ICD-10: Eight Months Later and What’s Next?

Pat Stricker, RN, MEd
Senior Vice President
TCS Healthcare Technologies

Remember the days leading up to the ICD-10 transition, when experts were predicting large-scale disruptions to the entire healthcare system? The "doom and gloom" predictions were similar to those of the "Y2K" (year 2000) transition in 1999.

However, when the ICD-10 launch finally occurred on October 1, 2015, it was easier than many had expected. The ICD-10 transition did not create volumes of rejected claims, financial losses, cash flow problems, physician retirements, or chaotic coding that healthcare experts predicted. There were reports of increased claim denials, reimbursement delays, decreased productivity, and some provider issues, but the problems were not as large or significant as originally predicted. A study by the American Health Information Management Association (AHIMA) and the AHIMA Foundation, one of the leading organizations tracking the transition, found that the ICD-10 transition showed minimal effects. And the Cooperative Exchange, which represents the healthcare clearinghouse industry, described the ICD-10 transition as a "non-event", just like Y2K.

In an effort to examine how ICD-10 has impacted coding accuracy and productivity, AHIMA and the AHIMA Foundation launched a study to analyze these at various provider settings. The study, released this month (June, 2016) includes responses from 156 AHIMA members listed as a "coding professional" who holds a corresponding credential. The survey consisted of 13 questions pertaining to demographics, type of work facility, and the respondents perceived impact of the ICD-10 implementation on coding productivity and accuracy (increase, decrease, or no change). The level of increase or decrease in both productivity and accuracy varied depending on where they worked (inpatient or outpatient setting), as well as their years of experience, level of education, and use of encoder or computer-assisted coding (CAC) products.

Overall, of those who responded:

• 38% indicated a change in accuracy

            o The average increase in accuracy was 24.50%; the average decrease was 12.52%.

            o Accuracy in inpatient settings decreased more than outpatient (13.25% vs. 10.58%)

            o Respondents with 11 to 15 years had the largest decrease (15.87%); those with 16 to 20 years of experience had the smallest (7%).

            o Those with Graduate degrees had the largest decrease (25.60%); those with Bachelor degrees had the smallest (7.62%).

            o Those who used a CAC experienced a 0.2% increase; those without a CAC experienced a 1.58% decrease.

• 74% indicated a change in productivity

            o The average increase in productivity was 30.63%; the average decrease was 23.89%.

            o Productivity in inpatient settings decreased more that outpatient (24.30% vs. 22.10%).

            o Respondents with 6-10 years of experience had the largest decrease (27.14%); those with 1-5 years of experience had the smallest (19.97%).

            o Those with Associate degrees had the largest decrease (26.76%); those with Graduate degrees had the smallest (20.71%).

            o Those who use CAC software experienced a 17.13% decrease; those without a CAC experienced an 11.92% decrease.

The study shows that the implementation of ICD-10 has led to a perceived decrease in productivity but has had no effect on accuracy of coding. However, further studies are needed to determine if productivity will revert to pre-ICD-10 levels and whether productivity and accuracy will increase as CACs become more ingrained into the coder workflow. Actual audits of records would also provide more objective data than the perceived, subjective data received from the respondents.

Since it has only been 8 months since the transition, it is too soon to determine the impact that ICD-10 has had on the quality of care, outcomes, and overall healthcare costs. At the end of the 1^st year, data will be compared with previous years to determine how ICD-10 has truly impacted these areas. However, in the past 8 months several studies and analyses have been done to determine the impact ICD-10 has had on accuracy of coding and workforce productivity. The following is a summary of Analyzing Eight Months of ICD-10 by Mary Butler, as well as other studies:

Accuracy

• Coding accuracy is similar to ICD-9.

• Coding accuracy is hovering around 62% now; the average productivity decline was only 14 percent, which is much lower than expected.

• 3M reported that over 5,000 hospitals are using their code-finding software are experiencing an average accuracy rate of 90%.

• An informal study of 30 participants representing 10 acute care hospitals and health systems revealed the accuracy of the 196 coders to be: 84.3% for diagnostic coders, 80.1% for procedural coders, and 81.9% for CPT coders.

Productivity

• 3M reported that hospitals using their code-finding software are experiencing a 20% decline in productivity, instead of the 30-40% drop that was expected.

• An informal productivity study of 30 participants representing 10 acute care hospitals and health systems revealed an average decrease in productivity of 36% for Inpatient, 22% for Ambulatory Surgery, and 21% for the Emergency Department.

• In a similar study, nearly half of all participants noted productivity declines. This decrease could be due to incomplete documentation and the need to spend more time to discover and document the details of the case.

• Some productivity decline was attributed to the addition of new CAC technology that was implemented at the time of the ICD-10 transition.
• Overall productivity has stabilized at about a 10% reduction.

Specificity

• According to an analysis conducted by AHIMA some of the specificity issues have been related to lack of clear documentation in the medical record to accurately determine the specific code. The most common issues cited include: identifying an initial visit from a subsequent visit, determining if a procedure was therapeutic or diagnostic, identifying whether dye was used in order to determine if the procedure was fluoroscopy or ultrasound, and confusion around devices, components, and grafting materials.

Providers

• The vast majority of the American Academy of Family Physicians (AAFP) members felt the transition was "pretty seamless." There was a slight decline in productivity and some minor payment delays, but not as much as originally predicted.

• Hospitals reported a decline in coder productivity of about 30% the first few weeks; however that leveled off and has been maintained at about a 10% since then.

Health Plans

• There were rumors prior to implementation suggesting that payers weren’t ready for ICD-10. However, health plans nationwide are touting widespread success in the implementation of ICD-10.

            o Most have seen no measurable differences and no significant payment issues.

            o Claim volume has been consistent.

            o Rejection rates have not increased.

            o Call volumes related to ICD-10 related questions have not increased.

            o They have seen a decline in the use of unspecified codes and an increase in coding accuracy with ICD-10. (This could be due to the fact the extensive ICD-10 training that was provided, which made them more knowledgeable and aware).

CMS (Center for Medicare and Medicaid Services)

• There were also rumors that CMS was not ready and would not be able to handle the transition. However these fears turned out to be unfounded:

            o CMS handled 4.6 million Medicare claims each day during the 4^th quarter of 2015, which was nearly identical to the historical 4^th quarter baselines.

            o In addition, about 1.9% of total claims submitted were rejected in the 4^th quarter of 2015, compared to a historical rejected benchmark of 2%.

Billing Companies

A survey conducted by the Healthcare Billing and Management Association showed the following results for the 38 companies who responded:

• 11 companies reported more accuracy with ICD-10; 14 companies reported coding accuracy was the same as with ICD-9.

• 2 companies reported an increase in coding errors.

• 22 companies reported productivity of up to 25% below pre-ICD-10 levels; 7 companies reported no decreased productivity.

• 8 companies reported no increase in denials; 22 companies experienced up to a 10% increase in denied claims.

• 3 revenue cycle companies went out of business due to problems implementing ICD-10.

Lessons Learned

• Organizations who experienced fewer problems or issues with the transition credit it to:

           o Providing extensive staff training and testing before Go-Live. Some used the implementation delays to provide more training and practice sessions for their staff.

           o Requiring all coding professionals to be on-site during the first weeks of the transition.

           o Support from senior leadership.

• Those who did not spend enough time training and testing saw more issues and problems, e.g. accounts receivable increased, coder productivity decreased, rework of claims edits increased, etc.

• Because the ICD-10 code set includes significantly more codes and codes that are much more specific than the ICD-9 code set, they offer a greater opportunity to capture a more detailed, specific, complete history of a patient’s condition, treatment options, procedures, and services. This also allows better tracking and analysis of care outcomes. However, if the person documenting does not include all the specifics of the patient’s condition and treatment in the medical record, the coders will be unable to capture the specificity of the patient’s condition and there will be little change in the value of the ICD-10 data compared to that of ICD-9 data. Therefore, it is critical to document the specificity, so the data can be analyzed to improve outcomes, develop better healthcare strategies, and help move the healthcare system to a value-based model of care.

So.... since the transition seems to have been a success, does that mean we can rest easy? No, not exactly.

There was an agreement between the American Medical Association (AMA) and CMS prior to Go-Live that said that for 12 months Medicare would not "deny physician or other practitioner claims billed under the Part B physician fee schedule through either automated medical review or complex medical record review based solely on the specificity of the ICD-10 diagnosis code, as long as the physician/practitioner used a valid code from the right family." This meant CMS would not deny ICD claims for the first year for lack of specificity. It gave the providers a year to get used to the ICD-10 codes, allowing them to submit unspecified codes. Currently, 25% of ICD-10-CM codes include the term "unspecified."

However, on October 1, 2016, that exception expires and specificity will be required. This will probably mean an increase in denials, unless good clinical documentation is included in the medical record that clearly explains the patient’s condition, acuity, treatments, and services. Organizations should be auditing records for accuracy and specificity of coding now, so they know what the issues and problems are. Training sessions should also be conducted for those who are documenting the cases, so they are sure to include the specifics, and for coders, so they know which specific codes to use. It is important to start now, in order to be ready for the changes on October 1, 2016.

One of the goals for using ICD-10 was the opportunity to have detailed data that could be used for population health and utilization management programs, as well as for research and quality improvement measurement. However, this relies on providers documenting clearly, precisely, and with enough detail so the coders can code with specificity. If this specificity is not there, reports will not have the level of detail needed to identify specific conditions, complications, treatment options, and services that will help us move to value-based care and payment. This specificity will provide a patient-centric view, rather than an episodic view, of the patient. As a result, coding will become less about payment and more about accurate patient information that leads to better tracking, trending, and patient management.

I also have some other good news/"bad" news for you. Since we haven’t had updates for the past 5 years, we will be getting annual updates again beginning October 1, 2016. The "bad" news is that this year’s update will contain 3,650 new ICD-10-PCS codes, about 1,900 new ICD-10-CM codes, in addition to nearly 500 revised ICD-10-PCS codes and 351 revised ICD-10-CM codes. The full list of the proposed new and revised codes can be viewed at the Centers for Disease Control website.

But don’t panic! This update should be easy now that you know the structure of the ICD-10 codes and have mastered the processes and workflows for finding and using the correct codes. What’s another 5,500+ codes? Case Managers can handle ANYTHING! Right?

Pat Stricker, RN, MEd, is senior vice president of Clinical Services at TCS Healthcare Technologies. She can be reached at pstricker@tcshealthcare.com.