A recent study of CME participants of Myelodysplatic Syndromes reviewing patient claims data claims data showed significant improvement in the mean time diagnosis of MDS patients from physicians who participated in courses on the topic.
The article CE Measure by Paul D. Walden, PhD entitled “Administrative Health Data to Assess Performance in a Myelodysplastic Syndromes CME Initiative,” used health claim data to evaluate a continuing medical education (CME) program designed to improve physician performance in the diagnosis of myelodysplastic syndromes (MDS). The study, which was funded by Celgene Corporation, concluded that “health insurance claims data can be used for performance evaluations of physicians participating in a CME activity who have adequate representation in an administrative healthcare database.”
“Myelodysplastic syndromes comprise a group of clonal hematopoietic stem cell disorders that frequently manifest as one or more peripheral cytopenias, often producing associated signs and symptoms such as anemia, bleeding, fatigue, pallor, infections, and bruising.” According to a recent report cited in the study, “the incidence of MDS in the Medicare population is 76,600/year, a figure 4 to 5 times higher than previous estimates from epidemiologic studies.
The median age at diagnosis in this population was 76 years, which was significant given that most cases of MDS present as anemia and the high prevalence of anemia in the elderly population. Due to incidence at an older age, the study noted that “it would not be surprising if many cases of MDS are overlooked, and that diagnosis can be further complicated due to the existence of overlap conditions.” Definitive diagnosis of suspected MDS requires cytogenetic as well as morphologic evaluation of a bone marrow sample, but only 57% of the diagnoses in the Medicare population were based on a bone marrow evaluation (43% were based on clinical impression).
Such challenges in the diagnosis of MDS can result in the delayed or inappropriate diagnosis, which reduces or eliminates the therapeutic window of opportunity to positively impact the natural progression of MDS to AML or death. As a result, the authors of the study addressed this challenge by starting a process to identify practice gaps; specify and validate International Classification of Diseases, 9th Revision (ICD-9), and Current Procedural Terminology (CPT) codes that can be used as measures of performance, development, and implementation of effective CME (context-based, interactive learning with practical application); and direct measurement of performance at baseline and post CME.
Once the process was established, the authors “hypothesized that health insurance administrative claims data can be used as a basis for assessing quality improvement/ performance improvement in, as well as identifying barriers to, change in CME.” To test this hypothesis, the study used administrative claims data from the Normative Health Informatics (NHI) database, which is owned and maintained by Ingenix as a research oriented data mart containing more than 25 million current members with UnitedHealthcare insurance policies.
From this data, the authors used their ongoing CME initiative in MDS as a case study, and then critically and systematically evaluated how administrative claims data could be effectively incorporated into a CME initiative designed to improve performance while at the same time maintaining full compliance with the Health Insurance Portability and Accountability Act (HIPAA). The primary learning objective in their current CME activities is to improve the time it takes for accurate diagnosis of MDS that is required for IPSS risk stratification and appropriate treatment.
The analysis then focused specifically on CME participants in calendar year 2006 who were matched to the NHI database based on (1) first name, (2) last name, (3) city, and (4) state to identify providers with sufficient data in the NHI database to permit evaluation under our criteria. A sliding scale window was them used so that claims submitted by the provider 6 months prior to participation in the CME activity was compared with the 6 months of activity after CME participation.
After data was gathered, 707 physicians practicing in 38 states participated in the MDS educational activities and applied for CME credit in calendar year 2006. Of those, 279 participants (40%) provided care to UnitedHealth Group (UHG) patients in 33 states. Hematologists (n = 118) and oncologists (n = 92) accounted for 75% of the total UHG physician subset.
Mean Time Diagnosis (MTD) among the 279 participants was compared with MTD among a matched control group of 279 physicians who did not attend the study’s CME activities, but who had similar demographics using matching variables of specialty, city, state, and patient panel size. All 279 CME participants in this analysis attended a single CME event.
It was noted that this comparison allowed for direct assessment of the impact their educational activities because when the study was conducted, other sources of CME on MDS for community physicians would have been very limited. In comparing the two groups, the study found that:
– There was a significant improvement in the number of patients with a first MDS diagnosis in the post-CME period seen by the CME physician participants compared with the matched non-CME (control) physicians between the January-June 2007 group; 271 versus 201 respectively;
– 2006 CME participants improved in time from initial diagnosis of anemia to first MDS diagnosis in the post-CME period;
– 7% more patients had an MDS diagnosis within 6 months of initial diagnosis of anemia in the 2006 CME participants group; and
– There was a 20% reduction in the MTD of MDS from the initial diagnosis of anemia among physicians who participated in the 2006 CME activities compared with the control group.
The authors did noted the need for measuring performance improvement in other disease states with larger numbers of patients to help account for using a narrower pre- and post-CME time period. The study also did not allow for the comparison of performance changes within the same group of physicians, only the different groups. It was also noted that standard limitations regarding the use of claims data applied to the use for CME outcomes analyses, and that future prospective assessments with sample chart audits will assist in characterizing such error in future studies.
Such limitations do not discount the considerable findings this research demonstrates. In particular, this study clearly demonstrates that CME improves patient outcomes and positively influences short- and long-term effects on the practice behavior of physicians. Moreover, the results demonstrate that when CME focuses on higher-level outcomes, plans and develops “context-based” content, provides interactive learning (through case presentations), and integrates learned information into practice, these methods significantly help physicians.
More research should be carried out in other disease states using health claims data to find further ways CME can continue helping physicians treat patients.