October 2017

HIM Challenges: Using Data to Improve Coding Performance
By Andrea Romero, RHIA, and Julia Hammerman, RHIA, CPHQ
For The Record
Vol. 29 No. 10 P. 5

In June 2017, Forbes Insights featured an article focused on the importance of data analytics in today's business world. Based on a Forbes Insights and Dun & Bradstreet survey of more than 300 executives across industries, the report shows that data analytics have moved from IT and finance to the majority of business functions.

Data analytics are proving to be the "secret sauce" businesses need to make better decisions and reduce costs. However, 38% of C-suite and senior leaders who participated in the survey agree their organizations can do more with analytics. For the health care industry, this includes HIM.

This article discusses the technology and data analytics processes that HIM leadership can use to make better decisions, advance coder performance, and boost overall coding quality.

ICD-10 Coding Performance: Room for Improvement
Data analytics can support strategies to improve coding performance and spur operational excellence through increased accuracy and productivity. Productivity is calculated by tracking key metrics and accuracy is ascertained via coding audits. According to available industry data, there is room for improvement in both.

A 2016 nationwide ICD-10 coding contest reported poor coding for inpatient, ambulatory, and emergency department coding. Based on contest findings, lower-than-expected accuracy resulted in an average loss of $1,877 per inpatient case due to diagnosis-related group inaccuracy.

On the coder productivity side, an internal himagine solutions study analyzed deidentified coding data from more than 250 hospitals to learn how the implementation of ICD-10 affected coder productivity. Not surprisingly, inpatient records showed the highest level of productivity decline (62%), although all coding types experienced some regression.

The expanded use of data analytics helps health care provider organizations continually monitor coding performance and spearhead ongoing improvements in coder productivity and quality.

Coding Productivity Data Analytics
Unlike many statisticians who merely look at numbers, HIM directors understand the nuances of coding data and how best to collate and report them. The typical elements affecting coder productivity include the following:

• impact of EHR implementation;

• number of systems accessed for coding workflow—each additional system a coder must access to complete a chart incrementally decreases productivity;

• clinical documentation improvement (CDI) process to streamline coding;

• turnaround time for missing documentation and queries; and

• noncoding tasks—for example, coders need a queue for sending charts missing key documentation.

Once these issues are addressed, coding productivity can be managed more effectively with data analytics. For optimal coder productivity monitoring, the following data must be tracked, entered, and analyzed:

• start and stop times for each coded record—by coder and chart type;

• average number of charts coded per hour by coder;

• percentage of charts taking more than the standard minutes to code—typically charts with long lengths of stay, high dollar amounts, or high case mix index; and

• types of cases each coder is processing every day.

It is also important to monitor denial rates. Compare denial rates with the total charts billed to determine how many charts are falling out, denying out, or editing out due to coding. There are multiple reasons for a case to edit out, including failed medical necessity and mismatched charge codes.

Careful review of key metrics in combination with regular coding audits is the next step in turning coding analytics into actionable HIM insights.

Coding Accuracy Data Analytics
Sometimes coding accuracy is compromised in pursuit of greater coding productivity—a risky short-term strategy. Decreases in quality result in increased denials, greater payer scrutiny, reimbursement issues related to undercoding or overcoding, and inaccurate capture of risk of mortality and severity of illness. Savvy coding leaders assess both data sets simultaneously and balance productivity with coding accuracy for optimal coding performance. The most common way to collect coding accuracy data is through coding audits.

For example, himagine solutions conducts routine coding accuracy audits for its 900 internal coders and analyzes audit data to target training, education, and other corrective action. Technology is a necessary component of this quality program. Sharp attention ensures data are recorded in ways that support back-end analysis, including the use of a proprietary coding audit database and manual spreadsheets. Results are analyzed at the individual coder and collective team levels.

How to Use Individual Results
Based on each coder's results, provide individual coaching at the conclusion of every audit. Include tips, recommendations, and resources to improve. These communications help coders understand the reasons for miscoding. In addition, give coders a rebuttal period to go back to the records and discuss cases with the auditor.

If a coder's accuracy data trend downward over several audits, work with the education team to target instruction and provide refresher coursework. The coding auditor should then conduct focused reaudits to ensure the necessary improvements are achieved.

How to Use Collective Results
Analysis of coding audit data across an entire team can identify patterns and trends in miscoding. Team data pinpoint where coders are struggling as a group or may be making mistakes. Review data by body system, diagnosis, and procedure type to identify opportunities for education and collaboration with other departments such as CDI to make improvements.

Finally, consider other resources that can be provided to coding staff to support quality. For example, a coding hotline or question queue can be established. This is particularly helpful for large coding teams working remotely and from different geographic areas. Track and aggregate questions received to create educational resources and identify topics for newsletters, blogs, and other knowledge sharing across the coding team.

Survey Says: Invest in Analytics Technology and Support
The Forbes Insights and Dun & Bradstreet survey found that sufficient technology and manpower were lacking to achieve data analytics excellence. Spreadsheets were being used as the primary tool for data analysis 23% of the time, and skill gaps were cited by 27% of participants. The need for better analytics technology, skills, and best practices across all industries is evident.

For HIM, the expansion of current coding technology applications is essential. Only a few of today's systems provide the comprehensive coding productivity, accuracy, and audit data required to drive performance improvement.

As with all expenditures, HIM leaders must secure adequate funding. CFOs and revenue cycle vice presidents require cost-justification numbers. The following strategies can help make a powerful case for adding coding analytics technology and staff:

• Calculate the number of dollars sitting in discharged not final billed (DNFB) and accounts receivable. Communicate how analytics support DNFB management and decrease outstanding receivables.

• Quantify denials and the associated dollars over time. Share how the system will support denials prevention while also increasing revenue through more accurate and productive coding.

• Use real-life data to show how coding data can be better manipulated at the individual and department levels. Emphasize that manual spreadsheets are outmoded in the era of EHRs. Data are more difficult to track and compile, expending management time and resources.

• Demonstrate how access to more data enables effective communication with the coding team. Insights can drive hands-on specificity of coder training as a best practice and increase coding quality for each team member. Communication strengthens relationships with coders while increasing job satisfaction and helping ensure a good return on investment for coder training.

Health care organizations are recognizing the benefits of reliable data to improve outcomes, allocate resources, and manage departmental costs. HIM directors collect much of the data required by executives for informed decision making. Sophisticated data analytics software ensures information is available for bottom line survival and future organizational growth.

— Andrea Romero, RHIA, is chief operating officer at himagine solutions.

— Julia Hammerman, RHIA, CPHQ, is director of education and compliance at himagine solutions.

TECHNOLOGY MUST-HAVES FOR EFFECTIVE CODING DATA ANALYTICS
• Data analytics programs with drill-down capabilities are imperative. They're expensive but worth the cost. Facilities can use these systems to effectively manage and prevent denials.

• Customized workflow management that can assign coders to queues based on their skillsets.

• Analytics tools to manage discharged not final billed daily and provide continual reporting.

• The ability to build rules to enable cases to go automatically to an audit queue based on specific factors (diagnosis, trend, Office of Inspector General, recovery audit contractor, etc).

• The ability to export and manipulate the data within other systems such as Excel.

• The ability to trend the data.

• Training on advanced manipulation of data (eg, pivot charts).

Also, every HIM department should have a copy of the AHIMA Data Analysis Toolkit. (It's free for AHIMA members.)

— AR and JH