The patient has COVID-19? How is the infection manifesting? You need a code for that detail.
The rapidity with which our medical world is evolving is unbelievable. I know that aiming for optimal documentation is a worthy goal, but I recognize that right now, my clinical colleagues have more important things to do than worry if they captured all the CCs and MCCs. They are in the biggest fights of their lives, clinically, personally, and societally. My consulting business and provider documentation modules with CME have been tabled so I can try to assist in whatever way I can with addressing this pandemic.
I have two contributions to make.
The first was developing a template for documentation of COVID-19 in the emergency department. The second is trying to sort out how to use the World Health Organization (WHO) ICD-10 code of U07.1 sensibly. In an unprecedented move for an unprecedented condition causing an unprecedented global pandemic, the CDC adopted U07.1, COVID-19 into ICD-10-CM in a matter of weeks for an off-cycle update. It is no April Fool’s joke that U07.1 became valid as of April 1.
Let’s back up for a moment and get some history and context.
This disease, COVID-19, is caused by a coronavirus. There are seven known human coronaviruses, which get their name from the appearance of a “corona” or halo surrounding the virion (the virus particle) under electron microscopy. Some of them cause pedestrian upper respiratory infections.
The coronavirus named MERS-CoV causes Middle East Respiratory Syndrome (MERS), first identified in 2012. SARS-CoV causes Severe Acute Respiratory Syndrome (SARS), which started in China in 2002. I remember this outbreak distinctly, because it was the first time I felt scared to be an emergency physician.
A “novel” organism is one which has not been previously identified and to which human beings have not previously been exposed. Recent novel viruses (e.g., H1N1 from 2009, SARS-CoV-2 from 2019) are believed to have arisen from pathogens that affected animals (e.g., bats) being introduced into the human population. Without previous immunity, our entire population is susceptible to infection.
A novel coronavirus was identified in Wuhan, China on Dec. 31, 2019. The virus was first referred to as 2019-nCoV (2019 = year first discovered; n = novel; co = coronavirus; V = virus), but then was transitioned to the international virus taxonomy nomenclature of SARS-CoV-2. This is also how the disease got its name: COVID-19 (CO = coronavirus; VI = virus; D = disease; 19 = 2019). Hence, the novel coronavirus, SARS-CoV-2, is the causative organism of COVID-19.
On Jan. 31, WHO convened an emergency meeting to discuss the creation of a code, and they established the emergency code of U07.1 for 2019-nCoV, later adopting COVID-19 as the official name of the disease.
In the U.S., we originally had no way of pinpointing this pandemic condition in codes. Until we had a dedicated, specific code, the ICD-10-CM Interim Coding Advice had us using the specific condition manifested, such as viral pneumonia or acute (viral) bronchitis (due to other specified organisms) as principal diagnosis (PDx), with a secondary code of B97.29, Other coronavirus as the cause of diseases classified elsewhere, to indicate the causative organism.
The CDC responded to the WHO creation of a dedicated code for COVID-19 by planning to incorporate it without clinical modification into ICD-10-CM in an extremely expedited fashion.
The section U00-U49 is for provisional assignment of new diseases of uncertain etiology or emergency use. There is an instruction to “use additional code to identify pneumonia or other manifestations.” There are also Excludes1 codes listed of B34.2, Coronavirus infection, unspecified; B97.2-, Coronavirus as the cause of diseases classified elsewhere; and J12.81, Pneumonia due to SARS-associated coronavirus.
Last week, there was confusion as to what to do with a patient who had sepsis due to COVID-19. People were misinterpreting the advice as a mandate for U07.1 (when available, i.e., after April 1) to be the principal diagnosis. On March 31, the CDC published final ICD-10-CM Official Coding and Reporting Guidelines for U07.1. The update indicates that sequencing sepsis first is acceptable, according to Section I.C.1.d., Sepsis, Severe Sepsis, and Septic Shock.
Let me break this all down for you:
- If a provider documents a diagnosis of COVID-19 without an uncertain qualifier, code U07.1. Do not question the provider, implying they need to perform a confirmatory test.
I don’t need to do a laboratory test for measles or mumps; if I see a patient and determine either is present due to the constellation of signs and symptoms in the context of a breakout, I diagnose it, and it should be coded. During this time of pandemic, COVID-19 is overwhelmingly prevalent. If your provider sees hypoxemia out of proportion to the appearance of dyspnea, they don’t need a test – they know that patient has COVID-19. In the future, if COVID-19 becomes seasonal or episodic, your providers may need to do confirmatory testing to be certain.
There are multiple tests out there, with varying degrees of false negatives. Good nasopharyngeal collection is challenging to achieve. Some of the statistics I have seen indicate that approximately 30 percent of negatives are false. If there is a negative test on the record, but the provider’s clinical acumen leads them to believe the patient has COVID-19, and they document it as such, it should be coded. Do not query.
However, if the clinical validity for the diagnosis seems questionable from the clinical indicators, follow your institution’s clinical validation query policy.
- At the beginning (you know, a month ago), the patient would get a preliminary test and then it would be repeated for confirmation. A positive test at the local or state level is considered “presumptive,” whereas the CDC public health validation renders it “confirmed.” “Presumptive” is not an uncertain diagnosis qualifier.
The guidance is that a “presumptive” positive is coded the same as a confirmed positive or a diagnosis based on clinical judgment. Presumptive positive COVID-19 tests are coded as confirmed – U07.1.
- Uncertain diagnoses (e.g., possible, probable, suspected, yet to be ruled out, inconclusive, etc.) of COVID-19 are not coded with U07.1. There is a second code WHO made for ICD-10 (not ICD-10-CM) of U07.2, COVID-19, virus not identified, intended to give the ability to capture these patients. It has not been imported into ICD-10-CM (yet), so the guidance is to code the signs or symptoms and/or Z20.828, Contact with and (suspected) exposure to other viral communicable diseases.
- COVID-19 will be the principal diagnosis if:
- it is present on admission;
- it is the reason, after study, that occasioned the admission to the hospital; and
- it is not superseded by another condition, as per typical ICD-10-CM guidelines.
The instructions always (i.e., even in the interim instructions, before we had U07.1) told us that, just like any other condition, if the patient is pregnant, an O code should be used as PDx. We are instructed in the final guidance to use O98.5-, Other viral diseases complicating pregnancy, childbirth, and the puerperium, but I think additionally O99.5-, Diseases of the respiratory system complicating pregnancy, childbirth, and the puerperium, might be apropos.
The final guidelines released on April 1 refer us to the sepsis section for instructions on how to handle sepsis due to COVID-19. My philosophy is that whenever there is a spectrum, the additional condition should not be considered integral to the underlying condition, and there should be a corresponding code. A patient may be asymptomatic, may have flu-like symptoms, or may have life-threatening acute hypoxic respiratory failure and sepsis from COVID-19. Sepsis needs to be captured to tell the story, and the sequencing is as per usual. If the patient presents with and is admitted for sepsis secondary to COVID-19, the principal diagnosis should be A41.89, Other specified sepsis, and U07.1 can serve as a secondary diagnosis.
If the patient is admitted for another condition and has concomitant but unrelated COVID-19 (e.g., acute appendicitis), the other condition is PDx and the COVID-19 (POA-Y) is a comorbidity.
If the patient is admitted for an unrelated other condition and develops COVID-19 during the hospitalization, U07.1 is a secondary diagnosis. You may need to query for POA status; it may depend on how long into the hospitalization the symptoms develop. The incubation of COVID-19 has a range of 2-13 days. The provider may need to use their judgment to determine POA status.
- I believe that you always use as many codes as you need to fully flesh out the patient’s situation. Once you have established that a patient has COVID-19, the question is: how is the infection manifesting? You need a code for that detail.
Before we had U07.1, that manifestation (e.g., viral pneumonia) was the PDx, and the code giving the detail that it was from a coronavirus was provided by B97.29, Other coronavirus as the cause of diseases classified elsewhere (causes of diseases classified elsewhere codes are never PDx). Now, use U07.1 as PDx and the detail of the presentation of the disease as a secondary diagnosis. For example, J12.89, Other viral pneumonia; J20.8, Acute bronchitis due to other specified organisms; and J80, Acute respiratory distress syndrome, will specify the manifestations of the COVID-19 illness.
It is preferable that your providers “go through the hASSLe” of detailing Acuity, Severity, Specificity, and Linkage, rather than using a diagnosis that would result in an unspecified code as exemplified in the guidelines of J40, Bronchitis, not specified as acute or chronic, or J98.8, Other specified respiratory disorders, for “respiratory infection.”
- If there has been known or suspected exposure to COVID-19 and the patient does not have a confirmed or presumptive case (that is, the patient has either a negative or pending laboratory test, or an uncertain diagnosis), use Z20.828, Contact with and (suspected) exposure to other viral communicable diseases as the reason for the encounter, with or without associated symptoms as additional diagnoses.
If there is known or suspected exposure and the patient tests positive or is deemed to have COVID-19 by clinical judgment, do not code Z20.828, as per Excludes1. That patient only is coded with U07.1 (plus manifestations, if applicable).
- If there is concern about possible exposure, but the exposure and COVID-19 are ruled out after evaluation, you use Z03.818, Encounter for observation for suspected exposure to other biological agents ruled out. Obviously, if the exposure is ruled out, but the patient turns out to test positive, the diagnosis is U07.1, COVID-19.
- If the patient is asymptomatic, has no known or suspected exposure, and they are deemed negative or have a pending test, you use a screening code. Z11.59, Encounter for screening for other viral diseases. Screening examinations, by definition, are only done on asymptomatic patients.
- Documentation of a positive test in an asymptomatic patient is coded with U07.1, COVID-19. Although their case is asymptomatic, they are considered infected and infectious. This is different than B20, HIV disease versus Z21, Asymptomatic HIV infection status.
In conclusion, think about the clinical situation and tell the story:
- If the provider documents COVID-19, code U07.1, plus the manifestations of the viral illness, such as pneumonia or ARDS.
- If the preliminary testing is positive and the provider documents presumptive COVID-19, code U07.1, plus the manifestations of the viral illness, such as pneumonia or ARDS.
- If the testing is negative, pending, or was not performed, but the provider documents “possible/probable/suspected/likely, etc. COVID-19,” code the signs or symptoms and/or Z20.828, Contact with and (suspected) exposure to other viral communicable diseases (depending on whether or not there was exposure).
- If there was documented exposure to COVID-19 but no diagnosis of U07.1, nor symptoms, capture Z20.828, Contact with and (suspected) exposure to other viral communicable diseases, as the reason for the encounter.
- Asymptomatic, “worried-well” patients without exposure who test negative, and who do not have a mental illness that could serve as principal or first-listed diagnosis, should be coded as Z11.59, Encounter for screening for other viral diseases.
- COVID-19 follows usual sequencing rules for obstetrical patients and sepsis.
- Have the providers remember to document comorbidities, which contribute to the complexity and severity of each patient.
I hope this eliminates confusion about how to code COVID-19. Comply with stay-at-home orders. Be well. Try to stay calm and stay sane. This will eventually pass, but at least now we have a code for it!