinicians who are only incentivized to use unspecified codes that are simple to find.
- Codes selected by clinicians may be more based on the habit of selecting their favorite codes, and not necessarily the most accurate codes, based on coding guidelines.
What will it take to leverage ICD-10 to improve data quality?
- Clinicians will need to be educated about the importance of complete, accurate, and specific coding, including explanations of the following:
- How this coding is interpreted by payors, reviewers, and auditors as being reflective of the risk severity and complexity of their patients’ condition
- How diagnostic coding factors into measures of quality, efficiency, and effectiveness
- The importance of consistent and accurate coding in ongoing analysis of population health
- The impact of coding on fraud, waste, and abuse investigations
- How proper coding reflects on payment methodologies associated with risk adjustment
- The evolving change in reimbursement methodologies that are more value-based and less fee-for-service based
- The adjustment for outcomes based on the risk and severity of each patient’s underlying condition and co-morbidities
- Improved audits to provide feedback to clinicians on their coding patterns
- Establishing new incentives that reward accurate, specific, and complete documentation and coding of the nature of each patient’s health condition
What will it take to leverage ICD-10 to improve data quality?
ICD-10 provides the “opportunity” to describe the nature of the patient condition, including risk, severity, and complexity, in a way that was not possible under ICD-9. Clinicians, however, can be just as non-specific in ICD-10 as they were in ICD-9. Without proper education and incentives, clinicians will continue old habits of picking the code that is easiest to find and still gets their claims paid.
