The purpose of this study is to examine if using genetics can improve statin adherence in patients who should be taking statins but are not because of prior side effects. This study will assist physicians/providers in making a personalized health care plan for prevention of cardiovascular disease.
Hydroxy-methylglutaryl-coenzyme A (HMG Co-A) reductase inhibitors ("statins") are commonly prescribed to lower low density lipoprotein cholesterol (LDLc) and to prevent cardiovascular disease (CVD), a leading cause of morbidity and mortality. Long-term adherence to statins in the primary care environment is challenging; consequences of statin non-adherence include higher LDLc levels, hospitalizations, costs, and death due to CVD. Medication non-adherence is complex and multifactorial and can be associated with a number of factors including medication cost, complexity of medication regimen, poor provider-patient relationship / communication, and adverse side effects. For statins, side effects such as muscle aches, cramping, and pain (referred to broadly as statin-related myopathy) are a frequent cause of non-adherence. These symptoms are non-specific and are frequent reasons for stopping statin therapy, due to patient or provider concern about the possibility of statin-related myopathy. Many patients may be needlessly deprived of the cardiovascular benefits of long-term statin use. A genetic risk factor for statin myopathy and subsequent non-adherence has recently been identified. In a genome-wide association study, a genetic variant (named SLCO1B1\*5) was a main contributor of statin myopathy. It was demonstrated that the SLCO1B1\*5 variant is not only a predictor of myopathy, but also of premature statin discontinuation. The risk with the \*5 allele is statin specific: greatest with simvastatin and atorvastatin use, the least with pravastatin or rosuvastatin. Therefore, the SLCO1B1\*5 variant is common, can predict myopathy, subsequent non-adherence, and due to its statin-specific effects creates a novel research paradigm for personalizing statins to an individual's genetic profile. Carriers of the SLCO1B1\*5 variant may do best on rosuvastatin, pravastatin, or fluvastatin whereas non-carriers may be treated with any statin. The objective of this study is to conduct a randomized trial comparing two strategies: 1. genetically guided statin therapy vs. 2. usual care (i.e., a strategy without genetics) on the effects of statin adherence and LDLc lowering. The overall hypothesis is that genetically guided statin therapy will lead to greater statin adherence and lower LDLc when compared to a non-guided strategy. The design of this trial will randomize primary care patients within Duke University Health System (DUHS) and travis Air Force Base (TAFB) clinics that are nonadherent to statins due to prior side-effects in an unblinded, 1:1 fashion, stratified by SLCO1B1\*5 genotype.
Study Type
INTERVENTIONAL
Allocation
RANDOMIZED
Purpose
PREVENTION
Masking
NONE
Enrollment
167
Genetic testing for SLCO1B1\*5 allele and reporting of results to patient and provider at randomization
Genetic testing for SLCO1B1\*5 allele and reporting of results to patient and provider at end of study
Blood test for SLCO1B1\*5 allele
David Grant US Air Force Medical Center
Travis Air Force Base, California, United States
Duke University
Durham, North Carolina, United States
Morisky Medication Adherence Scale (MMAS) Score
The Morisky Medication Adherence Scale (MMAS) is a self-reported measure of adherence, collected at baseline for general medication and at 3 and 8 months of followup for statin specific adherence. The eight-item MMAS survey will be used. This is a modified version of the original four-item MMAS capturing further aspects of adherence behavior. The survey includes 8 yes/no items that are summed to create an overall adherence score ranging from of 0 to 8, with higher scores indicating better adherence. The primary hypothesis is that the genetically guided statin therapy leads to greater adherence of statin therapy, corresponding to a higher MMAS score.
Time frame: 3 months and 8 months
Low Density Lipoprotein Cholesterol (LDLc) at Baseline, Month 3 and Month 8
The continuous outcomes LDLc will be modeled as a linear regression with arm and baseline LDL as predictors.
Time frame: Baseline, Month 3, Month 8
Medication Possession Ratio (MPR) From Baseline to Last Patient Follow-up
Medication possession ratio will be calculated based on number of statin medication refills over time from randomization to end of follow up. MPR is calculated as follows: 1.Sum of the days' supply of all statin medications is the sum of the number of pills dispensed for each statin prescription during follow up (taken from 3-month, 4-month and 8-month statin utilization review) 2.Sum of the days of follow up = date of 8-month follow up survey - date of randomization 3.MPR = #1/#2 MPR will be modeled as a linear regression with arm, genotype, and site as predictors.
Time frame: Baseline to Last patient follow-up in study (3 months or 8 months)
Number of Participants Reporting New Statin Prescriptions
The number of new prescriptions is binary and will be modeled with logistic regression with arm, genotype, and site as predictors. Any variables imbalanced between arms will also be included as covariates.
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.
Time frame: Baseline, Month 3, Month 8
Brief Pain Inventory (BPI) Score - Pain Severity at Month 3 and Month 8
Brief Pain Inventory data will be taken from 3 and 8-month follow up Patient Surveys. Pain severity and pain interference will be compared between groups. Both of these measures will be modeled as a linear regression with arm, genotype, and site as predictors. Transformations of the response may be explored depending on the distribution of the regression residuals. Baseline pain scores will also be included as a covariate to account for baseline variability. Scores range from 0-10. Higher scores indicate higher pain severity.
Time frame: Month 3 and Month 8
Brief Pain Inventory (BPI) Score - Pain Interference at Month 3 and Month 8
Brief Pain Inventory data will be taken from 3 and 8-month follow up Patient Surveys. -Pain severity and pain interference will be compared between groups -Both of these measures will be modeled as a linear regression with arm as predictor. Baseline pain scores will also be included as a covariate to account for baseline variability. Scores range from 0-10. Higher scores indicate higher pain interference with daily activities.
Time frame: Month 3 and Month 8
Change in Short Form -12 Item (SF-12) Health Survey - Physical Component (PC)
Month 3 and Month 8 SF12 scores for mental and physical health will be compared. Both of these measures will be modeled as a linear regression with arm as predictor. Baseline SF-12 scores will also be included as a covariate to account for baseline variability. Ranges from 0 to 100, where a zero score indicates the lowest level of physical health measured by the scales and 100 indicates the highest level of physical health
Time frame: Baseline, Month 3, Month 8
Change in Short Form -12 Item (SF-12) Health Survey - Mental Component (MC)
Month 3 and Month 8 SF12 scores for mental and physical health will be compared. Both of these measures will be modeled as a linear regression with arm as predictor. Baseline SF-12 scores will also be included as a covariate to account for baseline variability. Ranges from 0 to 100, where a zero score indicates the lowest level of mental health measured by the scales and 100 indicates the highest level of mental health.
Time frame: Baseline, Month 3, Month 8
Physical Activity Scale Score
Activity levels will be compared at the end of 8-months. Activity levels are defined by a five-level ordinal variable (0-4; higher level corresponding to higher activity). which was calculated based on survey answers. An ordinal logistic regression model will be used with arm as predictor. The assumption of proportional odds will be checked, and if it is not met, a multinomial regression model will be used. -Baseline physical activity will also be included as a covariate to account for baseline variability. Scale score (0-4): 0 - Inactivity, 1 - Ligh-intensity activity, 2 - moderate-intensity activity, 3 - Hard-intensity activity, 5 - very hard-intensity activity
Time frame: Baseline and Month 8
Beliefs About Medications (BMQ) Score at Baseline, Month 3 and Month 8
* Questionnaire administered at baseline, 3 months, and 8 months * This instrument assesses beliefs regarding necessity and concerns related to disease-specific medications * The score ranges from 5 to 25 representing the sum of 5 questions. This will be modeled with linear regression including treatment as predictor. Baseline BMQ scores will also be included as a covariate to account for baseline variability. Higher score corresponds to higher thought necessity and higher thought concerns about taking the medication. The higher the necessity score, the more the patient believed statins necessary for their health. The higher the concerns score, the more the patient was concerned about taking stains (side effects).
Time frame: Baseline, Month 3, Month 8