Final Report
Long Term Adherence to Hormone Therapy and Associated Outcomes in
Medicaid Patients with Breast Cancer
Clinical Incentive Pilot Program
May 16, 2011-May 15, 2013
Principal Investigator:
Jun Wu, PhD
Assistant Professor
Department of Clinical Pharmacy and Outcomes Sciences
College of Pharmacy, University of South Carolina
701 Grove Road
Health Sciences Administration Bldg. (MIPH)
Greenville, SC 29605
Tel: (864) 455-4729
Email: wujun@sccp.sc.edu
1
ABSTRACT
Objective: This study assessed long-term adherence to hormone therapy and associated
health care costs in women with breast cancer enrolled in Medicaid. Methods: We identified
women diagnosed with non-metastatic breast cancer (2000-2008) in the South Carolina Central
Cancer Registry and South Carolina Medicaid linked data set. All the patients were followed up
for 1, 2, 3, or 4 years after initiation of hormone therapy. Adherence rate was measured as
proportion of days covered (PDC) by tamoxifen or aromatase inhibitors (AIs) in a given interval
(1, 2, 3 or 4 years). Predictors of adherence to hormone therapy (PDC ≥ 80%) were identified
using generalized linear models for repeated measures. Yearly medical, pharmacy, and total
health care costs per patient were measured across 4 years. Generalized linear regression with
repeated measures was performed to examine the associations between medication adherence
and health care costs. Results: The population consisted of 612 eligible women who filled at
least one prescription for AIs (n=339, 55%) or tamoxifen (n=273, 45%) within 1 year of
diagnosis. The mean PDCs were 71% in the first year and 49% after 4 years. Only 25% of
women receiving hormone therapy maintained a PDC of at least 80% after 4 years. Age
differences in the likelihood of adherence were observed in hormone therapy. Race, cancer
stage, and comorbidity were not associated with long-term adherence to hormone therapy
significantly. Adherence to hormone therapy was associated with 31% decrease in medical
costs per year across 4 years but not associated with total health care cost savings.
Conclusions: Long-term use of hormone therapy remains low in the study population.
Adherence to hormone therapy in insured, low-income women with breast cancer declines
gradually over time. Suboptimal adherence to hormone therapy was associated with higher
medical costs across 4 year period. Early interventions must be undertaken to improve
adherence and achieve cost savings over the treatment period.
2
INTRODUCTION
Hormone therapy is a systemic treatment for hormone-receptor-positive breast cancers to
remove hormones or block their action and stop cancer cells from growing (1). Clinical studies
showed higher recurrence rates and worse survival with < 5 years of hormone treatment, thus
recognizing the potential for long-term regimens(2-7). However, knowledge on actual use of
hormone therapy outside of clinical trial populations, especially in low income patients, is limited.
Chlebowski et al, reviewed the literature on adherence to hormone therapy for breast cancer
and compared adherence in clinical trial and practice setting populations (8). Overall 23%-28%
of the participants in clinical trials discontinued therapy after 4 years and 30%-50% of tamoxifen
users discontinued treatments in clinical practice settings. Gotay et al reviewed most recent 5
year literature and found that 20%-50% of tamoxifen users and 30%-70% of aromatase
inhibitors (AIs) were adherent for 3-5 years in various practice settings (9). Overall the actual
long-term use of hormone therapy remains low and less than desirable. It is worth noting that
among those studies very few assessed long-term use of hormone therapy in low income
populations.
Health insurance status has been identified as a factor associated with breast cancer outcomes.
Previous studies indicated that women covered by Medicaid were more likely to have advanced
breast cancer and worse survival compared to those covered by private insurance (10-12).
However, it is not clear whether medication use behavior is associated with breast cancer
recurrence and survival in low income women (13). Understanding variation in the use of
hormone therapy and identifying associated predictors in Medicaid patients with breast cancer
would allow policymakers to identify obstacles to hormone therapy, estimate medication use
behavior, and develop effective disease management in this disadvantaged population.
The associations between medication nonadherence and risk for worsened clinical and
economic outcomes in chronic diseases and depression have been shown in previous studies
(14-19). How to improve medication adherence has been a promising strategy to achieve
desirable health outcomes and control health care costs.
However, it is not clear whether consistent results could be observed in patients receiving
hormone therapy to treat breast cancer. Long term use of hormone therapy could increase
experience of serious adverse events that may increase additional treatment costs. (20-23) On
the other hand, discontinuation of hormone therapy increases risk for disease progression that
may also increase costs to treat recurrent breast cancer. (24) To our knowledge, studies on the
3
association between adherence to hormone therapy and health care costs are very limited.
Understanding the impacts of long-term hormone therapy use behavior on health care costs
would help develop effective interventions to improve health and economic outcomes in breast
cancer care.
We created a linked database of South Carolina Central Cancer Registry (SCCCR) and South
Carolina Medicaid claims data and used a repeated measures method to assess intensity and
predictors of hormone therapy use and yearly health care service utilization and costs among a
cohort of Medicaid breast cancer women for up to 4 years after hormone therapy initiation. Our
specific aims were (1) to describe patient characteristics of low income women with breast
cancer using adjuvant hormone therapy, (2) to describe long-term adherence to hormone
therapy and identify associated predictors, and (3) examine the association between long-term
hormone therapy adherence and medical and total health care costs in low income, insured
women with breast cancer.
METHODS
Study population
This was a longitudinal cohort study. We used the SCCCR and South Carolina Medicaid
administrative database to identify women who were diagnosed with non-metastatic breast
cancer between 2000 and 2008. We restricted our sample to patients with hormone receptorpositive or unknown breast cancer receiving tamoxifen only or aromatase inhibitors (AIs) only
within 1 year of diagnosis. Tamoxifen and AIs (anastrozole, exemestane, and letrozole) were
identified by national drug codes (NDC). All patients were continuously enrolled in Medicaid for
1 year before and at least 1 year after the hormone therapy initiation. Eligible individuals were
followed up for 1, 2, 3 or 4 years after the index date (date of first prescription of hormone
therapy). The total number of patients for adherence rate measures in each interval could be
different due to the variation of enrollment eligibility. Patients with partial enrollment in the given
intervals (2, 3, or 4 years) were not included for adherence measures for the corresponding
intervals. Thus, each subject could have at least 1 and up to 4 adherence rates for the given
intervals. The Medicaid claims and SCCCR data were merged by South Carolina Office of
Research and Statistics using probabilistic match. Patient first name, last name, social security
number, and date of birth were used for the linkage. After a successful merge patient identifiers
were removed and the de-identified data was provided to us for analysis. The study protocol
was approved by the Institutional Review Board (IRB) at University of South Carolina.
4
Outcome measures
The outcomes of interest was medication adherence or the degree of prescription-filling in a
given interval (1, 2, 3 or 4 years), measured as proportion of days covered (PDC). The number
of days supplied from each filled prescription was used to calculate the proportion of days in
which a patient had hormone therapy available in each interval (25, 26). Adherence was
defined as a PDC ≥ 80% (26, 27)
We used the amount paid by Medicaid for all health care services in calculation of the costs for
each patient. Medical costs including inpatient, outpatient, emergency room (ER), and
physician office costs in each year were retrieved from medical claims. Pharmacy costs of all
prescription drugs use each year were obtained from pharmacy claims. Annual total health care
costs were a sum of medical and pharmacy costs associated with any conditions. Hormone
therapy cost in each year was a sum of tamoxifen or AIs cost. All costs were standardized to
U.S. dollar 2011. Medical utilization per year was identified from medical claims including total
numbers of physician office visits and hospital outpatient visits and frequencies of at least one
ER visit and inpatient service.
Potential predictors of adherence to hormone therapy and health care costs
Demographic and clinical characteristics were collected from the SCCCR-Medicaid linked data
set. Demographic variables included age and race. Stage categories from Surveillance
Epidemiology and End Results (SEER) were used to define breast cancer as localized or
regional. Charlson comorbidity index was used to create a weighted score (ranging from 1-6) to
assess the severity of comorbidity (28). International Classification of Disease, 9th revision
(ICD-9) codes were used to identify comorbid conditions from Medicaid claims data.
Data analysis
The mean and median PDC and the proportion of patients who were adherent to hormone
therapy were calculated for each interval (1, 2, 3, and 4 years). To identify significant predictors
associated with adherence, generalized linear models for repeated measures were used to
estimate likelihood of adherence.
Generalized linear models (log link and gamma distribution) with repeated measures were
performed to assess differences in annual total health care costs and medical costs between
adherence and nonadherence groups after adjusting for age, race, comorbidity, cancer stages
5
and type of hormone therapy. The level of statistical significance was set at α=0.05 for all the
analyses. All data were analyzed by using SAS 9.2 (SAS Institute, Cary, NC).
RESULTS
A total of 727 patients started hormone therapy within 1 year of diagnosis of non-metastatic
breast cancer were identified from the SCCCR-Medicaid linked data set (2000-2008). Excluding
those using both AIs and tamoxifen (n=115), 612 eligible women were included in the study.
Table 1 displays descriptive characteristics of the study population. More than 55% of the
women (n=339) used AIs only during study period and nearly 45% (n=273) used tamoxifen only.
Approximate 60% (n=358) of the population were younger than 65 years. The population was
mainly composed of blacks (47.2%) and whites (41%). More than 50% of the patients were
diagnosed with breast cancer between 2000 and 2003. According to the SEER cancer stages,
50% of the patients were local stage.
Figure 1A shows the mean, median, and interquartile range for the PDCs of hormone therapy
observed at each interval over the 4-year period. The use of hormone therapy declined
gradually after initiation of therapy, from a mean PDC of 71% in the first year (n=612) to 63% in
2 years (n=493), 56% in 3 years (n=395), and 43% in 4 years (n=331). The median PDC
declined from 83% in the first year to 74% in 2 years, 58% in 3 years and 47% in 4 years after
initiation of hormone therapy. As shown in Figure 1B, tamoxifen and AI users presented similar
trends in the adherence rate during the study period. However, the mean PDC of tamoxifen
declined more slowly than that of AIs. Figure 2 displays that the proportion of patients who were
adherent to hormone therapy was 56%, 44%, 34%, and 25% after 1, 2, 3, and 4 years.
6
Table 1 Characteristics of eligible patients with breast cancer who initiated adjuvant hormone therapy
within one year after diagnosis
Variable
Total (n=612)
Age at diagnosis, years, mean (SD)
61.8 (14.1)
18-44, n (%)
81 (13.2)
45-64, n (%)
277 (45.3)
≥65, n (%)
254 (41.5)
Race, n (%)
White
251 (41.0)
Black
289 (47.2)
Other
72 (11.8)
Stage, n (%)
In situ
62 (10.1)
Local
311 (50.8)
Regional
239 (39.0)
Tumor grade, n (%)
Low (1 or 2)
362 (59.2)
High (3 or 4)
195 (31.9)
Unknown
55 (9.0)
Breast conserving surgery, n (%)
399 (65.2)
Mastectomy, n (%)
200 (32.7)
Radiation, n (%)
275 (44.9)
Chemotherapy, n (%)
179 (29.2)
Charleson comorbidity index, Mean (SD)
2.5 (3.7)
0, n (%)
321 (52.5)
1-2, n (%)
100 (16.3)
3-6, n (%)
124 (20.3)
>6, n (%)
67 (11.0)
Year at diagnosis, n (%)
2000-2003
317 (51.8)
2004-2008
295 (48.2)
Hormone Therapy, n(%)
Aromatase inhibitors
339 (55.4)
Tamoxifen
273 (44.6)
7
1.00
PDC
0.75
0.50
0.25
0
1
2
3
Years after initiation of hormone therapy
Median
● Mean
4
1A
0.8
Aromatase inhibitors
Tamoxifen
PDC
0.6
0.4
0.2
0
0
1
2
3
4
Years after initiation of hormone therapy
5
1B
Figure 1 PDC of hormone therapy in each interval (1-4 years) in Medicaid enrolled women with breast
cancer
A: Mean, median, and interquartile range of PDC of hormone therapy. Boxes represent the interquartile
range; vertical whiskers extend to the maximum and minimum observation for each interval. The number
of subjects evaluated at each interval (1-4 years): n=612, 493, 395, 331.
8
B: Mean (95 confidence interval) of PDC by type of hormone therapy (aromatase inhibitors and
tamoxifen). The number of subjects evaluated in each interval (1-4 years): n (aromatase inhibitors) =339,
274, 215, and 177, n (tamoxifen) =273, 219, 180, and 154.
PDC: proportion of days covered
Table 2 Association between potential predictors and adherence to hormone therapy
Variable
Odds Ratio (95% confidence
interval)
Age, years
≥65
1.00
45-64
1.14 (0.81, 1.59)
< 45
0.31 (0.18, 0.55)***
Race
White
1.00
Black
0.49 (0.54 1.41)
Other
0.87 (0.57, 1.10)
Stage
Regional
1.00
Local
0.99 (0.70, 1.39)
In situ
0.81 (0.45, 1.46)
Chemotherapy
Yes
1.00
No
0.59 (0.39, 0.91)**
Radiation
Yes
1.00
No
1.59 (1.05, 2.41)*
Surgery
BCS
1.00
Mastectomy
0.81 (0.51, 1.28)
Hormone therapy
Tamoxifen
1.00
Aromatase inhibitors
0.62 (0.44, 0.87)**
Charleson comorbidity score
0.98 (0.93, 1.02)
Year
0.57 (0.53, 0.62)***
*: p<0.05, **: p<0.01, ***: p<0.001. Adherence was defined as having ≥ 80% of days with hormone
therapy available in a given interval.
Table 2 shows the results of the association between long-term hormone therapy adherence
and various predictors in the study population. Long-term hormone therapy adherence was low
for patients who were younger than 45 years, received adjuvant chemotherapy prior to hormone
therapy, and used AIs as hormone therapy agents. After adjusting for all other characteristics
studied, patients took AIs as hormone therapy were 38% less likely to be adherent to the
treatment during 4 years. Compare to the women older than 45 years, the odds of adherence to
hormone therapy were reduced by 69% in women younger than 45 years. In addition, the
9
likelihood of adherence to hormone therapy decreased by 43% with an increase in the years
after initiation of the therapy. Race, cancer stage, surgery type, and comorbidity were not
significantly associated with long-term adherence to hormone therapy.
Table 3 displays the medical utilization, and health care costs across 4 years. Mean PDCs (per
year) ranged from 0.71 to 0.75 and more than 50% women were adherent to hormone therapy
(PDC ≥ 0.8) each year during 4 year period. Mean total health care costs per year ranged from
$7993 to $12373 per patients across 4 years. Pharmacy costs per year accounted for 45%-60%
of total health care costs across 4 years. Table 3 displays unadjusted annual total health care
costs, medical costs, and pharmacy costs in adherence and nonadherence women across 4
years. In general, medical costs in the adherence group appeared to be lower than in the
nonadherence group during 4 year period. However, the similar trend was not observed in
yearly total health care costs during the study period.
Table 3 Hormone therapy adherence, medical utilization, and health care costs across 4 years in
Medicaid patients with breast cancer
Variable
Year 1
(n=410)
0.71 (0.30)
224 (54.6)
10.7 (12.8)
5.8 (7.8)
132 (32.2)
79 (19.3)
9752 (5728)
Year 2
(n=246)
0.71 (0.31)
133 (54.1)
7.8 (8.1)
3.8 (4.2)
86 (35.0)
48 (19.5)
12373 (6720)
Year 3
(n=146)
0.75 (0.30)
90 (61.6)
7.1 (7.5)
3.6 (5.9)
38 (26.0)
24 (16.4)
10001 (5871)
Year 4
(n=94)
0.71 (0.31)
52 (55.3)
6.0 (6.7)
3.5 (5.6)
20 (21.3)
14 (14.9)
7993 (5047)
PDC, mean (SD)
Adherence (PDC≥0.8), n(%)
Total office visits, mean (SD)
Total outpatient visits, mean (SD)
ER, n (%)
Hospitalization, n (%)
Total costs, mean (median), US $
All medical costs, mean (median),
4396 (738)
6669 (851)
4373 (728)
3064 (752)
US $
All pharmacy costs, mean (median),
5356 (4247) 5704 (4448)
5628 (4351)
4930 (3978)
US $
Hormone therapy costs, mean
1815 (1608) 1638 (1302)
1492 (892)
1266 (578)
(median), US $
AI costs, mean (median), US $
2742 (3059) 2490 (2826)
2687 (3224)
2606 (2901)
Tamoxifen costs, mean (median),
699 (446)
634 (502)
388 (457)
286 (358)
US $
AIs: aromatase inhibitors; PDC: proportion of days covered; SD: standard deviation, ER: emergency room;
The results of the association between adherence to hormone therapy and total or medical
costs by regression analyses are summarized in Table 4. After controlling for age, race,
comorbidity, cancer stage and type of hormone therapy, adherence to hormone therapy was
associated with 31% decrease in medical costs per year across 4 years. No significant
10
difference in total health care costs was observed between adherent and nonadherent women.
Figure 2 demonstrated the differences in mean medical costs per year between adherence and
nonadherence to hormone therapy across 4 years, after adjusting for confounders. A significant
reduction in medical costs was found in the first 2 years between adherence and nonadherence
patients and then the gap narrowed in the next 2 years.
Table 4 Associations between adherence to hormone therapy and health care costs across 4 years in
Medicaid patients with breast cancer (n=410)
Estimated coefficient (SE)
Variable
Adherence
No
Yes
Age (years)
Race
White
Black
Other
Charleson comorbidity score
Cancer stage
Regional
Local
In situ
Year
Hormone therapy
AIs
Tamoxifen
*: p<0.05, **:p<0.01, ***:p<0.001
log (total cost)
log (Medical costs)
0
0.16 (0.13)
-0.03 (0.01)***
0
-0.37 (0.24)**
-0.06 (0.01)***
0
-0.29 (0.11)**
-0.06 (0.27)
0.08 (0.01)***
-0.35 (0.21)
-0.08 (0.44)
0.16 (0.03)***
0
-0.08 (0.14)
-0.14 (0.24)
0.08 (0.05)
0
0.03 (0.24)
-0.22 (0.46)
0.16
0
-0.33 (0.12)**
0
-0.21 (0.27)
DISCUSSION
Cancer patients are generally expected to be highly persistent in treatment due to the lifethreatening disease. Despite clinical evidence of the effectiveness of hormone therapy in
reducing breast cancer recurrence and mortality, our study indicates that actual use in low
Medicaid enrolled women is low and that a decrease in long-term adherence to AIs is more than
the use of tamoxifen during 4 year period, thus likely reducing their potential benefits. A better
understanding of the magnitude and predictors of long-term adherence to hormone therapy may
provide health care professionals important information necessary to develop appropriate
interventions to improve adherence among low income individual patients, as well as design
and evaluation programs to improve population-level disease-free survival.
11
Adjusted Medical costs per year ($)
12000
Adherence
10000
Non-adherence
8000
6000
4000
2000
0
1
2
Year
3
4
Figure 2 Adjusted mean (95% confidence interval) medical costs per year (in US $) in adherence and
nonadherence groups across 4 years among women undergoing hormone therapy for breast cancer
The adherence rates in our retrospective analysis were far lower than that reported in clinical
trials, where 5-year discontinuation rates ranged from 12% to 24% for tamoxifen use and 2-5
year discontinuation rates ranged from 10% to 24% (29, 30). By contrast, we found that less
than 60% of all patients were adherent to hormone therapy after 1 year and less than 20% AI
users and nearly 35% tamoxifen users were adherent to hormone therapy after 4 years.
Adherence measures in clinical trials which are self-reported by participants or measured by
electronic monitoring or pill counts, are different from the adherence measure in our study
based on administrative claims data. Adherence measures in clinical trials may not reflect
actual medication use behavior in clinical practice settings. Trial participants could get
reminders or support to improve adherence rates from ancillary team members including
clinicians, researchers, and administrative managers (8). Thus, studies such as ours, which
assess long-term adherence to hormone therapy outside of clinical trials, are necessary to
improve outcomes of breast cancer treatment.
We found that the adherence rate of AIs decreased more rapidly than that of tamoxifen and AI
users were 38% less likely to be adherence to hormone therapy than tamxifen users during the
follow-up period, suggesting that different strategies might be considered to enhance long-term
use of tamoxifen and AIs. AI therapy has become a standard of care for the treatment of most
postmenopausal women with early-stage breast cancer (20). However, health belief and
12
medication side effects causing discontinuation of hormone therapy could not be identified
based on administrative claims data. It remains to be determined whether there is a difference
in discontinuation of hormone therapy associated with adverse events between long-term use of
AI and tamoxifen.
Comparing our findings with other studies of adjuvant hormone therapy use among Medicaid
population the mean adherence rates 1 year after an initiation of therapy (74% for AIs and 68%
for tamoxifen) were consistent with the rate (75%) found by Kimmick et al among North
Carolina Medicaid patients (1998-2002) (31). However, to our knowledge very few studies
assessed the adherence to hormone therapy (> 1 year) in low income women. Compared with
the results in women enrolled in private insurance plans or self-reported adherence, long-term
adherence to hormone therapy is poorer in low income women. During a 3 year study period,
among women with early-stage breast cancer enrolled in 3 commercial health plans adherence
rates of initial adjuvant anastrozole therapy decreased each year, ranging from 78% to 86% in
year 1 to 62% to 79% in year 3 (27). Another managed care claims data based study showed
that 49% of patients with early-stage breast cancer took adjuvant hormone therapy (tamoxifen
or AIs) continuously for the full 4.5 year duration (32). Continuous interventions to promote
adherence to hormone therapy after initiating treatment should be conducted particularly in low
income, insured women.
In terms of predictors of adherence, our study showed that long-term adherence to hormone
therapy was associated with age, treatment prior to hormone therapy, and type of hormone
therapy. We found that compared with women (≥ 65 years), women younger than 45 years were
69% less likely to use AIs or tamxifen continuously for a long term. Previous studies on
hormone therapy use in elderly (Medicare) populations or women enrolled in private health
plans (27, 32-35) indicated that generally younger (< 45 years) or extremely older (> 75 years)
patients were more likely to discontinue therapy and be suboptimally adherent. For chronic
diseases, race is a potential predictor to estimate medication use behavior; however we did not
find a racial difference in long-term use of hormone therapy between white and black women
with breast cancer. Finally we found that previous chemotherapy and radiation therapy had
different effects on the long-term use of hormone therapy. Further studies are needed to explore
the impacts of treatments prior to initiation of hormone therapy on their medication-taking
behavior in low income women.
13
To our knowledge, this is the one of the first studies to examine the relationship between
medication adherence and health care costs in breast cancer care. In this study, we analyzed
Medicaid claims data linked with cancer central registry data to explore the association between
long-term use of hormone therapy and resource utilization in low income, insured women with
breast cancer. Our study found that suboptimal adherence to hormone therapy was associated
with higher medical costs across 4 year period. This result is consistent with previous studies on
chronic diseases, such as diabetes, hyperlipidemia, and hypertension. (14-16) In Sokol et al’s
study, cost savings were observed for all-cause medical costs at high levels of medication
adherence (80% to 100%) in patients with diabetes, hypertension and hypercholesterolemia
covered by employer-sponsored health insurance plans. (15) Another study also reported
adherence to statin therapy associated with 15% decreased in medical costs in Medicaid
enrollees with type 2 diabetes. (16) Patients with continuous use of hormone therapy for 5
years have a lower risk for disease progression including recurrence of breast cancer and
mortality, so health services with high costs to treat more advanced breast cancer could be
reduced in the adherence group. It is worth noting that medical costs in our study varied across
4 years between adherence and nonadherence groups. A significant reduction in medical costs
was found between adherence and nonadherence women for first and second year of follow-up
and then the gap gradually narrowed for the third and fourth years. Clinical studies indicated
that early recurrent breast cancer occurs at the peak of approximately 2 years postsurgery and
most of early recurrence events are distant metastatic breast cancer. (36, 37) Therefore,
conducting early interventions to enhance continuous use of hormone therapy would be a
potential strategy to reduce risk for breast cancer recurrence and to control medical costs for
treating disease progression.
On the other hand, unlike diabetes, hypertension and hypercholesterolemia, total health care
cost savings were not observed in women adherent to hormone therapy to treat breast cancer
during the 4 year period. Studies on aforementioned lifestyle chronic diseases concluded that in
patients showing adherence to medications, the higher pharmacy costs might be more than
offset by medical cost reductions due to lower risks for medical utilization to treat diabetic
complications and coronary heart diseases, producing a net reduction in total health care costs.
However, breast cancer management costs are composed of costs for remaining on adjuvant
treatment and the costs for treating adverse events. Long term use of hormone therapy could
increase the probability to experience either minor or serious adverse events. Following the
occurrence of adverse events, if patients remained on the original hormone therapy, costs to
14
manage various adverse events would be added to total health care costs. Further studies are
needed to assess the association between long term adherence to hormone therapy and
adverse events-related costs.
Although claims data have been an important source to assess cost of health care and health
service utilization for various cohorts, these data are collected for administrative purposes and
several limitations should be considered. The study cohort consisted of patients enrolled in
South Carolina Medicaid. Although this population is representative of insured, low-income
women with breast cancer in SC, different results may have been observed in other states and
among privately insured patients. Studies based on a national Medicaid population are needed
in the future. Hormone therapy may have been discontinued by the prescriber for reasons
indiscernible using administrative data such as adverse drug events, lack of efficacy, or
conversion to other therapy. In clinical trials, 5-15% of patients had discontinued due to adverse
events by the end of 5-year intervention (29, 30). As is commonly done in studies of adherence
using administrative claims, we assumed a prescription filled was a prescription taken. Thus,
our results may better reflect prescribing patterns rather than actual use of AIs and tamoxifen.
Finally, the accuracy of resource utilization and completeness of claim submissions by health
care providers should be considered. The amount paid by Medicaid could vary due to miscoding
and missing information.
POLICY IMPLICATIONS AND FUTURE STUDIES
Findings from our study have important implications for health care professionals and other
decision makers responsible for the appropriate use of hormone therapy. They confirmed that
long-term adherence to hormone therapy in patients with breast cancer in clinical settings
cannot be assumed to approximate the levels observed in clinical trials. Interventions should be
initiated early in therapy to improve adherence, including educational, behavioral,
pharmacological and multidimensional approaches. Given low income, insured women with
breast cancer, interventions should focus on enhancing accessibility to health care, addressing
distress or side effects from the regimen, and improving patients’ understanding of clinical
benefits of long-term hormone therapy. Whether suggested interventions will improve
adherence deserves further studies.
The use of hormone therapy and health care costs across 4 year period in Medicaid
beneficiaries with breast cancer reported in this study demonstrated that adherence to hormone
15
therapy were associated with decreased medical costs. The greatest reduction of medical costs
occurred in the first 2 years after initiation of hormone therapy. The study highlights the
importance to identify women with breast cancer at risk for lower level of medication adherence
to hormone therapy and, in turn, higher medical costs owing to disease progression. Thus, early
adherence enhancing interventions may be suggested to help improve the use of hormone
therapy and reduce the risk for disease recurrence. In addition, analyses of patient medication
use behaviors may be a potential strategy to cost-effectively manage the breast cancer care in
women receiving hormone therapy in practicing settings. Further research is needed in larger
populations to fully explore the relationship between adherence to hormone therapy and costs
for treating adverse events and recurrent breast cancer.
PUBLICATIONS
Two original manuscripts were published or accepted by peer-reviewed journals.
1. Wu J, Stafkey-Mailey D, Bennett CL. Long-term adherence to hormone therapy in Medicaid
enrolled women with breast cancer. Health Outcomes Research in Medicine. 2012, 3 (4):e195e203.
2. Wu J, Lu ZK. Adherence to hormone therapy and costs in women with breast cancer. Am J
Pharm Benefits. November 2012 (Accepted).
REFERENCES
1. Breast cancer treatment: Treatment option overview [Internet]. Bethesda, MD: National Cancer Institute;
2011 [updated 11/04; cited 3/23/2012]. Available from:
http://www.cancer.gov/cancertopics/pdq/treatment/breast/Patient/page5#Keypoint24.
2. Early Breast Cancer Trialists' Collaborative Group (EBCTCG). Effects of chemotherapy and hormonal
therapy for early breast cancer on recurrence and 15-year survival: An overview of the randomised trials.
Lancet. 2005 May 14-20;365(9472):1687-717.
3. Goss PE, Ingle JN, Martino S, Robert NJ, Muss HB, Piccart MJ, et al. Randomized trial of letrozole
following tamoxifen as extended adjuvant therapy in receptor-positive breast cancer: Updated findings
from NCIC CTG MA.17. J Natl Cancer Inst. 2005 Sep 7;97(17):1262-71.
4. McCowan C, Shearer J, Donnan PT, Dewar JA, Crilly M, Thompson AM, et al. Cohort study examining
tamoxifen adherence and its relationship to mortality in women with breast cancer. Br J Cancer. 2008 Dec
2;99(11):1763-8.
5. Swedish Breast Cancer Cooperative Group. Randomized trial of two versus five years of adjuvant
tamoxifen for postmenopausal early stage breast cancer. swedish breast cancer cooperative group. J
Natl Cancer Inst. 1996 Nov 6;88(21):1543-9.
6. Geiger AM, Thwin SS, Lash TL, Buist DS, Prout MN, Wei F, et al. Recurrences and second primary
breast cancers in older women with initial early-stage disease. Cancer. 2007 Mar 1;109(5):966-74.
7. Howell A, Cuzick J, Baum M, Buzdar A, Dowsett M, Forbes JF, et al. Results of the ATAC (arimidex,
tamoxifen, alone or in combination) trial after completion of 5 years' adjuvant treatment for breast cancer.
Lancet. 2005 Jan 1-7;365(9453):60-2.
8. Chlebowski RT, Geller ML. Adherence to endocrine therapy for breast cancer. Oncology. 2006;71(12):1-9.
9. Gotay C, Dunn J. Adherence to long-term adjuvant hormonal therapy for breast cancer. Expert Rev
Pharmacoecon Outcomes Res. 2011 Dec;11(6):709-15.
16
10. Ayanian JZ, Kohler BA, Abe T, Epstein AM. The relation between health insurance coverage and
clinical outcomes among women with breast cancer. N Engl J Med. 1993 Jul 29;329(5):326-31.
11. Halpern MT, Bian J, Ward EM, Schrag NM, Chen AY. Insurance status and stage of cancer at
diagnosis among women with breast cancer. Cancer. 2007 Jul 15;110(2):403-11.
12. Bradley CJ, Given CW, Roberts C. Race, socioeconomic status, and breast cancer treatment and
survival. J Natl Cancer Inst. 2002 Apr 3;94(7):490-6.
13. Weaver KE, Camacho F, Hwang W, Anderson R, Kimmick G. Adherence to adjuvant hormonal
therapy and its relationship to breast cancer recurrence and survival among low-income women. Am J
Clin Oncol. 2012 Feb 6.
14. Balkrishnan R, Rajagopalan R, Camacho FT, Huston SA, Murray FT, Anderson RT. Predictors of
medication adherence and associated health care costs in an older population with type 2 diabetes
mellitus: A longitudinal cohort study. Clin Ther. 2003 Nov;25(11):2958-71.
15. Sokol MC, McGuigan KA, Verbrugge RR, Epstein RS. Impact of medication adherence on
hospitalization risk and healthcare cost. Med Care. 2005 Jun;43(6):521-30.
16. Wu J, Seiber E, Lacombe VA, Nahata MC, Balkrishnan R. Medical utilization and costs associated
with statin adherence in medicaid enrollees with type 2 diabetes. Ann Pharmacother. 2011
Mar;45(3):342-9.
17. Cantrell CR, Eaddy MT, Shah MB, Regan TS, Sokol MC. Methods for evaluating patient adherence to
antidepressant therapy: A real-world comparison of adherence and economic outcomes. Med Care. 2006
Apr;44(4):300-3.
18. Eaddy MT, Druss BG, Sarnes MW, Regan TS, Frankum LE. Relationship of total health care charges
to selective serotonin reuptake inhibitor utilization patterns including the length of antidepressant therapy-results from a managed care administrative claims database. J Manag Care Pharm. 2005 Mar;11(2):14550.
19. Sheehan DV, Keene MS, Eaddy M, Krulewicz S, Kraus JE, Carpenter DJ. Differences in medication
adherence and healthcare resource utilization patterns: Older versus newer antidepressant agents in
patients with depression and/or anxiety disorders. CNS Drugs. 2008;22(11):963-73.
20. Karnon J. Aromatase inhibitors in breast cancer: A review of cost considerations and cost
effectiveness. Pharmacoeconomics. 2006;24(3):215-32.
21. Doughty JC. When to start an aromatase inhibitor: Now or later? J Surg Oncol. 2011 Jun
1;103(7):730-8.
22. Duffy S, Jackson TL, Lansdown M, Philips K, Wells M, Clack G, et al. The ATAC adjuvant breastcancer trial: Six-year results of the endometrial subprotocol. J Obstet Gynaecol. 2010;30(6):596-604.
23. Arimidex, Tamoxifen, Alone or in Combination Trialists' Group, Buzdar A, Howell A, Cuzick J, Wale C,
Distler W, et al. Comprehensive side-effect profile of anastrozole and tamoxifen as adjuvant treatment for
early-stage breast cancer: Long-term safety analysis of the ATAC trial. Lancet Oncol. 2006 Aug;7(8):63343.
24. Subramanian S, Trogdon J, Ekwueme DU, Gardner JG, Whitmire JT, Rao C. Cost of breast cancer
treatment in medicaid: Implications for state programs providing coverage for low-income women. Med
Care. 2011 Jan;49(1):89-95.
25. Leslie SR, Gwadry-sridhar F, Thiebaud P, Patel BV. Calculating medication compliance, adherence
and persistence in administrative pharmamcy claims databases. Pharmaceutical Programming.
2008;1(1):13.
26. Sikka R, Xia F, Aubert RE. Estimating medication persistency using administrative claims data. Am J
Manag Care. 2005 Jul;11(7):449-57.
27. Partridge AH, LaFountain A, Mayer E, Taylor BS, Winer E, Asnis-Alibozek A. Adherence to initial
adjuvant anastrozole therapy among women with early-stage breast cancer. J Clin Oncol. 2008 Feb
1;26(4):556-62.
28. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity
in longitudinal studies: Development and validation. J Chronic Dis. 1987;40(5):373-83.
29. Fisher B, Dignam J, Bryant J, DeCillis A, Wickerham DL, Wolmark N, et al. Five versus more than five
years of tamoxifen therapy for breast cancer patients with negative lymph nodes and estrogen receptorpositive tumors. J Natl Cancer Inst. 1996 Nov 6;88(21):1529-42.
30. Fisher B, Costantino J, Redmond C, Poisson R, Bowman D, Couture J, et al. A randomized clinical
trial evaluating tamoxifen in the treatment of patients with node-negative breast cancer who have
estrogen-receptor-positive tumors. N Engl J Med. 1989 Feb 23;320(8):479-84.
17
31. Kimmick G, Anderson R, Camacho F, Bhosle M, Hwang W, Balkrishnan R. Adjuvant hormonal
therapy use among insured, low-income women with breast cancer. J Clin Oncol. 2009 Jul
20;27(21):3445-51.
32. Hershman DL, Shao T, Kushi LH, Buono D, Tsai WY, Fehrenbacher L, et al. Early discontinuation
and non-adherence to adjuvant hormonal therapy are associated with increased mortality in women with
breast cancer. Breast Cancer Res Treat. 2011 Apr;126(2):529-37.
33. Ziller V, Kalder M, Albert US, Holzhauer W, Ziller M, Wagner U, et al. Adherence to adjuvant
endocrine therapy in postmenopausal women with breast cancer. Ann Oncol. 2009 Mar;20(3):431-6.
34. Owusu C, Buist DS, Field TS, Lash TL, Thwin SS, Geiger AM, et al. Predictors of tamoxifen
discontinuation among older women with estrogen receptor-positive breast cancer. J Clin Oncol. 2008
Feb 1;26(4):549-55.
35. Lash TL, Fox MP, Westrup JL, Fink AK, Silliman RA. Adherence to tamoxifen over the five-year
course. Breast Cancer Res Treat. 2006 Sep;99(2):215-20.
36. Saphner T, Tormey DC, Gray R. Annual hazard rates of recurrence for breast cancer after primary
therapy. J Clin Oncol. 1996 Oct;14(10):2738-46.
37. Mansell J, Monypenny IJ, Skene AI, Abram P, Carpenter R, Gattuso JM, et al. Patterns and predictors
of early recurrence in postmenopausal women with estrogen receptor-positive early breast cancer. Breast
Cancer Res Treat. 2009 Sep;117(1):91-8.
18