A claims database analysis of persistence with alendronate therapy and fracture risk in post-menopausal women with osteoporosis.

Abstract

To explore the relationship between persistence with alendronate therapy and fracture rates in women with postmenopausal osteoporosis.
Claims data from a large US health plan database were used to examine persistence with therapy in postmenopausal women followed for 24 months. Persistence was defined as the time (in days) from the date of first fill to the run-out date of the last prescription with no lapses > 30 days after completion of the previous refill. A persistent cohort (length of persistence > or = 182 days) and a nonpersistent cohort (length of persistence < 182 days) were defined. The number of patients with a fracture claim in each cohort was determined. Cox-proportional hazards regression (HR) analysis was used to determine significant differences in fracture rates between the two cohorts.
4769 patients were followed for 24 months. Patients in the persistent cohort were significantly more likely to receive a treatment (vs. prevention) dose of alendronate (p = 0.03) and to be older than 65 years (p = 0.04). There was a trend toward more fractures in the non-persistent (4.9%) than in the persistent cohort (3.9%; p = 0.09). When controlled for other significant factors (including age and previous fractures) patients in the persistent cohort were 26% less likely to have a fracture diagnosis claim during the study period than those in the non-persistent cohort (HR = 0.74; 95% CI, 0.549-0.996; p = 0.045). Prescription fill data are an indirect measure of medication-taking behavior. The use of claims data to estimate persistence and identify fracture events prohibits the establishment of causality between these two variables.
Study results demonstrated that non-persistence with therapy, along with previous fracture and increasing age, was associated with a greater risk of fracture.

Full text

Page 1

ORIGINAL ARTICLE
A claims database analysis of
persistence with alendronate
therapy and fracture risk in
post-menopausal women with
osteoporosis
Deborah T. Gold
Frytak
a, Bradley C. Martin
b, Jennifer R.
c, Mayur M. Amonkar
d and Felicia Cosman
e
a Duke University Medical Center, Durham, NC, USA
b University of Arkansas for Medical Sciences, Little Rock, AR, USA
c i3 Innovus, Eden Prairie, MN, USA
d GlaxoSmithKline, Collegeville, PA, USA
e Helen Hayes Hospital, West Haverstraw, NY, and Columbia University 
Medical Center, New York, NY, USA
Address for correspondence:  Deborah T. Gold, PhD, Department of Psychiatry, Box 3003, Duke 
University Medical Center, Durham, NC 27710, USA. Tel.: +1 919 660 7530; Fax: +1 919 684 8569; 
email: dtg@geri.duke.edu
Key words: Alendronate – Bisphosphonates – Claims – Fracture risk – Medical database – Persistence
0300-7995
doi:10.1185/030079906X167615
All rights reserved: reproduction in whole or part not permitted
CuRRenT MeDiCAl ReseARCh AnD OpiniOn®
Vol. 23, No. 3, 2007, 585–594
© 2007 liBRAphARM liMiTeD
Paper 3604 585
Objective: To explore the relationship between
persistence with alendronate therapy and fracture
rates in women with postmenopausal osteoporosis.
Research design and methods: Claims data
from a large US health plan database were
used to examine persistence with therapy in
postmenopausal women followed for 24 months.
Persistence was defined as the time (in days)
from the date of first fill to the run-out date of the
last prescription with no lapses > 30 days after
completion of the previous refill. A persistent cohort
(length of persistence ≥ 182 days) and a non-
persistent cohort (length of persistence < 182 days)
were defined. The number of patients with a
fracture claim in each cohort was determined.
Cox-proportional hazards regression (HR) analysis
was used to determine significant differences in
fracture rates between the two cohorts.
Results: 4769 patients were followed for
24 months. Patients in the persistent cohort were
significantly more likely to receive a treatment
(vs. prevention) dose of alendronate (
to be older than 65 years (
trend toward more fractures in the non-persistent
(4.9%) than in the persistent cohort (3.9%;
p = 0.09). When controlled for other significant
factors (including age and previous fractures)
patients in the persistent cohort were 26% less
likely to have a fracture diagnosis claim during the
study period than those in the non-persistent cohort
(HR = 0.74; 95% CI, 0.549–0.996; p = 0.045).
Prescription fill data are an indirect measure of
medication-taking behavior. The use of claims data
to estimate persistence and identify fracture events
prohibits the establishment of causality between
these two variables.
Conclusion: Study results demonstrated that
non-persistence with therapy, along with previous
fracture and increasing age, was associated with
a greater risk of fracture.
p = 0.03) and
p = 0.04). There was a
A B S T R A C T

Page 2

586 Claims data: persistence with alendronate and fractures
© 2007 liBRAphARM lTD – Curr Med Res Opin 2007; 23(3)
Introduction
Osteoporosis, a highly prevalent condition
characterized by low bone mass, deterioration of
bone tissue, increased bone fragility, and a consequent
increased risk of fractures affects females and males in
a 4:1 ratio, respectively
two women and one in four men older than the age
of 50 years are likely to experience a fracture in their
lifetime
age of 50 years die within 1 year of experiencing a
hip fracture
care
fractures may be associated with increased mortality
Furthermore, women with at least 1 fracture have
a 2- to 5-fold greater risk for additional fractures
Thus, reducing the risk of bone fractures can result in
improved patient outcomes and reduced healthcare
burden
Bisphosphonates are a class of compounds widely
used to treat postmenopausal osteoporosis (PMO)
Multiple clinical trials have demonstrated an association
between treatment of PMO with bisphosphonates
and reductions in fracture risk
literature review concluded that bisphosphonates were
associated with the largest reduction in fracture risk
when compared with other therapies such as calcium
and vitamin D supplementation, estrogen or hormone
therapy, selective estrogen receptor modulators, and
calcitonin
agreement in the literature that patient persistence
and compliance with daily and weekly bisphosphonate
dosing regimens seen in clinical practice is suboptimal
with dosing frequency inversely proportional to
compliance
the fear of side effects can also negatively influence
compliance
Evidence shows that compared with non-persistent
and non-compliant patients, those who continue to
take their osteoporosis therapies experience improved
outcomes, such as gains in bone mineral density
(BMD)
those studies may have not been bisphosphonate-
specific (a commonly accepted therapy for
osteoporosis
short time period (e.g., 1 year)
increasing burden of osteoporosis in the United States
and the need to look at the impact of persistence over
longer time periods, the current study was conducted
to assess the relationship between persistence with
bisphosphonate therapy and 2-year fracture risk for
women with postmenopausal osteoporosis using
retrospective administrative claims data from a large
US health plan.
1,2. It is estimated that one in
1. As many as 25% of patients older than the
3 and an equal number require long-term
4. Additionally, studies have shown that vertebral
5,6.
7–9.
10.
2,11–19.
13,14,20–27. A recent
28. Despite proven efficacy, there is general
29–36,
33,37. Actual medication side effects or
29,30.
38–40 and lower fracture rates
32,41,42. However,
32,41), and/or assessed outcomes over a
32. Given the high and
Patient population and
methods
Data source
Data on eligibility, medical, and pharmacy claims from
a large, geographically well-distributed, proprietary, US
healthcare organization’s Health Insurance Portability
and Accountability Act (HIPAA) – compliant research
database, collected between July 1996 and December
31, 2003, were analyzed retrospectively. The total
number of individuals with medical and pharmacy
benefits included in the administrative health care
claims research database ranged from 3.4 million
in 1997 to 13.5 million in 2001. The database is
comprised of data from 30–35 health plans. All plans
associated with the organization are discounted, fee-
for-service, independent practice association models.
These plans provide coverage for physician, hospital,
and prescription drug services. Submitted claims (paid
or unpaid) from Medicare/Medicaid patients were
included if the patients had both medical and pharmacy
benefits. Medicare/Medicaid patients accounted for
approximately 10–12% of the analyzed population.
Eligibility and claims data in the research database
are updated monthly. The updates include application
of claims adjustment logic against all claims. This
results in the removal of reversed and duplicate claim
records.
patient population
Subjects for this study included females with commercial
coverage, 45 years of age or older, who were new
users of the bisphosphonate alendronate, approved for
the treatment of osteoporosis. The index date for
study subjects was identified as the first occurrence of
a study medication during the subject selection period
between January 1, 1997, and December 31, 2001
(Figure 1).
The following inclusion and exclusion criteria were
applied.
Inclusion criteria
At least 6 months of baseline enrollment data and
24 months of follow-up health plan eligibility.
•
Figure 1. Time periods for study
07.01.96
Subject selection
Baseline period
Follow-up period
01.01.9712.31.03 12.31.01

Page 3

© 2007 liBRAphARM lTD – Curr Med Res Opin 2007; 23(3)
Claims data: persistence with alendronate and fractures Gold et al. 587
At least one prescription for alendronate (not
including the 40
for Paget’s disease) during the subject selection
period.
mg tablet, which is indicated
Exclusion criteria
Presence of a pharmacy claim for alendronate,
risedronate, or etidronate during the 6-month
baseline period so as to exclude patients with a
history of bisphosphonate treatment.
Patients receiving risedronate during the follow-
up period were excluded due to small sample size
(n = 121).
Patients receiving etidronate during the follow-up
period were excluded because it does not have an
indication for osteoporosis.
Presence of a medical claim with an ICD-9 code
for drug-induced osteoporosis (733.09) during the
study period.
Presence of a medical claim with an ICD-9 code
for malignant cancer (140.xx–208.xx) or human
immunodeficiency virus (HIV) (042.xx) during the
study period.
Presence of a medical claim with an ICD-9 code
for Paget’s disease (731.xx) during the study
period.
Crossover of cohort (e.g., daily subjects receiving a
weekly prescription during the follow-up).
Claim for a study medication with greater than
1-month days’ supply during the follow-up so as to
avoid overestimating the length of persistence.
Days’ supply longer than 1 month (e.g., mail
order prescriptions) are less likely to mimic actual
medication-taking behavior such as the patient’s
motivation to return to the pharmacy to refill
medications.
study cohort
Subjects were grouped into the following age groups:
45–54, 55–64, > 65 years. The age of the subject was
calculated as the year of study entry minus the year
of the subject’s birth. Co-morbidities were assessed
during the 6-month baseline period. Subjects were
assigned to one of two cohorts on the basis of length of
persistence. A subject was considered to be persistent
with therapy at any time point in which a prescription
for the index medication was filled within 30 days of
the run-out date of the previous prescription (run-out
date equals the date of fill + days supply). Therefore,
a subject was considered non-persistent at the time
in which there was greater than a 30-day gap in
medication coverage. Once a period of non-persistence
was identified, the subject’s last date of persistence was
•
•
•
•
•
•
•
•
•
•
set at the run-out date of the last prescription for the
index medication. Days of persistence were calculated
as the number of days between the index date and
the last day of persistence. This length of persistence
was used to define two cohorts for analysis – a cohort
with persistence ≥ 182 days (≥ 6 months) and a second
cohort with persistence < 182 days (< 6 months).
This threshold for persistence was selected based on
previous persistence data showing a steep drop in
persistence at 1–3 months with a flattened but steady
decline at 6 months
reductions in fractures with bisphosphonates as early
as 6 months
33, and clinical data demonstrating
43–45.
presence of a fracture
During their 24-month follow-up period, subjects
in the two cohorts were examined for a claim with a
fracture diagnosis. In order to maximize the probability
of identifying osteoporotic (fragility) fractures, only
closed fractures affecting all sites were examined in
these analyses. Open wound fractures were assumed to
be traumatic and were not included. A claims diagnosis
of osteoporosis-related fracture was identified by
specific ICD-9 codes (Appendix I).
A dichotomous variable was assigned to each subject
to indicate the presence or absence of a fracture in the
follow-up period. The same methodology was applied
for a baseline fracture, defined as whether or not the
subject had a medical claim with a diagnosis of fracture
in the 6 months prior to their index date. Claims with a
Current Procedural Technology code of 70010–79999
were excluded from the definition of fractures in order
to prevent misclassifying a potential rule-out diagnosis
as a fracture. In addition, the number of days from
the subject’s index date until the first fracture was
recorded.
statistical methods
The number and percentage of subjects in each cohort
with a fracture during the follow-up period were
calculated and analyzed. Cox-proportional hazards
regression (HR) was used to examine the relationship
between therapy persistence and fracture risk by
modeling time to fracture over the 24-month follow-
up period while controlling for potential confounding
covariates. Covariates were chosen on the basis of
clinical relevance and their availability in the claims
database and included age, certain disease conditions
(Appendix II), prior fracture, treatment versus
prevention dose, days’ supply of other osteoporosis
medications in the follow-up period, and number of
pre-index osteoporosis-related office visits.

Page 4

588 Claims data: persistence with alendronate and fractures
© 2007 liBRAphARM lTD – Curr Med Res Opin 2007; 23(3)
Results
patient population characteristics
For the 4769 subjects who met the inclusion and
exclusion criteria and completed 24 months of follow-
up, as shown in Figure 2, information on age, type of
dose (treatment vs. prevention), dosing regimen, pre-
index fractures, and geographic distribution is presented
in Table 1. Patients in the persistent cohort were
significantly more likely to have received a treatment
(vs. prevention) dose (
to have been older than age 65 years (
non-persistent patients were more likely to be on the
daily regimen compared with the weekly regimen
(
p = 0.03) of alendronate and
p = 0.04). Also,
p < 0.0001).
persistence rates
Of the combined cohorts studied, 57.45% (2740/4769)
were not persistent with therapy during the 24-month
follow-up, whereas 42.55% (2029/4769) of patients
were persistent. Over 24 months of follow-up, the
mean length of therapy for all subjects was 261 days
and the median length was 121 days. Over 24 months
of follow-up, the mean length of therapy for all subjects
was 261 days and the median length was 121 days. The
mean length of therapy for patients in the persistent
cohort was 533 days (median = 613 days). For patients
who were not persistent at 6 months, the mean
length of persistence was 59 days (
persistent patients) and the median was 31 days.
p < 0.0001 vs.
Table 1. Baseline characteristics of patient population
Demographic
n
Percent persistent
Total population
Type of dose*
Treatment dose
Prevention dose
Dosing regimen
Daily
Weekly
Age category (years)
45–54
55–64
> 65
Pre-index (baseline) fracture diagnosis
Did have fracture in baseline period
Did not have a fracture in baseline period
Geographic region
Northeast
Midwest
South
West
Endocrine nutritional and metabolic
Diseases of the nervous system
Neoplasms
Postmenopausal osteoporosis
Number of pre-index osteoporosis office visits
0
1
2
3+
4769 42.5

43.4
39.7

37.9
49.0

40.6
43.3
46.2

43.5
42.5

45.4
44.9
39.7
43.9
42.7
41.4
45.7
46.2

38.9
45.3
49.2
44.2

3711
1058

2784
1985

1936
2294
539

124
4645

302
2116
2178
173
2161
1356
514
1090

2379
1367
652
371
*Treatment dose = alendronate 10 mg/day or 70 mg once weekly; prevention dose =
5 mg/day or 35 mg once weekly

Page 5

© 2007 liBRAphARM lTD – Curr Med Res Opin 2007; 23(3)
Claims data: persistence with alendronate and fractures Gold et al. 589
Kaplan–Meier curves are presented in Figure 3 to
demonstrate the persistence trends for the daily and
weekly cohort.
Fracture rates
There were a total of 214 subjects with a fracture during
the 24-month follow-up period, of whom 135 (4.93%)
were in the non-persistent cohort and 79 (3.89%)
were in the persistent cohort (chi-square p = 0.09). On
the basis of the univariate analysis, the differences in
the fracture rates between the two cohorts were not
statistically significant, yet there was a trend toward
fewer fractures in the persistent cohort compared with
the non-persistent cohort.
proportional hazards regression
The relationship between persistence and fracture
risk was tested in a multivariate analysis using a Cox-
proportional hazards model. The results are presented
in Table 2. A total of 20 subjects (7 persistent and
13 non-persistent) had a first fracture claim on their
index date, which resulted in a time-to-fracture of
0 days. These subjects were dropped from the Cox-
proportional hazards model. Controlling for other
factors, subjects who were persistent with alendronate
therapy were approximately 26% less likely to have
a fracture during the follow-up period compared
with non-persistent subjects (HR = 0.74; 95% CI,
0.549–0.996; p = 0.045). Subjects in the age groups
45–54 years and 55–64 years had a lower fracture risk
compared with subjects in the ≥ 65-year-old group.
Baseline co-morbidities associated with an increased
hazard of fracture included endocrine, nutritional,
metabolic, and nervous system diseases and neoplasms.
A claims diagnosis of fracture during the baseline period
(HR = 16.36; p < 0.001) had the largest association
Figure 2. Patient selection flowchart
Patients who filled at least one prescription for alendronate
between July 1996 and December 31, 2003
n = 49951
At least 6 months of baseline enrollment and 24 months of
follow-up enrollment
n = 30397
No claims diagnosis of drug-induced osteoporosis, cancer,
HIV, or Paget’s disease
n = 8319
No prescription for index medication with > 34 days’ supply
during the 24-month follow-up period
No crossover during the 24-month follow-up period
n = 4769
Women ≥ 45 years
n = 12409
No prescription for bisphosphonate during the 6-month
baseline period
No prescription for etidronate during the 24-month
follow-up period
n = 11919
Figure 3. Kaplan–Meier curves of time to discontinuation for the daily and weekly cohorts

Page 6

590 Claims data: persistence with alendronate and fractures
© 2007 liBRAphARM lTD – Curr Med Res Opin 2007; 23(3)
with a diagnosis of fracture per the claims data.
Baseline osteoporosis-related office visits (HR = 1.086;
p < 0.01) also were associated with an increased hazard
of fracture. The global test of proportionality showed
the model was not constant over time (
the hazard ratio (HR) coefficient for the persistence
cohort is interpreted as the average hazard of fracture
during the follow-up period rather than a proportional
hazard of fracture at each point in time during the
follow-up period.
p
< 0.01). Thus,
Discussion
This retrospective claims database analysis of the effect
of persistence with newly initiated bisphosphonate
therapy on long-term fracture risk demonstrated
that poor persistence with therapy resulted in a 26%
increased risk of fractures. These findings are consistent
with other studies which reported significant increases
in fracture risk when compliance with osteoporosis
therapy was poor
among 11
who were poorly compliant with therapy, including
bisphosphonates, were at a significantly greater risk
(16%) of fractures than compliant patients. In the
Canadian Database of Osteoporosis and Osteopenia
(CANDOO) study
greater 10-year fracture risk (
taking bisphosphonates inconsistently compared with
those taking therapy as directed. McCombs et al.
demonstrated 12-month reductions of 62% and 40%
32,38,41. Caro et al.
249 women with osteoporosis, those
41 reported that
38, there was a trend towards a 27%
p = 0.18) among patients
in hip and vertebral fractures, respectively, with
uninterrupted osteoporosis therapy
Results presented here revealed that older women,
those who had experienced a previous fracture, and
those with certain co-morbid conditions (endocrine,
nutritional, metabolic, and nervous system diseases,
and neoplasms) were more likely to have an increased
risk of fracture. The need for persistence with
postmenopausal therapies of all kinds among older
women has been noted
of poor persistence on clinical outcomes is likely to
be greater for older compared with younger patients,
for several reasons. First, the incidence of osteoporosis
increases with age and is estimated at 22% among
women aged 60–69 years; 30% among women aged
70–79 years, and 70% among women aged 80 years
and older
rise in fracture rates for every decade of life
that there is a greater likelihood of poor therapeutic
outcomes resulting from poor persistence among older
patients.
The relationship between previous and future
fractures confirms and is consistent with previous
findings
condition resulting in general fatigue, muscle weakness,
gait abnormalities, poor balance, or impaired sensory
processing could influence the likelihood of falls. The
relationship between neoplasms and fracture risk
is likely to be complex and hence requires further
investigation, particularly in light of concomitant
medications (steroids, antimitotic agents) that may
affect bone health in patients with neoplastic disease.
32.
46,47. Furthermore, the impact
48. This increase, combined with the 30%
49 suggests
7–9. With respect to co-morbid conditions, any
Table 2. Cox-proportional hazards model of time to fracture
Characteristic Hazard ratio
p-value
0.045*
0.001**
0.005**
0.005**
0.006**
0.018*
0.267
< 0.001**
0.831
0.246
0.309
0.932
0.960
< 0.001**
Persistent at 6 months
Ages 45–54†
Ages 55–64†
Endocrine, nutritional, and metabolic disease
Diseases of the nervous system
Neoplasms
Postmenopausal osteoporosis
Pre-index fracture diagnosis
Treatment dose‡
Follow-up hormone therapy days’ supply
Follow-up calcitonin days’ supply
Follow-up raloxifene days’ supply
Follow-up oral steroids days’ supply
Number of pre-index osteoporosis office visits
0.739
0.481
0.583
1.507
1.515
1.651
1.205
16.359
1.041
0.990
1.025
1.002
1.001
1.075
*p-value < 0.05; **p-value < 0.01
Reference groups: †Age > 65 years; ‡prevention dose

Page 7

© 2007 liBRAphARM lTD – Curr Med Res Opin 2007; 23(3)
Claims data: persistence with alendronate and fractures Gold et al. 591
While there is an increasing number of effective
therapeutic agents for treating osteoporosis, the success
of these regimens is reliant on patient persistence and
compliance to therapy. Given the growing evidence
that persistence with therapy is associated with
better health outcomes, several investigators have
examined strategies for improving patient persistence
to osteoporosis medication. These include clinical
interventions (educational materials, referral for bone
densitometry, and consultation)
medication costs
and new dosing options
A systematic review of the literature revealed that
more convenient dosing regimens are associated with
greater patient preference and increased medication
adherence. In a multicenter, randomized, open-label,
crossover study in which subjects with osteoporosis
were treated with bisphosphonates, 86% of patients
preferred once-weekly doses compared with 9.2%
who preferred once-daily doses (
proportion of patients believed that once-weekly dosing
would allow them to be more compliant with therapy,
particularly because the regimen was associated with
decreased gastrointestinal irritation and was more
convenient
randomized, open-label, crossover study, patient
preference was assessed for once-monthly ibandronate
versus once-weekly bisphosphonates, and it was shown
that 66%, as compared with 26.5%, preferred the
once-monthly regimen to the once-weekly regimen
With regard to medication adherence, several
database studies have demonstrated that less frequent
bisphosphonate dosing (weekly vs. daily) results in
improved adherence, although rates are still low and
as many as 50% of patients remain nonadherent
Newer, less-frequent (e.g., once-monthly) oral bisphos-
phonate dosing regimens are now available and may be
expected to positively impact the number of patients
realizing clinical benefit from treatment.
Because this study is retrospective and observational
in nature there are certain limitations the reader
must consider when interpreting the findings. Claims
data were used to estimate persistence and ICD-9-
CM markers, to identify fracture events; thus, it is
not possible to establish causality between these two
variables, since patients are not randomly assigned
to the study cohorts. Another complicating factor is
that it is possible that patients with a fracture claim
during the 24 months follow-up had it before 182 days
(6 months) and hence became more persistent after
the event so as to be classified as a persistent patient.
Underlying differences may be present in the study
populations and these may affect persistence and
clinical outcomes. Without the ability to control for
such unidentified variables or covariates (such as BMD
50, financial subsidies for
51, bone marker and staff monitoring
53,54.
52,
p < 0.001). A similar
53. Additionally, in another multicenter,
55.
33,37,56.
levels, height, weight, alcohol intake, smoking status,
vitamin D intake, and patient’s beliefs and attitudes)
some selection bias may exist. The inability to adjust
for the numerous potential co-morbid conditions that
are clinically relevant and hence could be potential
confounders is a further limitation of the study. Instead,
the broad categories of endocrine, nutritional and
metabolic diseases were used as covariates. Similarly,
other risk factors that can impact the likelihood of
fracture such as race and the use of blood thinners and
diuretics were not included in the analysis.
Even though claims data have demonstrated good
correlation with drug exposure and are commonly used
for estimating persistence
primarily for administrative purposes; hence billing
and coding errors cannot be ruled out. Prescription
fill data are an indirect measure of medication-taking
behavior, and the presence of a prescription claim
does not necessarily imply that the medication was
taken
who received only one month’s supply of medication at
a time removed 1879 patients from the analysis. While
this prevented an overestimation of patient persistence,
it reduced the ability to generalize the results to all
patients receiving study medication.
57, such data are collected
58,59. Restricting the patient population to those
Conclusion
This retrospective analysis from a large US health
plan database showed that lack of persistence with
alendronate was one of several factors associated with
an increased risk of fracture over the 2-year duration of
the study. The Cox-proportional hazards model demon-
strated that persistent patients were approximately 26%
less likely to have a fracture during the follow-up period
compared with non-persistent subjects. In light of the
continued reports that long-term adherence to osteo-
porosis medications is poor, it is imperative that subjects
receiving bisphosphonates be encouraged to remain
persistent and compliant with their prescribed therapy.
Acknowledgment
Declaration of interest: This study was supported and
funded by F. Hoffmann-LaRoche, Ltd. and GlaxoSmith-
Kline. The authors acknowledge Ellen Lewis, PhD and
Linda Gorman, PhD for assistance with research and
preparation of this manuscript. DTG is a consultant for
Procter & Gamble, sanofi-aventis, GlaxoSmithKline,
Roche, and Eli Lilly and Company. JRF is an employee of
i3 Innovus. MMA is an employee of GlaxoSmithKline.
FC is a consultant for Novartis, Eli Lilly, Merck, Roche,
Procter & Gamble, and Pfizer.

Page 8

592 Claims data: persistence with alendronate and fractures
© 2007 liBRAphARM lTD – Curr Med Res Opin 2007; 23(3)
References
1. Melton 3rd LJ. Epidemiology worldwide. Endocrinol Metab
Clin North Am 2003;32:1-13
2. Tuck SP, Francis RM. Osteoporosis. Postgrad Med J 2002;78:
526-32
3. US Congress OoTA. Hip fracture outcomes in people aged fifty
and over [online]. Available from http://www.wws.princeton.
edu/cgi-bin/byteserv.prl/~ota/disk1/1994/9413/9413.PDF
[May 10, 2005]
4. Cummings SR, Black D. Should perimenopausal women
be screened for osteoporosis? Ann Intern Med 1986;104:
817-23
5. Center JR, Nguyen TV, Schneider D, et al. Mortality after all
major types of osteoporotic fracture in men and women: an
observational study. Lancet 1999;353:878-82
6. Kado DM, Browner WS, Palermo L, et al. Vertebral fractures
and mortality in older women: a prospective study [Study of
Osteoporotic Fractures Research Group]. Arch Intern Med
1999;159:1215-20
7. Klotzbuecher CM, Ross PD, Landsman PB, et al. Patients with
prior fractures have an increased risk of future fractures: a
summary of the literature and statistical synthesis. J Bone Miner
Res 2000;15:721-39
8. Lindsay R, Silverman SL, Cooper C, et al. Risk of new vertebral
fracture in the year following a fracture. J Am Med Assoc
2001;285:320-3
9. Chevalley T, Hoffmeyer P, Bonjour JP, et al. An osteoporosis
clinical pathway for the medical management of patients with
low-trauma fracture. Osteoporosis Int 2002;13:450-5
10. Johnell O. The socioeconomic burden of fractures: today and in
the 21st century. Am J Med 1997;103:S20-S25
11. Body JJ. Management of primary osteoporosis. Acta Clin Belg
2002;57:277-83
12. Boonen S, Vanderschueren D. Fracture protection in osteoporosis
with risedronate. Hosp Med 2004;65:535-40
13. Chesnut IC, Skag A, Christiansen C, et al. Effects of oral
ibandronate administered daily or intermittently on fracture
risk in postmenopausal osteoporosis. J Bone Miner Res
2004;19:1241-9
14. Rozkydal Z, Janicek P. The effect of alendronate in the
treatment of postmenopausal osteoporosis. Bratisl Lek Listy
2003;104:309-13
15. Sartori L, Adami S, Filipponi P, et al. Injectable bisphosphonates
in the treatment of postmenopausal osteoporosis. Aging Clin
Exp Res 2003;15:271-83
16. Black DM, Greenspan SL, Ensrud KE, et al. The effects of
parathyroid hormone and alendronate alone or in combination
in postmenopausal osteoporosis. New Engl J Med 2003;349:
1207-15
17. Morabito N, Lasco A, Gaudio A, et al. Bisphosphonates in the
treatment of thalassemia-induced osteoporosis. Osteoporosis Int
2002;13:644-9
18. Stephen AB, Wallace WA. The management of osteoporosis.
J Bone Jt Surg Br 2001;83:316-23
19. Cole RP, Palushock S, Haboubi A. Osteoporosis management:
physicians’ recommendations and womens’ compliance
following osteoporosis testing. Women Health 1999;29:101-15
20. Emkey R. Alendronate and risedronate for the treatment of
postmenopausal osteoporosis: clinical profiles of the once-
weekly and once-daily dosing formulations. Med Gen Med
2004;6:6
21. Liberman UA, Weiss SR, Broll J, et al. Effect of oral alendronate
on bone mineral density and the incidence of fractures in
postmenopausal osteoporosis [The Alendronate Phase III
Osteoporosis Treatment Study Group]. New Engl J Med
1995;333:1437-43
22. Ensrud KE, Black DM, Palermo L, et al. Treatment with
alendronate prevents fractures in women at highest risk:
results from the Fracture Intervention Trial. Arch Intern Med
1997;157:2617-24
23. Karpf DB, Shapiro DR, Seeman E, et al. Prevention of
nonvertebral fractures by alendronate. A meta-analysis
[Alendronate Osteoporosis Treatment Study Groups]. J Am
Med Assoc 1997;277:1159-64
24. Cummings SR, Black DM, Thompson DE, et al. Effect of
alendronate on risk of fracture in women with low bone density
but without vertebral fractures: results from the Fracture Inter-
vention Trial. J Am Med Assoc 1998;280:2077-82
25. Harris ST, Watts NB, Genant HK, et al. Effects of risedronate
treatment on vertebral and nonvertebral fractures in women
with postmenopausal osteoporosis: a randomized controlled
trial [Vertebral Efficacy with Risedronate Therapy (VERT)
Study Group]. J Am Med Assoc 1999;282:1344-52
26. Reginster J, Minne HW, Sorensen OH, et al. Randomized trial of
the effects of risedronate on vertebral fractures in women with
established postmenopausal osteoporosis [Vertebral Efficacy
with Risedronate Therapy (VERT) Study Group]. Osteoporosis
Int 2000;11:83-91
27. McClung MR, Geusens P, Miller PD, et al. Effect of risedronate
on the risk of hip fracture in elderly women [Hip Intervention
Program Study Group]. New Engl J Med 2001;344:333-40
28. Eichner SF, Lloyd KB, Timpe EM. Comparing therapies for
postmenopausal osteoporosis prevention and treatment. Ann
Pharmacother 2003;37:711-24
29. Tosteson AN, Grove MR, Hammond CS, et al. Early discontinua-
tion of treatment for osteoporosis. Am J Med 2003;115:209-16
30. Segal E, Tamir A, Ish-Shalom S. Compliance of osteoporotic
patients with different treatment regimens. Isr Med Assoc J
2003;5:859-62
31. Hamilton B, McCoy K, Taggart H. Tolerability and compliance
with risedronate in clinical practice. Osteoporosis Int
2003;14:259-62
32. McCombs JS, Thiebaud P, McLaughlin-Miley C, et al. Com-
pliance with drug therapies for the treatment and prevention of
osteoporosis. Maturitas 2004;48:271-87
33. Cramer JA, Amonkar MM, Hebborn A, et al. Compliance
and persistence with bisphosphonate dosing regimens among
women with postmenopausal osteoporosis. Curr Med Res Opin
2005;21:1453-60
34. Sunyecz J, Gallagher R, MacCosbe P. Persistence with medication
in women taking daily versus weekly bisphosphonates for osteo-
porosis. The Female Patient 2006;31:21-8
35. Lo JC, Pressman AR, Omar MA, et al. Persistence with
weekly alendronate therapy among postmenopausal women.
Osteoporosis Int 2006;17:922-8
36. Siris E, Harris S, Rosen C, et al. Adherence to bisphosphonate
therapy and fracture rates in osteoporotic women: relationship
to vertebral and nonvertebral fractures from 2 US claims
databases. Mayo Clin Proc 2006;81:1013-22
37. Recker RR, Gallagher R, MacCosbe PE. Effect of dosing
frequency on bisphosphonate medication adherence in a large
longitudinal cohort of women. Mayo Clin Proc 2005;80:
856-61
38. Sebaldt RJ, Shane LG, Pham B, et al. Longer term effectiveness
outcomes of noncompliance and nonpersistence with daily
regimen bisphosphonate therapy in patients with osteoporosis
treated in tertiary specialist care. Osteoporosis Int 2004;15:
S107
39. Yood RA, Emani S, Reed JI, et al. Compliance with pharmac-
ologic therapy for osteoporosis. Osteoporos Int 2003;14:965-8
40. Finigan J, Bainbridge P, Eastell R. Adherence to osteoporosis
therapies. Osteoporosis Int 2001;12:[abstract P.110]
41. Caro JJ, Ishak KJ, Huybrechts KF, et al. The impact of
compliance with osteoporosis therapy on fracture rates in actual
practice. Osteoporosis Int 2004;15:1003-8
42. Siris ES, Harris ST, Rosen CJ, et al. Adherence to bisphosphonate
therapy and fracture rates in osteoporotic women: relationship
to vertebral and nonvertebral fractures from 2 US claims
databases. Mayo Clin Proc 2006;81:1013-22
43. Watts NB, Worley K, Solis A, et al. Comparison of risedronate
to alendronate and calcitonin for early reduction of nonvertebral
fracture risk: results from a managed care administrative claims
database. J Managed Care Pharm 2004;10:142-51
44. Watts NB. Risedronate reduces the risk of clinical vertebral
fractures in just 6 months. J Bone Miner Res 2001;16:S407
45. Harrington JT, Ste-Marie LG, Brandi ML, et al. Risedronate
rapidly reduces the risk for nonvertebral fractures in women
with postmenopausal osteoporosis. Calcif Tissue Int 2004;74:
129-35

Page 9

© 2007 liBRAphARM lTD – Curr Med Res Opin 2007; 23(3)
Claims data: persistence with alendronate and fractures Gold et al. 593
46. Schiff I, Rebar R, Cramer J, et al. Achieving long-term continu-
ance of menopausal ERT/HRT: Consensus Opinion of the
North American Menopause Society. Menopause 1998;5:
69-76
47. Cramer JA. Enhancing patient compliance in the elderly.
Role of packaging aids and monitoring. Drugs Aging 1998;12:
7-15
48. Prestwood KM. Osteoporosis: up-to-date strategies for
prevention and treatment [Interview by Marc E. Weksler].
Geriatrics 1997;52:92-4, 7-8
49. Gallagher JC. Pathophysiology of osteoporosis. Semin Nephrol
1992;12:109-15
50. Cuddihy MT, Amadio PC, Gabriel SE, et al. A prospective
clinical practice intervention to improve osteoporosis
management following distal forearm fracture. Osteoporosis Int
2004;15:695-700
51. Liel Y, Castel H, Bonneh DY. Impact of subsidizing effective anti-
osteoporosis drugs on compliance with management guidelines
in patients following low-impact fractures. Osteoporosis Int
2003;14:490-5
52. Clowes JA, Peel NF, Eastell R. The impact of monitoring on
adherence and persistence with antiresorptive treatment for
postmenopausal osteoporosis: a randomized controlled trial.
J Clin Endocrinol Metab 2004;89:1117-23
53. Claxton AJ, Cramer J, Pierce C. A systematic review of the
associations between dose regimens and medication compliance.
Clin Ther 2001;23:1296-310
54. Simon JA, Lewiecki EM, Smith ME, et al. Patient preference for
once-weekly alendronate 70
10
Clin Ther 2002;24:1871-86
55. Emkey R, Koltun W, Beusterien K, et al. Patient preference
for once-monthly ibandronate versus once-weekly alendronate
in a randomized, open-label, cross-over trial: the Boniva
Alendronate Trial in Osteoporosis (BALTO). Curr Med Res
Opin 2005;21:1895-903
56. Ettinger MP, Gallagher R, MacCosbe PE. Medication persistence
with weekly versus daily doses of orally administered bisphos-
phonates. Endocr Pract 2006;12:522-8
57. Steiner JF, Prochazka AV. The assessment of refill compliance
using pharmacy records: methods, validity, and applications.
J Clin Epidemiol 1997;50:105-16
58. Steiner JF, Koepsell TD, Fihn SD, et al. A general method of
compliance assessment using centralized pharmacy records.
Description and validation. Med Care 1988;26:814-23
59. Farmer KC. Methods for measuring and monitoring medication
regimen adherence in clinical trials and clinical practice. Clin
Ther 1999;21:1074-90
mg versus once-daily alendronate
mg: a multicenter, randomized, open-label, crossover study.
Appendix I. ICD-9 codes used to identify fractures
Fracture type
Clavicle
Closed
Closed/open not indicated
Femur, other/unspecified (closed)
Pathologic
Shaft/unspecified
Lower end
Closed/open not indicated
Forearm/wrist (closed)
Pathologic
Upper end
ICD-9 code

810.0x
810

733.15
821.0x
821.2x
821

733.12
813.0x












Fracture type
Humerus (closed)
Pathologic
Upper end
Shaft/unspecified
Lower end
Closed/open not indicated
Pelvis (closed)
Acetabulum
Pubis
Other specified
ICD-9 code

733.11
812.0x
812.2x
812.4x
812

808.0x
808.2x
808.4x
Unspecified 808.8x
Shaft 813.2x
Closed/open not indicated 808
Lower end 813.4x
Tibia/fibula (closed)
Unspecified 813.8x
Pathologic 733.16
Closed/open not indicated 813
Upper end 823.0x
Hip (closed)
Shaft 823.2x
Pathologic 733.14
Unspecified 823.8x
Transcervical 820.0x
Closed/open not indicated 823
Pertrochanteric 820.2x
Vertebral, osteoporosis related 733.13


Unspecified
Closed/open not indicated
820.8x
820

Page 10

594 Claims data: persistence with alendronate and fractures
© 2007 liBRAphARM lTD – Curr Med Res Opin 2007; 23(3)
Appendix II. list of ICD-9 codes for co-morbidities
Co-morbid category
Infectious and parasitic diseases
Neoplasms
Endocrine, nutritional, and metabolic disease
Diseases of the blood and blood forming organs
Mental disorders
Diseases of the nervous system
Diseases of the circulatory system
Diseases of the respiratory system
Diseases of the digestive system
Diseases of the genitourinary system
Complications of pregnancy, childbirth and the puerperi
Diseases of the skin and subcutaneous tissue
Diseases of the musculoskeletal system
Congenital anomalies
Certain conditions originating in the perinatal period
Symptoms, signs, and ill-defined conditions
Injury and poisoning
ICD-9 diagnosis code
001.XX–139.XX
140.XX–239.XX
240.XX–279.XX
280.XX–289.XX
290.XX–319.XX
320.XX–389.XX
390.XX–459.XX
460.XX–519.XX
520.XX–578.XX
580.XX–629.XX
630.XX–677.XX
680.XX–709.XX
710.XX–739.XX
740.XX–759.XX
760.XX–779.XX
780.XX–799.XX
800.XX–999.XX
CrossRef links are available in the online published version of this paper:
http://www.cmrojournal.com
Paper CMRO-3604_3, Accepted for publication: 09 January 2007
Published Online: 09 February 2007
doi:10.1185/030079906X167615