Discussion 4q1 statistics

nonexperimentalstudy.pdf

Home >Nursing homework help >Discussion 4q1 statistics

Impact of Adherence to Interferons in the Treatment of Multiple Sclerosis A Non-Experimental, Retrospective, Cohort Study

Stephanie C. Steinberg,1 Richard J. Faris,2 Cyril F. Chang,3 Andrew Chan4

and Mark A. Tankersley2

1 Accredo Health Group, Inc., The University of Memphis, Memphis, Tennessee, USA

2 Health Outcomes, Accredo Health Group, Inc., Memphis, Tennessee, USA

3 Methodist-Le Bonheur Center for Healthcare Economics, The University of Memphis, Memphis,

Tennessee, USA

4 Department of Neurology, St Josef-Hospital, Ruhr-University Bochum, Bochum, Germany

Abstract Background: Relapsing-remitting multiple sclerosis (RRMS) is a chronic disease affecting about 400 000 people in the US characterized by increasing

patient disability and burden on society. While there is no cure for multiple

sclerosis (MS), pharmaceutical treatments exist that can limit the number of

relapses a patient experiences, and slow disease progression. One such class of

agents used to treat RRMS are the interferons: interferon-b-1a (Rebif� and Avonex�) and interferon-b-1b (Betaseron� and Extavia�). Patients must take these injectable medications regularly to achieve the optimal outcomes.

However, patient issues and potential adverse effects of the medication may

prevent the patient from taking the medication as directed and lower

adherence. To date, limited evidence exists regarding the effect of patient

adherence to interferon-b therapies on clinical and economic outcomes. Objective: The purpose of this study was to explore the impact of patient

adherence to interferon-b therapy on MS relapse rates and healthcare re- source utilization.

Methods: Using a non-experimental, retrospective cohort design, a sample

population (n = 1606) was drawn from patients identified in a database that includes both pharmacy and medical claims data. The study population was

separated into two groups based on a measure of medication possession ratio

(MPR)-adherent and non-adherent patients, and adherence was defined as

MPR ‡85% in a given year during the study period (2006–8). Key outcome variables included MS relapses and healthcare resource utilization. Data were

analysed using parametric and non-parametric statistics, and regression

modeling.

Results: During the study period, the average MPR for all patients on

interferon-b therapy varied from 72% to 76%. Only 27–41% of patients in

each year were considered adherent (i.e. MPR ‡85%) and only 4% of patients had an MPR of ‡85% throughout the 3-year study period (2006–8). Patients who were adherent tended to have a lower risk of relapses over 3 years than non-

adherent patients. A significantly lower risk of relapses was found in 2006 (risk

ratio [RR] 0.89; 95% CI 0.81, 0.97). Furthermore, an increasingly larger effect emerged between adherence and relapses when comparing adherent patients

(MPR ‡85%) with subgroups of non-adherent patients (<80%, <75%, <70%, <65% and <60%). The impact of adherence on emergency room (ER) visits also tended to suggest a lower risk during 2006, 2007 and 2006–8. During 2008, the

risk for an ER visit was significantly lower for patients adherent in 2007

(RR 0.78; 95% CI 0.61, 0.99). Inpatient admissions followed the ER trends, as patients considered adherent in 2006 and 2007 tended to have a lower

risk over 3 years. This result was significant for patients adherent in 2007

(RR 0.79; 95% CI 0.65, 0.98). Conclusion: The findings of low patient adherence and the impact of

adherence on relapses and healthcare resource utilization strongly suggest

opportunities to reduce healthcare resource utilization and healthcare costs

among RRMS patients taking interferon-b therapy. Efforts should be un- dertaken to understand and improve medication-taking behaviour in this

population so as to minimize the negative impacts of RRMS on patients while

reducing unnecessary direct and indirect costs to treat disease exacerbations.

Background

Relapsing-remitting multiple sclerosis (RRMS) is a chronic disease characterized by relapses that over time may lead to accumulating disability.[1] As the disease progresses, patient impairment and the burden on society also increase. While there is no cure for MS, which affects about 400000 people in the US alone, pharmaceutical agents exist that can limit the number of relapses a patient experiences and slow disease progression.[2] One such class of agents approved to treat RRMS are known as in- terferon-b products: interferon-b-1a (Rebif� [EMD Serono, Rockland, MA, USA] and Avonex�

[Biogen Idec, Wellesley, MA, USA]) and inter- feron-b-1b (Betaseron� [Bayer Healthcare, Mont- ville, NJ, USA] and Extavia� [Novartis, East Hanover, NJ, USA]).[3]

Achieving optimal clinical, economic and humanistic outcomes from interferon-b therapy requires that RRMS patients take the medication as prescribed. Additionally, head-to-head com- parisons with the same interferon-b therapy show

strong indications for a dose and frequency dependency of the therapeutic effects.[4] The ob- jectives of this study were: (i) to assess the impact of RRMS patient adherence to interferon-bs on relapse rates; and (ii) to determine the impact of adherence to interferon-bs on healthcare resource utilization.

Previous Research

Since the mid-to-late 1990s, disease-modifying therapies (interferon-b and non-interferons, e.g. glatiramer acetate) have become the standard first-line treatment for RRMS. Early treatment with disease-modifying therapy, particularly interferon-b and glatiramer acetate, may offer several beneficial outcomes for MS patients (i.e. clinically isolated syndrome and early MS), including: a reduction in clinical relapses, prevention of new lesion formation on magnetic resonance imaging (MRI), and presumably slower disability progression rates.[5] More recent studies have also shown that early and continued

90 Steinberg et al.

injectable disease-modifying drug therapy may reduce healthcare costs for patients.[6-9]

However, adherence to therapy and remaining on therapy are important factors in achieving the benefits of interferon therapy in this chronic disease. The unpredictable and potentially devas- tating effects of MS are evident beyond the clinical realm.[10-12] An examination of an RRMS patient’s utilization and related costs for healthcare re- sources demonstrates the tremendous financial impact of RRMS on society, the individual and the family. A study by Pope and colleagues found that healthcare utilization among insured individuals with MS (private insurance and government plans) was two to three times higher than that for in- dividuals without MS.[13] Among MS patients, healthcare costs included a higher use of healthcare resources in proportion to the patient’s increased disability (according to the Kurtzke Expanded Disability Status Scale [EDSS]) and intermittent relapses.[14,15] A more recent study of RRMS patients found the cost of a relapse can range from $US248 for a mild relapse to $US12 870 for a severe relapse (year of costing 2002), depending on the pharmacological treatment and need for in- patient care.[16] These factors, coupled with reduced or lost wages, can devastate the patient in terms of physical, vocational and social well-being, loss of employment and social contacts, higher healthcare costs, and a lower quality of life.[14-16]

Adherence to interferon-b therapy can be prob- lematic for patients. Discontinuation rates for non-adherent patients in previous studies ranged from <20% to 50% within the first 2 years.[17-20] The reasons for discontinuation may include psy- chological factors, such as depression and anxiety, clinical factors, e.g. adverse effects, cognitive im- pairment, fatigue and disease progression, and financial and physical factors, such as co-payments and the need to inject agents. Specific determinants include the patient’s perception of control over the disease as a result of using the prescribed drug, a high level of optimism and no previous use of other disease-modifying drugs.[7,18,21]

Studies linking drug adherence to clinical outcomes exist for some of the most prevalent chronic diseases such as hypertension and dia- betes mellitus.[22,23] Although there is evidence of

a dose-frequency dependency for therapeutic ef- ficacy of interferon-b in MS,[24] far fewer studies of the effects of drug adherence on therapeutic outcome have appeared in the literature. A rare exception is a recent observational study that describes the initial experience in Italy with the use of three interferon-b therapies (Rebif�, Avonex� and Betaseron�).[25] This 3-year study also investigated the effects of adherence on the frequency of acute relapses among RRMS pa- tients but the results are still preliminary accord- ing to the investigators.

Interferon-b therapies have been shown to be effective in reducing the frequency of relapses in RRMS patients.[26,27] To date, however, scant evidence exists concerning the intricate relation- ship between the level of patient adherence to interferon-b therapies and the frequency of re- lapses. Also lacking are data on the effects of adherence on RRMS patients’ use of healthcare resources such as emergency room (ER) visits, hospitalizations and visits to specialists.

Data and Methods

Data Sources

Claims data for the years 2006, 2007 and 2008 from a Pharmacy Benefit Management company with a specialty component were used in this study. The database included both pharmacy and medical claims. The reliability and validity of prescription claims data for measuring adherence have been examined in numerous studies in the US and Canada.[28-34] Study results found phar- macy claims to be a reliable and stable measure, albeit indirect, of adherence.

Design

This study used a non-experimental, retro- spective cohort design. The balanced study popu- lation (n = 1606) was drawn from all patients identified with at least two claims for an interferon- b product (i.e. interferon-b-1b [Betaseron�] or interferon-b-1a [Rebif� or Avonex�]) during the base period of 2005. Inclusion criteria included patients who were continuously eligible for benefits and had prescription filling opportunity for the

Adherence to Interferons in Multiple Sclerosis 91

entire study period of 2006–8. Extavia� (inter- feron-b-1b) was not considered due to its US licensure in 2009.

Description and Definitions of Key Variables

The key outcome variables in this study were: (i) relapses, and (ii) utilization of healthcare re- sources. The key treatment variable was adherence.

Adherence

Adherence is a global term that represents three distinct subcategories: acceptance, persis- tence and compliance.[35,36] Many studies in the literature use the term compliance synonymously with adherence, which provides only a limited view. This study used an adherence measure that incorporates both compliance and persistence, since patient acceptance is assumed by the pre- sence of a prescription claim. A fixed-interval method was used to calculate adherence, meaning that patients were followed over a standard (or fixed) period of time, and their medication pos- session ratio (MPR) was calculated per the amount of medication on hand over this fixed period. Because the method incorporates both compliance and persistence (which variable-in- terval methods do not), lower adherence esti- mates are often seen.[37] The formula for the fixed-interval method is described below:

MPR ¼

ðTotal days supply dispensed 360 days from the index prescriptionÞ

360 days

Adherence was used in this study as both a continuous and dichotomous variable. As a con- tinuous variable, each patient’s MPR was calculated and used to determine the ‘average’ MPR for the study population. Each patient’s adherence rate was then converted to a dichot- omous variable by assessing whether the patient was adherent or not. The cutoff point for asses- sing adherence was 85% for all analyses. Selection of the cutoff point was based upon a sensitivity analysis that compared the risk ratios for relapses over the 3-year study period. Results from the sensitivity analysis in figure 1 indicated that pa- tients considered adherent at 85% tended to have a lower risk of relapses versus subpopulations not

adherent below 80%. Patients adherent at 90% did not have a lower risk for a relapse than pa- tients adherent at less than 85%. This suggests that an 85% adherence threshold maximizes a patient’s benefits from interferon-b therapy.

Relapses

Relapses are a common clinical outcome for MS patients that can be detected in claims data.[16] Pharmacy and medical claims informa- tion were used to identify and categorize relapses based upon the most intensive site of care for the patient’s treatment. Several key assumptions ap- plied to this study include: (i) a relapse was de- fined as a 90-day period during which a patient experienced an exacerbation of existing symp- toms or the appearance of new symptoms; (ii) the period between relapses, or ‘clean period’, was at least 30 days; (iii) the place where healthcare was delivered and the scope of services provided defined the level of severity of a relapse. Specifi- cally, the site and intensity of care were char- acterized as follows: (i) a hospitalization and subsequent care signified high intensity manage- ment (severe relapse); (ii) ER and/or physician office visits signified moderate intensity manage- ment (moderate relapse); and (iii) physician office visits and symptom-related medications signified low intensity management (low relapse).[16,38-42]

Next, a relapse variable was constructed for each study year by incorporating previously published definitions.[16] Patients were then classified ac- cording to: (i) whether they experienced a relapse or not, and (ii) the total number of relapses.

0.90

0.95

1.00

1.05

R e

la p

se r

is k

ra tio 1.10

1.15

<80 <75 <70 <65 <60 MPR (%)

≥85

Fig. 1. Comparison of multiple sclerosis relapse risk ratios by level of adherence. MPR = medication possession ratio.

92 Steinberg et al.

Utilization of Healthcare Resources

The utilization of healthcare resources encom- passed a second outcome variable. The primary resources used to treat MS patients included ER visits, hospital admissions and physician office visits. ER visits, hospital admissions and physi- cian office visits were defined as binary variables to indicate the occurrence or non-occurrence of the event during each study year.

Independent Variables

Healthcare theory and pharmacoeconomic studies suggest that relapses and healthcare re- source utilization are a function of age, sex, co- morbidities, sociodemographic factors (education, income and residence), and patient out-of-pocket expenses (co-payments).[13-15,19] By integrating variable selection methodologies from previous MS studies, the following variables were identified from the pharmacy and medical claims data for inclusion in this study. The key independent

variables included: age or age group, sex, anti- depressant use, patient co-pay amount, chronic disease score (CDS), geographic region in the US, and interferon-b use. Each is defined below: � Age was defined by patient for each year of the

study. A continuous variable form, for averages, and an ordinal variable form by age group were used. Age group categories were con- structed using the following intervals: age group 1 (0–34 years), age group 2 (35–44 years), age group 3 (45–54 years), age group 4 (55–64 years), and age group 5 (‡65 years).

� Sex was defined as a binary variable to denote female (yes = 1 and no = 0).

� Antidepressant use was defined as a binary variable to denote a patient’s use (yes = 1 and no = 0) during each study year.

� Patient co-payment amount was defined as a continuous variable to denote out-of-pocket payment for interferon-bs.

� CDS was a continuous variable with a mini- mum of zero to denote disease acuity measure for each study year. Scores increased with the number of chronic conditions present. There- fore, a higher CDS implied more co-morbid- ities, or greater costs due to greater healthcare utilization.[43]

� Geographic region was categorized by South, Northeast, West and Midwest USA based on the patient’s place of residence at the start of the study period.

� The three interferons (Betaseron�, Rebif� and Avonex�) were defined by a binary variable (filled prescription = 1 and no fill = 0) to repre- sent whether a patient filled an interferon prescription.

Table I. Study population (n = 1606)

Characteristic Value

Sex

Female (%) 76

Age (average – SD) [y]

2006 51.84 – 9.81

2007 52.84 – 9.81

2008 53.84 – 9.81

Age group categories (%)

Group 1 (0–34 y) 5

Group 2 (35–44 y) 18

Group 3 (45–54 y) 38

Group 4 (55–64 y) 31

Group 5 (‡65 y) 8

Antidepressant use (%)

2006 46

2007 47

2008 47

Patients by geographic region (%)

South USA 39

Northeast USA 17

West USA 17

Midwest USA 27

SD = standard deviation.

Table II. Adherence to interferon-b therapy

MPR (average – SD)

2006 0.72 – 0.20

2007 0.75 – 0.21

2008 0.76 – 0.20

Patients adherent (MPR ‡85%) [%]

2006 27

2007 40

2008 41

2006–8 4

MPR = medication possession ratio; SD = standard deviation.

Adherence to Interferons in Multiple Sclerosis 93

Descriptive statistics such as the average values, standard deviations (SDs) and confidence intervals (CIs) for all outcome and independent variables are summarized in tables I and II.

Analytical Strategy

A variety of appropriate statistical techniques were used for each of the research questions as outlined below using SAS� version 9.1.3 (SAS Institute, Cary, NC, USA).

Impact of Adherence on Relapses and Healthcare Resource Utilization

A two-part analysis was conducted to determine whether adherent patients experienced fewer and/or lower healthcare resource utilization than non-adherent patients over the study period. First, the average number of relapses and health- care resource utilization were compared by study year (2006, 2007 and 2008) for the adherent and non-adherent populations using a Kruskal-Wallis test. Second, multivariate logistic general estimat- ing equation (GEE) regression models were con- structed to examine the impact of adherence on relapses and healthcare resource utilization. Re- gression analyses also included combinations of current year variables (i.e. dependent and inde- pendent variables), and time-lagged indepen- dent variables. This allowed for a more rigorous examination of any current and time-lagged as-

sociations between adherence and relapses. Both explanatory and dependent variables represented values for 2006, 2007 and 2008.

Impact of Non-Adherence on Relapse

The analyses mentioned in the previous sec- tion were constructed to examine the risk of re- lapse for adherent and non-adherent patients at an 85% cutoff. Individuals with MPRs <85% comprised the single non-adherent group. A sub- analysis was conducted to determine the presence of an incremental effect of decreased levels of adherence on the risk of relapse. The subanalysis, shown in figure 1, defined five levels of compar- ison between patients adherent at ‡85% and pa- tients adherent at lower thresholds (i.e. <80%, <75%, <70%, <65% and <60%) in relationship to the risk of a relapse over the 3-year study period.

Results

Characteristics of the Study Population

The key characteristics of the study popula- tion are given in table I. During the baseline year (2006), patients were predominately females (76%), had an average age of 52 years (69% were aged 45–64 years) and most frequently resided in the South (39%). Also notable was the prevalence of antidepressant use in the population (i.e. 46%). The average adherence rate for the study population

Table III. Average (– SD) relapses in patients adherent (MPR ‡85%, fixed-interval method) or non-adherent to interferon-b therapy (2006–8)

Year Adherent Non-adherent p-Value

2006 1.03 – 1.14 1.18 – 1.13 0.01*

2007 1.30 – 1.12 1.30 – 1.12 0.71

2008 1.26 – 1.13 1.22 – 1.15 0.49 MPR = medication possession ratio; SD = standard deviation; * p < 0.05 by Kruskal-Wallis test.

Table IV. Association between adherence (MPR ‡85%, fixed-interval method) and relapses

Year of patient

adherence

Relapses

2006 >0 RR (95% CI)

Relapses

2007 >0 RR (95% CI)

Relapses

2008 >0 RR (95% CI)

Relapses

2008 >2 RR (95% CI)

Relapses

2006–8 >0 RR (95% CI)

Relapses

2006–8 >2 RR (95% CI)

2006 0.89 (0.81, 0.97)* 0.92 (0.92, 1.07) 0.99 (0.91, 1.06) 0.83 (0.61, 1.12) 0.97 (0.91, 1.02) 0.83 (0.61, 1.13)

2007 Not relevant 0.98 (0.92, 1.05) 1.03 (0.96, 1.10) 0.68 (0.47, 0.99)* 0.78 (0.58, 1.05) 0.81 (0.62, 1.06)

2008 Not relevant Not relevant 1.06 (0.99, 1.13) 0.88 (0.66, 1.17) 1.04 (0.78, 1.39) 0.91 (0.69, 1.19)

2006–8 Not relevant Not relevant 1.06 (0.89, 1.25) 0.87 (0.74, 1.01) 0.97 (0.94, 1.00)* 0.87 (0.74, 1.01)

CI = confidence interval; MPR = medication possession ratio; RR = risk ratio; * p < 0.05.

94 Steinberg et al.

revealed an increasing trend from 72% to 76% over 3 years as shown in table II. When using the 85% cutoff point to classify patient adherence, the rates ranged from 27% to 41% each year. However, over all 3 years, only 4% of patients were classified as adherent.

Impact of Adherence on Relapses

Adherent patients experienced fewer relapses on average over the study period compared with non-adherent patients, as shown in table III. This difference was statistically significant (p < 0.05) in 2006. However, over the 3-year period, average relapses increased for both adherent and non- adherent patients – an indication that a portion of adherent patients may continue to experience relapses.[16]

Lack of adherence to drug therapy is not the only factor to examine in relation to an MS relapse. Many other prognostic factors, such as the pa- tient’s age, sex, stage of disease, development of neutralizing antibodies, etc., may impact on a pa- tient’s risk of relapse.[8] Furthermore, current and lagged relationships may exist between adherence, and the risk of relapses experienced by a patient over time. A multivariate GEE logistic regression analysis conducted using relapses (relapse = 1 and no relapse = 0) in a given year as the dependent variable was estimated to test the significance of the relationship between relapse and adherence to interferon-bs in 2006, 2007 and 2008. Results for the regression models are shown in table IV, which reports only the risk ratios for MPR, the main variable of interest. A strong association was re- vealed between adherence and relapses during 2006. Adherent patients had an 11% lower risk of relapses (p < 0.05) in 2006 compared with non- adherent patients. In all regression models, patients who were adherent in 2006 tended to have a lower

risk of relapses in 2007, 2008 and over the 3-year study period (2006–8). Results for patients con- sidered adherent in 2007 were generally not statis- tically significant, but did tend to suggest a lower risk of a relapse in current and subsequent years. Patients adherent in 2007, however, did have a significantly lower risk (p < 0.05) for greater than two relapses in 2008, while patients adherent over 2006–8 had a 3% lower risk of relapses (p < 0.05) over the 3-year study period.

Impact of Non-Adherence on Relapses

The previous analyses examined the association between adherence with an 85% cutoff point and the risk of a relapse during 2006–8. Adherent pa- tients tended to have a lower risk for at least one relapse and at least two relapses during 2006–8. However, this analysis did not address the incre- mental effect of adherence at levels <85%. To measure the incremental effects of lower adherence threshold levels, five multivariate logistic regres- sion models were run to compare adherence at <80%, <75%, <70%, <65% and <60%, respectively, with adherence at ‡85% as the comparison (i.e. risk ratio = 1). The regression results shown in table V and figure 1 indicated an increased risk for a re- lapse as adherence decreased by 5% increments. Significantly higher risks (p < 0.05) for a relapse were found for adherence at levels <70%, <65% and <60%. These results also highlighted the levels

Table V. Sensitivity analysis of incremental effects of lower adherence threshold levels vs adherence cutoff threshold of MPR ‡85% on risk ratio of relapse

MPR 2006–8

‡85% <80% <75% <70% <65% <60%

Risk ratio (95% CI) 1.00 1.03 (0.971, 1.08) 1.03 (0.974, 1.09) 1.12 (1.03, 1.23)* 1.10 (1.02, 1.18)* 1.13 (1.03, 1.24)*

CI = confidence interval; MPR = medication possession ratio; * p < 0.05.

Table VI. Average (– SD) inpatient admissions in patients adherent (MPR ‡85%, fixed-interval method) or non-adherent to interferon-b therapy (2006–8)

Year Adherent Non-adherent p-Value

2006 0.09 – 0.38 0.11 – 0.40 0.24

2007 0.12 – 0.56 0.10 – 0.42 0.38

2008 0.12 – 0.54 0.11 – 0.42 0.88 MPR = medication possession ratio; SD = standard deviation.

Adherence to Interferons in Multiple Sclerosis 95

of non-adherence that offered the greatest oppor- tunities to improve adherence and to lower the risk of relapse.

Healthcare Resource Utilization

Hospital Admissions

Adherent patients experienced fewer hospital admissions during 2006 and a slightly higher rate in 2007 and 2008 compared with non-adherent patients (table VI). Average differences between adherent and non-adherent patients were not statistically significant during the study period.

The association and statistical significance be- tween adherence and hospital inpatient admissions were examined using a series of multivariate logis- tic GEE regression models (table VII). Overall, adherent patients tended to have a lower risk for an inpatient admission over all 3 years. Significantly lower risks (p < 0.05) for inpatient admissions, however, were identified for patients adherent in 2006 and 2007. Adherence in 2006 was associated with a 40% lower risk of an inpatient admission during 2006. Patient adherence in 2007 and over the 3-year study period was also associated with a lower risk of inpatient admissions in 2008 (21% and 84% reductions, respectively), while patients adherent in 2008 tended to have a lower risk of inpatient admissions during 2008 and over 3 years.

Emergency Room Visits

On average, adherent patients had fewer ER visits than non-adherent patients over the 3-year study period, as shown in table VIII. Specifi- cally, adherent patients had fewer ER visits in the first 2 years and a slightly higher rate in the third year compared with non-adherent patients. Average comparisons between adherent and non-

adherent patients were not statistically significant during the study period.

Multivariate regression models were run to examine the association between adherence and ER visits in more detail. The results shown in table IX indicated that patients adherent during the study period tended to have a lower risk for ER visits during 2006 and over all 3 years. Pa- tients who were adherent in 2007 had a 22% lower risk (p < 0.05) of an ER visit in 2008. This result was consistent with the overall trend of a lower risk for ER visits in 2007 and over 3 years. Simi- larly, patients who were adherent in 2008 tended to have a lower risk of an ER visit over 3 years.

Office Visits

Adherent patients had fewer physician office visits during 2006 and 2007 compared with their non-adherent counterparts, as shown in table X. However, in 2008, a higher rate of office visits was evident for adherent patients (p < 0.05). Overall, office visits rates increased for adherent patients during the study period. Over 3 years, non-adherent patients had a lower rate of office visits in 2006, but a higher rate in 2007 than ad- herent patients. With the exception of 2008, the average comparisons were not significant at p < 0.05.

Table VII. Association between adherence (MPR ‡85%, fixed-interval method) and inpatient admissions

Year of patient

adherence

Inpatient

admissions

2006 >0

Inpatient

admissions

2007 >0

Inpatient

admissions

2007 >1

Inpatient

admissions

2008 >0

Inpatient

admissions

2006–8 >0

RR (95% CI) RR (95% CI) RR (95% CI) RR (95% CI) RR (95% CI)

2006 0.61 (0.36, 1.01) 1.24 (0.99, 1.54) 1.85 (1.22, 2.79)* 1.05 (0.85, 1.30) 0.98 (0.73, 1.30)

2007 Not relevant 1.22 (0.86, 1.74) 0.75 (0.35, 1.58) 0.79 (0.65, 0.98)* 0.16 (0.03, 0.93)*

2008 Not relevant Not relevant Not relevant 0.79 (0.44, 1.46) 0.97 (0.77, 1.22)

2006–8 Not relevant Not relevant Not relevant 0.92 (0.74, 1.45) 0.74 (0.30, 1.82)

CI = confidence interval; MPR = medication possession ratio; RR = risk ratio; * p < 0.05.

Table VIII. Average (– SD) emergency room visits in patients adherent (MPR ‡85%, fixed-interval method) or non-adherent to interferon-b therapy (2006–8)

Year Adherent Non-adherent p-Value

2006 0.05 – 0.35 0.05 – 0.25 0.33

2007 0.06 – 0.33 0.07 – 0.34 0.75

2008 0.08 – 0.41 0.07 – 0.32 0.44 MPR = medication possession ratio; SD = standard deviation.

96 Steinberg et al.

A secondary multivariate analysis examined the association between adherence and office visits in more detail (table XI). Patients adherent over 3 years tended to have a lower risk of at least one office visit and more than three office visits. Adherence in 2006 was a significant predictor of patients having a lower risk of office visits over all 3 years of the study. Specifically, patients adherent in 2006 had a significantly lower risk (p < 0.05) of at least one and more than three office visits over 3 years (26% and 16% reductions, respectively). Patients adherent in 2007 also experienced a lower risk of office visits in 2007 (10% lower risk, p < 0.05). A mixed result was apparent for pa- tients adherent during 2008, with these patients tending to have a lower risk for at least one office visit during 2008 and over the 3-year study period, and a higher risk (p < 0.05) for more than three office visits during 2008 and over 3 years.

Discussion

Patient adherence to MS drug therapy is a multifaceted phenomenon. Previous studies have focused mostly on the common barriers to ad- herence in MS patients that include adverse drug events, perceived lack of drug efficacy, injection problems, complacency (time in remission), treatment fatigue, depression and other clinical and financial circumstances.[18,42,44] In contrast, this study advances the research on adherence in MS patients beyond a mere identification of barriers by suggesting that adherence affects relapses and healthcare resource utilization.

A major finding of this study was that MS patients adherent to interferon-b therapy had better outcomes, including a lower risk of re- lapses, inpatient admissions, ER visits and office

visits than those who were not adherent. A sig- nificant secondary finding of our study was the lower level of MS patient adherence with inter- feron-b therapy over a 3-year period that, by implication, resulted in unnecessary and poten- tially avoidable healthcare resource utilization among our study population. It is therefore im- portant to explore ways to implement patient adherence programmes for this population. These programmes should, at a minimum, in- clude patient education at the time of the new prescription and ongoing education and moni- toring to assist patients in overcoming any po- tential barriers to adherence. Adherence can be further enhanced by the use of specialty phar- macy services that employ dedicated pharmacists and nurses to increase adherence rates in MS patients.[45]

Another contribution of this study was the use of a fixed-interval adherence calculation method- ology and the application of a clinically optimal cutoff adherence point for interferon-b therapy at 85%. Previous adherence studies of RRMS pa- tients have relied upon an arbitrary adherence cutoff of 80%, and a variable-interval MPR cal- culation. However, as the variable-interval method does not incorporate the broader definition of adherence (i.e. compliance plus persistence), this

Table IX. Association between adherence (MPR ‡85%, fixed-interval method) and emergency room (ER) visits

Year of patient

adherence

ER visits 2006 >0 ER visits 2007 >0 ER visits 2008 >0 ER visits 2006–8 >0

RR (95% CI) RR (95% CI) RR (95% CI) RR (95% CI)

2006 0.73 (0.41, 1.30) 1.13 (0.88, 1.44) 1.08 (0.84, 1.38) 0.96 (0.67, 1.36)

2007 Not relevant 0.94 (0.63, 1.41) 0.78 (0.61, 0.99)* 0.92 (0.66, 1.27)

2008 Not relevant Not relevant 1.17 (0.79, 1.73) 0.92 (0.71, 1.21)

2006–8 Not relevant Not relevant 0.99 (0.78, 1.25) 0.91 (0.67, 1.24)

CI = confidence interval; MPR = medication possession ratio; RR = risk ratio; * p < 0.05.

Table X. Average (– SD) physician office visits in patients adherent (MPR ‡85%, fixed-interval method) or non-adherent to interferon-b therapy (2006–8)

Year Adherent Non-adherent p-Value

2006 2.51 – 3.35 2.69 – 3.23 0.20

2007 2.77 – 3.39 2.91 – 3.62 0.48

2008 2.82 – 3.41 2.48 – 2.94 0.02*

MPR = medication possession ratio; SD = standard deviation; * p < 0.05 by Kruskal-Wallis test.

Adherence to Interferons in Multiple Sclerosis 97

study opted for the more comprehensive fixed- interval method.

The decision to determine an adherence cut- off point for the study population was influenced by current discussions within the pharmacoeco- nomic research community. Recently, the deli- neation of the MPR cutoff point for specific disease states, such as HIV, has been explored by researchers, and the issue has attracted the attention of the International Society for Pharmacoeconomic and Outcomes Research (ISPOR) because of inconsistencies in adherence calculation methodologies.[38] Confirmation of the need to examine adherence by disease state is evi- denced by a study of HIV patients.[46] In this study, researchers concluded that an MPR ‡95% was clinically optimal for viral suppression ther- apy.[34,46-49] The results of this study provide the foundation for ongoing discussion and research to evaluate the methodology to determine a clinically appropriate adherence cutoff point in MS.

Our study has several limitations. First, the insurance claims data lacked potentially relevant variables such as disease progression, EDSS score, therapy expectations, specialty care avail- ability, and adverse drug events. The omission of these factors from the analyses may have ex- plained some of the variations in results across the sample.[18,20] Secondly, determination of the stage of disease progression and disability status among patients was not possible by any means other than proxy measures. Consequently, the relationship between adherence, relapses and healthcare utilization may differ. Lastly, given the older age and higher antidepressant usage in the study population (i.e. commercially insured only), the adherence rates reported here

are presumably not generalizable to all MS patients.[20,50-53]

Conclusions

As demonstrated in this study, interferon-b adherence among RRMS patients follows an unpredictable pattern. Therefore, findings of a positive impact of adherence on RRMS patient relapses and the substantial portion of study pa- tients non-adherent to their prescribed interferon therapies strongly suggest that opportunities ex- ist for improving clinical outcomes and reducing healthcare costs in patients with RRMS.

Acknowledgements

Funding for this study was provided by Merck-Serono. Richard Faris, Stephanie Steinberg and Mark Tankersley are employed by Accredo Health Group, Inc. Dr Cyril Chang is employed by the University of Memphis, and was retained as a consultant for design, analysis and review of this project. Dr Andrew Chan is the Deputy Head of the Department of Neurology at St Josef-Hospital, Ruhr-University Bochum. Dr Chan has served on advisory boards and has received research support and speakers honoraria from Bayer Schering, Biogen Idec, Merck-Serono, Novartis, Sanofi-Aventis and Teva.

References 1. Rudick RA, Goodkin DE, editors. Multiple sclerosis ther-

apeutics. London: Martin Dunitz Ltd, 2000

2. National multiple sclerosis society [online]. Available from URL: http://www.nationalmssociety.org/about-multiple- sclerosis [Accessed 2009 Oct 12]

3. Goodin DS, Frohman EM, Garmany GP, et al. Disease modifying therapies in multiple sclerosis: report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology and the MS Council for Clinical Practice Guidelines. Neurology 2002 Jan; 58 (2): 169-78

Table XI. Association between adherence (MPR ‡85%, fixed-interval method) and physician office visits

Year of patient

adherence

Office visits

2006 >0 Office visits

2007 >0 Office visits

2008 >0 Office visits

2008 >3 Office visits

2006–8 >0 Office visits

2006–8 >3

RR (95% CI) RR (95% CI) RR (95% CI) RR (95% CI) RR (95% CI) RR (95% CI)

2006 0.92 (0.74, 1.13) 1.01 (0.97, 1.06) 0.98 (0.95, 1.00) 0.74 (0.60, 0.94)* 1.00 (0.96, 1.04) 0.84 (0.72, 0.97)*

2007 Not relevant 0.90 (0.81, 0.99)* 0.99 (0.96, 1.01) 0.89 (0.74, 1.07) 0.98 (0.95, 1.02) 0.99 (0.87, 1.13)

2008 Not relevant Not relevant 0.99 (0.95, 1.04) 1.22 (1.01, 1.47)* 0.99 (0.94, 1.04) 1.16 (1.02, 1.32)*

2006–8 Not relevant Not relevant 0.99 (0.96, 1.02) 0.91 (0.77, 1.01) 0.91 (0.81, 1.02) 0.86 (0.65, 1.15)

CI = confidence interval; MPR = medication possession ratio; RR = risk ratio; * p < 0.05.

98 Steinberg et al.

4. Manfredonia F, Pasquali L, Dardano A, et al. Review of the clinical evidence for interferon B 1a (Rebif) in the treat- ment of multiple sclerosis. Neuropsychiatr Dis Treat 2008 Apr 4 (2): 321-36

5. Lublin F, Baier M, Cutter G. Effect of relapses on devel- opment of residual deficit in multiple sclerosis. Neurology 2003; 61: 1528-32

6. Johnson KL, Kuehn CM, Yorkston KM, et al. Patient perspectives on disease-modifying therapy in multiple sclerosis. Int J MS Care 2006; 8: 11-8

7. Treadway K, Cutter G, Salter A, et al. Factors that influence adherence with disease-modifying therapy in MS. J Neurol 2009 April; 256 (4): 568-76

8. Flachenecker P, Rieckmann P. Early intervention in multi- ple sclerosis: better outcomes for patient and society? Drugs 2003; 63: 525-33

9. Comi G. Shifting the paradigm toward earlier treatment of multiple sclerosis with interferon beta. Clin Ther 2009 Jun; 31 (6): 1142-57

10. Jewell S. Issues in adherence: initiating and maintaining patients on immunotherapy. Multiple Sclerosis Pro 2001 [online]. Available from URL: http://www.multiplesclero sispro.com/admin/templates/Template_Nurse_Issues.as px?artic. [Accessed 2009 Oct 18]

11. National Institute of Neurological Disorders and Stroke: 2009 [online]. Available from URL: http://www.ninds.nih. gov/disorders/multiplesclerosis/detail_multiple_sclerosis.htm [Accessed 2009 Nov 10]

12. Kolbelt G, Lindgren P, Smala A, et al. Costs and quality of life in multiple sclerosis: an observational study in Germany. HEPAC 2001; 2: 60-8

13. Pope GC, Urato CJ, Kulas ED, et al. Prevalence, ex- penditures, utilization, and payment for persons with MS in insured populations. Neurology 2002; 58: 37-43

14. Grima DT, Torrance GW, Francis G, et al. Cost and health- related quality of life consequences of multiple sclerosis. Mult Scler 2000; 6 (2): 91-8

15. Miltenburger C, Kobelt G. Quality of life and cost of mul- tiple sclerosis. Clin Neurol Neurosurg 2002; 104 (3): 272-5

16. O’Brien JA, Ward AJ, Patrick AR, et al. Cost of managing an episode of relapse in multiple sclerosis in the United States. BMC Health Serv Res 2003 Sept; 3: 17

17. Portaccio E, Zipoli V, Siracusa G, et al. Long-term ad- herence to interferon b therapy in relapsing-remitting multiple sclerosis. Eur Neurol 2008; 59 (3-4): 131-5

18. Recognizing nonadherence in patients with MS: defining adherence [online]. Available from URL: http://www.medscape. com/view article/579661 3 [Accessed 2009 Aug 2]

19. Vercellino M, Romagnolo A, Mattioda S, et al. Multiple sclerosis relapses: a multivariable analysis of residual disability determinants. Acta Neurol Scand 2009 Feb; 119 (2): 126-30

20. Klauer T, Zette UK. Compliance, adherence, and the treatment of multiple sclerosis. J Neurol 2008 Dec; 255 Suppl. 6: 87-92

21. Tremlett H, Van der Mei I, Pittas F, et al. Adherence to the immunomodulatory drugs for multiple sclerosis: contrast- ing factors affect stopping drug and missing doses. Phar- macoepidemiol Drug Saf 2008 Apr; 17 (6): 565-76

22. Zycynski TM, Coyne KS. Hypertension and current issues in compliance and patient outcomes. Curr Hypertens Rep 2000; 2 (6): 510-4

23. Cramer JA, Benedict A, Muszbeck N, et al. The significance of compliance and persistence in the treatment of diabetes, hypertension and dysplipidaemia: a review. Int J Clin Practice 2008 Jan; 62 (1): 76-87

24. Revel M. Interferon-B in the treatment of relapsing- remitting multiple sclerosis. Pharmacol Therap 2003 Oct; 100 (1): 49-62

25. Portaccio E, Amato MP. Improving compliance with inter- feron-beta therapy in patients with multiple sclerosis. CNS Drugs 2009; 23 (6): 453-62

26. MacLean HJ, Freedman MS. Immunologic therapy for relapsing-remitting multiple sclerosis. Curr Neurol Neuro- sci Rep 2001 May; 1 (3): 277-85

27. McCormack PL, Scott LJ. Interferon-beta-1b: a review of its use in relapsing-remitting and secondary progressive mul- tiple sclerosis. CNS Drugs 2004; 18 (8): 521-46

28. Grymonpre R, Cheang M, Fraser M, et al. Validity of a prescription claims database to estimate medication ad- herence in older persons. Med Care 2006 May; 44 (5): 471-7

29. Fairman K, Motheral B. Evaluating medication adherence: which measure is right for your program? J Manag Care Pharm 2000; 6: 499-506

30. Friedman DS, Quigley HA, Gelb L, et al. Using pharmacy claims data to study adherence to glaucoma medications: methodology and findings of the Glaucoma Adherence and Persistence Study (GAPS). Invest Ophthalmol Vis Sci 2007; 48: 5052-7

31. Johnson RE, Volmer WM. Comparing sources of drug data about the elderly. J Am Geriatr Soc 1991; 39: 1079-84

32. Kirking DM, Ammann MA, Harrington CA. Comparison of medical records and prescription claims files in doc- umenting prescription medication therapy. J Pharmaco- epidemiol 1996; 5 (1): 3-15

33. Lau HS, De Boer A, Beuning KS, et al. Validation of pharmacy records in drug exposure assessment. J Clin Epidemiol 1997; 50 (5): 619-25

34. Tamblyn R, Lavoie G, Petrella L, et al. The use of pre- scription claims databases in pharmacoepidemiological research: the accuracy and comprehensiveness of the pre- scription claims database in Quebec. J Clin Epidemiol 1995 Aug; 48 (8): 999-1009

35. Sikka R, Xia F, Aubert RE. Estimating medication persis- tency using administrative claims data. Am J Manag Care 2005 Jul; 7 (11): 449-57

36. Osterberg L, Blaschke T. Adherence to medication. N Eng J Med 2005 Aug; 132: 487-97

37. Steiner JF, Earnest MA. The language of medication taking. Ann Intern Med 2000; 132: 926-30

38. Peterson AM, Nau DP, Cramer JA, et al. A checklist for medication compliance and persistence using retrospective databases. Value Health 2007 Jan-Feb; 10 (1): 3-12

39. Ruggieri RM, Settipiani N, Viviano L, et al. Long-term interferon-b treatment for multiple sclerosis. Neurol Sci 2003 Dec; 24 (5): 361-4

40. Schumacker GA, Beebe G, Kebler RF, et al. Problems of experimental trials of therapy in multiple sclerosis: report by the panel on the evaluation of experimental trials of therapy in multiple sclerosis. Ann NY Acad Sci 1965; 122: 552-68

Adherence to Interferons in Multiple Sclerosis 99

41. Goldberg P, Fleming MC, Picard EH. Multiple sclerosis: decreased relapse rate through dietary supplementation with calcium, magnesium, and vitamin D. Med Hypotheses 1986 Oct; 21: 193-200

42. Buljevac D, Hop WC, Reedeker AC, et al. Self reported stressful life events and exacerbations in multiple sclerosis: prospective study. BMJ 2003; 327: 646-9

43. Putnam KG, Buist DS, Fishman P, et al. Chronic disease score as a predictor of hospitalization. Epidemiology 2002 May; 13 (3): 340-6

44. Jacobs LD, Cookfair DL, Rudick RA, et al. Intramuscular interferon beta-1a for disease progression in relapsing multiple sclerosis. Ann Neurol 2004 Oct; 39 (3): 285-94

45. Tang J, Faris RJ. Exploring the impact of dispensing channel on medication adherence among multiple sclerosis patients. ISPOR 14th Annual International Meeting; 2009 May 16-20; Orlando (FL)

46. Garcia DO, Knobel H, Carmona A, et al. Impact of ad- herence and highly active antiretroviral therapy on survival in HIV-infected patients. J Acquir Immune Defic Syndr 2002; 30 (1): 105-10

47. Brigido LF, Rodrigues R, Casseb J, et al. Impact of adherence to antiretroviral therapy in HIV-1-infected patients at a university public service in Brazil. AIDS Patient Care STDS 2001 Nov; 15 (11): 587-93

48. DeMasi RA, Graham NM, Tolson JM, et al. Correlation between self-reported adherence to highly active anti- retroviral therapy (HAART) and virologic outcome. Adv Ther 2001; 18 (4): 163-73

49. Gwadry-Sridhar F, Manias E, Zhang Y, et al. A framework for planning and critiquing medication compliance and persistence research using prospective study designs. Clin Ther 2009 Feb; 31 (2): 421-35

50. Daugherty KK, Butler JS, Mattingly M, et al. Factors leading patients to discontinue multiple sclerosis therapies. J Am Pharm Assoc 2005 May-Jun; 45 (3): 371-5

51. Munschauer FE, Tyree J. Persistency in patient with multiple sclerosis. Paper presented at the Annual Meeting of the Consortium of MS Centers (CMSC) 2004 Jun 4; Toronto (ON)

52. Tremlett HL, Oger J. Interrupted therapy: stopping and switching of the beta-interferons prescribed for MS. Neuro- logy 2003; 61: 551-4

53. Kreuger KP, Berger BA, Felkey B. Medication adherence and persistence: a comprehensive review. Adv Ther 2005 Jul-Aug; 22 (4): 313-56

Correspondence: Richard J. Faris, PharmD and PhD, Accredo, a Medco Company, 1670 Century Center Parkway, Memphis, TN 38134, USA.

100 Steinberg et al.

be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.