SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. REFERENCES

Objective

There are many effective osteoporosis (OP) medications with a variety of dosing intervals and delivery options, but even when diagnosed, OP is often undertreated. We sought to determine the bone density consequences of the decision to initiate and comply with therapy for OP.

Methods

We identified 243 women who received a dual x-ray absorptiometry (DXA) evaluation and fulfilled the World Health Organization criteria for OP. One year later, the patients were asked to return for a followup DXA. Administrative electronic health records were used to identify prescription drug use.

Results

A total of 142 women (58%) initiated pharmacologic therapy for OP during the year after the initial DXA; 144 returned for a followup DXA after 1 year. For those women with ≥66% of days receiving therapy, the mean annual change in spine bone mineral density (BMD) was 4.5% compared with 2.0% for those with <66% of days receiving therapy and 0.8% for those not receiving OP therapy (P < 0.001). For those women with ≥66% of days receiving therapy, the mean change in hip BMD was 2.3% compared with 0.3% for those with <66% of days receiving therapy and −0.8% for those not receiving OP therapy (P < 0.001).

Conclusion

We found significant bone density consequences of the decision to initiate and comply with therapy in the first year after diagnosis of OP. Improvement in both initiation rates of treatment as well as compliance are needed in order to reduce the frequency of osteoporotic fractures.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. REFERENCES

Compliance with therapy for osteoporosis (OP) is suboptimal; approximately half of those who initiate pharmaceutical treatment discontinue therapy within 1 year (1–5). Studies using large administrative databases have shown that patients with poor compliance have a higher fracture rate and higher costs than those with better compliance (4–12). In these studies, the triggering event for analysis was the first prescription for an OP medication, thereby excluding patients who do not initiate treatment after diagnosis. There is little information available on the impact of the decision to initiate therapy on bone density after the diagnosis of OP.

We sought to evaluate the bone density consequences of the decision to initiate therapy and the number of days receiving therapy in the first year after OP diagnosis among women receiving usual care.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. REFERENCES

The study population for this report has been previously described (13). From February 2006 through March 2007, we identified 465 consecutive female members of a managed care plan with pharmaceutical coverage who received a dual x-ray absorptiometry (DXA) evaluation at the multispecialty medical group practice and fulfilled the World Health Organization criteria for OP by a T score less than −2.5. Patients were excluded if they received OP prescription medications in the prior 6 months. Approximately 12 months after the date of the initial DXA, the patients were asked to respond to a mailed questionnaire and return for a followup DXA to determine whether we could document any impact of initiation and compliance within 1 year of diagnosis. Administrative electronic health records were used to identify prescription drug use and health care utilization in the 2 years prior to the baseline bone density study through the date of the followup bone density study. Diagnoses were captured based on International Classification of Diseases, Ninth Revision codes used by physicians for administrative purposes. All of the patients included in this study had prescription drug coverage through the managed care plan. The group's administrative data include all claims submitted for reimbursement for pharmaceuticals. The Charlson comorbidity index was determined based on all of the diagnoses recorded in the 2 years prior to the bone density study (14, 15). Additional demographic characteristics and lifestyle factors were assessed using self-reported questionnaires that were mailed approximately 1 month after the baseline DXA.

We used the terms compliance and adherence as synonyms, as defined by the International Society for Pharmacoeconomics and Outcomes Research (16). Compliance (synonym: adherence) is defined as “the extent to which a patient acts in accordance with the prescribed interval and dose of a dosing regimen” (16). We assessed the percentage of days receiving OP medication therapy between the baseline and followup DXA by summing the day's supply of OP medication (a bisphosphonate, estrogen, calcitonin, teriparatide, or raloxifene) dispensed according to pharmacy records divided by the number of days between the baseline and followup DXA. This is similar to the medication possession ratio (MPR) calculation, except that for the MPR calculation the time begins with the first drug prescription. The value of the day's supply was truncated if the supply extended beyond the time period of observation. Consistent with previous studies (1, 17) for women with any days receiving therapy, the percentage of days receiving OP therapy was dichotomized into categories of ≥66% of days receiving therapy and <66% of days receiving therapy among women receiving OP therapy during the period between the baseline and followup DXA. A separate category of women receiving no OP therapy during the observation period was also created (women with 0 days receiving therapy). Change in bone mineral density (BMD) was calculated as the change in BMD (gm/cm2) from the first to the second study at the lumbar spine and total hip.

Differences in characteristics of women according to the percentage of days receiving OP therapy (≥66% of days receiving therapy, <66% of days receiving therapy, and no OP therapy) were compared using a one-way analysis of variance (ANOVA) for continuous variables and chi-square tests or Fisher's exact test for categorical variables. Bivariate analyses were then performed using ANOVA to evaluate if the change in BMD was associated with the percentage of days receiving OP therapy, and t-tests were performed to evaluate other patient characteristics, including categories of patient age, education, smoking status, weight, exercise, alcohol use, comorbidity, and daily calcium intake. The ANOVA was used to evaluate the effects of variables (and interactions) for those factors that bivariate results suggested were related to the change in BMD (i.e., P value for the associated statistical test of 0.20 or less), with final models including variables significant at the 0.05 level. Linear regression models were also constructed to evaluate the associations between patient characteristics and change in BMD, analyzing days receiving therapy and patient age as continuous variables; the results were similar to those reported using the ANOVA and are not reported.

This study was approved by the Saint Vincent/Fallon Clinic/Fallon Community Health Plan Institutional Review Board. All of the analyses were performed using the SAS statistical package, version 9.1.3 (SAS Institute).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. REFERENCES

A total of 243 women of 465 contacted for participation returned research authorization forms either at the time of the original DXA or at 1-year followup, and thereby enrolled in the study. One hundred forty-two (58%) of the 243 received OP drug therapy in the 1 year following the initial DXA. A total of 144 women (59% of those enrolled) returned for a followup DXA; the majority of women were age ≥65 years (63%; n = 91) and were of white race (95%; n = 137). Most women reported that they had been diagnosed with OP (89%) and that a doctor or nurse had recommended prescription medication for the treatment of OP (88%). The women who returned for a followup DXA did not differ clinically in age, education, smoking or alcohol history, weight, exercise, comorbidities, or baseline bone density from those who did not (data not shown).

Of the 144 women who answered the questionnaire and returned for a followup DXA, 80 (56%) had received OP drug therapy in the 1 year following the initial DXA. Seventy-two had received oral bisphosphonate therapy, 3 received intravenous bisphosphonate therapy, 5 received raloxifene, 3 received calcitonin, and 1 received teriparatide. Three patients had received more than one class of drug and 1 patient received both oral and intravenous bisphosphonate therapy. The 3 patients receiving intravenous therapy were omitted from subsequent analyses.

For those 77 women who had started drug therapy, the mean ± SD percentage of days receiving therapy between the initial and followup DXA was 57% ± 34%. Table 1 describes the characteristics of the women according to the percentage of days receiving therapy between the baseline and followup DXA. The groups were similar with regard to most demographic and lifestyle factors, except the proportion who reported exercising at least 3 days per week, which was lowest among those women with <66% of days receiving therapy (P = 0.02) (Table 1).

Table 1. Comparison of patients with 2 BMDs according to percentage of days receiving therapy*
Characteristic≥66% of days receiving therapy (n = 35)<66% of days receiving therapy (n = 42)No OP therapy (n = 64)
  • *

    Values are the number (percentage) unless otherwise indicated. In some cases, totals do not add up to 141 patients due to patient nonresponse to specific survey items or use of multiple medications during the observation period. BMD = bone mineral density; OP = osteoporosis; DXA = dual x-ray absorptiometry.

  • P < 0.05.

Age, years   
 35–6415 (42.9)19 (45.2)19 (29.7)
 ≥6520 (57.1)23 (54.8)45 (70.3)
Education   
 High school or less18 (52.9)22 (52.4)34 (53.1)
 Some college or more16 (47.1)20 (47.6)30 (46.9)
Smoking status   
 Current smoker3 (8.6)4 (9.5)8 (12.5)
 Nonsmoker32 (91.4)38 (90.5)56 (87.5)
Weight, pounds   
 <12711 (32.4)14 (34.2)14 (21.9)
 ≥12723 (67.7)27 (65.9)50 (78.1)
Exercise, days per week   
 <313 (37.1)25 (61.0)22 (34.4)
 ≥322 (62.9)16 (39.0)42 (65.6)
Alcohol   
 0 days with ≥2 drinks28 (80.0)30 (71.4)40 (62.5)
 At least 1 day with ≥2 drinks7 (20.0)12 (28.6)24 (37.5)
Charlson comorbidity index   
 017 (48.6)16 (38.1)34 (53.1)
 16 (17.1)17 (40.5)19 (29.7)
 ≥212 (34.3)9 (21.4)11 (17.2)
Daily calcium   
 Yes25 (73.5)31 (75.6)46 (76.7)
 No9 (26.5)10 (24.4)14 (23.3)
Fracture after baseline DXA   
 Yes4 (11.3)1 (2.4)6 (9.7)
 No31 (88.6)41 (97.6)56 (90.3)
Baseline BMD, mean ± SD gm/cm2   
 Hip0.715 ± 0.0990.732 ± 0.1070.710 ± 0.096
 Lumbar spine0.798 ± 0.1330.762 ± 0.0830.762 ± 0.087

For those women with ≥66% of days receiving therapy, the mean change in spine BMD was 4.5% compared with 2.0% for those with <66% of days receiving therapy and 0.8% for those not receiving OP therapy (P < 0.001) (Table 2). For those women with ≥66% of days receiving therapy, the mean change in hip BMD was 2.3% compared with 0.3% for those with <66% of days receiving therapy and −0.8% for those not receiving OP therapy (P < 0.001). The only other patient characteristic that was associated with a difference in the mean change in BMD was exercise, where women reporting that they exercised at least 3 times per week had a mean change in hip BMD of 0.9% compared with a mean change of −0.5% among those reporting that they exercised less than 3 times per week (P = 0.02). However, change in lumbar spine BMD was similar for the 2 groups. Age was associated at the 0.20 significance level with a difference in the mean change in hip BMD (Table 2); other factors associated at the 0.20 significance level with a difference in the mean change in spine BMD included education, and comorbidity (as measured by the Charlson comorbidity index).

Table 2. Change in bone density
CharacteristicLumbar spine, mean ± SD %Hip, mean ± SD %
  • *

    P < 0.05 for the association between change in bone mineral density and the characteristic.

Total2.04 ± 4.340.32 ± 3.37
Osteoporosis medication  
 ≥66% of days receiving therapy4.46 ± 3.04*2.30 ± 3.72*
 <66% of days receiving therapy2.04 ± 4.89*0.27 ± 2.78*
 None0.76 ± 4.03*−0.76 ± 3.05*
Age, years  
 35–642.52 ± 3.390.91 ± 2.52
 ≥651.76 ± 4.81−0.03 ± 3.76
Education  
 High school or less1.47 ± 4.260.12 ± 2.84
 Some college or more2.54 ± 4.250.53 ± 3.90
Smoking status  
 Current smoker0.80 ± 3.611.31 ± 2.90
 Nonsmoker2.19 ± 4.400.20 ± 3.41
Weight, pounds  
 <1271.81 ± 4.830.29 ± 4.11
 ≥1272.21 ± 4.160.32 ± 3.08
Exercise, days per week  
 <31.99 ± 4.16−0.51 ± 3.22*
 ≥31.99 ± 4.420.88 ± 3.33*
Alcohol  
 0 days with ≥2 drinks2.33 ± 4.610.25 ± 3.57
 At least 1 day with ≥2 drinks1.37 ± 3.560.49 ± 2.88
Charlson comorbidity index  
 02.52 ± 3.730.56 ± 2.40
 10.91 ± 4.49−0.19 ± 3.16
 ≥22.55 ± 5.100.51 ± 5.12
Daily calcium  
 Yes2.08 ± 4.310.53 ± 3.48
 No1.97 ± 4.62−0.07 ± 3.05

Similar associations were found in multivariate analyses; the only factors significantly associated with the mean change in hip BMD were days receiving OP therapy (P < 0.001) and exercise (P = 0.01). The only factor significantly associated with the mean change in spine BMD was days receiving OP therapy (P < 0.001).

The spine bone density decreased at 1 year more than the least significant change for the bone densitometer in 20 of 141 women. Of these, none were compliant ≥66% of days receiving treatment, 6 were compliant <66% of days receiving treatment, and 14 never initiated treatment.

In response to the questionnaire, patients not currently taking prescription OP medication after 1 year indicated the reasons shown in Table 3.

Table 3. Reasons for not taking medication according to percentage of days receiving therapy*
 ≥66% of days receiving therapy (n = 35)<66% of days receiving therapy (n = 42)No OP therapy (n = 64)
  • *

    Values are the number (percentage). In some cases, totals do not add to 141 patients due to patient nonresponse to specific survey items or use of multiple medications during the observation period. OP = osteoporosis.

If you are not currently taking prescription OP medication, please tell us why.   
 My doctor never recommended that I start1 (2.86)2 (4.8)7 (11.1)
 I didn't think I needed to take it (or take it any longer)007 (11.1)
 I took it as long as I needed to1 (2.86)00
 I was taking too many medications004 (6.3)
 I was concerned about possible long-term side effects1 (2.86)5 (11.9)23 (36.5)
 I was concerned about the cost07 (16.7)10 (15.9)
 I don't like taking medications03 (7.1)18 (28.6)
 I didn't like the inconvenience of taking the medication005 (7.9)
 I want to use diet and/or exercise instead07 (16.7)23 (36.5)
 I wanted to use calcium and vitamin D012 (28.6)31 (49.2)
 I wanted to use other nonprescription treatments instead1 (2.86)3 (7.1)13 (20.6)
 I have a hard time swallowing pills02 (4.8)5 (7.9)
 I didn't think my OP was bad enough for me to need medication05 (11.9)14 (22.2)
 I didn't think it was working000
 My doctor told me to stop03 (7.1)2 (3.2)
 I had a hard time remembering to take it002 (3.2)
 I never wanted to take it to begin with02 (4.8)7 (11.1)

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. REFERENCES

We demonstrate that in a population of previously untreated women with OP diagnosed as part of routine clinical care, only 54% of those who returned for a followup bone density study initiated treatment within 1 year of diagnosis. All of the women in our study were enrolled in a managed care plan with prescription drug coverage and 88% reported that they had been advised to initiate OP therapy. As shown in Table 1, the clinical baseline characteristics of those who initiated therapy were generally similar to those who did not do so. We previously reported that the decision to start OP treatment appears to be related to a patient's beliefs in the effectiveness of OP medications and distrust of medications (13). We asked patients not currently taking prescription OP medication at the end of one year to tell us why (Table 3). Factors associated with the decision not to initiate therapy included a dislike of taking medications, fear of long-term side effects, belief that the OP was not bad enough to take prescription medication, and a desire to treat OP with nonprescription treatments, including calcium and vitamin D, diet, and exercise. In the patients who initiated therapy, these same factors appear to be associated with poor compliance.

We were able to document an impact of the decision to initiate and comply with therapy on bone density within 1 year. Patients with more days receiving therapy had improved bone density response compared with those who obtained fewer days of therapy or had no therapy (Table 2 and Figure 1). Our analysis of change in bone density based on days receiving therapy is different than comparing change in bone density based on an MPR. The MPR is the day's supply of medication dispensed during the year following initiation of treatment divided by the observation period, and is a reliable measure of compliance (18). We calculated days receiving therapy beginning with the diagnostic bone density study and ending with the second DXA approximately 1 year later. Accordingly, we were able to include in our analysis patients who never initiated therapy. We divided users of therapy into those with less than 66% of days receiving therapy and those with more based on previous studies (1, 13). Recently, Cotté et al have shown that the optimal MPR threshold for predicting fracture risk was ≥68% (11). In our study, all of the women who lost bone density over 1 year had compliance <66%; 14 of the 21 women who lost bone density never initiated therapy. This suggests that clinicians should be very attentive to the possibility of poor compliance in all of the women who lose bone density.

thumbnail image

Figure 1. Annualized change in bone density by days.

Download figure to PowerPoint

Previous studies have documented that at least half of the patients started on treatment for OP discontinue therapy before they are likely to have any benefit (5–11). Feldstein et al demonstrated that in a community cohort of postmenopausal women at risk, the fracture risk of patients who received bisphosphonates did not differ significantly from those who did not initiate bisphosphonate therapy (19). These studies identified patients based on initiation of therapy. We document that these studies miss nearly half of the patients with OP diagnosed by bone density study as part of routine care, and we show that there are bone density consequences to the decision not to initiate therapy as well as consequences for poor compliance. Although our population was too small to evaluate fracture rates, it is likely that these low rates of initiation and compliance would have an impact on fracture risk.

Our study has several limitations. It was based in one managed care plan, so the results may not be generalizable to other populations. However, the compliance rates we document are similar to other reports using large administrative databases. Only 59% of women enrolled in the study returned for a followup DXA, but we were still able to document bone density differences. Since those patients who returned for a followup DXA did not differ clinically from those who did not, and since the initiation rate in those with a second DXA was similar to that of the overall population (54% versus 57%), it is likely that our findings are reflective of the entire population studied. It is possible that some patients initiated treatment after our observation period, but we followed patients for a full year after the original DXA study. Although it is possible that some of the patients had not been advised to start treatment, 88% reported that they had been advised to initiate treatment, and chart review documented that nearly all of the patients who did not start treatment had been advised to do so (13).

In conclusion, we were able to document a low initiation rate of therapy as well as significant bone density consequences of the decision to initiate therapy in the first year after OP diagnosis among women receiving usual care. It appears that the unmet need for treatment is even larger than previously documented in the population of patients with newly diagnosed OP. Improvement in both initiation rates of treatment as well as compliance are needed to reduce the frequency of osteoporotic fractures.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. REFERENCES

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Yood had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Yood, Andrade, Mazor, Chan, Kahler.

Acquisition of data. Yood.

Analysis and interpretation of data. Yood, Andrade, Mazor, Fouayzi, Chan.

ROLE OF THE STUDY SPONSOR

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. REFERENCES

Novartis Pharmaceuticals Corporation had no role in the study design, data collection, data analysis, and writing of the manuscript, as well as the approval of the content of the submitted manuscript. Publication of this article was not contingent on the approval of Novartis Pharmaceuticals Corporation.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. REFERENCES