Tips for improving outpatient coding, CDI and billing.
In moving from inpatient clinical documentation improvement (CDI) to outpatient CDI, I have been discovering some unique educational pointers. I know we are all very busy, so let’s just dive right into it.
- Doctors and coders, you should not have diagnoses that are mutually exclusive to one another on the same date of service. I am not talking about excludes 1; I am talking about situations that just make no sense. Let’s look at some examples:
- A patient should not have both CKD stage 2 and CKD stage 3 for the same outpatient encounter.
- A patient should not have diabetes with no complications and diabetes with specified complications on the same visit. Watch your electronic medical record (EMR); improper use is likely generating this contradictory documentation. And there is no such thing as “Schrödinger’s complications” when it comes to reporting diabetic manifestations.
- Morbid Obesity and malnutrition on an outpatient encounter? For that to stand, I believe you will have to elaborate on your rationale a bit more, rather than just listing it in your dx list.
- Don’t document diagnoses you don’t code. I regularly will see situations in which the outpatient provider does some form of monitoring, evaluation, treatment, or diagnostic testing (and even medical decision-making) for diagnoses that end up not being reported. Even changing up the medications or ordering a referral doesn’t always result in a code for the reported condition. Things like liver cirrhosis, thrombocytopenia, AAA without rupture, gastrostomy status, diabetic complications such as chronic kidney disease (CKD), transplant or amputation status, and even histories of substance abuse may receive clear attention from the provider in the note, yet they are not reported as diagnoses for the encounter.
- Code with specificity. Using checklists or some form of the old “superbill,” I have seen oncologists list “cancer of head and neck” as the ICD-10 code – yet clearly, the patient had cancer of the tonsil. With that discrepancy, you have lost your HCC and some of your severity weight. CKD is frequently reported as unspecified in the coding, although the stage is clearly present in the documentation. Pressure and non-pressure wounds, and even abscesses are often documented as “wounds,” and end up getting reported as traumatic injuries. Long-term use of insulin may be obviously present from the record, but no ICD-10 code is reported, again to the loss of an HCC. I have even seen pathological fractures reported as traumatic when the patient clearly had a neoplasm in the bone at the site of the fracture. I also see atherosclerosis of a bypass graft listed on the same record when no native disease is present. Pro tip: while we might assume native disease must be present in such a case, the data doesn’t get reported that way unless it is written and coded. Here is a good question to ask yourself; why is your ortho still documenting arthropathy and Charcot when it is clearly from diabetic foot, and not relating it to diabetes anywhere in his note? There are combination codes for just about all diabetic manifestations, and they should be reported as a unit 100 percent of the time. Another somewhat common issue is the failure to follow coding rules for the “with” associations. You might see a patient with hypertension, heart disease, and CKD, but the ICD code on the claim is only for hypertensive CKD. In a record with no explanation that heart failure isn’t resulting from hypertension, such reporting isn’t appropriate at all. One last pet peeve of mine: hemorrhagic thrombocytopenia being reported in a patient whose history reflects frequent DVTs, and hyper-coagulopathy with no explanation for the obvious discrepancy. You see this even on the inpatient side, where a patient has a history of hyper-coagulopathy and is on coumadin, but the coding staff refuses to report it because they are confusing the clinical information with the old coding rules, which dictate that a “high PT/INR is expected when you are on coumadin.” My response is simple: “no.” Hyper-coagulopathy is the diagnosis you are treating with the coumadin, and in most cases, it should be reported. Why are you quoting rules about hypo-coagulopathy as an expected response to coumadin? I hate to sound this way, but if you don’t know the difference between clotting too fast and not clotting fast enough, start there before continuing this discussion.
A few more pet peeves, for the record:
- Diagnoses reported with no substantiation in the note. We call this “clinical validation.” The code for “diabetic PVD” is on the claim form, while the entire note only says “diabetes, stable.” Patients whose cancer is clearly the definition of a “history only” (resolved, not receiving treatment, no recurrence, etc.) are frequently coded by the doctor, who notes that the patient actually still has cancer, as they use the active cancer codes. I have seen the combination code for hypertensive heart failure when the note mentions nothing about the patient even having heart failure in the first place.
- Body mass index (BMI) on the record with no clinically significant body habitus codes. What can I say?
- Per the Official Coding Guidelines, page 94: “BMI codes should only be assigned when there is an associated, reportable diagnosis (such as obesity).”
- Per Coding Clinic, the edition published for the fourth quarter of 2018, pages 81 and 82: “BMI codes are not intended for routine capture unless there is provider documentation of an associated diagnosis (such as for overweight or obesity).”
- Dropped HCCs. The old joke is ” did the amputated leg grow back this year?” Of course not. Add to that joke the fact that I do not believe that the patient’s diabetic CKD reversed itself, or that the severe systolic heart failure spontaneously resolved. Nor did the severe chronic obstructive pulmonary disease (COPD) and its associated chronic respiratory failure. A PCP should be addressing all chronic conditions such as these on every visit, period.
- Not connecting the dots. If you have a patient with severe degeneration of the spine with paresthesia and weakness, who needs prescription-strength anti-inflammatories or even steroids, you can get to an appropriate HCC by naming the diagnosis as inflammatory spondylitis/spondylopathy. If it is present and clinically accurate, why short-change yourself by describing the condition, yet not naming it?
- Logic disconnects. I will end with one more quip. When you state that your patient has major depression, unspecified, the code that gets reported is for major depression, single episode. How do you think it looks when someone sees that you believe your patient, who has had depression for 10 years, is having a “single episode?” Could it possibly be MDD recurrent at that point?
I believe there is a real need for some no-nonsense discussion about how outpatient reporting is done: from my perspective, not very well as an industry.