Let’s trade: I’ll eat a peach if you’ll select a specific diagnosis code

When are two diagnosis codes better than one?  And when is one enough?

 Physicians get conflicting advice about how many diagnosis codes they should document and report.  When working in the hospital, Health Information tells them to document every sneeze, because hospital payment is often based on the severity of the patient’s condition and underlying medical problems.  In the office, most physicians are paid based on the CPT code and associated fee, and not on the diagnosis code.  However, more practices have risk based contracts that vary payment incentives based on the acuity of the practices panel of patients and Medicare’s shared savings programs will include acuity as one of the factors in its risk based incentive and payment programs.  This increases the importance of accurate and complete diagnosis coding on physician claims.

 Here are the guidelines for diagnosis coding for office/outpatient physician services:

  1. Use the ICD-9-CM codes that describe the patient’s diagnosis, symptom, complaint, condition, or problem.
  2. Use the ICD-9-CM code that is chiefly responsible for the item or service provided.
  3. Assign codes to the highest level of specificity. Use the fourth and fifth digits when indicated as necessary in your ICD-9-CM volumes.
  4. Do not code suspected diagnoses in the outpatient setting. Code only the diagnosis symptom, complaint, condition, or problem reported. Medical records, not claim forms, should reflect that the services were provided for “rule out” purposes.
  5. Code a chronic condition as often as applicable to the patient’s treatment.
  6. Code all documented conditions which coexist at the time of the visit that require or affect patient care or treatment. (Do not code conditions which no longer exist)

 Do not unbundle diagnosis codes, and don’t use two diagnosis codes when one is a symptom of the otherTwo heads are better than one, aren’t they?  So, two diagnosis codes must be better than one, too.  Not necessarily.  If the patient presents with joint pain, and tendonitis is diagnosed as the cause of the joint pain, use only tendonitis on the physician claim.  Joint pain is a symptom of the condition and would not be separately reported.  If the definitive diagnosis is known, use it without adding the symptoms of the condition.  If the definitive diagnosis is not known, code the symptoms.

Do not free text a few question marks in front of a diagnosis code and then select the code.   Such as this:  “?? Other venereal disease due to chlamydia trachomatis lower genitourinary sites 099.53.”  Doing this assigns the diagnosis to the patient.  Use the patient’s symptoms, if any. If the patient does not have symptoms, use a screening diagnosis such as V74.5 Screening examination for venereal disease.  Also, do not free text, “possible” “probable” or “rule out” in front of the diagnosis code in the electronic health record.  In many EHRs, this assigns the code that goes on the claim form.  Note the instructions, “Medical records, not claim forms, should reflect that the services were provided for “rule out” purposes.”

 Do assign underlying conditions if considered or assessed during the visit.  For example, a patient presents to General Surgery for gall bladder disease.  The patient’s underlying medical conditions include sleep apnea, diabetes and hypertension.   The surgeon wants a medical clearance before scheduling the surgery.  The surgeon is not treating the medical conditions, but these affect the surgeon’s treatment.  On the initial encounter, report the surgical condition in the first position and the medical conditions in the second and subsequent positions on the claim form. 

 Do or do not report underlying conditions?  ICD-9 instructions tell us to report those conditions that “coexist at the time of the visit that require or affect patient care or treatment.”  This instruction would tell us not to code conditions that were not assessed or which did not affect patient care.  However, many at risk or shared savings contracts base payments on the acuity level of the population of patients in the panel.  How does the payer know how sick the panel of patients is?  First, by analyzing the panel’s age and gender distribution.  Second, by the diagnosis codes submitted on claim forms.  If your practice is entering into those types of agreements, reporting underlying conditions is important. 

 Do select specific codes  You know how doctors tell their patients to exercise more and eat more fruits and vegetables, and we all ignore the advice?  We coders tell doctors to select more specific diagnosis codes, and the doctors all ignore our advice.  It’s a stalemate.  Perhaps we could trade: I’ll eat a peach if you’ll select a specific diagnosis code.    Specificity now will make the transition to ICD-10 much easier.  If your group will participate in any risk based contracts or shared savings programs, it is imperative that you accurately report diagnoses.

 Don’t tell anyone, but I’ve been a lazy diagnosis coder.  I knew that the practice would be paid the same amount if I reported one diagnosis code or four.  And, I knew the claim form (then) only allowed four diagnosis codes.  But, I’ve turned over a new leaf and have adopted a new attitude.  This old dog is learning new tricks.  Block that metaphor? 

One thought on “Let’s trade: I’ll eat a peach if you’ll select a specific diagnosis code

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s