Introduction

Patient compliance and persistence with treatments for chronic conditions is often low [1, 2]. Among patients receiving treatment for osteoporosis, approximately half discontinue therapy within the first 6 months [3, 4]. Analyses of administrative data suggest that more than 50% compliance and persistence with osteoporosis therapy is required for antifracture efficacy, and it is likely that 75% to 80% persistence with therapy is necessary for patients to experience fracture risk reduction consistent with what is seen in clinical trials of osteoporosis medications [5]. Poor compliance and persistence may increase the risk of fracture [5–7], thus adversely affecting patient outcomes and increasing health resource utilization [8–11]. Understanding patient perceptions and preferences of therapies may help in developing strategies to improve compliance and persistence with treatment.

The bisphosphonate alendronate is the most commonly prescribed antiresorptive agent for the treatment of postmenopausal bone loss [12, 13] and is available as a daily or weekly tablet as well as an oral solution. Alendronate binds to the mineralized surface of the bone and inhibits the bone-resorbing activity of mature osteoclasts, thus decreasing bone turnover, increasing bone mineral density (BMD), and reducing the risk for fracture [14, 15]. Recent preference studies with alendronate have shown that patients prefer weekly dosing over daily dosing [16, 17].

Denosumab, a fully human monoclonal antibody, is an antiresorptive agent in late-stage clinical development that neutralizes RANKL, thereby inhibiting osteoclast-mediated bone resorption, by affecting osteoclast development, function, and survival [18–20]. Administered as a subcutaneous injection every 6 months, denosumab has been shown to decrease bone turnover and increase BMD in postmenopausal women with low bone mass or osteoporosis and reduce fracture risk in postmenopausal women with osteoporosis [21–25].

Published studies have described patient preference for once weekly, once monthly, and once annual therapies for osteoporosis [16, 17, 26–28]. However, there have been no data that characterize patient preference and satisfaction with 6-month subcutaneous injections. The Preference and Satisfaction Questionnaire (PSQ) was developed specifically to evaluate patient preference and satisfaction with two different modes and frequency of dosing for the treatment of postmenopausal bone loss: a subcutaneous injection administered every 6 months (Q6M) and an oral tablet taken once weekly (QW) [29]. We report the results of patient responses to the PSQ from two blinded phase 3 head-to-head randomized controlled trials that directly compared denosumab with alendronate [30, 31].

Methods

Patients

Both the Determining Efficacy: Comparison of Initiating Denosumab versus AlEndronate (DECIDE) trial and the Study of Transitioning from AleNdronate to Denosumab (STAND) trial were international, double-blind, double-dummy, randomized, phase 3 head-to-head trials comparing denosumab with alendronate. Details of the design and inclusion and exclusion criteria of these studies have been published [30, 31] and are summarized below. The primary endpoint for both studies was the percent change from baseline in BMD at the proximal femur (total hip) after 12 months of treatment. These studies were approved by the Institutional Review Board or Ethics Committee at each site and were conducted in accordance with the ethical standards of the Declaration of Helsinki.

The DECIDE study was conducted from April 2006 to December 2007 and compared the efficacy and safety of denosumab with weekly oral branded alendronate (Fosamax® 70 mg; Merck) in postmenopausal women with low bone mass (T-score at the total hip or lumbar spine ≤−2.0) with no or very limited previous bisphosphonate use. Patients were excluded if they had a disease or condition known to affect bone metabolism or used drugs known to affect bone metabolism including any prior intravenous bisphosphonate use, more than 3 months of oral bisphosphonate use within the past 2 years, or 1 month or more of oral bisphosphonate use within the past year. In contrast, the STAND study (conducted October 2006 to March 2008) evaluated the efficacy and safety of denosumab in postmenopausal women with low bone mass (T-score at the total hip or lumbar spine between ≤−2.0 and ≥−4.0) who were receiving bisphosphonate therapy for a minimum of 6 months prior to entering the study. Patients with a disease or condition known to affect bone metabolism were excluded, as well as those with prior intravenous bisphosphonate use. Patients enrolled in STAND had a 1-month run-in period of weekly oral alendronate before being either transitioned from weekly alendronate therapy (Fosamax® 70 mg; Merck) to denosumab (60 mg Q6M) or continued on alendronate therapy. Patients enrolled in both studies were asked to complete the PSQ after 12 months of treatment or upon study discontinuation. Study participants self-administered the tablet (alendronate or placebo), while the injection (denosumab or placebo) was administered by study staff at the scheduled study visits.

PSQ aims and design

The PSQ is a 34-item questionnaire that measures patient preference and satisfaction with a Q6M injection versus weekly tablet for the treatment of low bone mass. Development and validation of the PSQ is reported separately [29]. Briefly, the PSQ was developed from a pool of potential questions derived from a review of the published literature and input from topic experts. Questions were then refined based on feedback from in-depth cognitive interviews with two separate focus groups in the United States. The refined PSQ was then translated from US English into 17 languages and given to all trial participants. The PSQ was designed to compare whether the proportion of subjects who preferred or were more satisfied with the injection was the same as with the tablet within each treatment group. Since participants were blinded to treatment allocation, questions specific to perceptions about denosumab or alendronate treatment were not incorporated into this questionnaire. For the PSQ, preference was defined as the medication or treatment choice made by a patient, based on specific attributes of the medication or treatment. Satisfaction is the degree to which the attributes of a specific medication or treatment actually meets the expectations that the patient had for the medication or treatment. Finally, bother is the degree to which a patient is annoyed or disturbed by certain attributes of the medication. Overall preference was assessed from the question, “Which do you prefer? The weekly pill (tablet), the 6-month injection, or no preference.” Satisfaction with the frequency of treatment administration was assessed using the question, “With which frequency of administration have you been more satisfied? The weekly pill, the 6-month injection, or I am not satisfied with one frequency of administration over the other.” Six additional preference or satisfaction measures were also evaluated in the PSQ: overall satisfaction, satisfaction with mode of treatment administration, lifestyle fit, convenience, choice for long-term use, and choice for continuation. Subjects were also queried about whether they were bothered by the weekly tablet or the 6-month injection.

Statistical analysis

Patients who completed one or more items of the PSQ were included in the analysis. A complete case approach was used to analyze the PSQ data. Missing data were not imputed and were not included in the within-treatment and between-treatment group analyses. An asymptotic test was used to evaluate whether the proportion of patients who preferred or were more satisfied with the 6-month injection was equal to the proportion of patients who preferred or were more satisfied with the weekly tablet within each treatment group. The Mantel–Haenszel test was used to determine if there were significant differences in the proportion of patients between the two treatment groups (tablet versus injection) adjusting for study (DECIDE versus STAND) on preference and satisfaction items. The same statistical methodologies were also applied to subgroup analyses (such as age [<65 versus ≥65 years old] and prior fracture [yes versus no]) on preference and satisfaction. Analyses of other preference and satisfaction endpoints were exploratory and were summarized using descriptive statistics.

Results

Baseline demographics and disease characteristics

Overall, the demographics of patients enrolled in the two studies were similar. Patients who enrolled in the DECIDE study were younger (mean [SD] age 64.4 years [8.5 years]) than those who enrolled in the STAND study (67.6 years [7.8 years]) and had a shorter average time since menopause (Table 1). Most of the patients enrolled in both studies were Caucasian (84% for DECIDE versus 93% for STAND), and there was a greater percentage of patients who self-identified as Hispanic or Latino in the DECIDE study compared with the patients in the STAND study. Baseline T-scores at the lumbar spine and total hip were similar between studies.

Table 1 Baseline demographics and disease characteristics for the overall study populations
Full size table

The majority of patients completed 12 months of the study (94% DECIDE and 95% STAND). Reasons for discontinuation were similar, with withdrawal of consent being the most common. Over 93% of patients enrolled in the DECIDE and STAND studies completed at least one item of the PSQ (Fig. 1). Demographics or baseline characteristics were very similar between patients who took the PSQ and those who did not.

Fig. 1
figure 1

Patient disposition for the DECIDE and STAND trials

Full size image

Patient-reported outcomes

The overall study design, medications used, inclusion and exclusion criteria (other than the difference in prior oral bisphosphonate exposure), study length, dates the studies were conducted, and demographics of the study participants were very similar for both the DECIDE and STAND studies. Furthermore, the results of the PSQ given to patients in both studies were very similar. We, therefore, present the combined data analysis.

Preference and satisfaction with dosing frequency

Nearly all patients completed the question regarding preference (99%). Some patients in each treatment group did not indicate a preference for either treatment (16% of denosumab-treated patients; 17% of alendronate-treated patients). Among patients who reported a preference, significantly more patients preferred the 6-month injection (65% of denosumab group; 63% of alendronate group) to the weekly tablet (19% for both treatment groups; P < 0.0001; Fig. 2a). The proportion of patients choosing the tablet, the injection, or no preference was similar among patients randomized to receive the active injection (denosumab group) or the placebo injection (alendronate group; P = 0.4500).

Fig. 2
figure 2

Patient-reported preference (a) and satisfaction with frequency of treatment administration (b). The values shown are combined data from the DECIDE and STAND trials.*P < 0.0001 for 6-month injection versus weekly tablet

Full size image

The majority of patients also responded to the question regarding satisfaction with the frequency that the treatments were administered (99%). Twenty percent of patients in both treatment groups were not more satisfied with one dosing frequency over the other. Significantly more patients in both the denosumab (64% versus 16%) and alendronate (63% versus 16%) groups were more satisfied with the dosing frequency of the 6-month injection over the weekly tablet (P < 0.0001; Fig. 2b). For both the denosumab and alendronate treatment groups, similar proportions of patients selected the weekly tablet, 6-month injection, or not satisfied with one over the other (P = 0.9907).

Additional measures of preference and satisfaction

Six additional measures of preference and satisfaction were also evaluated by the PSQ and are summarized in Fig. 3. Most patients were more satisfied overall and found the 6-month injection to be more convenient and fit better with their lifestyle than the weekly tablet. In addition, more patients indicated that they would choose the 6-month injection for long-term use or continuation of treatment. The findings were similar among patients that received either the active (denosumab group) or placebo (alendronate group) injection.

Fig. 3
figure 3

Patient-reported outcomes for additional preference and satisfaction measures evaluated in the PSQ. The values shown are combined data from the DECIDE and STAND trials.*P < 0.0001 for 6-month injection versus weekly tablet

Full size image

Preference and satisfaction by demographic, ethnic, and regional differences

Patient responses to the questions about preference and satisfaction with frequency of treatment administration were evaluated by age at enrollment (<65 versus ≥65 years old) and history of fracture. Both younger (<65 years) and older (≥65 years) patients indicated that they preferred and were more satisfied with the dosing frequency of the 6-month injection over the weekly tablet (P < 0.0001; data not shown). Similarly, patients with or without a prior fracture also preferred and were more satisfied with the frequency of the 6-month injection than the weekly tablet (P < 0.0001).

The PSQ was administered to patients in 14 countries in 18 languages. The US English version of the PSQ was the most commonly completed (33%), followed by the Danish (11%) and Canadian English (11%) versions [29]. In all 14 countries, more patients preferred the injection over the tablet (Supplementary Table 1). In addition, patients from most countries were also more satisfied with the frequency of the 6-month injection over the weekly tablet (Supplementary Table 1).

The PSQ was given in two languages in Canada, Belgium, and the United States. No differences were noted for preference and satisfaction with frequency of administration between patients who took the Canadian English or Canadian French versions of the PSQ. In Belgium, a somewhat greater percentage of patients completing the Dutch version of the PSQ (n = 17) preferred and were more satisfied with the frequency of administration of the injection than patients completing the French version (n = 33), but the sample size was too small for comparison. The largest differences were noted in the United States between patients who took the US English versus the US Spanish version of the PSQ. Over 70% of patients who took the US English version of the PSQ preferred and were more satisfied with the administration frequency of the injection, while slightly more than 50% of patients who took the US Spanish version preferred and were more satisfied with the administration frequency of the injection. However, the English-speaking group was much larger (n = 520) than the Spanish-speaking group (n = 66), making comparison difficult.

Treatment bothersomeness

Nearly half of the patients in both studies indicated that one treatment was not more bothersome than the other (44% in denosumab group, 45% in alendronate group). However, among patients who reported a difference in bother between treatments, most indicated the weekly tablet (45% of denosumab-treated patients and 44% of alendronate-treated patients) was more bothersome than the 6-month injection (5% of patients in both treatment groups; Fig. 4). Patients were less likely to claim bother with the injection, if they reported greater satisfaction with the frequency of the injection (data not shown).

Fig. 4
figure 4

Patient-reported bothersomeness of treatments. The values shown are combined data from the DECIDE and STAND trials

Full size image

The reasons for bother with the 6-month injection or weekly tablet are described in Table 2. Overall, more patients indicated that following the correct routine for the 6-month injection was not bothersome at all compared with following the correct routine for the weekly tablet. Greater numbers of patients reported that they found the weekly tablet to be minimally or moderately bothersome compared with the 6-month injection. In addition, more patients reported experiencing at least some stomach upset after taking the weekly tablet than after receiving the 6-month injection (Table 2). Stomach upset was reported by similar numbers of patients taking placebo and active alendronate tablets.

Table 2 Reasons for bother with treatment
Full size table

Approximately 25% of patients reported that they had at least some dislike of the needle. Most patients reported that dislike of the needle was minimal, although a small number of subjects (n = 48, 3%) did report that they were quite or severely bothered by the needle. Most patients did not report any pain associated with the injections; however, nine (0.6%) patients did report they had pain that bothered them quite a bit or severely after receiving the injections.

Discussion

In two large randomized, double-blind, double-dummy phase 3 trials of postmenopausal women with bone loss, among patients who expressed a preference, significantly more preferred to receive and were more satisfied with the frequency of administration of a 6-month injection over a weekly tablet after 12 months of treatment. Many patients indicated that they were more likely to continue the 6-month injection long term and that it was better fitted to their lifestyles. Among patients indicating bother with treatments, more patients found the weekly tablet bothersome than the 6-month injection. Although in both trials denosumab treatment significantly increased BMD compared with alendronate [30, 31], patients were blinded to their treatment randomization and BMD results at the time they answered the PSQ. Furthermore, there were no significant differences in the fracture rates or incidence of adverse events in these studies, making it unlikely that patients could distinguish which active therapy they were receiving.

Compliance and persistence with therapy are major concerns in the treatment of osteoporosis due to the asymptomatic nature of the disease. While currently available therapies have demonstrated efficacy and safety in the controlled setting of a clinical trial, their use in the community setting has been suboptimal. Adverse effects of treatment, along with complacency with the disease process, may contribute to poor compliance and persistence with treatment regimens, and the resultant loss of antifracture efficacy among patients who discontinue therapy [32–34]. Thus, despite the availability of generally well-tolerated and efficacious therapies, an unmet need exists in the management of osteoporosis. New therapies for osteoporosis will not only need to demonstrate antifracture efficacy, tolerability, and safety in the clinical trial setting, but also convenient administration so that the efficacy may be maximized in the community setting.

Understanding patient preference for the mode and frequency of administration for treatments plays an important role in designing strategies to optimize compliance and persistence with therapy. Considerations that shape patient preference are likely multifactorial and may include perceived treatment effectiveness and safety, specific method of administration (oral versus injection), adverse effects associated with treatment, length of time since regulatory approval for use, physician recommendation, patient experience with and convenience of dosing regimen, and influence on lifestyle [4, 35]. Older patients or individuals with comorbidities who require multiple medications may view a twice yearly subcutaneous injection as more desirable than a weekly oral tablet. Injection of osteoporosis medication may involve a visit to a health care professional, which may be perceived as added value to some patients, but as an inconvenience to others. As all patients in these studies received an injection from a health care provider, we cannot comment on how the interaction with a health care provider might impact patient preference. Furthermore, cultural differences also seem to influence patient preference. For these studies, patients in all countries preferred the injection over the tablet, yet the proportion varied by country. Additionally, differences in percentages of patients preferring the injection were noted among patients in the same country, but who took the PSQ in a different language. The sample size for most countries and ethnic populations were not large enough to allow comparison. Also, no questions included in the PSQ are designed to evaluate how cultural differences may influence patient response. Further studies are needed to understand how ethnic differences may impact preference and satisfaction with mode of treatment administration.

To date, the clinical implications of preference studies remain largely uninvestigated. It has been shown that patients prefer and are more likely to adhere to treatments that are dosed once per day rather than multiple times per day [36]. For the treatment of osteoporosis, patients prefer [16, 17] and have increased compliance and persistence with treatments dosed weekly rather than daily [37–40]. However, compliance and persistence with weekly treatment is still suboptimal [5]. While limited data is available that suggests patients prefer monthly dosing intervals [26, 27, 41], there is little evidence that patients comply and persist with monthly dosing to a greater extent than weekly dosing [42]. When annual intravenous infusion therapy was compared with weekly oral therapy, patients preferred the once yearly treatment [28, 43]. At this time, it is not known if the availability of an annual treatment for postmenopausal bone loss will improve long-term compliance and persistence in the community setting.

In the two studies reported here, overall levels of preference and satisfaction with the injection may be higher than in the community setting based on the self-selection of patients in these studies. As with all studies of preference and satisfaction, the willingness of subjects to accept the treatments offered can potentially confound the interpretation of the results. Subjects in the STAND study were likely accepting and tolerant of bisphosphonate therapy, as ≥6 months prior oral bisphosphonate use and a 1-month run-in period of weekly oral alendronate were requirements for study entry. Similarly, patients in both the DECIDE and STAND studies were aware that they would be receiving two injections. As such, patients with a fear of needles or injections might have declined to participate. Furthermore, some potential prescribers might be concerned that many patients would want to avoid treatments administered by injection [44]. However, patient aversion to injections seems to be more prevalent prior to the first injection, and as patients become familiar with injections, they may be more likely to select this treatment option [45]. Finally, the PSQ was administered at the end of the studies, after patients had had the opportunity to reflect on their experiences with both the injections and tablets.

In summary, patients in these trials indicated that they preferred and had greater satisfaction with an osteoporosis treatment that was administered less frequently. The high prevalence of expressed preference and satisfaction with a 6-month injection over weekly oral medication suggests the acceptability of this route. Positive perceptions of the mode by which treatment for postmenopausal bone loss is administered may increase patient compliance and persistence with and subsequent benefit from prescribed therapy. Based on the results from the DECIDE and STAND studies, the PSQ is being further refined for use in other clinical trials. Additional research is needed to evaluate whether PSQ results are consistent with increased compliance and persistence to medication in the community setting and improved subsequent clinical outcomes.