July 16, 2012
Canada’s Experience With Coding Diabetes in ICD-10
By Doris Gemmell, BSc, MBA, CHIM, and Deb Tetreault, RHIT, CHIM
For The Record
Vol. 24 No. 13 P. 8
According to the World Health Organization, nearly 350 million people worldwide have diabetes, a number expected to continue to grow. Diabetes is a serious chronic condition creating a major disease management problem. Consider the following:
• The Canadian Institute for Health Information (CIHI) reports that 80% of adults with diabetes die from heart disease and stroke.
• A British Columbia study showed that adults with diabetes used an average of 2.4 times the health resources of the general population.
• Life expectancy for adults with diabetes is reduced by up to 15 years for type 1 and by five to 10 years for type 2.
To make matters worse, type 2 diabetes is being diagnosed in more children as the obesity epidemic expands.
With the United States moving to ICD-10, some have questioned whether the deeper level of coding granularity offered in the new codes will help remedy diabetes problems, better inform clinicians and researchers, and improve patient outcomes. ICD-10 delivers a more accurate and refined data set that can improve the disease management process across the care continuum and result in reduced costs associated with this patient population. While ICD-10 cannot cure diabetes, it can arm clinicians with accurate information and provide a solid foundation for treatment and ongoing monitoring.
To achieve these goals, physician documentation for diabetes must be more detailed, and coding professionals must understand diabetes pathology, including its progression and manifestations. With ICD-10 tentatively set to go live in the United States in October 2014, there is no time to waste in getting diabetes documentation and coding up to par. The first stop for HIM professionals is clinical documentation improvement (CDI).
Physician Documentation Gets Granular
Coding diabetes presents an excellent working example of the effort required to implement ICD-10. In ICD-10, there are approximately 300 diabetes codes, compared with roughly 20 in ICD-9, making it more challenging to correctly code. Physician documentation requires a substantially higher degree of specificity in the following areas:
• diabetes type;
• insulin dependency;
• complications (eg, neuropathy, retinopathy);
• kidney disease stage (1 through 5); and
• diabetes in pregnancy (gestational or preexisting).
Additionally, borderline diabetes is no longer a code; patients are diabetic or not.
As with any CDI initiative, physicians must be shown that improved documentation processes will ultimately benefit their patients and make the additional effort worthwhile. In Canada, the following practices helped improve diabetes documentation success rates:
• Conduct short educational sessions (10 to 15 minutes) with a small group of clinicians.
• Post brief (10 words or fewer) white board tips where doctors document or dictate.
• Work closely with physician champions and the chief of staff because doctors teaching doctors is the best scenario.
• Include possible codes and modifiers on chart checklists.
Our experience shows that HIM professionals in smaller rural hospitals have closer relationships with their physicians and report greater CDI success. HIM professionals in larger teaching hospitals have more distant relationships with their medical staff, and physicians tend to be less receptive or unaware of documentation requirements. In light of these trends, organizations should tailor physician educational programs to their unique audience. One size does not fit all when it comes to ICD-10.
Coding Education Expands
At a national level, Canada performs regular reabstraction studies to analyze documentation and the associated ICD-10 codes. Through these studies, many inconsistencies and inaccuracies with diabetes coding were identified. As a result, tighter standards for diabetes coding and expanded coder education programs were established.
In the United States, coding is governed by correct coding initiatives. In Canada, however, the coding industry shifted from correct coding guidelines to mandatory data standards based on the reabstraction study results. This change has greatly increased the consistency and validity of coded data. However, the effort must include perpetual training for clinical coders, particularly for those coding diabetes.
Canadian coders required upgraded education in anatomy and physiology to accurately code diabetes in ICD-10. A deeper understanding of diabetes origins is essential. Also, since so many other conditions are secondary to diabetes, all the various manifestations and correct coding of each should be clearly understood.
The CIHI, Canada’s coding governing body, has developed various e-learning modules featuring remote access for all coders. Currently, there are at least two specific modules for diabetes, each requiring six to eight hours to complete. In addition, the CIHI has developed a national e-query tool available to all HIM professionals contributing to the coding databases. Many of the tool’s Q & As are specific to diabetes coding.
The studies also exposed a large volume of incorrect and unspecified diabetes codes due to either poor documentation or improper code assignment. Although code quality and consistency continues to improve through the coding improvement program, documentation issues remain a problem. Doctors believe they are there to treat patients, not necessarily to document care or help coders.
Don’t Overlook Outpatient
Diabetes is typically treated in an outpatient setting. In this environment, nonphysician care providers, such as chronic care nurses, nutritionists, and dietitians, frequently see patients. This has led ancillary clinicians to take on more documentation responsibilities, making it imperative that healthcare organizations include them in any CDI educational programs.
In Canada, ancillary clinicians are greatly improving diabetes management. However, because coders cannot rely solely on their notes, physicians must still be engaged. Perhaps this will change, but for now, physician documentation remains the key to ICD-10 success.
One lesson learned when coding diabetes in Canada was the impact other factors had on the disease process. To successfully track and manage diabetic patients, additional data were collected and coupled with diagnosis codes. However, collecting these codes can be problematic if they are not consistently documented and coded. Coding secondary codes requires additional resources and should be undertaken only if the information will be used clinically.
For example, in ICD-10 other factors that may contribute to diabetes can be coded. Socioeconomic statuses, such as homelessness, and lifestyle factors, such as smoking and drug abuse, all have ICD-10 codes.
In Canada, best practices are recurrently reviewed to ensure relevance. Canadians continue to learn through their ICD-10 journey and have gone through multiple iterations of best practices with regards to diabetes documentation and coding. Knowing that secondary factors exist and successfully managing them can improve diabetes management and have a significant impact on patient outcomes.
The Turning Point
The long-term goal of ICD-10 is to improve population health through higher-quality data for research and forecasting. This requires that providers go beyond the initial coding and financial issues and use the accumulated data as a research tool for evidence-based care. For diabetic populations, continual fine-tuning will be required. Though many providers will not make this leap, those who do can differentiate themselves as high-quality providers and diabetes specialists.
— Doris Gemmell, BSc, MBA, CHIM, is director of coding services at Accentus.
— Deb Tetreault, RHIT, CHIM, is manager of coding, data quality, and consulting services at Accentus.
Lessons Learned in Coding Diabetes in ICD-10
• Accurate coding depends on physician documentation, which must be more specific.
• Ancillary clinician notes are also important.
• Train physicians in short segments and small, specialized groups.
• Coders require in-depth anatomy and physiology training specifically around diabetes.
• Disease manifestation and secondary condition coding is critical.
• Socioeconomic/lifestyle factors may add value to disease management.
— DG, DT