I know for a fact that most of you who code using ICD-10 in public and private hospitals in Malaysia, are conscientious, dedicated, hard-working, and detail-oriented as medical records professionals.

I am sure when erroneous mordbitidy and mortality statistical reports shows discrepancies and weird facts, and when errors in your ICD-10 coding work are discovered, most of you are extremely upset with yourselves, and you would sensibly work even harder to improve your coding skills.

Although as humans we inevitably make occasional mistakes.

What is lacking I think is, an auditing process in Malaysia, where an analysis of common errors found in auditing inpatient records can be done.

I diged into my journal and notes of coding experiences, and I like to post and share this tuesday morning, 22 May 2012 what I think are the 5 most reasons as outlined below, why coding errors are made. This post addresses some of the common coding errors and suggests some ways to prevent them, as I see it. An insight (knowing) where the “traps” lie dormant, should help you to avoid them.

1. Rush to get the work done:
When you choose productivity or rushing over your job of coding inpatient records as a priority over quality, this can cause you  to rush through a medical record without thoroughly reading all available documentation. Additionally, the distractions and disruptions that occur in you workplace environment may result in errors.

2. Assigning diagnosis codes from memorising:
I know the many experienced amongst you who cannot help but memorise many code assignments after using them repeatedly. Sometimes, however, our memories fail and the direct entry of memorised codes may lead to error.

3. Incomplete or inadequate documentation:
When documentation is incomplete or conflicting, it is difficult for you to code completely and accurately. Since we code before discharge summaries or other dictated reports are available (correct me if I am wrong), final conclusions/diagnoses may differ from those determined by the you in reviewing History & Physical Examination reports and progress notes alone.

4. Incorrect principal diagnosis selection:
Errors in selecting the principal diagnosis may be the result of a lack of knowledge of basic coding principles and terminology. The quality of your initial training program and/or “on-the-job experience” is fundamental to building your  expertise, as is your ability to stay abreast of current coding guidelines. Misunderstanding or misinterpreting a coding guideline may also occur by failing to read inclusion and exclusion terms, and coding references during the coding process. Common examples of incorrect principal diagnosis selection including :

• Coding a condition when a complication code should have been selected instead
• Coding a symptom or sign rather than the definitive diagnosis.
• Assuming a diagnosis without definitive documentation of a condition
• Coding from a discharge summary alone.
• Incorrectly applying the coding guidelines for principal diagnosis, especially in a situation where the coder selects the diagnoses when two or more diagnoses equally meet the definition of principal diagnosis.

5. Incorrect or missing secondary diagnoses:
Secondary diagnoses are frequently coded when they do not meet the criteria for reporting secondary diagnoses. Some of the “traps” in coding secondary diagnoses are found in the doctor’s documentation.

Examples include:  (1) Using the term “history of” for conditions that are currently under treatment, as well as for those that have been resolved prior to admission; (2) Misusing the term “coagulopathy.” It is often documented when a patient on anticoagulant therapy has an expected prolonged prothrombin time, rather than a true coagulopathy.   Secondary diagnoses may be missed by when you attempt to code from a discharge summary alone without reviewing all documentation.

RECOMMENDATIONS :

1. Focus on quality, not just productivity. The quality of coded data is more critical This fact justifies taking the time to focus on coding accuracy and reading medical record documentation thoroughly. Try to eliminate as much of the daily distractions and disruptions in the workplace as possible.
2. Query conflicting and incomplete documentation. When a record has been coded without a final discharge summary, a process should be developed for reviewing them when it is complete.
3. Apply critical thinking skills when reviewing documentation and code assignments.
4. Always refer to the ICD-10 Instruction Manual to understand the official WHO coding guidelines for principal diagnosis. When multiple conditions may be present or suspected on admission, it is especially challenging to determine if the guideline for two or more diagnoses meeting the definition of principal diagnosis may be applied.
5. Review all questionable code assignments with your senior or another person who also codes using ICD-10; sometimes a discussion with another ICD-10 user  is enough to clarify your questions.
6. If you need to discuss with the doctor making the final diagnosis, query as necessary; be clear and concise and avoid “leading” the doctor to alter a diagnosis (this is sensitive material, however I think the how-to is covered in the ICD-10 Instruction Manual, you can check).

7. Exercise care when coding secondary diagnoses from the History & Physical Examination. Remember that the definition of “other diagnoses” for reporting purposes is conditions that affect patient care in terms of requiring clinical evaluation, therapeutic treatment, diagnostic procedures, extended length of hospital stay, or increased nursing care and/or monitoring. A condition that meets only one element of this definition may be coded; conditions that do not meet this definition should not be coded.

Pals, I am no expert in ICD-10, but I think you out there as responsible medical records staff, must realise the importance of accuracy in coding that cannot be underestimated. I am sure we do not wish our statistical reports on morbidity and mortality to be rediculed. Perhaps it is the time to assess your coding skills and use all resources available to improve them to ensure coded data of the highest quality.