Why Don't Hematologists Follow Chronic Myeloid Leukemia Monitoring Guidelines? A Survey Of Practioners In New Jersey and Indiana

Successful treatment of chronic myeloid leukemia (CML) requires monitoring of cytogenetic/molecular responses during tyrosine kinase inhibitor therapy.Both the National Comprehensive Cancer Network and European LeukemiaNet have published monitoring guidelines. However our group has reported <30% of 1200 CML ptsin 2 insurance databases underwent polymerase chain reaction (pcr) monitoring during their 1st year as recommended (Chen ASCO 2013). These inadequately monitored pts experienced inferior outcomes, more hospitalizations and higher costs. We have also conducted a >400 pt chart review from 38 practices and noted over half did not undergo appropriate molecular monitoring, leading to higher rates of progression and mortality (Goldberg Curr Med Res Opin 2013). Barriers to physician adherence with CML guidelines have not been previously explored.

An anonymous 3 page survey was mailed to 515 hematologist-oncologists in New Jersey (n=359) and Indiana (n=156) in June 2013. Questions about barriers to guidelines were taken from a prior validated general guideline study (Taba, BMC Health Services Research 2012).

96 physicians (19%) responded [NJ=68 (19%); IN=28 (18%)]. Respondents worked 11% in academic centers and 89%community practice; 18% solo practice, 65% heme/onc group, 17% multi-specialty. Physicians were in practice 6% <5, 11% 6-10, and 83% >10 years. 17% saw primarily hematology pts, 8% solid tumors, 75% mixed heme/onc. 92% saw pts with CML and 19% referred. For new CML pts, respondents felt the following tests should be done: CBC 100%,spleen measurement byhand 100%, spleen imaging 37%, flow cytometry peripheral blood (PB) 54%, bone marrow (BM) biopsy or aspirate for morphology 73%, PB karyotype 25%, PB FISH 42%, PB pcr bcr-abl 73%, PB abl kinase mutation 15%, BM karyotype 73%,BM FISH 56%, BM pcr bcr-abl 44%, BM abl kinase mutation 10%, (italic items recommended by NCCN). 77% correctly identified routine PB pcr testing as quarterly (6 % biannual, 2% at one year). PB FISH was chosen 38% quarterly. The 3 month evaluation was queried: 73% recommended obtaining a PB pcr which is a treatment decision point by NCCN guidelines, 38% recommended a PB FISH. 79% utilized a lab performing pcr testing by International Standard, 6% used non-IS lab, 15% did not know what their lab reported and 2% were unfamiliar with IS. 77% stated they were familiar with NCCN guidelines and used them, 19% were familiar but did not use, and 4% were not familiar. Using 5-point Likert scale, 98% strongly or somewhat agreed “treatment guidelines are evidenced based”, 98% agreed “guidelines are useful in daily clinical work and improve quality of treatment”, 90% agreed “guidelines include different aspects of the disease and are a good tool for confirming diagnoses, starting initial treatment and managing complications” and 92% agreed “guidelines are convenient and the information is easy to find.” Barriers to guidelines: 56% disagreed “guidelines are hard to implement due to lack of medical resources”, 48% disagreed “guidelines are hard to implement due to lack of patient resources”, 65% disagreed “there is no time to search for information’, 77% disagreed “guidelines are not accessible”, 92% disagreed “guidelines are too complicated and it is difficult to find the information”, 71% disagreed “guidelines reduce doctors autonomy”, 77% disagreed “guidelines limit treatment options”, 67% disagreed “guidelines limit flexibility and individual approach’, 85% disagreed “there is no need for guidelines as treatment routines exist”, and 83% disagreed “patients do not want doctors to conform to treatment guidelines.” Facilitators included: 90% agreed “an easy to find online database”, 46% agreed “special training courses”, 77% agreed “published materials”, 52% agreed “information through professional societies”, and 67% agreed “available consultation to answer questions about the guidelines.”

This survey, the 1st onCML guidelines barriers, suggests that guidelines are considered useful (especially on-line or published) without major system, resource or attitudinal barriers. The low response ratecautions firm conclusions, especially since >70% respondents correctly identified pcr monitoring practices (indicating a more motivated physician cohort).Given that outcomes improve when CML pts are monitored per published guidelines, efforts to encourage usage are needed.

No relevant conflicts of interest to declare.