Diabetic Kidney Disease and the Third Quarter Coding Clinic

A clinician’s perspective on the AHA’s third-quarter coding clinic.

I spoke at the inaugural combined Oregon Health Information Management Association (OHIMA) and ACDIS conference in Albany, Ore. last week. One of the comments I hear often when I speak to coders involves how much they appreciate getting a clinician’s perspective. This is why I feel compelled to share with you my thoughts on the Coding Clinic. I recently reviewed the edition published for the third quarter, and I am going to recognize National Diabetes Month now, too.

The first response was not clinically accurate. The questioner stated that the diagnoses were documented as “ESRD due to diabetic nephropathy on dialysis, diabetic retinopathy, diabetic peripheral neuropathy, and hypertension.” First, kudos to whoever trained that doctor to give such clear linkage and specificity! That is an excellent diagnosis list.

The question posed was whether the automatic linkage to hypertensive chronic kidney disease was appropriate, in this instance. Chronic kidney disease (CKD) is often multifactorial, and the combination of diabetes and hypertension often leads to CKD. In fact, control of blood pressure in the presence of diabetes is considered more important than glycemic control (see Diabetic Kidney Disease: Chronic Kidney Disease and Diabetes, by Jerry Yee, https://doi.org/10.2337/diaspect.21.1.8).

If the provider had documented “ESRD due to diabetic nephropathy and hypertension,” the appropriate codes would have been E11.21, Type 2 diabetes with diabetic nephropathy, I12.0 Hypertensive chronic kidney disease with stage 5 chronic kidney disease, and N18.6, End-stage renal disease. It is often difficult to sort out which disease process is the dominant cause, and likely, they both have an impact and share responsibility. In this case, however, the provider linked the ESRD to diabetes and listed hypertension as a separate diagnosis, so I agree with not assuming linkage.

The incorrect portion of the response came as an aside at the end, where it was stated that “it would be redundant to assign codes for both diabetic nephropathy (E11.21) and diabetic chronic kidney disease (E11.22), as diabetic chronic kidney disease is a more specific condition.”

It is true you wouldn’t code both. Diabetic nephropathy is a specific subset of CKD. It is an advanced renal disease due to microvascular damage from hyperglycemia, manifested by proteinuria. I again refer you to the article referenced above; diabetic kidney disease includes diabetic nephropathy and other parenchymal kidney diseases, without proteinuria. If your provider has done their due diligence and given you the specificity of diabetic nephropathy, you should be coding diabetic nephropathy.

More on the Coding Clinic
A question was also asked regarding the progression of pre-eclampsia from mild to severe, and what the appropriate code assignment and present-on-admission (POA) indicator would be. The advice was to pick up the most severe stage during the encounter and to assign POA-Y because the pre-eclampsia had been present on admission.

It offends my sensibilities that coding rules are not applied consistently. Why should there be two different stage codes with different POA indicators for pressure ulcers, but this is the advice for pre-eclampsia? Either always have two codes with the appropriate severity stages and the corresponding POA indicators, or always pick up the most severe stage, and the POA indicator reflects whether or not the condition was present on admission initially, regardless of progression.

At the very least, if a patient has a pressure ulcer in the same site, present on admission, if it progresses to a higher stage, the HAC designation should be nullified.

On page 13, there are two questions regarding pyelonephritis and kidney stones, with apparently conflicting advice. This is a coding-clinical disconnect.

In the first case, “pyelonephritis with bilateral nonobstructive renal calculi” is noted on CT scan. The indexing for “pyelonephritis” with acuity unspecified mandates exclusively using the code for the calculus and the questioner is puzzled. The stone was not the reason for admission; the kidney infection was the reason. Although serving as a nidus for infection, surely kidney stone as the sole diagnosis does not tell the whole story.

In the next related question, the patient has acute pyelonephritis and nephrolithiasis, and the advice is to use two codes: N10, Acute pyelonephritis, and N20.0, Calculus of kidney. Coding Clinic points out that there is no Excludes1 prohibition.

These are the correct diagnoses for both patients, but the coding instructions prevent obtaining the correct codes in the first case. If guidance leads you to a code that seems wrong, you should do more research, and sometimes may need provider clarification. Our goal is to accurately depict the patient encounter. The proper action for the coder or clinical documentation improvement specialist isn’t to just pick up the N20.0 because the guidance says so; they should query for the acuity of the pyelonephritis, thereby permitting the accurate capture of both conditions and codes.

On page 16, we have more advice that makes no sense clinically. A patient who underwent total knee arthroplasty undergoes a wisdom tooth extraction and is diagnosed with a hematogenous prosthetic knee infection. The CC interpretation was that “the infection of the knee is not due to the prosthetic joint, but occurred secondary to another source.” They recommended using an M code for arthritis due to other bacteria, and the Z code for the presence of a left artificial knee joint.

Having an infection in a joint or body part that has a foreign body (e.g., joint prosthesis, artificial valve, implant, catheter, calculus, etc.) is always related. “Hematogenous” suggests the germs spread from a distant site. The dental intervention caused bacteremia, and the bacteria settled in the joint because there is a prosthesis. It isn’t a coincidence. An explanation may be required to eliminate the infection.

T84.54XA, Infection and inflammatory reaction due to internal left knee prosthesis, initial encounter, is most assuredly appropriate for this patient. I hope the Coding Clinic prints a retraction or correction as they did for the next situation.

They rescinded the advice to pick up “lobar pneumonia” when the provider notes there is pneumonia in a specific location, i.e., right upper lobe. We are instructed to use the code, J18.1, Lobar pneumonia, unspecified organism, only when the provider documents “lobar pneumonia,” typically for consolidation of a lobe, not for patchy infiltrates. Good call.

It would be best for the American Hospital Association (AHA) and us all to remember that the Coding Clinic has both “coding” and “clinic(al)” in it. I really wish that the advice always reflected clinical practice, to get to accurate codes.

Programming Note: Listen to Dr. Remer’s live during Talk Ten Tuesday, 10-10:30 a.m. And to read more of her reporting, visit ICD10monitor.com immediately following today’s Talk Ten Tuesday broadcast.