ICD-10 could bring with it a 15% drop in provider revenue due to incomplete charts and claims rejections. Thankfully, a few basic principles can keep you up to date, and [hopefully] in the black.
Dear Director: ICD-10 just went live and while our group has done some basic training, my providers keep reminding me they are clinicians, not coders (or bean counters). Any pointers to help decipher the code?
When my Health Information Management department asked me to start getting educated on this a year ago, I wasn’t that interested. I had enough issues on my desk without one with a hypothetical start date a year away. But now that ICD-10 is live (went into affect October 1st), it’s time to get everyone educated and onboard. After all, the implementation and utilization of ICD-10 codes is as equally relevant to the boots on the ground clinician as it is to the hospital or group administrator and treasurer of your practice.
Created by the World Health Organization in 1992 and implemented by many countries around the world in the 1990s, the U.S. is finally joining the ICD-10 platform after years of delays and deferrals. While the U.S. government voted to adapt ICD-10 in 2009, CMS delayed implementation for some time. The initial timeline had ICD-10 starting earlier this year without any grace period; CMS has now finalized the start date and is allowing for a 1-year grace period for government payers. ICD-9 has been limited to approximately 14,000 codes and has reached saturation; this problem goes away with the near 69,000 codes available with ICD-10. The main difference is specificity, and this is the key word here; while de-emphasizing the NOS (not otherwise specified) or chief complaint driven diagnoses, ICD-10 is more focused on the source of the problem. This not only allows for more accurate diagnoses, but also is intended to assist data tracking of medical care and research throughout the world. Using an ankle sprain as an example, ICD-9 contains four codes to cover the clinical condition, while ICD-10 contains 72 codes, indicating the specific ligament, the side of the injury, and the nature of the encounter, whether initial or follow-up. Since we typically won’t have an MRI of an ankle when seeing the patient with these injuries in the ED, we can make some diagnoses clinically or use the unspecified codes (ankle sprain of unspecified ligament, left ankle, initial encounter code S93.402A) Making your best clinical diagnosis of which ligament is involved that is suported by your medical documentation also won’t be accounted against you. Of course, if you’re like me, you use the NOS diagnoses a lot for common complaints like headache, chest pain, and abdominal pain. Don’t worry, there will be some symptom based diagnoses, just asking for more details, like atypical or non-cardiac for chest pain or the location of the pain for abdominal pain. The start date for ICD-10 implementation was October 1, 2015. If that date sounds familiar, it should. It’s also the same date that we began the SEP-1 bundle. Two major CMS changes activated on the same day? The truth is, while the SEP-1 presents a number of challenges, the conversion to ICD-10 can be managed with a few basic principles.
Some industry experts predict a 10-15% drop in revenue as we adjust to the new codes. I certainly can’t afford that kind of pay cut, nor can my group. This potential drop in revenue is because we anticipate issues with incomplete charts and increases in denials and claims rejections. Therefore, we need a solid plan in place to implement an accurate use of ICD-10 correctly as soon as possible. From a big picture point of view, your EMR should be updated with the new codes and now you need to educate your docs. Afterwards, you’ll need audits and review of data to see where the issues are and then refocus efforts on individuals who are lagging behind.
Since specificity is the key, first off recognize the information needed on your diagnosis line is already present in other places on your chart. Despite more than four times as many codes, medicine is not more complex that it was last month. At the heart of the coding is an increased need for specificity with enhanced and detailed documentation required in your note. (Hint: if you don’t have scribes, now’s the time to build the return on investment argument.) For example, a patient who has a mechanical fall with a subsequent wrist fracture would historically have those two diagnoses, fall and wrist fracture. Now, with greater specificity, the “fall” turns into “tripping with fall over animal (W01.0) or carpet/rug (W18.09) or ice or snow (W00.0).” The wrist fracture turns into “fracture of lower end of left distal radius, initial encounter, of closed fracture.” While this may appear overwhelming at first blush, remember that these details already appear in your HPI and radiology report. It’s our job to pull those facts down and place them in the diagnosis line. Recognizing this simple principle will make all the difference for your groups practice. A coder cannot extrapolate this information from your chart or from the radiology report. You must put the details in the diagnosis line yourself.
Some general principles to guide your coders to the right code:
- If the injury or illness involves something we have two of, we need to mention which one (laterality).
- For traumatic injuries, it’s important to describe the condition with the most detail. A fracture, for example, include laterality, location, type of fracture, union vs malunion, and type of visit (initial, subsequent, or sequela)
- For chronic conditions, you will also need to include as much detail as possible (CHF, acute vs chronic, left vs right heart failure, diastolic vs systolic dysfunction; COPD, emphysema vs chronic bronchitis vs asthma, acute vs chronic).
- Also with chronic conditions, you will need to identify othe current state of health to the chronic condition. For example, for a patient with DKA, you would add secondary to insulin dependent diabetes, secondary to pneumonia vs non-compliance vs any other cause.
- There is some flexibility with a STEMI, and acute CVA. Using the root tree, there are sections for artery (STEMI) and for CVA (embolic, thrombosis, occlusion and then which artery), but often times that information is not available at the time of diagnosis. In that case, where the artery may not be known for the STEMI or CVA, you could give a diagnosis such as STEMI, unspecified or Cerebral Infarction, unspecified.
What about this grace period I keep hearing about? Will our providers be given a benefit of the doubt without bankrupting our practice?
On July 6th 2015, CMS and the AMA made a joint statement allowing for a one-year grace period preventing automatic claim rejections based on specificity alone. This statement focuses on government payers only, leaving less benefit of the doubt afforded to us by the private insurers. The caveat here is that you provide the coders with enough detail to get in the right family of codes. The grace period extends beyond the E/M codes, and enters the realm of meaningful use (MU) and PQRS programs. Finally, it is the expectation that the Medicare Audit Contractors (MAC) and Recovery Audit Contractors (RAC) also acknowledge the grace period. With all that said, and extending beyond the scope of medicine, my experience with Government agencies tells me that extra attention to detail and front-end management can be a saving grace from long-term complications. As it pertains to chart submission, an insurer can certainly return the chart for clarifications. While not an outright rejection, the delay in payments for services already rendered can have significant downstream effects on your revenue stream. Keeping in mind that there is a fixed amount of time we have to submit charts for payment, since insurance companies won’t be rejecting charts (but asking for clarification), it may take additional FTEs on the coding/billing/documentation side to manage this additional workload. And considering the private payers have not agreed to the grace period, it’s even more important that we do a good job from the beginning. The one constant to keep in mind from a management perspective is that our docs will need their paychecks to come in on time despite a likely delay or decrease in our revenue cycle.
Any thoughts about charts being rejected for other reasons?
If you haven’t started already, now is the time to rev up your EOB appeals staff or communicate with your coding company about how they plan on navigating the transition. While I have no doubt that Medicare and Medicaid will honor the grace period, remember that this applies to rejections only. How the private firms manage this transition is still a bit of an unknown. I would certainly expect carriers to seek clarification or request more records, causing a significant and immediate effect on reimbursement. It’s safe to assume there will be delays in reimbursement.
Where do you see the biggest pitfalls with ICD-10 implementation?
I’d have to say that most hospitals and hospital systems have already started training providers and given them some form of an EMR/EHR upgrade that will allow you pick the closest and most reasonable diagnosis used in emergency medicine. The biggest pitfall I see is with those systems still on paper charts. Without being prompted, the potential exists for providers to list diagnoses that are not ICD-10 compatible. An interesting fix here would be to incorporate a mobile app or some form of tech adjunct to make sure you hit the mark. After that, the next most proximate landmine would be legibility. As most groups have seen, the number of audits nationwide has increased exponentially over the past few years, and even with the most bulletproof chart, there is just no defense for illegible charts. The same would apply to the diagnosis line. Penmanship matters, even in the digital world
The conversion to ICD-10 represents a significant overhaul of our coding philosophy. It is not just a coding change, it is a process change. With the right training and understanding of basic principles of diagnoses, this conversion doesn’t have to cripple the practicing clinician. Just like when you implemented your latest EMR, change leads to certain problems. Some you can anticipate while others you discover along the way. There is a learning curve that makes for a painful month or two but ultimately, you get your ED machine running full speed again. Since the rest of the world has been able to transition to ICD-10, I anticipate that we’ll be fine down the road. Make sure your docs and staff are ready. Remember, specificity is the key so include as much detail as possible. For traumatic injuries, make sure to include what you would typically see in a radiology report (location, laterality, type of fracture, and type of visit). If it’s a strain or sprain, make sure to identify the muscle, ligament, or tendon involved. For medical conditions, remember to mention the type of condition, level of chronicity, and whether the current condition is related to a chronic problem (linkage). A lot of the information in the diagnosis line can be obtained from other areas of the chart, but recognize that the coders can’t use it unless you put it there. The good news is, we now have a way to accurately identify and code some of those common scenarios we see in the ED: “Problems in relationship with in-laws “(Z63.1), “bizarre personal appearance” (R46.1) and “burn due to water skis, subsequent encounter (Y93.D:V91.07XD).” Finally, make sure to capture those rare second events, such as “sucked into a jet engine, subsequent encounter” – there is a code for that as well.