Using quantile regression to investigate racial disparities in medication nonadherence
 Mulugeta Gebregziabher^{2, 3},
 Cheryl P Lynch^{1, 2},
 Martina Mueller^{3, 4},
 Gregory E Gilbert^{1},
 Carrae Echols^{1},
 Yumin Zhao^{2} and
 Leonard E Egede^{1, 2}Email author
DOI: 10.1186/147122881188
© Gebregziabher et al; licensee BioMed Central Ltd. 2011
Received: 7 September 2010
Accepted: 6 June 2011
Published: 6 June 2011
Abstract
Background
Many studies have investigated racial/ethnic disparities in medication nonadherence in patients with type 2 diabetes using common measures such as medication possession ratio (MPR) or gaps between refills. All these measures including MPR are quasicontinuous and bounded and their distribution is usually skewed. Analysis of such measures using traditional regression methods that model mean changes in the dependent variable may fail to provide a full picture about differential patterns in nonadherence between groups.
Methods
A retrospective cohort of 11,272 veterans with type 2 diabetes was assembled from Veterans Administration datasets from April 1996 to May 2006. The main outcome measure was MPR with quantile cutoffs Q1Q4 taking values of 0.4, 0.6, 0.8 and 0.9. Quantileregression (QReg) was used to model the association between MPR and race/ethnicity after adjusting for covariates. Comparison was made with commonly used ordinaryleastsquares (OLS) and generalized linear mixed models (GLMM).
Results
Quantileregression showed that NonHispanicBlack (NHB) had statistically significantly lower MPR compared to NonHispanicWhite (NHW) holding all other variables constant across all quantiles with estimates and pvalues given as 3.4% (p = 0.11), 5.4% (p = 0.01), 3.1% (p = 0.001), and 2.00% (p = 0.001) for Q1 to Q4, respectively. Other racial/ethnic groups had lower adherence than NHW only in the lowest quantile (Q1) of about 6.3% (p = 0.003). In contrast, OLS and GLMM only showed differences in mean MPR between NHB and NHW while the mean MPR difference between other racial groups and NHW was not significant.
Conclusion
Quantile regression is recommended for analysis of data that are heterogeneous such that the tails and the central location of the conditional distributions vary differently with the covariates. QReg provides a comprehensive view of the relationships between independent and dependent variables (i.e. not just centrally but also in the tails of the conditional distribution of the dependent variable). Indeed, without performing QReg at different quantiles, an investigator would have no way of assessing whether a difference in these relationships might exist.
Keywords
Medication adherence Quantile regression Diabetes Health disparitiesBackground
Diabetes is a chronic debilitating illness that affects approximately 24 million people in the United States [1]. Medication adherence is an important component of good diabetes care and medication nonadherence is associated with poor glycemic control [2, 3], increased health utilization [4, 5], increased health care costs [6, 7], and increased risk of death [5]. African Americans and other ethnic minority groups have higher prevalence of diabetes and are at increased risk for poor outcomes from diabetes [1]. Multiple recent studies have shown that ethnic minority groups with diabetes have poorer glycemic, lipid, and blood pressure control compared to Whites [8]. There are also data that suggest a correlation between ethnic differences in diabetes outcomes (e.g., glycemic, lipid, and blood pressure control) and ethnic differences in medication adherence [9]. Therefore, medication nonadherence is an important risk factor for poor diabetes outcomes, especially in ethnic minority groups.
Several methods exist to assess medication adherence including patient selfreport, pill counts, physician/nurse report, pharmacy refill data, electronic monitoring, and biological assays [10]. The most commonly used methods use pharmacy refill data and provide reliable estimates of medication adherence [10]. Common methods for assessing medication nonadherence with pharmacy refill data include continuous measure of medication acquisition (CMA), continuous multiple intervals of oversupply (CMOS), medication possession ratio (MPR), and medication refill adherence (MRA), which have all been shown to be identical in terms of measuring adherence to prescription refills over a study period [11].
While the literature on ethnic/racial disparities on medication adherence is scant, some studies using pharmacy refill data from administrative databases have documented ethnic differences in medication adherence among individuals with diabetes [12–14]. However, the magnitude of these racial/ethnic differences is unclear, especially across ranges of medication adherence (e.g. 40% vs. 60% vs. 80%). In addition, it is not clear if the findings of prior studies are reliable given some methodological weaknesses. For example, most prior studies used traditional regression methods that may not be valid if certain assumptions are not satisfied. Some studies used linear regression, which requires the residuals to be normally distributed and homoscedastic [5, 9]. Others have used logistic regression after categorization of the outcome [4, 12, 14], which could lead to arbitrary choice of categories such that results could be sensitive to choice of cutoff values. These methods also may not capture the effect of covariates on the entire distribution of the response variable.
While both linear and logistic regression focus on differences in means associated with covariates, quantile regression allows for studying different directions of the effects of a covariate on different parts of the distribution (lower and upper tails, middle part). Furthermore, quantile regression makes use of the full information of data in contrast to logistic regression, which is usually associated with a loss of information due to transformation of the response MPR into a categorical variable (e.g., binary variable with cutoff at 80%). More importantly, MPR is a quasicontinuous variable that takes on values that are bounded (i.e., have lower and/or upper bounds) and hence traditional methods that use mean changes of the dependent variable with changes in the independent variables may fail to discern differential patterns in nonadherence across racial/ethnic groups. Therefore, the aims of this study were twofold. First, was to examine racial differences in medication nonadherence using quantile regression. Second, was to demonstrate through empirical evidence how choice of a regression method (e.g., QReg, OLS or GLMM) could result in different conclusions for response variables like MPR, which usually have skewed distributions and take on bounded values. We hypothesized that QReg provides estimates of the effect of covariates on the conditional quantiles of MPR, leading to a more complete picture of the differences between race/ethnicity groups over the entire distribution of MPR including the tails and center of the conditional distribution.
Methods
We created a cohort of veterans with type 2 diabetes from a Veterans Administration (VA) facility in the Southeastern United States using multiple patient and administrative files from the Veterans Health Administration (VHA) Decision Support System (DSS) files linked by Social Security Number (SSN). The study period was from April 1996 to May 2006 with an average follow up period of 5.4 years. The datasets were merged, cleaned and then used as the final dataset for analysis. Veterans with type 2 diabetes were identified based on having at least two ICD9 codes for diabetes (250.xx) in either outpatient or inpatient files and having two or more visits each year since diagnosis based on a previously validated algorithm [15]. The datasets were merged to create a subset that only included individuals with complete adherence data, resulting in a cohort of 11,272 veterans with type 2 diabetes, of which 5,307 were nonHispanic White (NHW), 3,061 were nonHispanic Black (NHB), 51 were Hispanic and 1,879 were identified as Other ethnic/racial group. There were also 974 (8.6%) with missing or unknown race/ethnicity information. The study was approved by our institutional review board (IRB) and local VA Research and Development committee.
Outcome Measures
The primary outcome was the mean medication possession ratio (MPR). MPR informs patient medication adherence by providing the ratio of the number of days of medication supplied within a refill interval to the number of days in a specified refill interval [16, 17]. We calculated the number of eligible days per medication within each 90day refill period per patient. We considered supply of insulin and oral hypoglycemic agents (VA classes HS501 and HS502, respectively). The sum of eligible days served as the denominator for the MPR calculation [18]. The average MPR was calculated over the follow up period from 19962006. Prescriptions that became inactive during that time period did not contribute to the MPR calculation. We chose 90day intervals because veterans typically have a 90day of supply of medications mailed to their homes. If the MPR exceeded 100%, it was set to 100%.
Primary Covariate
The primary covariate of interest was race/ethnicity classified as NHW, NHB, and Other (including unknown and missing).
Demographic Variables
We controlled for three demographic variables in addition to the primary covariate. Age at baseline was treated as a continuous variable and centered at its mean value. Marital status was classified as never married, married (reference category), or separated/widowed/divorced. Employment was classified as employed, not employed (reference category), or retired.
Medical Comorbidity
Cancer, congestive heart failure (CHF), coronary heart disease (CHD), hypertension, and stroke were defined based on enhanced ICD9 codes using validated algorithms [19] and coded as 0 or 1 based on presence or absence of history of the disease at baseline.
Psychiatric Comorbidity
Six psychiatric comorbidities including bipolar disorder, generalized anxiety disorder, major depressive disorder, posttraumatic stress disorder, psychotic disorders, and substance use disorder were defined as present (1) or absent (0) at baseline based on enhanced ICD9 codes using validated algorithms [19].
Statistical analysis
First, we examined the characteristics of the sample through univariate analysis. This step was followed by premodel building analysis, which included testing whether each covariate was individually associated with the outcome. To assess whether the relationship between age and MPR was nonlinear, we examined the significance of a quadratic term for age. Next, a final model investigating the association between MPR and race/ethnicity was developed adjusting for all covariates such as demographics, medical comorbidities, and psychiatric comorbidities.
For quantile regression analysis, the response variable, MPR, was defined as the quantile of the mean medication possession ratio for each individual averaged over the study period. The specifications of the unconditional quantiles were made in two different ways: Scenario 1) the quantiles were specified based on clinically meaningful specific MPR cutoff values: Q1 = 0.40, Q2 = 0.60, Q3 = 0.80, Q4 = 0.90 where the values corresponded to the 2^{nd}, 4^{th}, 15^{th} and 27^{th} percentiles of the distribution of MPR and Scenario 2) the quantiles were based on the distribution of MPR values where the 5^{th}, 10^{th}, 15^{th}, 25^{th} and 50^{th} percentiles were considered. These unconditional percentiles corresponded to MPR cutoff values of Q1 = 0.66, Q2 = 0.75, Q3 = 0.80, and Q4 = 0.88 and Q5 = 0.97, respectively.
Quantile regression is used to model the effects of covariates on the conditional quantiles of a response variable [20]. This approach is a robust method that makes no distributional assumption about the error term in a model. It is also robust to extreme points in the response space (outliers) but not to extreme points in the covariate space (leverage points). Confidence intervals for the estimated parameters in QReg are based on inversion of a rank test [21, 22].
Quantile Regression Model
The special case τ = 0.5 is equivalent to median regression. We used the finite smoothing algorithm [23, 24] to compute the solution of this equation so that the NewtonRaphson algorithm could be used iteratively to obtain the solution after a finite number of loops. The regression coefficient at a given quantile (β _{ τ } ) indicates the effect on Y of a unit change in X, assuming that the other factors are fixed.
Both unadjusted and covariate adjusted models were fitted with MPR as the response variable and race/ethnicity as primary variable of interest. Since our sample size is sufficiently large, the final model was adjusted for all covariates including demographic variables such as age, gender, marital status, employment status and medical and psychiatric comorbidities [25]. All models were assessed for goodnessoffit using residual analysis. In addition, QReg was assessed using robust multivariate location and scale estimates for leverage point detection [26].
PROC QUANTREG in SAS 9.2 (SAS Institute Inc., Cary NC) was used to compute the regression models and to conduct statistical inferences on the estimated parameters. Verification for all QReg models was performed using the R [27] quantreg package.
Ordinary Least Squares (OLS)
SAS Proc GLM was used to estimate the parameters of a multiple regression model where the errors for different observations were assumed to be uncorrelated with identical variances (homoscedastic). Under these assumptions, OLS provides estimates of the linear parameters that are unbiased and have minimum variance among linear estimators. Residual plots were used to assess these assumptions but they did not hold true for our data.
Generalized linear Mixed Model (GLMM)
This model extends the above model by allowing a more flexible specification of the covariance matrix of the error terms. In other words, it allows for both correlation and heterogeneous variances, although requires normality assumption [28] which did not hold true for our data. SAS Proc GLIMMIX was used to estimate the parameters of a linear mixed model with a random intercept. This specification allowed different subjects to have different baseline MPR values. The same sets of covariates were used in OLS, GLMM and QReg.
Comparison of statistical methods (QReg, OLS, GLMM)
The second aim was addressed using empirical studies based on resampling of the data with replacement. Traditionally, MonteCarlo simulation studies based on data generated from statistical models have been used for this kind of comparative study. Resampling has the advantage that the data in resampled datasets are based on observations from real patients [29] and thus reflect the appropriate level of diversity and variability found in realistic populations [30, 31]. Sampling with replacement was used since our dataset can be considered large to permit numerous samples of reasonable size to obtain stable conclusions within the smaller samples. Each dataset in the resampling study consisted of 5,000 patients, which represents many of the typical studies that use regional VA data. In order to robustly and accurately estimate the parameters, a total of 10,000 bootstrap replications were performed. The final estimates of the parameters and their standard errors were obtained using means and standard deviations of the 10,000 parameter estimates. Additionally, we computed exact percentiles (e.g., 97.5%; 2.5%) for constructing empirical confidence intervals.
Results
Sample Characteristics by Race and Ethnicity (n = 11,272)
Race/Ethnicity Category  

Variable  All  NHW  NHB  Other 
Age (years mean, sd)  66 (11.6)  68 (10.7)  64 (12.3)  66 (11.7) 
Male  97.3  97.8  96.6  97.2 
Female  2.7  2.2  3.4  2.8 
Married  65.2  67.2  58.3  68.8 
Divorced  28.6  28.1  31.6  26.3 
Never Married  6.2  4.7  10.1  4.8 
Unemployed  48.2  48.8  53.0  42.0 
Retired  30.8  32.7  25.4  33.1 
Employed  20.8  18.4  21.6  24.5 
Cancer  5.0  5.2  7.5  2.1 
CHD  13.9  20.0  12.4  4.6 
CHF  8.0  10.0  9.3  3.1 
Hypertension  25.7  29.2  33.8  10.7 
Stroke  3.0  4.1  3.1  1.0 
HbA1c 8+  11.6  10.0  15.5  10.3 
HbA1c (% mean, sd)  7 (0.9)  7 (0.9)  7 (1)  7 (0.9) 
Bipolar Disorder  1.9  2.2  2.5  0.6 
Generalized Anxiety Disorder  2.2  3.1  1.9  0.7 
Major Depressive Disorder  7.8  8.8  10.5  3.0 
Post Traumatic Stress Disorder  5.1  4.6  7.9  3.0 
Psychotic Disorder  2.4  1.7  4.9  1.1 
Substance Use Disorder  14.4  14.7  21.4  6.4 
Dead  16.4  19.0  16.5  11.7 
MPR (mean, sd)  90.4 (0.2)  91.2 (0.2)  88.7 (0.3)  90.7 (0.3) 
MPR (median, IQR)  97.1 (13)  97.1 (11)  95.0 (16)  100 (11.9) 
Adjusted parameter estimates (β) and pvalues for quantile regression, ordinary leastsquares regression, and the generalized linear mixed model
QReg  OLS  GLMM  

Quantile 1  Quantile 2  Quantile 3  Quantile 4  
Parameter  β  P  β  P  β  P  β  P  β  P  β  P 
Intercept  38.2  0.00  49.2  0.00  72.1  0.00  82.6  0.00  88.2  0.00  89.5  0.00 
NHB  3.42  0.11  5.42  0.01  3.1  0.00  2.0  0.00  1.4  0.00  1.3  0.00 
Other  6.33  0.00  3.75  0.09  0.9  0.34  0.5  0.36  0.1  0.74  0.1  0.70 
NHW (REF)  
Male  11.4  0.04  7.00  0.22  0.4  0.86  1.1  0.42  0.5  0.58  0.4  0.62 
Age  0.24  0.01  0.27  0.00  0.2  0.00  0.1  0.00  0.1  0.00  0.1  0.00 
Agesquared  0.01  0.14  0.01  0.04  0.0  0.11  0.0  0.00  0.0  0.00  0.0  0.00 
Never Married  2.84  0.45  7.62  0.05  0.5  0.79  1.8  0.06  0.9  0.16  0.9  0.13 
Divorced  3.37  0.09  5.35  0.01  2.9  0.00  2.6  0.00  1.4  0.00  1.2  0.00 
Married (REF)  
Employed  5.49  0.02  4.52  0.07  2.8  0.01  1.1  0.06  1.3  0.00  0.6  0.15 
Retired  1.03  0.61  2.04  0.34  1.9  0.04  1.3  0.01  0.6  0.06  0.6  0.04 
Unemployed (REF)  
Cancer  8.24  0.05  13.0  0.00  3.1  0.10  0.4  0.70  1.4  0.05  1.0  0.11 
CHD  14.8  0.00  8.2  0.01  0.0  1.00  0.4  0.65  0.7  0.19  0.5  0.31 
CHF  1.37  0.71  0.3  0.93  1.0  0.54  2.0  0.04  0.9  0.16  0.9  0.08 
Hypertension  0.46  0.86  0.5  0.85  0.8  0.51  1.6  0.02  1.1  0.01  0.9  0.02 
Poor HbA1c  12.2  0.00  7.1  0.01  0.6  0.58  2.7  0.00  1.0  0.02  5.0  0.00 
Stroke  1.60  0.75  4.7  0.37  0.5  0.81  0.7  0.58  0.6  0.47  0.5  0.47 
Bipolar  4.16  0.54  5.1  0.47  2.3  0.45  0.6  0.72  0.2  0.84  0.1  0.93 
GAD  1.53  0.79  1.1  0.86  3.7  0.16  1.6  0.28  0.9  0.36  0.6  0.45 
Psychoses  7.28  0.23  6.8  0.28  3.4  0.21  0.9  0.57  0.0  1.00  0.5  0.53 
PTSD  3.30  0.40  1.6  0.70  2.6  0.14  1.4  0.18  0.7  0.30  0.9  0.12 
Substance Use  7.00  0.01  5.78  0.03  2.6  0.02  1.5  0.02  1.0  0.02  0.8  0.03 
Mean parameter estimates (β) with corresponding 2.5% and 97.5% quantiles from a bootstrap study of 10,000 replications with sample size n = 5000
QReg  

Quantile1  Quantile2  Quantile3  Quantile4  OLS  GLMM  
Parameter  β (95%CI)  β (95%CI)  β (95%CI)  β (95%CI)  β (95%CI)  β (95%CI) 
Intercept  39.83 (14.7,61.3)  48.8(28.4,69.2)  71.5(60.5,81.5)  82.5(76.4,88.8)  88.23(84.4,91.9)  92.21(88.7,95.2) 
NHB  3.54(11.2,3.43)  4.55(10.9,1.7)  2.91(5.6,0.19)  1.87(3.3,0.54)  1.38(2.31,0.46)  1.42(2.10,0.77) 
Other  5.94(14.4,2.2)  4.38(13.2,3.1)  1.05(3.8,1.60)  0.44(1.2,2.1)  0.12(1.10,0.85)  0.65(1.49,0.11) 
NHW (REF)  
Male  11.8(22.8,0.39)  6.78(16.6,3.6)  0.61(4.9,7.8)  1.13(3.2,5.5)  0.51(2.88,2.01)  0.55(2.50,1.50) 
Age  0.24(0.01,0.55)  0.27(0.01,0.52)  0.17(0.05,0.29)  0.13(0.06,0.19)  0.08(0.03,0.12)  0.04(0.01,0.08) 
Age^{2}  0.01(0.02,0.01)  0.01(0.02,0.01)  0.01(0.02,0.002)  0.01(0.01,0.001)  0.01(0.01,0.00)  0.01(0.01,0.001) 
Never Married  3.51(12.9,6.9)  5.69(15.9,6.6)  0.76(5.3,2.9)  2.09(5.1,0.91)  0.88(2.71,0.88)  0.83(0.41,2.12) 
Divorced  3.45(9.7,2.2)  4.67(10.9,1.2)  3.13(5.9,0.57)  2.57(4.0,1.1)  1.44(2.36,0.56)  0.91(0.24,1.59) 
Married (REF)  
Employed  5.10(1,8,13.1)  4.58(2.8,11.4)  2.60(0.17,5.42)  1.16(4.57,2.9)  1.27(0.20,2.33)  0.08(0.85,0.97) 
Retired  0.06(6.9,7.8)  1.42(5.2,7.5)  1.68(0.91,4.36)  1.22(0.15,2.6)  0.63(0.27,1.53)  0.65(1.35,0.00) 
Unemployed (REF)  
Cancer  8.20(19.2,2.7)  11.4(23.5,1.7)  3.14(9.9,2.99)  0.45(3.6,2.2)  1.37(3.45,0.57)  0.45(1.89,0.81) 
CHD  11.8(0.82,21.2)  8.27(0.94,15.6)  0.65(2.6,4.32)  0.20(1.9,2.2)  0.70(0.64,2.04)  0.14(1.04,0.79) 
CHF  0.68(10.4,11.2)  0.29(7.8,7.4)  1.05(4.7,2.46)  2.05(4.7,0.28)  0.87(2.45,0.66)  0.79(1.93,0.27) 
Hypertension (ICD)  1.36(6.7,11.2)  0.89(8.6,6.8)  1.30(4.7,1.54)  1.69(3.5,0.05)  1.13(2.33,0.05)  0.09(0.88,0.67) 
Poor HbA1c  10.2(0.5,18.2)  6.45(0.34,12.0)  0.84(3.6,1.72)  2.59(4.3,0.99)  1.00(2.03,0.02)  4.13(4.89,3.34) 
Stroke  0.76(12.3,10.9)  4.04(13.3,5.9)  1.18(10.3,4.29)  0.48(4.2,2.6)  0.60(2.87,1.55)  0.05(1.47,1.27) 
Bipolar Disorder  3.68(24.3,19.9)  1.57(22.7,18.3)  1.49(4.5,7.1)  0.35(4.9,3.9)  0.24(3.35,2.50)  0.31(1.71,2.06) 
GAD  3.89 (9.5,18.9)  0.50(14.7,13.2)  2.36(5.9,7.9)  1.57(1.5,3.9)  0.88(1.58,3.12)  0.74(0.86,2.36) 
Psychoses  5.90(12.0,22.4)  4.72(12.7, 17.1)  2.38(8.9,4.6)  0.07(6.1,3.7)  0.001(2.7,2.6)  0.65(2.78,1.23) 
PTSD  0.13(11.7,12.5)  0.75(13.2, 9.5)  2.30(3.1,6.5)  1.50(0.72,3.5)  0.68(1.01,2.20)  0.58(0.66,1.86) 
Substance Use  5.65(3.1,15.2)  5.53(2.5,12.7)  2.90(0.001,5.7)  1.41(0.06,2.8)  1.00(0.01,2.00)  0.26(0.48,0.99) 
Discussion
The findings of this study show that the choice of regression methods in the study of nonnormal, semicontinuous and bounded responses can influence whether disparities between different racial groups are uncovered. In this large cohort of Veterans with diabetes, differences in the lower tails of the distribution of MPR by race and comorbidities such as CHD may not have been discovered using OLS or GLMM methods, but were identified using quantile regression. While the regression coefficients of race in both, OLS and GLMM, only indicate the differences in mean MPR (i.e. covariate effect in the central portion of the MPR distribution), the most clinically relevant differences that were found in the tails of the distribution of MPR (those that are low or high in adherence) were only detected through testing of the significance of the regression coefficients in the lower and upper quantiles of the QReg model.
This study used a large cohort of veterans and appropriate statistical methodology permitting a more comprehensive assessment of differences in medication nonadherence by race/ethnicity. Ordinary least squares regression, logistic regression (after categorization) and general linear mixed models assume that covariates affect only the location of the conditional distribution of the response, and not its scale or any other aspect of its distributional shape, while quantile regression has the flexibility for modeling of data with heterogeneous conditional distributions. QReg provides a complete picture of the covariate effect when a set of percentiles is modeled, and thus offers the capability to capture important features of the data possibly missed by models that average over the conditional distribution. One other recent approach that might be able to capture the effect of covariates on the entire density of MPR is Bayesian density regression (BDR) [32, 33]. Like QReg, BDR avoids the assumption of normality and linearity. However, this approach is not as easy to understand and implement as QReg. Other approaches include Quasilikelihood [32], BoxCox transformation to normality [33] and robust regression [34, 35]. However, each of these methods has its own limitations [30].
Research on medication adherence patterns has consistently shown greater nonadherence to antihyperglycemic agents among NHB with type 2 diabetes compared to NHW [5, 9, 12–14]. Consistent with prior studies, this study found that NHB were more likely to be medication nonadherent across each of the quantiles. Potential reasons for the difference in medication adherence by race/ethnicity group have been studied and seem to suggest that Blacks express more concern about drug side effects [36], medication dependency, reduced quality of life [37], and issues related to cost of medications [7, 36, 38–40]. For example, among an insured cohort with pharmacy benefits, an increased patient cost share of $5/month led to a 15% decrease in the odds of medication adherence and worsened glycemic control [38]. However, in the VA system where cost of medications is less of an issue because copays are very low, other factors beyond cost of medications are likely to explain the observed differences. Potential explanatory factors that were not available in our dataset include patientlevel factors such as health literacy, numeracy, selfefficacy, cultural beliefs and attitudes about medications, and social support. The contribution of these and other factors need to be explored in future studies.
Despite the strengths of our data and methodology, there were limitations that need mentioning. The dataset did not include information to determine the duration of diabetes as a way to distinguish between new and regular users of diabetes medication, thus, we were not able to assess its impact on medication adherence rates. However, we created a 'new users' group who did not use medication within the first year of the study and their proportions were not different from the overall sample proportion either by race or other demographic factors (see additional file 1, tables S5). Due to the age and gender distribution of our sample, our results should be interpreted with caution in women and younger aged individuals. In addition, our findings could have been biased by the 8.6% of veterans with missing race data. While we believe that the unreported race information is missing at random, we also performed a sensitivity analysis via multiple imputation and found that the results were not different from what is reported in this paper. While the conclusions are mainly applicable to skewed and bounded outcomes from cross sectional studies, the message is easily transferable to the analysis of longitudinal skewed and bounded outcomes via longitudinal quantile regression.
Conclusions
In conclusion, quantile regression allowed modeling the differential patterns in medication adherence between the racial/ethnic groups that would have been missed using traditional regression methods. QReg is a very useful tool for data that are heterogeneous in the sense that the tails and the central location of the conditional distributions vary differently with the covariates. Indeed, without performing quantile regression at different quantiles, an investigator would be unable to assess whether there might be a difference in these relationships. This method is also robust as it makes no distributional assumption about the error term in the model. Future studies need to be cautious when using traditional regression methods in modeling quasicontinuous and bounded outcome such as MPR.
Conflict of interests
The authors declare that they have no competing interests.
Abbreviations
 HbA1c:

Hemoglobin A1c
 VA:

Veterans Administration
 CVD:

Cardiovascular disease
 CHD:

Coronary heart Disease
 CHF:

Congestive Heart Failure
 MDD:

Major Depressive Disorder
 ICD9:

International Classification of Diseases, Ninth Revision
 VHA:

Veterans Health Administration
 DSS:

Decision Support System
 SSN:

Social Security Number
 DRG:

Diagnostic Related Group
 IRB:

Institutional review board
 CI:

Confidence Interval
 VADT:

Veterans Affairs Diabetes Trial
 MPR:

Medication Possession Ratio
 GAP:

Gap between refills
 CMA:

Continuous Measure Of Medication Acquisition
 CMOS:

Continuous Multiple Interval Of Oversupply
 MRA:

Medication Refill Adherence
 NIDDK:

National Institute of Diabetes and Digestive and Kidney Diseases
 OLS:

Ordinary Least Squares
 GLMM:

General Linear Mixed Model
 QReg:

Quantile Regression
 NHB:

Non Hispanic Black
 NHW:

Non Hispanic White
Declarations
Acknowledgements
This study was supported by Grant # REA 08261, Center for Disease Prevention and Health Interventions for Diverse Populations funded by Veterans Affairs Health Services Research and Development (PI  Leonard Egede).
Authors’ Affiliations
References
 Centers for Disease Control and Prevention: National diabetes fact sheet: general information and national estimates on diabetes in the United States, 2007. 2008, Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and PreventionGoogle Scholar
 Rozenfeld Y, Hunt JS, Plauschinat C, Wong KS: Oral antidiabetic medication adherence and glycemic control in managed care. Am J Manag Care. 2008, 14 (2): 7175.PubMedGoogle Scholar
 Pladevall M, Williams LK, Potts LA, Divine G, Xi H, Lafata JE: Clinical outcomes and adherence to medications measured by claims data in patients with diabetes. Diabetes Care. 2004, 27 (12): 28002805. 10.2337/diacare.27.12.2800.View ArticlePubMedPubMed CentralGoogle Scholar
 Lau DT, Nau DP: Oral antihyperglycemic medication nonadherence and subsequent hospitalization among individuals with type 2 diabetes. Diabetes Care. 2004, 27 (9): 21492153. 10.2337/diacare.27.9.2149.View ArticlePubMedGoogle Scholar
 Ho PM, Rumsfeld JS, Masoudi FA, et al: Effect of medication nonadherence on hospitalization and mortality among patients with diabetes mellitus. Arch Intern Med. 2006, 166 (17): 18361841. 10.1001/archinte.166.17.1836.View ArticlePubMedGoogle Scholar
 Balkrishnan R, Rajagopalan R, Camacho FT, Huston SA, Murray FT, Anderson RT: Predictors of medication adherence and associated health care costs in an older population with type 2 diabetes mellitus: a longitudinal cohort study. Clin Ther. 2003, 25 (11): 29582971. 10.1016/S01492918(03)803478.View ArticlePubMedGoogle Scholar
 Lee WC, Balu S, Cobden D, Joshi AV, Pashos CL: Prevalence and economic consequences of medication adherence in diabetes: a systematic literature review. Manag Care Interface. 2006, 19 (7): 3141.PubMedGoogle Scholar
 Kirk JK, D'Agostino RB, Bell RA, et al: Disparities in HbA1c levels between AfricanAmerican and nonHispanic white adults with diabetes: a metaanalysis. Diabetes Care. 2006, 29 (9): 21302136. 10.2337/dc051973.View ArticlePubMedPubMed CentralGoogle Scholar
 Adams AS, Trinacty CM, Zhang F, et al: Medication adherence and racial differences in A1C control. Diabetes Care. 2008, 31 (5): 916921. 10.2337/dc071924.View ArticlePubMedPubMed CentralGoogle Scholar
 Farmer KC: Methods for measuring and monitoring medication regimen adherence in clinical trials and clinical practice. Clin Ther. 1999, 21 (6): 10741090. 10.1016/S01492918(99)800265. discussion 1073View ArticlePubMedGoogle Scholar
 Hess LM, Raebel MA, Conner DA, Malone DC: Measurment of adherence in pharmacy administrative databases: a proposal for standard defnitions and preferred measures. Annals of Pharmacotherapy. 2006, 40: 12801287. 10.1345/aph.1H018.View ArticlePubMedGoogle Scholar
 Shenolikar RA, Balkrishnan R, Camacho FT, Whitmire JT, Anderson RT: Race and medication adherence in Medicaid enrollees with type2 diabetes. J Natl Med Assoc. 2006, 98 (7): 10711077.PubMedPubMed CentralGoogle Scholar
 Hertz RP, Unger AN, Lustik MB: Adherence with pharmacotherapy for type 2 diabetes: a retrospective cohort study of adults with employersponsored health insurance. Clin Ther. 2005, 27 (7): 10641073. 10.1016/j.clinthera.2005.07.009.View ArticlePubMedGoogle Scholar
 Yang Y, Thumula V, Pace PF, Banahan BF, Wilkin NE, Lobb WB: Predictors of medication nonadherence among patients with diabetes in Medicare Part D programs. A retrospective cohort study. Clin Ther. 2009, 31 (10): 21782188. 10.1016/j.clinthera.2009.10.002.View ArticlePubMedGoogle Scholar
 Miller DR, Safford MM, Pogach LM: Who has diabetes? Best estimates of diabetes prevalence in the Department of Veterans Affairs based on computerized patient data. Diabetes Care. 2004, 27 (Suppl 2): B10B21.View ArticlePubMedGoogle Scholar
 Karve S, Cleves MA, Helm M, Hudson TJ, West DS, Martin BC: An empirical basis for standardizing adherence measures derived from administrative claims data among diabetic patients. Med Care. 2008, 46 (11): 11251133. 10.1097/MLR.0b013e31817924d2.View ArticlePubMedGoogle Scholar
 Peterson AM, Nau DP, Cramer JA, Benner J, GwadrySridhar F, Nichol M: A checklist for medication compliance and persistence studies using retrospective databases. Value Health. 2007, 10 (1): 312. 10.1111/j.15244733.2006.00139.x.View ArticlePubMedGoogle Scholar
 Scott Leslie R, GwadrySridhar F, Thiebaud P, Patel B: Calculating medication compliance, adherence and persistence in administrative pharmacy claims databases. Pharmaceutical Programming. 2008, 1: 1319. 10.1179/175709208X334614.View ArticleGoogle Scholar
 Quan H, Sundararajan V, Halfon P, et al: Coding algorithms for defining comorbidities in ICD9CM and ICD10 administrative data. Med Care. 2005, 43 (11): 11301139. 10.1097/01.mlr.0000182534.19832.83.View ArticlePubMedGoogle Scholar
 Koenker RW: Quantile Regression. 2005, Cambridge Univ PressView ArticleGoogle Scholar
 Koenker RW: Confidence intervals for regression quantiles. Asymptotic Statistics, Proceedings of the Fifth Prague Symposium. Edited by: Mandl P, Hu skov'a M. 1994, Springer, Heidelberg, 34959.Google Scholar
 Hao L, Naiman DQ: Quantile Regression. 2007, Sage Publication IncView ArticleGoogle Scholar
 Chen C: An adaptive algorithm for quantile regression. Theory and applications of recent robust methods. Edited by: Hubert M, Pison G, Struyf A, Van Aelst S. 2004, Series: Statistics for Industry and Technology, Birkhauser, Basel, 3948.View ArticleGoogle Scholar
 Madsen K, Nielsen HB: A Finite Smoothing Algorithm for Linear Estimation. SIAM Journal on Optimization. 1993, 3: 223235. 10.1137/0803010.View ArticleGoogle Scholar
 Harrell FE: Regression Modeling Strategies. 2001, New York: SpringerView ArticleGoogle Scholar
 Rousseeuw PJ, Van Driessen KA: Fast Algorithm for the Minimum Covariance Determinant Estimator. Technometrics. 1999, 41: 212223.View ArticleGoogle Scholar
 Koenker R: quantreg: Quantile Regression. R package version 4.44. 2009, [http://CRAN.Rproject.org/package=quantreg]Google Scholar
 Diggle P, Liang K, Zeger S: Analysis of longitudinal data. 2002, New York: Oxford University Press, 25: 2Google Scholar
 Rubin DB: Multiple Imputation for Nonresponse in Surveys. 2004, New York: John Wiley and SonsGoogle Scholar
 Royston P, Altman DG: Regression using fractional polynomials of continuous covariates: parsimonious parametric modelling. Journal of the Royal Statistical Society Series CApplied Statistics. 1994, 43 (3): 429467.Google Scholar
 Marshall A, Altman D, Holder R: Comparison of imputation methods for handling missing covariate data when fitting Coxproportional hazards model: a resampling study. BMC Medical Research Methodology. 2010, 10 (1): 11210.1186/1471228810112.View ArticlePubMedPubMed CentralGoogle Scholar
 Dunson DB: Empirical Bayes density regression. Statistica Sinica. 2007, 17: 481504.Google Scholar
 Dunson DB, Pillai NS, Park JH: Bayesian density regression. Journal of the Royal Statistical Society B. 2007, 69: 163183. 10.1111/j.14679868.2007.00582.x.View ArticleGoogle Scholar
 McCullagh P, Nelder JA: Generalized Linear Models. 1989, London: Chapman and Hall, 2View ArticleGoogle Scholar
 Box George EP, Cox DR: An analysis of transformations. Journal of the Royal Statistical Society, Series B. 1964, 26: 211252.c.Google Scholar
 Holland P, Welsch R: Robust Regression Using Interactively Reweighted LeastSquares. Commun Statist Theor Meth. 1977, 6: 813827. 10.1080/03610927708827533.View ArticleGoogle Scholar
 Chen C: Robust Regression and Outlier Detection with the ROBUSTREG Procedure. Proceedings of the Twentyseventh Annual SAS Users Group International Conference. 2002, Cary, NC: SAS Institute IncGoogle Scholar
 Aikens JE, Piette JD: Diabetic patients' medication underuse, illness outcomes, and beliefs about antihyperglycemic and antihypertensive treatments. Diabetes Care. 2009, 32 (1): 1924. 10.2337/dc081533.View ArticlePubMedPubMed CentralGoogle Scholar
 Huang ES, Brown SE, Thakur N, et al: Racial/ethnic differences in concerns about current and future medications among patients with type 2 diabetes. Diabetes Care. 2009, 32 (2): 311316.View ArticlePubMedPubMed CentralGoogle Scholar
 Kurlander JE, Kerr EA, Krein S, Heisler M, Piette JD: Costrelated nonadherence to medications among patients with diabetes and chronic pain: factors beyond finances. Diabetes Care. 2009, 32 (12): 21432148. 10.2337/dc091059.View ArticlePubMedPubMed CentralGoogle Scholar
 The prepublication history for this paper can be accessed here:http://www.biomedcentral.com/14712288/11/88/prepub
Prepublication history
Copyright
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.