Proper coding is an essential skill necessary to capture the appropriate level of care provided by the physician. Both undercoding and overcoding are problematic, and few residency programs adequately educate trainees in this skill. A key coding difficurty is differentiating between level 2 and level 3 daily care charges. Although educational sessions are important to establishing a knowledge base, effective behavioral change requires ongoing strategies such as feedback. Feedback can be effective when given individually, but some evidence suggests that combining this with group feedback may be more successful. Over the past 2 years, our program has used individual feedback and in this current year has added group feedback (ensuring confidentiality) so individuals can compare themselves to the rest of the group. Because there is no measure of a “correct” ratio of level 2 or 3 codes, we chose to examine variability of ratios between hospitalists. Because all members of this group care for similar patients, we believe this ratio should be stable within a single provider longitudinally and similar across providers, regardless of the absolute value of the ratio.
The purpose of this study was to determine (a) if individual feedback reduces variability over time and (b) if feedback at a group level lowers intragroup variability.
Using our professional fee database, we compiled on a monthly basis percentages of level 3 daily charges compared with level 2 daily charges and provided these percentages to the individual for the past 2 academic years (2008‐2010). For 2009‐2010. we made the data available to the group as well. Outliers were provided brief written descriptions of how they differed from others. For 2008‐2009, the greatest variability in coding techniques was in the first quarter. In the fourth quarter, provider variability decreased. The greatest reduction in variability with 1 provider was a first‐quarter range from 7% to 97% level 3 charges per month, but in quarter 4 the level 3 range was only 49%‐54%. However, there remains a great degree of variability when looking at year‐end data between providers, as ranges varied from 16% to 59%. For 2009‐2010, individual variability decreased again as the year progressed. However there was less variability between group members (8 of 10 providers were within the 43%‐53% range) compared to 2008‐2009, when group feedback was not given.
As expected, giving feedback lowered a provider's variability in coding. Giving group feedback also had the effect of reducing variability between providers. Although coding is very dependent on patient acuity, we have an admission system that assures similarities in acuity. By giving individual and group feedback, we have a developed a system to minimize over‐ and undercoding. We are interested in using these methods in other areas besides coding including quality metrics and compliance with documentation expectations.
A. Ansari, none; C. Whelan, none.
To cite this abstract:Ansari A, Whelan C. Reducing Inappropriate Variability in Coding Through the Use of a Stepped Feedback System. Abstract published at Hospital Medicine 2010, April 8-11, Washington, D.C. Abstract 155. Journal of Hospital Medicine. 2010; 5 (suppl 1). https://www.shmabstracts.com/abstract/reducing-inappropriate-variability-in-coding-through-the-use-of-a-stepped-feedback-system/. Accessed September 18, 2019.