A Clinician’s Comments on Coding Clinic

The Second Quarter Coding Clinic was posted recently by the AHA.

The American Hospital Association (AHA) Coding Clinic for the second quarter of 2022 came out several weeks ago. I reviewed the publication, and it will come as no surprise to you that I have thoughts on the advice. Remember that I do not offer any sort of official guidance in the coding hierarchy – but here are my comments anyway!

On page 5, there is a question regarding postpartum sepsis due to a urinary tract infection. This highlights the coding-clinical disconnect regarding temporal references. Just like “postoperative” to a provider means “in the period of time following an operation,” “postpartum” to the clinician means in the period of time following delivery.

Puerperal sepsis, however, is a specific entity. It accounts for 10-15 percent of maternal deaths in the postpartum period, and is often caused by group A streptococcus. The World Health Organization (WHO) defines puerperal sepsis as an infection of the genital tract occurring at any time following rupture of membranes, to the 42nd day following delivery, with specific signs including pelvic pain, fever, abnormal vaginal discharge, and delay in uterine involution (shrinkage of uterus after delivery).

The latest Coding Clinic explains that a urinary tract infection, even if it causes sepsis one week after a vaginal delivery, is not considered “puerperal sepsis” and shouldn’t be coded as such. I would welcome the proposed expansion of O85 to add a code for postpartum systemic sepsis without infection of the genital tract.

On page 7, a question is posed regarding a pressure injury to the mucosal membrane of the penis from an indwelling urethral catheter. This type of injury is not coded with a L89 code, indicating pressure injury of skin and subcutaneous tissue. The advice is to use T83.511A, Infection and inflammatory reaction due to indwelling urethral catheter, plus N36.8, Other specified disorders of urethra. It seems to me as though T83.89XA, Other specified complication of genitourinary prosthetic device, implant and graft, would be more accurate, because a pressure injury is neither an infection nor an inflammatory reaction. In my opinion, it is a mechanical erosion from the pressure causing direct damage to the mucosal tissue.

The next question is related, and the advice explains that mucosal pressure injury is “an inflammatory reaction caused by the pressure of the device on the affected area.” However, they select an “other specified complications of surgical and medical care, not elsewhere classified” because they do not designate an endotracheal tube as an internal prosthetic device, implant, or graft (I’m not sure that a Foley catheter is strictly a genitourinary prosthetic device, implant, or graft either, but that is where those complications are housed). Perhaps there should be consistency, and we should use T88.8XXA, Other specified complications of surgical and medical care, not elsewhere classified, for all mucosal pressure injuries, with an additional code to signify the site involved.

Speaking of pressure injuries, my interpretation of the next advice on page 8 is that deep-tissue pressure injury (DTPI) present on admission (POA) that is subsequently revealed to be a specific stage during the admission can be coded as the higher Stage IV because of how the documentation was done in this particular instance. It does not seem to be generalizable that DTPI can always go to the final stage revealed during an encounter and be considered POA-Y. I think this is unfortunate, because I believe there are PIs deemed quality issues that were most likely already brewing on admission, but not able to be accurately classified at the onset.

On page 9, Coding Clinic gives dispensation to take documentation of “Class 3 obesity” to E66.01, Morbid obesity due to excess calories. The important takeaway to me would be if a patient has “Class 2 obesity,” defined as a BMI of 35.0 to 39.9, a query may be indicated to sort out if it is really E66.01. Having a BMI of 35 to 39.9 with an obesity-related comorbid condition (such as hypertension or Type 2 diabetes) is also considered morbid obesity.

On page 10, there is a question regarding a patient who suffers a serious adverse effect from treatment of hyponatremia. Dr. Ronald Hirsch’s sensibilities were offended when Coding Clinic used a term that has evolved since he and I started practicing a hundred years ago: central pontine myelinolysis. It is more common to refer to the clinical entity as osmotic demyelination syndrome (ODS) now, because it doesn’t seem to be confined to the pons. However, if the patient developed locked-in syndrome, it indicates pontine involvement, and G37.2, Central pontine myelinolysis, does seem to be the best code.

The question regarding serotonin syndrome, which has no unique code, is a good reminder that we use codes for the manifestations of the syndrome as a surrogate. Remember, a code indicating adverse effects of medications is coded after the clinical manifestations, as opposed to manifestations being listed after poisoning codes.

Let’s go to page 14. A patient has an automatic implantable cardioverter defibrillator (AICD), but the cardiologist documents that there is no ventricular fibrillation (VF) noted at evaluation. Is VF considered a chronic condition meeting reporting requirements? Previous Coding Clinic advice was referenced. The guidance is that VF should not be captured in this instance. If the patient had an episode of VF triggering the AICD, then it would be. VF is an intermittent dysrhythmia – patients do not walk around in chronic VF (that is known as “death.”) If it occurs initiating or during the encounter, then it should be coded. The distinction with a pacemaker for sick sinus syndrome (SSS) is that SSS is a chronic condition. If the pacemaker were turned off, bradycardia and/or pauses would be noted.

The question at the bottom of page 16 highlights the difference between maternal care O codes and newborn P codes. Say a pregnancy patient undergoes a caesarean section at 38 weeks due to placenta previa, which is complicated by light meconium-stained fluid. The Apgar score is positive. The question paraphrased is “do you use O77.0, Labor and delivery complicated by meconium in amniotic fluid even if there hasn’t been fetal distress and the maternal care is affected?” I agree with the response. The maternal care is affected by the fact that there was meconium staining, which could cause aspiration and fetal distress and needs to be evaluated, so O77.0 is appropriate. Now, the next question is whether a P24.- code is indicated. That hinges on whether there was meconium aspiration on the part of the neonate, and did it cause respiratory distress or not? O for Obstetrics, P for Perinatal.

On page 27, Coding Clinic explains how to code EVUSHELD. This is a medication to prevent COVID-19, usually used in patients who are immunocompromised. Z29.8, Encounter for other specified prophylactic measures, is the prescribed code.

On page 28, the advice is correct, but the reasoning is faulty. The question is whether to code Z20.822, Contact with and (suspected) exposure to COVID-19, for a term newborn whose mother had COVID-19 during her second trimester. It isn’t picked up because the infant wasn’t affected “(e.g., small for gestational age),” as offered by CC; and it isn’t picked up because at the time of delivery, the mother no longer had COVID-19. If the father was in the delivery room with known COVID-19 and held his baby, they could clearly code it, because there would be contact with COVID-19. If the baby doesn’t need to be tested (i.e., clinically evaluated), I guess you could consider it “implications for future healthcare needs,” wherein the baby would need to be watched for signs of infection.

The advice regarding an immunocompromised patient with sickle cell disease who presents in painful crisis triggered by a COVID-19 infection is generalizable. The COVID-19 infection is recorded with U07.1. Any condition triggered by COVID-19 would also be coded (e.g., sickle cell crisis, acute myocardial infarction, pulmonary embolism), and the sequencing would depend on the circumstances of the admission. This is different from pneumonia from SARS-CoV-2, because that is a direct manifestation caused by a viral attack on lung tissue. The others are secondarily caused by the viral infection (or possibly its resultant hypoxemia or hypercoagulability).

Finally, I disagree with the response regarding the request for clarification regarding a mental disorder during an ED visit. The question was, should a mental disorder be coded even if it was not treated during the current encounter, and there was no documentation to support that the condition affected patient care or management? This is more nuanced than the response, which was that the advice did not conflict with the Official Guidelines (Section IV.J.), which state “code all documented conditions that coexist at the time of the encounter/visit and require or affect patient care, treatment or management.”

There are conditions that affect the patient evaluation or treatment, but the provider does a lousy job of connecting those dots. For instance, I posit that there are very few “incidental” pregnancies. If I see a patient in the ED and they need a medication for their unrelated condition, I consider the fact that they are eight weeks pregnant quite relevant, even if I don’t document it explicitly (which, of course, I would!)

If a patient has a mental disorder like anxiety or depression, well-controlled by meds, and they present for an ankle sprain, that might not be relevant. However, if they have schizophrenia and it takes my nurses 20 minutes to coax them into allowing a blood draw, or if I have to consider their maintenance psychiatric medication when I select a new medication to minimize the chance of interaction, their mental disorder is eligible for coding. Providers do a lot of mental calculation that, frankly, they don’t have the time or inclination to spell out in every record. If it is important to capture that diagnosis, the coder/clinical documentation integrity specialist (CDIS) may need to query for clarification. But stating out of hand that a mental disorder shouldn’t be coded because it isn’t being treated or documented as affecting patient care or management doesn’t take into account that the documentation may just be lacking.

Those are my thoughts on AHA Coding Clinic, Volume 9, Number 2, for the second quarter of 2022. If you haven’t had time to read it, I recommend you do so. There were other topics on which I did not comment.

I hope you appreciate all the hard work Nelly Leon-Chisen and her folks do at AHA as much as I do.

Programming Note: Listen to Dr. Erica Remer every Tuesday on Talk Ten Tuesdays when she cohosts the broadcast with Chuck Buck, 10 Eastern.