E-News Exclusive

How I Learned to Code Clinically

By Ben Sayabovorn, RHIT, CCS

The more coders are aware of the rationale and motivation behind how and why clinicians document, the more coders can code clinically.

Through conversations with coders, nurses, midlevel providers, and physicians over the years on clinical documentation improvement (CDI) projects, I began to see key differences between how coders and clinicians view documentation and how understanding those differences eventually improved my ability to code. Coders see the medical record as a document that needs to be translated to codes. Clinicians see the medical record as a tool for communication and a data repository for patient information. This article will illustrate examples of some of the differences between the coding and clinical perspectives.

Treatment vs Monitoring
A patient was found to have a sodium level of 130, and no specific treatment (eg, IV fluids) was ordered. Should a query be sent clarifying whether the patient has hyponatremia even though it wasn’t treated?

Per coding guidelines for capturing diagnoses, a condition can be coded if it is monitored, evaluated, or treated. So, if repeat labs are ordered to monitor the sodium, the requirement is met. Whenever I query physicians, I ask whether a condition is being “managed” rather than “treated.” Treatment implies that a more definitive intervention was performed. Alternatively, I would ask whether their medical decision making was altered in any way due to a clinical finding (eg, sodium of 130). If the answer is yes, then I would ask that they document a diagnosis to reflect the clinical effort required to manage the patient’s condition.

Another good example of a condition that is not treated directly (unless surgery is performed), but impacts patient management is morbid obesity. Many times internal medicine doctors say that they will not document morbid obesity because they are not treating it. This is when I ask whether the morbid obesity is impacting their management of the patient; the answer is often “yes.”

Symptoms
I noticed that physicians consistently document symptoms in place of or in addition to definitive diagnoses. For example, they would document sepsis and fever and then document only fever, without mentioning sepsis. This inconsistent documentation would often result in a physician query to ask whether sepsis was ruled in or out.

One of the reasons for documenting symptoms is for physician billing purposes (professional fee). Medical doctors bill for various evaluation and management (E&M) levels of service; this requires that they document the following three components: history, physical exam, and medical decision making (documenting at least the symptoms is necessary to meet E&M documentation requirements). The specificity of the diagnosis documented does not significantly impact the E&M level billed and therefore is not a primary focus for physicians (documenting symptoms can be sufficient for their billing). Knowing that physicians are documenting symptoms for their own billing purposes and not necessarily as the principal diagnosis encouraged me to seek clarification when documentation was unclear. So, if physicians stop documenting a specific diagnosis and only document symptoms, a query clarifying whether a “condition has resolved” may be a good approach to get stronger documentation.

Physicians Orders
A way I have approached cases where there is a lack of diagnosis specificity/consistency is to look more closely at the physician orders to get insight into what a physician is thinking. We always encourage our CDI specialists to pay very close attention to what is ordered and the timing of orders. For example, if a physician suspects a patient has pneumonia, he or she will typically order standard labs (eg, complete blood count), an X-ray and/or CT of the chest, antibiotics, and a pulmonary consultation. So, if a coder is unsure whether a condition should be coded or a query is necessary, the physician orders may offer clues as to the physician’s thought process. If the physician does not order tests or any other workup for a condition, a query may not be necessary.

The physician orders are also a good place to confirm whether a condition is still being treated. For example, if a physician believes a patient has pneumonia, he or she will order antibiotics and keep the patient on them for the full course (typically seven to 14 days). If the antibiotic is discontinued within a day or two of being ordered, it is likely pneumonia was ruled out.

Sepsis
If a patient with sepsis is discharged after just two days, does that mean the patient did not have sepsis? Should a query be placed to confirm whether the patient had sepsis?

It may seem counterintuitive, but the hospital is the worst place to be when sick. Patients who have an illness are more likely to contract hospital-acquired conditions (eg, Clostridium difficile colitis) than healthy patients. Physicians will send patients home to finish their course of antibiotics as soon as they believe acute treatment/monitoring is no longer necessary. Just because a patient goes home quickly does not mean the patient was not very ill. A query is not necessary if the clinical picture, treatment, and documentation for sepsis are evident.

Learning how to code is a never-ending process. The more exposure coders have to CDI specialists and practicing clinicians, the more they can build their ability to code clinically.

— Ben Sayabovorn, RHIT, CCS, is a manager at Provident Consulting with more than 10 years of experience with HIM, inpatient coding, and clinical documentation improvement.