Previous month:
July 2014
Next month:
January 2015

September 2014

Mandating Physicians to Use SNOMED codes Has Higher Potential to Improve Healthcare Than ICD-10 Mandate

I have consistently advocated for skipping ICD-10 and initiating an unprecedented effort to accelerate the development of ICD-11-CM.  Although I still believe this strategy to be the one best aligned with quality care, I fear that the sunken ICD-10 costs are now so large that skipping ICD-10 is unpalatable for most organizations, even for some physicians, and is politically perilous.  Since we must do something, I have been thinking more about the proposal to replace ICD-9 with SNOMED in physician practices.  Leveraging SNOMED to improve care, lower costs and remove physician practices from the ICD conversion melees should be a serious national conversation at this point.  
 
After several delays CMS has established October 1, 2015 as the new implementation date for the replacement of ICD-9 code sets used by medical coders and billers to report healthcare diagnoses and procedures with ICD-10 codes   But another postponement remains a possibility--especially when one considers the unclear reasons for action taken by Congress earlier this year to call off the 2014 implementation.  ICD conversion delays are costly to the healthcare industry and action should be taken to address the impediments that increase the risk of such delays.   One of the major impediments to address is the adverse impact ICD conversions have on individual physician practices.    
 
So let's jump out of the box of conventional charged impulses propagating across our cerebrums (thinking) to consider how to make ICD-10 optional for physician practices while still achieving our goal of dispensing with obsolete ICD-9 code sets.  One alternative is to mandate physicians replace ICD-9 codes sets with SNOMED code sets and require EHRs to incorporate translator technology that converts SNOMED to ICD codes in the background.  Since it would not be practicable to expect EHR vendors to incorporate the translator technology into their products by October 1, 2015, there would need to be an interim period where physician practices are exempt from the requirement to use ICD-10 codes sets until their EHR incorporates the translator technology.    This alternative mandate allows the ICD-10 conversion to proceed for the rest of the healthcare industry including any physician practices who see value in completing their conversion.  This mandate would reduce the current and future adverse impacts that ICD conversions have on physician practices, has higher potential to improve care, is more cost effective, helps EHRs be more user-friendly to physicians and mitigates the risk of further delays to ICD-10 as well as future ICD-X conversions.
 
I would anticipate a two-year transitional period where the ICD-10 conversion would be optional for physician practices based on an assumption that EHR vendors will need until 2017 to upgrade their products.  
 
Some opposition among physicians is likely to be encountered due to their lack of familiarity with SNOMED as well as questions about how this alternative strategy adds value to patient care.   I base that on the responses I heard from some respected colleagues at this weekend's Texas Medical Association meeting.  The unfamiliarity issue can be addressed by pointing out that many of us are already using SNOMED, but that we just do not know it.  CMS mandates that the problem lists in EHRs use SNOMED codes, so when one selects "Exercise-induced asthma" from a pick list of problems in their certified EHR, they are actually using SNOMED. 
 
More difficult to articulate to physicians is how this proposal to convert from using ICD-9 to SNOMED codes in our EHRs would improve healthcare, how it would improve their work flow and how this is more cost effective for physicians as compared to complying with the current mandate.  So I have developed the following bullet list to use when describing this to my colleagues:
 
  • Informatics experts are in agreement that ICD-9 is obsolete, and that although ICD-10 has potential to improve healthcare, ICD-11 and SNOMED have higher potential to improve healthcare. 
  • SNOMED, which is interwoven in ICD-11's development, is inherently compatible with ICD-11 and is already required by CMS to be incorporated into certified EHRs for Problem Lists--thus, mandating use of SNOMED is not really new to physicians and will not result in an added cost to physicians
  • EHRs can be built with technology that automatically converts SNOMED codes into ICD codes--thus, mandating use of SNOMED is agnostic to the version of ICD-X being used; the cost to physicians of using the translating technology is very small as compared to the cost of finishing the conversion to ICD-10 and then converting to ICD-11 in the next 15 years.
  • After we convert to ICD-10 in 2015, discussions about implementing ICD-11 will ensue; since ICD-10 is over 20 years old and is less sophisticated than ICD-11, it will become apparent rather quickly that we need to convert to ICD-11 as soon as possible in order to improve healthcare (i.e. today's argument about ICD-9)
  • It takes the U.S. 7-10 years to refine the international version of ICD codes into the U.S. version we use--since the ICD-11 international version is expected to be completed in 2017, the earliest conversion to ICD-11 in the U.S. would be 2024 unless an unprecedented effort to accelerate development took place
  • In any case, converting to ICD-10 in 2015 will result in two ICD conversions in physician practices over the next 15 years.  The proposed alternative strategy to convert physicians one time from ICD-9 to SNOMED results in just one conversion with all future ICD conversions occurring in the background without significant impact on physician practices--thus, mandating the use of SNOMED to replace ICD-9  would be a significant cost savings to physicians.
  • SNOMED codes have been developed for the purpose of clinical input; ICD codes are developed for important administrative and financial output purposes-- thus, use of SNOMED codes for input will improve physician work flow because SNOMED is more intuitive to use for physicians to describe clinical encounters; this also preserves the use of ICD code sets for the important administrative and financial functions that our healthcare system currently depends on.
I believe that if CMS is going to maintain their mandate to move off of ICD-9, then we should move on to an available coding system that has the most potential to improve healthcare at the lowest cost:
 
  • Informatics experts agree that ICD-11 is more sophisticated and has more potential to improve healthcare than ICD-10, but the earliest that a US version of ICD-11 could be available is 2024 unless an unprecedented effort to accelerate development occurs
  • On the other hand, SNOMED is already incorporated in EHRs and being used by physicians
Thus, I believe the mandate to convert off of ICD-9 is more likely to improve healthcare, improve physician work flow and impose the lowest costs if we make the 2015 conversion to ICD-10 optional for physician practices and mandate physicians start using SNOMED (with the translator technology incorporated in EHRs) in 2017.    There will inevitably be tactical challenges involving diverse groups of healthcare stakeholders to work on, but if we remain aligned to the goal of improving quality care, I am confident we will find mutually agreeable solutions. 
 
cook children's healthcare system