Predicting ICD-10 losses and tracking key performance indicators after October 1

After several delays, ICD-10 implementation is finally upon us. The healthcare industry has spent years planning, training, and testing?and now the moment we have all been waiting for has arrived. But don't breathe a sigh of relief just yet.

While planning to go live is a thing of the past, there is still much work to be done. Specifically, organizations should now focus on setting performance benchmarks to determine the true impact of ICD-10 implementation.

After ICD-10 go-live, many in the industry have predicted that there will be a 90-day stabilization period, which represents the amount of time payers will need to stabilize systems and make determinations about claims, says Kelly Whittle, MS, principal of ICD-10 Advisory and Whittle Advisors.

However, the overall impact of the transition on healthcare organizations may go beyond 90 days, Whittle says. Naturally, it will take healthcare ­organizations time to adjust to the new coding system. It's no secret that most will experience a decrease in productivity, but what is less known is how much productivity will drop and how long it will take to operate at full capacity once again.

For this reason, Whittle and her colleagues decided to take a closer look at the effect ICD-10 has on productivity.

"We began to realize that at any point within the revenue cycle, if and when we experienced a productivity loss, it's not likely that the productivity loss or impact would be cleared up in a 30-day period," Whittle says.

She and her colleagues designed a model that examines productivity loss across the revenue cycle and goes beyond the concept of a 90-day stabilization period. What they found was surprising.

"We saw the productivity impact lasting all the way to June or July 2016," Whittle says. "That's a really different model than what we are hearing about in today's environment."

Predicted losses

The model developed by Whittle and her colleagues is based on the impact ICD-10 will have on hospitals. It may hold up for physician practices, but Whittle anticipates that they may not take quite as long to recover from the pains of ICD-10.

Physician practices may take a hit when it comes to coder productivity as well as the time it takes for providers to identify the right diagnosis code as part of their workflow. Whittle has noticed a 10%-15% coder productivity loss with some physician practices, as well as a 15%-20% productivity loss among providers using EHRs for the first time or using a new EHR to work through patient encounters. The latter often boils down to the skillset and comfort level of the provider and whether a medical assistant is present to work with the EHR as the provider meets with patients, she says.

Whittle calculated that if a provider has a 10% productivity loss per patient, each patient encounter would increase by at least 70 seconds. If the provider works eight hours a day and sees four patients per hour, the practice could potentially miss out on 700 patient visits in one year if the provider's hours are not increased. "This is a significant impact to revenue," she says.

Draw from this example to determine the overall impact ICD-10 may have on your facility over the next year or so. The model begins by examining accounts receivable (AR) days just before ICD-10 and after. It looks at a 13-week daily average of the starting AR days. It also examines starting AR dollar values.

"That's really important, because dollar values directly impact the amount of cash or liquidity in the system, and that's where some hospital systems are going to get in trouble," Whittle says.

Next, Whittle took a look at the impact on coding, patient financial systems, and claims management.

In terms of coding, some in the industry have predicted as much as a 50% productivity loss. "The projections are all over the place," Whittle says.

In one instance, while working with a facility that had started dual coding prior to go-live, Whittle predicted a 35% decrease in coder productivity from October through the end of 2015. The facility's productivity decline is projected to be at 30% in January 2016 and 25% in spring 2016. By July 2016, she anticipates a 25% hit to coder productivity, which is equivalent to an additional 1.8 days in AR. But facilities need to take a look at the big picture, not just the impact of ICD-10 on coders.

"That's just the coding impact," she says. "When we start to add the billing and the follow-up or claims management impact, we predict in June 2016 an increase in AR days by 4.65. That's millions and millions of dollars for your average healthcare system."

Not surprisingly, coder productivity loss has the biggest impact on AR days?1.8 of the 4.65 Whittle predicted?but that doesn't mean that other aspects of the revenue cycle will not be problematic for facilities in a post-ICD-10 implementation world. At the height of productivity loss, Whittle predicts that billing could add approximately 1.7 AR days and claims management could add 1.1 AR days.

Bear in mind, these predictions are based on facilities that were dual coding in ICD-9 and ICD-10 prior to October 1. The overall productivity loss is likely to be higher?perhaps even closer to 50% rather than 35%?for facilities that did not opt for dual coding, Whittle says.

"I absolutely agree with all of the above predictions and have seen facilities with months of dual coding experience not regain their productivity," says Monica Pappas, RHIA, president of MPA Consulting, Inc., in Long Beach, California. "In spite of preparation, I am not sure facilities have thought through all the revenue cycle and clinical documentation impacts."

Pappas, an editorial advisory board member for sister publication Medical Records Briefing, also notes that Medicaid programs in four states will be using a translation tool in order to pay claims (California, Louisiana, Maryland, and Montana), likely resulting in added back-end review. "This is all an opportunity for HIM professionals to demonstrate leadership in their facilities," she says.

Key performance indicators to watch

It's the question that has been on the minds of many people in the healthcare industry for some time: Can we match ICD-9 benchmarks in ICD-10?

"If your systems--your technology systems, level of staffing--remain the same, the answer is no," Whittle says. Technology is advancing rapidly, with new offerings popping up every day, and healthcare organizations need to keep up.

Resources, technologies, and processes that worked in meeting ICD-9 key performance indicators (KPI) are not likely to produce the same result in an ICD-10 world, Whittle says. However, if organizations bulk up resources in high-traffic areas (e.g., patient registration), there is hope that they can one day match or come close to ICD-9 levels of productivity. In addition, if your systems are not up to date, it is advisable to upgrade those that are integral to ICD-10.

Set goals for meeting KPIs in ICD-10. In terms of operational KPIs, hospitals must monitor their case-mix index, Documented Not Final Coded, Documented Not Final Billed, query rates, and query response rates. "For HIM, queries are probably the No. 1 tool for connecting to the documenters to improve the documentation specificity, both midlevels and providers," Whittle says.

Of course, facilities must also monitor coder productivity and coder quality, as these will be major issues throughout 2016 and 2017, Whittle says. "Coders have a huge learning curve, obviously, but you want to make sure they're maintaining quality and not picking up poor habits throughout the first year as they're learning and regaining their productivity," she says.

Pay attention to complicating or comorbid conditions (CC) and major complicating or comorbid conditions (MCC). Many codes classified as CCs in ICD-9 are now MCCs in ICD-10, or vice versa. Capture the conditions that result in a change in revenue, Whittle says.

As mentioned previously, AR days will take a significant hit now that ICD-10 is a reality. In addition to monitoring the financial impact of AR days, be sure to track denials. "Industry professionals are talking about the huge potential for increases in denial rates," Whittle says.

CMS at one time predicted that ICD-10 may result in a 100%?200% increase in denials, and Whittle sat down to do the math. First, she considered the average number of Medicare denials a facility may have in a month, and increased that figure by 100%. In her sample case, Whittle estimated that if a facility had an additional 40,000 Medicare denials per month, it may need to hire five or six additional billers just to manage the uptick in denials?and that's not including any increase in denials from commercial payers.

The denials process is often labor-intensive because it involves so many manual processes. Billers need to triage incoming denials. Service line leads, clinicians, and coders may also need to review denials. This effort can add up, which can result in additional operational costs and eroding margins.

Minimizing the impact on KPIs

To help minimize the impact ICD-10 can have on KPIs, start by ensuring you have the right workforce members touching the claims the first time, which may minimize denials and avoid rework.

"Put your strongest Medicare folks on the denial work as soon as you start to see them," Whittle says.

If you haven't done so already, explore options for hiring backfill coders or other workforce members. This can be costly, but it goes a long way in ensuring claims get out the door sooner, Whittle says.

Analyze documentation to identify trends and patterns in ICD-9 that may or may not be carried over to ICD-10. For example, the coding definition of atrial fibrillation may differ from the ways in which providers commonly document it, and this should be communicated to physicians so they can document conditions in a way that will help with accurate code assignment in ICD-10, Whittle says.

"Honing in on the known documentation gaps and giving your clinicians a definition or an understanding of how to apply the new specificity will help them move quickly through their documentation process," she says. This may not be something facilities can do for every code, but if they can identify the top 10 or 15 codes that are most commonly used or misused and provide physicians with education about documenting for these conditions, it's a step in the right direction.

ICD-10 impacts so many roles, but means something different to everyone in a facility. Naturally, a facility's structure and planning will affect how it aligns with Whittle's predictive model. "Facilities can expect not to hit their ICD-9 benchmarks, and should work toward controlling the amount of variance from ICD-9 to ICD-10 benchmarks," she says.