Using Diagnoses From Prior Encounters

One of the most common questions we get is whether or not a diagnosis from an earlier encounter can be used for a current encounter.

Standard coding advice has always been that a diagnosis cannot be "pulled forward" to a new encounter. Coding Clinic, Q3, 2013 has addressed this issue.  In part, the Coding Clinic advice says:

"Conditions documented on previous encounters may not be clinically relevant on the current encounter.... However, if the condition is not documented in the current health record, it would be inappropriate to go back to previous encounters to retrieve a diagnosis without physician confirmation."

Coding Clinic indicated this advice applies to both ICD-9 and ICD-10 coding.  We hope this advice will clear up this frequent question.

Gearing Up for ICD-10

There are thousands of articles out there on getting ready for ICD-10 implementation.  Most of them are very broad in their scope.  One item that sometimes gets short shrift is planning for new superbills in the physician office setting.

Maybe because it seems overwhelming--or maybe because the writers often come from a hospital setting.  In the hospital setting, coding is much different.  The query process means that you can start off with documentation that is of low specificity, and end up with more specificity to improve code selection.  In all but the largest of physician clinic settings, there is no query process.  The documentation is what it is.

So, what advice can we give to the practicing physician, who may be overwhelmed with the large number of codes available?  I think the most important thing for physicians to do is be realistic about their documentation (while taking the opportunity to improve it!), and analyze what they really use in ICD-9 as a starting point.  Your superbill may contain 100 ICD-9 codes, that translate to 2,500 possible ICD-10 codes.  But how many of those ICD-9 codes do you really use and document in your day to day practice? That old clinical saw about horses vs. zebras will serve you in good stead here too.

Use your practice management software to analyze the top 100 codes you use today--not what is on your superbill.  Look at the frequency of use over the course of a year.  If 25 of those codes make up 80% of your ICD-9 volume, turn to the General Equivalence Mappings (GEMs) to help you find crosswalks for those codes.  Allow plenty of space for write in- diagnoses in the beginning--and track those codes.  Take the top 25-50 of your write- in codes, and add them to your superbill.   By the time you have 75 codes on your superbills, MOST specialties will have a superbill that helps you code 80% or more of your daily visits.

Some specialties (orthopedic surgery, for example) will be forced to create much larger superbills, unless they limit their practice to a specific body part.  But the basic premise is the same.  However, because of the specificity of ICD-10, and the few areas of the code book that do not contain a "Not Otherwise Specified" code, you'll be forced to change your documentation style, or be unable to find codes for some conditions.  Remember that anatomy and laterality are key in ICD-10, along with dominance (mostly for neurological conditions).  Your documentation should indicate R vs L, along with whether the patient is R or L hand dominant.  You might have been able to find a code for "R arm fracture" in the past, but now you'll need to document the bone(s) involved, including the area of the bone involved.  This does mean some increase in documentation, but these increases have the potential to improve documentation greatly.  Continuity of care can be improved with these changes, and exchange of important clinical information between specialties.

Whatever approach you plan to take to create superbills in ICD-10, the time to start is now.   Today, there are 426 days until ICD-10 submission is required.  Waiting until then to see what is required is too late.  Starting now will allow you the necessary time to make any changes needed in your documentation, and begin to understand the differences in ICD-10 language and rules.  Take the time now to read the official guidelines (even if you've never read the ICD-9 guidelines!) so you know what the requirements are to choose a code.

Remember to check the Presentation and ICD-10 pages on HCC University for information on ICD-10 coding.  And let us know what types of information you need on ICD-10.

Code Numbers Instead of a Narrative Diagnosis

All too often, we see a diagnosis code written in the medical record, in lieu of a narrative diagnosis.  In those instances, we cannot code what has been written for two important reasons:

• Coding is done based on the narrative documentation in the medical record--with no narrative, no coding can take place
• There's no way of telling if the diagnosis code in the chart is correct (i.e., what the provider meant to code)

We get this question at least once a year--and sometimes it leads to lively exchanges.  Thankfully, Coding Clinic, Q1 2012, has decided to address it.  Their answer, in part reads:

"There are regulatory and accreditation directives that require providers to supply documentation in order to support code assignment. Providers need to have the ability to specifically document the patient's diagnosis, condition, and/or problem. Therefore, it is not appropriate for providers to list the code number or select a code number from a list of codes in place of a written diagnosis."

So--the next time you're asked if a provider can just write the code in the chart--you'll know where to point them for clear guidance.