Editor's note: H&HN Daily, in partnership with the College of Healthcare Information Management Executives, is pleased to present "ICD-10 In Real Time." In this 12-month blog series, three leading CIOs share their experiences on ICD-10 implementation, physician engagement, productivity, payer readiness and more. The blog will run on the first Wednesday of every month.
In my July post, I danced around the issues of testing and training for ICD-10. If you have not pondered the implications of testing, ask yourself these questions: What happens to my hospital if I cannot submit claims at all for 30 days or more? What if my software fails to properly create an ICD-10-compliant 837-5010 on Oct. 1, 2014? How many days cash on hand do we have to sustain a prolonged shortfall of incoming money from my payers? What if my payers cannot accept ICD-10-compliant 837-5010 claims? Can our reserves reasonably carry us through a crisis in cash flow? What happens to you, the health information technology leader, if this crisis of cash flow occurs as a result of your failure to test and prepare?
I don't like, even remotely, to think about the implications of those questions, especially the last one, but without planning and preparation, they are very real.
Now another set of questions: Will my medical staff be ready to document in a fashion that allows coders to work in ICD-10? Does our organization fully understand the implications of Systematized Nomenclature of Medicine, or SNOMED, coming into play? What training do we have planned for staff, and exactly when are we going to perform it? Who needs that training? How much time is going to be required and how much is it going to cost?
Our organization chose to engage early and stay the course even after Health & Human Services announced the one-year delay in ICD-10 implementation. We have planned the work and are working the plan.
To begin, we conducted a gap analysis of all systems and work processes that utilized coding. Fortunately, as a small rural facility, that list is not terribly long, but we had some surprises. For example, we discovered a homegrown Access database that used data from coded claims to render certain financial projections. Several quarters of the hospital rely on those projections, so that database got onto our list to remediate. Had we not done the gap analysis, we wouldn't have known of its existence until it was too late.
The next step was to consult with our software and systems partners and document ICD-10 compliance plans. Actually, in our case that part was fairly easy. Our primary electronic health record vendor, CPSI, has released its ICD-10-compliant version, which we are about to begin beta testing. Our clearinghouse is also ready to begin testing. Payers, on the other hand present a current challenge. While our primary payers assure us they are on the cusp of being prepared, we do not have definite testing dates yet. That gives one things to ponder on sleepless nights.
We were fortunate in that 5010 went off without a hitch. We experienced little or no claim delay with that process and have been running successfully since January. Our advice on ICD-10 testing is to make sure all the glitches with 5010 have been fully resolved first and that the system is running in production without any delays or problems with reimbursement. 5010 is a precursor to ICD-10 claims and if problems remain there, ICD-10 testing is a futile effort. You would not know the source of any problems encountered — 5010 or ICD-10?
Our test plans call for setting one of our test databases to 5010 and to run parallel processes for at least 90 days. Of course, the key will be what we learn from the initial testing. We believe it will be pretty straightforward, to at least test the formats. Once we install the beta version for CPSI, coincidentally scheduled to go into our test environment as I write this, we will be in a position to drop ICD-10 technically compliant claims. That will be our first test: Can we produce and edit compliant claims. That does not mean however, that we will be complaint just because claims drop. I see the technical side of this to be the easy part. The formats either work or they do not; you find the errors and fix them. Straightforward, logical and a great beginning, but, are they coded correctly to get proper reimbursement?
Our original plan provided for training our coder trainers in Q1 2012. We did not waiver from that schedule when the delay was announced. We stayed our course. Further, Q2 of 2012 called for training all existing coders. We brought outside help in to assist our trainers and that went quite well. So we are ready to begin test coding parallel to ICD-9 production with the documentation presented to us. That is the sticking point. We believe that we could code about 80 percent of claims to some level of reimbursement. However, we do believe that over a third of those are improperly coded for the services actually performed. Some of those happen to be our highest dollar items — formative and attention-grabbing task at best and at worst, our nightmare scenario.
Sometimes being small is a blessing. We have the chance to meet regularly with all medical staff, and have been preaching the ICD-10 mantra for more than a year. Topics such as granularity, laterality, and specificity all have high awareness among the medical staff. We have actually conducted mini-ICD-10 demonstrations at medical staff meetings wherein we have taken a claim and its documentation, and walked through the coding process in front of them. Prior to the demonstration we had a medical staff member correct his documentation to ICD-10 standards, and then present the coding exercise to the group with the appropriate documentation. Unfortunately, all we could do was provide our best estimates of the reimbursement difference, as we do not have complete information from payers, but estimates were attention-grabbing, ranging from 20 to 100 percent differences. With this visible presentation to all medical staff, they agreed to detailed training. We currently have that planned for the latter half of 2013. Our point has been made, recognized and reluctantly accepted. We do not want to train too far in advance of actually being able to use the skills, but want to get far enough ahead of the deadline to be reasonably comfortable. We will determine that exact timeline when we get more solid direction from our major payers. It would be our intent to begin submitting ICD-10 claims as far in advance of Oct. 14, 2014, as possible. Why waste valuable time when the crunch will inevitably come?
To summarize: get close to your partners and work with them on their timelines. Talk to your payers and attempt to nail down their timelines. Educate the coding staff with whom you have more influence, and get them involved immediately. Educate your medical staff and obtain buy-in for a training plan. Convince them of the magnitude and implications of what is coming, because without them you fail.
Of course, this presumes you have done the gap analysis and laid out a plan to mitigate the gaps. Another advantage of being small is the enormity of our tasks is not the same as some of yours. Our numbers work to our advantage, with one exception: resources. That one rubs the same regardless of size. More size begets more resources but, at the same time, more systems and issues to address. Testing and training will be critical to success. My advice remains the same, if you have not started, start now. If you have begun, do not allow the delay to spark complacency. Time is our enemy here and it waits for no one.
Stephen M. Stewart is CIO of Henry County Health Center, a critical access hospital in southwest Iowa. He is a regular contributor to H&HN Daily's ICD-10 series.