Background:

Many providers find E/M coding rules confusing and frustrating. Tasking hospitalists with completing charge tickets often leads to delinquent or inaccurate charges. Hospitalist groups have incorporated various education methods and incentives to improve documentation, billing and collections with varying success. Conversely, systems in which coders retrieve and code notes may be subject to delays in charge entry. Additionally, group leaders and providers worry that documentation will be “down coded” by coders, adversely impacting productivity reports and practice revenue. Previous studies have demonstrated considerable variability among certified coders. No best practice for roles in the coding process has been established.

Purpose:

To assess the impact of a change from provider to coder assignment of E/M levels for inpatient visits on code distribution, charge lag and collections.

Description:

Faculty members in a rapidly growing academic medical center hospitalist practice were educated on provider coding in small groups in 2009. Thereafter, physicians submitted their own codes daily leading to a dramatic increase in practice revenue. The adoption of typed EMR progress notes made review by coders much easier than paper charts. In May 2010, the academic practice group decided to have coders assign levels to visits. In an effort to keep hospitalist faculty engaged in the provider billing process, they were incentivized to submit daily patient lists and respond to documentation completion notices (i.e. dates, signature) within one month. Department leadership and faculty are eager to learn how this impacted coding levels, charge lag, denials and total clinical revenue. Charge entry and payment data for the hospitalist practice recorded in the Anodyne and FPSC databases were reviewed for the periods of provider and coder charge assignment. There was a significant increase in the total number of charges entered and wRVU’s. Subsequently, the coding distribution patterns for initial and subsequent visits were compared. There were negligible increases in 99221 initial visits charged and 99223 initial visits. More surprising was an increase in level 99233 subsequent visits (from 39% to 52%). Also, the average charge lag decreased (from 16 to 14 days). Most significantly, collections increased 18%. The charge per wRVU also increased by 5%.

Conclusions:

The change from provider to coder designation of E/M codes in an academic medical center hospitalist practice resulted in a change of coding distribution. Though practice leadership was worried about a decrease in clinical revenue from lower code assignment, it led to an increase in charges entered and collections. To our surprise charges were also entered sooner. The increase in collections more than paid for the coder’s salary. Given provider opinions about coding duties, it may be best to reduce providers’ role in the billing process.