Eric Houvenagel, PhD

Many effective treatments are currently available for the management of post menopausal osteoporosis. Large-scale clinical trials have proved the ability of these treatments to significantly reduce the risk of fracture; however, this evidence of efficacy has been collected from pivotal studies in settings not necessarily equivalent to routine clinical practice. It has, for instance, been established that adherence to treatment for osteoporosis is largely insufficient

 and that this subsequently has a negative impact on treatment efficacy.

 Several strategies to enhance adherence have been examined.

 The objective of this study was to evaluate the contribution of phone follow-up to improve adherence to oral antiosteoporosis treatment among post menopausal women presenting with osteoporosis-related fractures.

This was a prospective, single-blinded, randomized controlled trial conducted in 1 center in France to compare adherence to oral antiosteoporosis treatments between groups of patients with or without a telephone follow-up. Inclusions began after the study protocol had been approved by the independent regional ethics committee. There was a 2-year inclusion period, and the follow-up period lasted 1 year. The protocol was single-blinded, as the person evaluating the outcome was unaware of the patient’s group (with or without telephone follow-up).

Patients were recruited within the framework of a dedicated healthcare setting, Fracture Liaison Service, designed to screen for and treat osteoporosis-related fractures. The patients included were women aged 50 years or more who had had a documented osteoporosis-related fracture warranting initiation of an oral antiosteoporosis treatment (other than hormone replacement therapy for menopause). They were ineligible if they previously used an osteoporosis treatment.

Screening for osteoporosis and initiation of treatment were conducted during regular outpatient visits or during day hospitalizations. A physical evaluation and bone densitometry were performed in both settings. A multidisciplinary management was proposed during the day hospitalization, formulated with collective or individual workshops with the participation of different specialists such as a rheumatologist, a dietitian, and a health and fitness advisor. The objective was to inform patients about the pathology of osteoporosis and treatment options, and also to give advice about dietary strategies and the importance of physical activity. The choice between the 2 settings (outpatient visit or day hospital) depended on the patient’s preference. For treatment choice, different factors were taken into account, especially contraindications and patients’ preference. Several providers were responsible for treating patients. A letter summarizing the therapeutic propositions was sent to each patient’s primary care physician.

The investigator informed the patients of the study objectives and protocol, and patients also received an information document. The ethics committee did not impose a written consent for patients to be included in the study. Patients could leave the study at any time during the follow-up.

For each recruitment setting (outpatient visit or day hospital), patients were randomly assigned to 1 of 2 groups: intervention (telephone follow-up) or control (no telephone follow-up) (

). The person who performed the randomization was not a member of the investigating team.

Two medical secretaries were in charge of calling patients in the intervention group every 2 months. Before starting the calls, the secretaries participated in a 2-hour training session led by a rheumatologist who explained the basic principles underlying the management of osteoporosis, the types of treatments prescribed, and dosing modalities.

The main objectives of the phone calls were to motivate patients to maintain good adherence to the treatment, to detect any difficulties in compliance with the prescription using non-incriminating questions (

), and to recall the importance of continuing the treatment as prescribed, rather than just assessing adherence alone. The telephone interview was short (less than 10 minutes). If poor adherence was detected, the secretary encouraged the patient to consult her primary care physician.

If the patient was unable to manage her treatment alone (eg, due to cognitive impairment), the secretary contacted the resource person—in charge of procuring the pills, managing the pill box, and administrating the pills&mdash;whose telephone number had been recorded at inclusion.

For all patients, the outcome was assessed 1 year after treatment onset. The 1-year evaluation was conducted during a regular medical visit, or during a telephone interview if the patient was unable to attend the 1-year follow-up visit at the appropriate time. The same practitioner, who was blinded to the patient’s group, conducted both face-to-face and telephone interviews. Patients were informed before the 1-year evaluation to not reveal their study group (intervention or control) to the evaluating physician. Shortly before the evaluation, each patient completed the self-administered Morisky adherence questionnaire,

 which uses nonincriminating questions to detect any difficulty in adhering to treatment. The physician then applied a nondichotomic response grid (

) to fine-tune the responses to the Morisky questionnaire using a more directive interview designed to obtain a precise assessment of how the patient took her medications (regularity, missed doses) and identify any causes of poor adherence.

The main outcome was the number of adherent patients to an oral antiosteoporosis treatment 1 year after its institution (the prescribed drug could be changed during the year).

A patient was considered to be adherent if, during the 2 months preceding the evaluation, medications were taken as prescribed and the medication possession ratio (MPR) was ≥80%.

The statistical analysis was performed with Statistica 9.0 (Statsoft Inc, Tulsa, Oklahoma) using Pearson’s χ

 test. P <.05 was accepted as statistically significant. An intention-to-treat analysis was performed to compare the characteristic features of the patients and the outcomes in each group after randomization. The potential positive effect of the telephone interviews was also assessed using a per protocol analysis and an odds ratio analysis.

According to the data found in the literature, the proportion of compliant osteoporotic patients is low at 1 year: around 40% to 50%.

 Little data exist so far on the impact on compliance of a phone follow-up. In our study, our starting assumption was that compliance would be 45% in our control group, and that telephone follow-up would provide a 25% gain in the study group, with 10% of patients lost to follow-up. This required including approximately 160 patients to obtain a statistically significant difference.

Characteristic Features of the Included Patients

One hundred sixty-eight patients were randomized to the 2 groups. Four of these patients asked to withdraw from the study shortly after their inclusion; all 4 had been recruited during day hospital stays, and they had not yet received a follow-up call when they stated their desire to drop out of the study. The analysis thus included 164 patients (

All 164 patients had a history of at least 1 osteoporosis-related fracture; the mean age was 70.4 years. These patients were recruited during outpatient visits (n = 67) or day hospitalizations (n = 97). An oral treatment was prescribed for nearly all patients: weekly bisphosphonate (n = 101, 61.6%); strontium ranelate (n = 62, 37.8%). The number of attempts to reach the patients by phone in the intervention group was not recorded. In fact, all patients (except those lost from the study) were contacted on the scheduled dates. Treatment was modified during the course of the study year in 22 patients. The final outcome was not available for 15 patients (9.1%) who were lost to follow-up, had withdrawn from the follow-up, or declined to participate in the final evaluation. The 1-year evaluation was thus performed in 149 patients, either during a face-to-face interview (n = 117 patients, 78.5%) or a telephone interview (n = 32 patients, 21.5%).

Study of Patients Recruited During an Outpatient Visit

The mean age of patients recruited during an outpatient visit was 72.9 years. There were 33 patients in the intervention group and 34 in the control group (Figure 4). The 2 groups were comparable in age, type of treatment prescribed, number of patients lost to follow-up, and mode of evaluation (face-to-face or telephone interview). At 1 year, 24 patients in the intervention group were taking their treatment persistently (although not necessarily correctly) (72.7%) versus 16 in the control group (47%). Twenty-two patients in the intervention group were considered to be adherent (66.6%)—with the remaining 2 patients of the 24 total having been persistent but not adherent, as they were taking their treatment but not correctly&mdash;versus only 10 in the control group (29.4%). The difference was statistically significant (P <.01). The per protocol analysis showed that the proportion of adherent patients was 71% in the intervention group versus 33.3% in the control group (P <.01) (

). The risk of discontinuing treatment was 4.57-fold greater among patients without telephone follow-up (odds ratio [OR], 4.57; 95% CI, 1.66-12.58).

Study of Patients Recruited During a Day Hospital Stay

The mean age of patients recruited during a day hospital stay was 68.7 years. There were 46 patients in the intervention group and 51 in the control group (Figure 4). The 2 groups were comparable regarding age, type of treatment prescribed, number of patients lost to follow-up, and mode of evaluation (face-to-face or telephone interview). At 1 year, 33 patients in the intervention group were taking their treatment (71.7%) versus 27 in the control group (52.9%). Twenty-nine patients in the intervention group were considered to be adherent (63%) versus only 18 in the control group (35.3%); the difference was statistically significant (P <.01). The per protocol analysis showed that the proportion of adherent patients was 70.7% in the intervention group versus 38.3% in the control group (P <.01) (

). The risk of discontinuing treatment was 2.55-fold greater among patients who did not have the phone follow-up (OR, 2.55; 95% CI, 1.13-6.75).

A comparison of the patients who had a telephone follow-up versus those who did not, regardless of the recruitment setting (outpatient visit or day hospital), showed that the 2 groups were comparable regarding age, type of treatment prescribed, number of patients lost to follow-up, and mode of evaluation (face-to-face or telephone interview). Among the 72 patients in the intervention group who were evaluated at 1 year, 57 were taking their treatment (72.6%) and 51 were considered adherent (64.6%). Among the 77 patients in the control group evaluated at 1 year, 43 were taking their treatment (50.6%) and 28 were considered adherent (32.9%). The per protocol analysis showed that the proportion of adherent patients was 70.8% in the intervention group versus 36.4% in the control group (P <.01) (

). The risk of discontinuing treatment was 3.65-fold greater among patients who did not have the telephone follow-up (OR, 3.65; 95% CI, 1.92-6.92).

The reasons most often reported by patients to explain why they had discontinued their treatment were lack of motivation to continue and nonrenewal of the prescription by their primary care physician. In most cases, patients cited several reasons. By order of frequency, the reasons cited were: lack of motivation (n = 22), nonrenewal of prescription (n = 13), fear of potential adverse effects (n = 9), and multiple medications (n = 5).

Reasons of discontinuation for nonrenewal of the prescription were unclear. We could not elucidate whether the primary care physician decided to stop the prescription for medical reasons or because the patient did not meet regularly with the physician.

The results of this study argue in favor of telephone interviews to enhance the adherence of osteoporosis patients to their therapeutic regimen.

Management of patients with chronic disease stresses the importance of adherence to the therapeutic program

—a real problem being encountered with osteoporosis,

 which is a particularly common and serious disease, making it a major public health concern.

 Although much progress has been made over the last few years in terms of treatments proposed for osteoporosis, they can be fully effective only if taken as prescribed for a relatively long period of time. It has been demonstrated that good therapeutic adherence is associated with greater gain in bone mineral density, greater inhibition of bone remodeling, and most importantly, reduced risk of fracture.

 Inversely, poor adherence can lead to an increased risk of adverse effects and also greater overall cost for the society, given that many fracture-related expenditures can be avoided, as can the cost of adverse effects, and the cost of the treatment that was not taken correctly.

Several strategies have been explored in an attempt to improve adherence to antiosteoporosis treatments. However, results have been somewhat disappointing, often limited, and inconsistent.

 Regular patient follow-up would be an effective strategy, especially during the “at risk” period during the first few months following the first prescription.

 This observation prompted us to examine the effectiveness of a telephone interview follow-up performed by non-physicians in order to improve adherence to oral antiosteoporosis treatments in post menopausal women. A few reports in the literature describe the usefulness of this type of follow-up in patients with osteoporosis and chronic diseases.

 For the first time, our study utilized  a randomized control trial to test the efficacy of this strategy in osteoporosis patients receiving care in a dedicated healthcare setting.

We measured the 2 main elements classically examined to study the way patients take their medications: patient adherence and persistence with their treatment. Persistence can be defined as continuing to take a medication for a given period of time, whereas a patient is considered adherent if they take the medication exactly as prescribed. Assessment of these elements requires counting the number of days the medication is actually taken as prescribed by the physician during the course of a defined period; correct dosing and timing of intake are essential. The same is true concerning the way the treatment is ingested, since it may have an important impact on the bioavailability of such treatments. This is particularly true for antiosteoporosis treatments, especially bisphosphonates. An 80% MPR is widely accepted as “satisfactory,” and this 80% MPR has been established as the optimal level of adhesion in antihypertension trials. It has been demonstrated that the rate of osteoporosis-related fractures increases significantly above this cutoff.

 MPR is commonly calculated as the number of pills dispensed in an observation period divided by the total time in the observation period. In our study, the MPR referred to the number of pills effectively taken by the patient divided by the total number of pills patients were theoretically supposed to take in a given period, which in our case was the last 2 months of follow-up.

The purpose of a telephone interview follow-up goes beyond increasing the MPR. Its goal is to help as many patients as possible reach the 80% level. We therefore chose the number of adherent patients as the main outcome rather than the MPR observed in each group of patients. Indeed, a significant increase in MPR could theoretically be achieved with the telephone interview follow-up despite an insufficient proportion of adherent patients (ie, patients whose MPR would be high enough to ensure the best possible treatment efficacy).

It is difficult to ascertain objective assessments of adherence other than direct reports from patients. Counting prescription refills does not give the number of pills actually taken by the patient or any evaluation of dosing or timing. At present, there is no validated questionnaire to measure therapeutic adherence in patients with osteoporosis, but nevertheless, history-taking is a validated and effective method for evaluating therapeutic adherence.

 In addition, our evaluations were made in a homogeneous manner—utilizing a standardized interview, single physician conducting all evaluation interviews, simple-blinded evaluation protocol&mdash;which limits bias in interpreting the patient’s potentially subjective report.

We did not use markers of bone remodeling, though variations in such markers can, in certain cases, provide information on the effect of medications, and thus their intake. Such results would have been difficult to interpret in the population under study (ie, recent fracture for most patients, minimal predictable variability observed in patients taking strontium ranelate). Nevertheless, in the framework of routine clinical practice, it is quite possible that an analysis of bone remodeling markers could provide an interesting contribution in combination with telephone interviews, particularly when there is doubt about poor adherence in a patient with an antiresorption medication who has not had a recent fracture.

All patients in our study were informed about the objective of the study at inclusion. This might have had a positive impact on adherence since patients were informed that their behavior (adherence) would be evaluated at 1 year, but this was true for both the groups receiving and not receiving follow-ups, so this did not represent a possible bias.

The results of our study confirm the poor therapeutic adherence observed in patients treated for osteoporosis. In the control groups that did not have a regular telephone follow-up, just slightly over half of the patients were taking their treatment at 1 year, and less than 40% could be considered adherent in that they were taking it just as prescribed—patients may still have experienced frequent oversights, incompliance with the schedule, and other issues or mistakes that would have prevented actual adherence. These figures are comparable to previous data in the literature, generally, concerning adherence to prescriptions of bisphosphonates in osteoporotic patients who did not necessarily have a history of 1 or more fractures.

 According to our findings, a history of 1 or more fractures does not appear to guarantee better adherence. Similarly, from the results observed in the groups without regular follow-up, more “intensive” management during a day hospitalization where the patients received precise information about the disease and its treatments did not appear to guarantee better adherence.

Conversely, the proportion of adherent patients increased very significantly with regular telephone interviews. This was observed both after an initial outpatient visit and after an initial day hospitalization. This better adherence observed in the intervention groups appears to be linked to the regular telephone follow-up. The different groups compared were quite similar, and there were no distinctive characteristics that might have explained the better adherence observed in patients in the intervention groups. In particular, there was no significant difference regarding age and antiosteoporosis treatment. Nevertheless, other elements which might also have an effect on therapeutic adherence (such as comorbid illnesses, number of other medications, educational level, socioeconomic level, or multiple medications) were not studied.

Articles

The authors studied the usefulness of a regular telephone follow-up to improve adherence to oral antiosteoporosis treatments. Their results support such follow-up and the use of this strategy in routine clinical practice.