A Benchmark for Coding Productivity: Accounting for Multiple Variables

HIM professionals know that coding accuracy must be monitored and corrected when found to be below standard. At the same time, the speed at which records are coded must be efficient as healthcare organizations are faced with declining reimbursement and increased data collection. A balance between coding quickly and coding accurately must be achieved. However, what is the best method for establishing fair and realistic expectations for coder performance today?

Benchmarks for coding productivity have been examined in the past and historical productivity standards have been based on data gathered from HIM professionals. Calculating coding productivity is not difficult in theory, you simply divide the number of records coded by the number of hours worked. But it has long been recognized that a mechanism is needed to address additional factors impacting coders’ work, including for example:

[nz_gap height=”15″ /]
[nz_il type=”none” icon=”icon-checkmark” icon_color=”#f0634c” background_color=”” ]

  • Hospital CMI
  • Average Length of Stay
  • Work types (inpatient, outpatient)
  • EHR performance
  • Noncoding duties


[nz_gap height=”15″ /]

Published coder productivity data typically acknowledges that standard coding productivity rates are difficult to establish due to various factors that must be taken into account. In fact, reliable comparative data, which is necessary to establish and maintain meaningful standards, is extremely difficult to gather. Therefore, it is not uncommon for today’s coding productivity standards to be based on ICD-9-CM data and outdated coding processes. Now, nearly two years since the implementation of ICD-10-CM/PCS, it’s time for organizations to re-set coding productivity standards. Not only has the code set standard changed, but also numerous factors have impacted coding productivity, both positively and negatively, in recent years. For instance, monetary incentives to adopt electronic health records means that coders have been dealing with hybrid and evolving electronic health record formats. In addition, the primary coding software, the encoder, has also evolved and now often employs computer-assisted coding (CAC) capabilities. At the same time healthcare reforms, and particularly reimbursement reform, have expanded the data elements that a medical coder has to consider and capture in coded form negatively impacting coding productivity.

In preparing for ICD-10-CM/PCS implementation, the industry anticipated that coding productivity would drop, at least during the initial stages of implementation. Indeed, it did drop, but not long after implementation of ICD-10-CM/PCS, the coding community was feeling more optimistic about what productivity might look like with ICD-10-CM/PCS. Still, what is a fair and realistic expectation for coder performance today?

UASI has the privilege of coding a wide mixture of health records, with hundreds of UASI coders remotely coding hundreds of thousands of health records for UASI clients throughout the United States. We consistently monitor both quality and productivity on all UASI coding professionals. We also continue to monitor and track the variables that can impact coding productivity. Given this rich data source, we compared actual inpatient ICD-10-CM/PCS productivity to inpatient ICD-9-CM productivity, using ICD-9-CM as a benchmark to gauge today’s coding productivity.

Direct comparison between individual hospitals cannot be made, due to the previously mentioned varying elements that impact facility-specific process and work effort. However, general trending variances in the speed of coding over time is an extremely useful comparison. The results of this data are presented here for the first time.

The ICD-9-CM benchmark productivity level was based on actual UASI coding productivity 6 months prior to ICD-10-CM/PCS implementation. ICD-10-CM/PCS coding productivity data was then sampled twice since the transition of ICD-10-CM/PCS coding. The first sample reflected coding ICD-10-CM/PCS productivity during the first full quarter of coding. The most recent data sample was then collected at the end of the first quarter (January-March) of 2017. Statistical results revealed an overall productivity reduction, compared to ICD-9-CM productivity as the benchmark, at the following rates:

  • ICD-10 sample #1 (first quarter 2016): compared to the ICD-9-CM benchmark resulted in an 11.18% overall average reduction in productivity
  • ICD-10 sample #2 (first quarter 2017): compared to the ICD-9-CM benchmark resulted in an 8.68% overall average reduction in productivity
  • A comparison of Sample #1 to Sample #2 resulted in an increase in productivity of 2.25%

This data suggests that coding productivity today is likely 8.68% less than it was in mid 2015. The transition to ICD-10-CM/PCS is clearly the most significant reason for this overall decrease, however, we observed additional variables that significantly impact coding productivity today. While it’s commonly recognized that the size and complexity of a facility impact coding productivity, additional variables include the use of CAC, the extent of abstracting done by the coder, the type of query process in place, and whether a discharge summary was required for coding.

Coding with CAC: Our sample groups showed that coders using CAC experienced less drop in productivity in sample #1 but only improved <1% more in sample #2. Those in the group that coded without a CAC application had lower productivity than average in both samples #1 and #2. In sample #2, the productivity reduction for those coding without CAC was approximately 10% (compared to an 8.5% overall average).

Detailed Abstracting Procedure: Data from the subset of facilities where the coder performed detailed abstracting during the coding process showed a greater loss in productivity than average in both samples #1 and #2. In addition, coder productivity did not rebound from sample #1 to sample #2.

Query Completion: Query processes vary greatly from facility to facility, in terms of the amount of responsibility that lies with the coders themselves. However, our analysis revealed that productivity, at facilities where the coders are significantly involved in the query process, was not impacted more or less favorably in ICD-10-CM/PCS compared to the ICD-9-CM benchmark. These results remained essentially consistent with the overall average from both samples.

Waiting for Discharge Summary: Facilities that waited for discharge summaries prior to coding had a larger drop in productivity than average in sample #1. However, by sample #2, this group had rebounded consistent with the subgroup of facilities that did not wait for discharge summaries.

In summary, it does appear that noteworthy losses, with the initial transition to ICD-10-CM/PCS, rebounded relatively quickly and then leveled out, with smaller incremental gains over the course of the following year.

It is clear that multiple factors impact coding productivity with increasing complexity. For this reason, useful comparisons from one facility to the next, or to national benchmarks, is nearly impossible. Instead, healthcare organizations should track and trend coding productivity and develop and test best practice methodologies to establish facility-specific production standards that are repeatable and can be used to evaluate coding professionals.

Realistic medical coding productivity and quality standards are vital to achieve and maintain efficient and effective coding operations. With the myriad of regulatory and technological changes impacting coding operations today, it is imperative that healthcare organizations re-evaluate their coding productivity and reset their standards. In addition, organizations should continue to monitor and reaffirm this standard on an ongoing basis as healthcare data capture efforts continue to evolve.

UASI is committed to the highest quality coding and recognizes that coding productivity is a number one priority. As a leader in medical coding, we are frequently asked for opinions and advice on coding productivity and therefore will plan to continue to share our data with periodic updates on actual coder productivity.

About the Authors:

Marcy Blitch, (marcy.blitch@uasisolutions.com) is an Education Manager, UASI Coding Services. Gwen Summers (gwen.summers@uasisolutions.com) is a Senior Manager, UASI Remote Coding Services. United Audit Solutions, Inc. (UASI) is a leading national provider of revenue cycle solutions designed to help healthcare facilities achieve correct reimbursement in the quickest possible time.

Download a free copy of this article here.

Posted in