Electronic Health Records

Use & Optimization News

EHR Problem Lists Not Accurate Enough for Risk Adjustment

EHR problem list data is a poor predictor of comorbidity and lacks the accuracy necessary for risk adjustment.

EHR problem lists are not accurate enough for risk adjustment.

Source: Thinkstock

By Kate Monica

- Comorbidity data derived from EHR problem lists is not accurate enough for risk adjustment, according to a study published in The American Journal of Managed Care (AJMC).

EHR problem lists comprise patient diagnoses entered into EHR systems by clinicians during visits. Outpatient health records often rely on EHR problem lists to identify conditions. These lists are occasionally updated by clinicians.

To assess whether EHR problem lists can be used as a source of comorbidity data, Daskivich et al. conducted a retrospective cohort study of 1,596 men diagnosed with prostate cancer between 1998 and 2004 with long-term follow-up at two Southern California VA care sites.

Researchers compared EHR problem list-based comorbidity assessments with manual review of EHR free-text notes to see how sensitively and specifically each method identified major comorbidities. The team also compared Charlson Comorbidity Index (CCI) scores for both EHR-based and free-text based comorbidity assessments to determine which method more accurately predicted long-term mortality among patients.

“Comorbidity is a key component of the risk adjustment needed for fair comparisons of measures of quality, as comorbid disease burden affects readmissions, complications, quality of life, and mortality,” wrote researchers in the report.

Study findings indicated EHR problem-list based comorbidity assessments had poor sensitivity for identifying major comorbidities. Specifically, EHR problem list-based comorbidity assessments detected myocardial infarction with a sensitivity value of only 8 percent. Furthermore, EHR problem lists only assisted in detecting liver disease with a sensitivity value of 1 percent.

“Despite interest in capitalizing on readily available problem list data in the EHR for purposes of risk adjustment, our findings suggest that these data should be validated prior to application to performance assessment,” wrote researchers. “The sensitivity of the VA problem list for identifying common major comorbidities was poor, ranging from 1 percent to 46 percent, compared with manual free-text note abstraction.”

The inaccuracy of EHR problem lists in identifying major comorbidities for risk adjustment could have financial consequences for physicians.

Physicians and physician groups participating in accountable care organizations (ACOs) report certain quality measures to CMS as part of value-based care.

“Absent a valid method of adjustment for comorbidity, it is not possible to confidently distinguish between physicians or groups who provide poor care and those who disproportionately see patients with greater disease burden,” stated researchers.

“Because measures of quality of care are now being tied to compensation in programs like value-based purchasing, the stakes are higher and the consequences of errors in performance assessment are much more substantial,” they continued.

Given the low sensitivity of EHR problem list-based assessments in identifying major comorbidities, researchers advised providers take other kinds of data into account for risk adjustment. For example, gathering standardized, quantifiable data from patients could help to improve accuracy in risk adjustment.

“This strategy has been shown in numerous studies to be accurate and reliable in the assessment of comorbidity,” wrote researchers.

Additionally, researchers recommended EHR-based comorbidity data be validated before being applied to risk adjustment to ensure accuracy.

Improving risk adjustment accuracy will be increasingly important as the healthcare industry continues the transition to value-based care and ACOs receive incentives based on their ability to reduce 30-day readmissions and mortality rates.

“The stakes are high for quality indicators to be unbiased and fair in order to avoid unduly penalizing providers and organizations who care for the sickest patients,” concluded researchers.

Using a variety of data types can also be useful for improving the accuracy of predictive analytics, according to a recent study by researchers at Google.

Researchers found representations of comprehensive patient EHRs using Fast Healthcare Interoperability Resources (FHIR) can improve the accuracy of predictive analytics by allowing clinicians to incorporate more information about patients in each prediction. 



Sign up to continue reading and gain Free Access to all our resources.

Sign up for our free newsletter and join 60,000 of your peers to stay up to date with tips and advice on:

EHR Optimization
EHR Interoperability

White Papers, Webcasts, Featured Articles and Exclusive Interviews

Our privacy policy

no, thanks

Continue to site...