Imatinib mesylate (Gleevec®), has been found both effective and cost-effective for treatment of chronic myeloid leukemia (CML) and gastrointestinal stromal tumor (GIST).This study examines the relationship between compliance and health care utilization and costs. This retrospective study used claims from a large national US health plan. Patients (pts) had ≥2 imatinib claims from 6/1/01–3/31/05; ≥1 yr follow-up (F/U) after the 1st imatinib claim; diagnosis of CML (ICD9 205.1 ×) or GIST (ICD9 159.0, 159.8, 159.9 or evaluation of claims by independent hematologist-oncologist). Compliance was measured each in F/U Yrs 1 and 2 with medication possession ratio (MPR: ((days supply of imatinib)/365)*100). Compliance categories were good (≥90%); medium (70–89.9%); poor (<70%). Comorbidity measure was Charlson Comorbidity Index (CCI). Outcomes were number (#) of inpatient hospital stays (IP), procedures (procs), lab tests, length of IP stay (LOS), costs of IP, procedures, labs, total medical care, and total health care (medical+pharmacy). Rates of Yr2 outcomes/person-yr within cohort and compliance category were bootstrapped (100 repetitions) to adjust for varying F/U and provide standard errors. Outcomes were compared within cohort, across compliance categories; t-tests compared mean utilization and costs. 374 CML and 91 GIST had mean age 50.8±14.1 and 57.5±12.0, respectively. 218 (58.3%) CML and 56 (61.5%) GIST pts were male. CML and GIST pts had mean F/U of 29.5±13.2 and 24.9±10.6 months, respectively. GIST pts were sicker with mean CCI 6.5±3.2 vs CCI 2.4±1.0 for CML. For almost all outcomes, pts with poor compliance had significant higher utilization and costs than pts with good compliance (Table 1,2). GIST pts with good compliance had some higher costs, likely due to small sample, high comorbidity, and wide variation in costs (vs CML). For CML pts, mean medical costs were 87% lower for good compliance vs poor in both years; mean health care costs were 45% and 54% lower for good compliance in Years 1 and 2. Pts with poor compliance had 14 times longer IP LOS, 1.4 times more labs, 2.1 times higher lab costs than good compliance, on average. Overall, good imatinib compliance was significantly associated with lower health care resource consumption.

Table 1:

CML Outcomes (Means)

Year 1Year 2
ComplianceGood N=166Medium N=79Poor N=129Good N=143Medium N=84Poor N=147
*0.01<p<0.05, **p<0.01 compared with good compliance 
# IP 0.1 0.3 1.1** 0.2 0.2 1.1** 
IP LOS 0.6 3.1 14.4** 0.8 2.0** 9.9** 
# Procs 18.4 28.3* 50.3** 18.6 18.2 56.2** 
# Labs 49.9 50.6 71.1* 32.2 37.5** 65.7** 
$ IP 307 1737 4360** 1127 392** 5745** 
$ Procs 5512 16483* 67328** 4613 7428** 52620** 
$ Labs 2570 2621 5589** 2140 1778** 4527** 
$ Medical 10427 23659* 83685** 9080 11101** 70940** 
$ Health care 47384 52693 102745** 47603 39044** 86197** 
Year 1Year 2
ComplianceGood N=166Medium N=79Poor N=129Good N=143Medium N=84Poor N=147
*0.01<p<0.05, **p<0.01 compared with good compliance 
# IP 0.1 0.3 1.1** 0.2 0.2 1.1** 
IP LOS 0.6 3.1 14.4** 0.8 2.0** 9.9** 
# Procs 18.4 28.3* 50.3** 18.6 18.2 56.2** 
# Labs 49.9 50.6 71.1* 32.2 37.5** 65.7** 
$ IP 307 1737 4360** 1127 392** 5745** 
$ Procs 5512 16483* 67328** 4613 7428** 52620** 
$ Labs 2570 2621 5589** 2140 1778** 4527** 
$ Medical 10427 23659* 83685** 9080 11101** 70940** 
$ Health care 47384 52693 102745** 47603 39044** 86197** 
View Large
Table 2:

GIST Outcomes (Means)

Year 1Year 2
ComplianceGood N=27Medium N=20Poor N=44Good N=32Medium N=18Poor N=41
*0.01<p<0.05, **p<0.01 compared with good compliance 
# IP 0.1 0.3 1.2** 0.3 0.8** 1.5** 
IP LOS 0.6 1.6 14.1 1.6 2.7** 21.3** 
# Procs 28.8 41.1 49.1 32.1 34.5** 51.8** 
# Labs 34.3 29.2 33.6 31.2 25.9** 25.1** 
$ IP 414 324 3505 524 2163** 5064** 
$ Procs 12409 15223 33222* 12947 12585 26686** 
$ Labs 755 885 913 695 734** 953** 
$ Medical 15088 17991 39743* 15275 16915** 35664** 
$ Health care 55211 50435 55579 57745 48270** 45659** 
Year 1Year 2
ComplianceGood N=27Medium N=20Poor N=44Good N=32Medium N=18Poor N=41
*0.01<p<0.05, **p<0.01 compared with good compliance 
# IP 0.1 0.3 1.2** 0.3 0.8** 1.5** 
IP LOS 0.6 1.6 14.1 1.6 2.7** 21.3** 
# Procs 28.8 41.1 49.1 32.1 34.5** 51.8** 
# Labs 34.3 29.2 33.6 31.2 25.9** 25.1** 
$ IP 414 324 3505 524 2163** 5064** 
$ Procs 12409 15223 33222* 12947 12585 26686** 
$ Labs 755 885 913 695 734** 953** 
$ Medical 15088 17991 39743* 15275 16915** 35664** 
$ Health care 55211 50435 55579 57745 48270** 45659** 
View Large

Topics:
gastrointestinal stromal tumor, imatinib mesylate, health resources, brachial plexus neuritis, laboratory techniques and procedures, laboratory test finding, follow-up, leukemia, myelocytic, chronic, standard error, international classification of diseases

Author notes

Disclosure:Employment: Ms. Barghout and Dr. Williams are full-time employees of Novartis Pharmaceuticals Corporation, which manufactures Gleevec. Dr. Halpern conducted the analysis.

2007, The American Society of Hematology
2007
Sign in via your Institution