C l a i m D x

The health­care indus­try is mak­ing sig­nif­i­cant strides toward the adop­tion of elec­tronic health records (EHRs). How­ever, increased deploy­ment of EHR tech­nol­ogy only par­tially addresses the cru­cial need for bet­ter health infor­ma­tion in the U.S. The full ben­e­fits of an EHR can only be real­ized if we improve the qual­ity of data that EHRs are designed to manage.

The cur­rent clas­si­fi­ca­tion cod­ing sys­tem used in the U.S., ICD-9-CM, is a 30-year old sys­tem and can no longer accu­rately describe today’s prac­tice of med­i­cine. Con­tin­u­ing to use this sys­tem jeop­ar­dizes the abil­ity to effec­tively col­lect and use accu­rate, detailed health­care data and infor­ma­tion for the bet­ter­ment of domes­tic and global health­care. By fail­ing to upgrade, we could find our­selves build­ing an infra­struc­ture that does not pro­vide the infor­ma­tion nec­es­sary to meet the health­care demands of the 21st century.

Despite annual updates, ICD-9-CM does not meet our health­care needs. Ter­mi­nol­ogy and clas­si­fi­ca­tion of numer­ous con­di­tions and pro­ce­dures are out­dated and incon­sis­tent with cur­rent med­ical knowl­edge and appli­ca­tion. ICD-9-CM also can­not address the increas­ing pres­sure for more spe­cific codes, espe­cially codes that rep­re­sent new tech­nol­ogy. Out­dated codes pro­vide inac­cu­rate or lim­ited data and insuf­fi­cient detail related to health diag­noses, pro­ce­dures and technologies.

Med­ical care has changed dras­ti­cally since ICD-9-CM was imple­mented. Many new health­care diag­noses and ser­vices can no longer be described accu­rately using ICD-9-CM. For exam­ple, the ter­mi­nol­ogy used to clas­sify asthma in ICD-9-CM is out­dated, result­ing in an inabil­ity to assess the effi­cacy of treat­ment pro­to­cols that are based on the cur­rent clin­i­cal clas­si­fi­ca­tion of asthma types.

Spe­cific Gains

Upgrad­ing ICD-9-CM to ICD-10-CM, a U.S. ver­sion of ICD-10 devel­oped by the World Health Orga­ni­za­tion, and ICD-10-PCS, a pro­ce­dural cod­ing sys­tem designed by the Cen­ters for Medicare and Med­ic­aid to replace the cur­rent inpa­tient pro­ce­dural cod­ing sys­tem included in ICD-9-CM, will pro­vide supe­rior infor­ma­tion for mea­sur­ing health­care ser­vice qual­ity, safety and the effi­cacy of care. This data impacts qual­ity reports and report cards, out­comes and per­for­mance mea­sure­ment, clin­i­cal research, case man­age­ment and pay-for-performance programs.

So why is the U.S. health­care sys­tem the only devel­oped coun­try that has not adopted ICD-10? Per­ceived cost is a huge fac­tor. How­ever, delay­ing its imple­men­ta­tion will actu­ally increase future imple­men­ta­tion costs as imple­ment­ing a new cod­ing sys­tem will require sys­tems and appli­ca­tion upgrades. This can be avoided if imple­men­ta­tion is planned with other sys­tem changes to max­i­mize impact and reduce complexity.

Not only does ICD-10 lever­age our invest­ment in EHR and mod­ern tech­nolo­gies, but it also will give the U.S. health­care system:

* Bet­ter data for patient safety, qual­ity of care analy­sis and reimbursement;

* Increased capac­ity to iden­tify and respond to pub­lic health or bio­log­i­cal threats;

* The abil­ity to achieve full ben­e­fit from the use of SNOMED-CT as the clin­i­cal lan­guage of medicine.

Although reim­burse­ment con­sid­er­a­tions should not drive code set revi­sions, accu­rate and spe­cific coded data should be used to sup­port appro­pri­ate reim­burse­ment. The level of speci­ficity in ICD-10-PCS will pro­vide pay­ers, pol­icy mak­ers and providers with more detailed infor­ma­tion for estab­lish­ing appro­pri­ate reim­burse­ment rates, eval­u­at­ing and improv­ing the qual­ity of patient care, improv­ing effi­cien­cies in health­care deliv­ery, reduc­ing health­care costs, and effec­tively mon­i­tor­ing resource and ser­vice utilization.

For exam­ple, reduced health­care costs will result if a more spe­cific cod­ing sys­tem is employed, facil­i­tat­ing pre­ven­tion and iden­ti­fi­ca­tion of fraud and abuse or the speci­ficity needed to con­duct good qual­ity improve­ment and error reduc­tion pro­grams. The exchange of addi­tional data beyond the basic claim, and the time it takes to gather and process such detail, will sig­nif­i­cantly be reduced due to the more spe­cific detail con­tained in the ICD-10-CM and ICD-10-PCS codes.

An Invest­ment That Pays

By being the only coun­try with a mod­ern health­care sys­tem that has not adopted ICD-10 (more than 100 coun­tries already have), we are also mak­ing it more dif­fi­cult to share dis­ease and mor­tal­ity data at a time when such global shar­ing is crit­i­cal for pub­lic health. With recent pub­lic health and bioter­ror­ism events, our abil­ity to track and respond to inter­na­tional threats in real time is lim­ited by our reliance on ICD-9-CM.

For exam­ple, ICD-10-CM would have bet­ter doc­u­mented the West Nile Virus and SARS com­plexes for ear­lier detec­tion and bet­ter track­ing. Our national data sys­tems are being ren­dered use­less. The U.S. must be part of the world health data com­mu­nity. Many coun­tries rely on U.S. ana­lyt­i­cal tools based on codes and groupers. This mar­ket edge is being lost by the inex­plic­a­ble delay in com­ing inline with con­tem­po­rary code sets.

The adop­tion of SNOMED-CT as the Con­sol­i­dated Health Infor­mat­ics ini­tia­tive and license agree­ment between the National Library of Med­i­cine and SNOMED Inter­na­tional posi­tion it as the clin­i­cal lan­guage of med­i­cine in the U.S. Even­tual use of SNOMED-CT will not elim­i­nate the need for clas­si­fi­ca­tion sys­tems. Even with fully deployed ter­mi­nol­ogy sys­tems, most health infor­mat­ics pro­fes­sion­als point to the ongo­ing need for clas­si­fi­ca­tions to per­mit aggre­ga­tion of clin­i­cal data for admin­is­tra­tive pur­poses includ­ing reim­burse­ment, sta­tis­ti­cal and epi­demi­o­log­i­cal analy­ses, and health pol­icy decisions.

Con­tin­ued use of the out­dated ver­sion of ICD dimin­ishes the U.S. invest­ment in SNOMED-CT. The antic­i­pated ben­e­fits of an EHR will not be fully achieved if the ref­er­ence ter­mi­nol­ogy employed, such as SNOMED-CT, is aggre­gated into a 30-year old clas­si­fi­ca­tion sys­tem for admin­is­tra­tive use and index­ing. Valid maps are urgently needed to link from a highly spe­cific ter­mi­nol­ogy to a clas­si­fi­ca­tion sys­tem so that infor­ma­tion cap­tured in the ref­er­ence ter­mi­nol­ogy can uti­lize the power of sum­mary required for health­care report­ing and index­ing offered by the clas­si­fi­ca­tion sys­tems. The ICD-10 med­ical cod­ing sys­tem facil­i­tates more robust map­ping from SNOMED-CT in the EHR due to its gran­u­lar­ity and use of cur­rent clin­i­cal terminology.

In 2003, the National Com­mit­tee for Vital and Health Sta­tis­tics (NCVHS) rec­om­mended that the Depart­ment of Health and Human Ser­vices ini­ti­ate the rule­mak­ing process for adop­tion of ICD-10-CM and ICD-10-PCS. NCVHS spent two years care­fully lis­ten­ing to argu­ments and weigh­ing the evi­dence before pro­ceed­ing with their pos­i­tive rec­om­men­da­tion to the full committee.

ICD-10-CM and ICD-10-PCS needs to be imple­mented by Octo­ber 2008. That means imme­di­ate adop­tion through the HIPAA rule mak­ing process in 2005 is crit­i­cal to per­mit the health­care indus­try to man­age this tran­si­tion and ensure con­ti­nu­ity in its health­care data reporting.

Linda Kloss, R.H.I.A., C.A.E., is the CEO of the Amer­i­can Health Infor­ma­tion Man­age­ment Asso­ci­a­tion, head­quar­tered in Chicago.