Managed Care - February 2009 - (Page 42) TABLE 2 Simple linear regression models of HEDIS score with analog insulin, human insulin, branded oral, and generic oral prescription share Correlation with HEDIS 2005 (N = 194)* −0.22 −0.09 0.22 0.07 Correlation with HEDIS 2006 (N = 211)* −0.30 −0.10 0.326 0.12 Drug Class Analog insulin share Human insulin share OAD share (brand ) OAD share (generic) Model coefficient -0.319 -0.218 0.378 0.910 P value† 0.002 0.197 0.002 0.321 Model coefficient −0.424 −0.257 0.573 0.180 P value† < .001 0.128 9 percent) HEDIS scores and higher oral product prescription share. This unfavorable correlation was statistically significant in the case of branded oral product prescription share in 2005 and 2006 (p = 0.002 and p<0.001, respectively). The unfavorable correlation was not significant for generic oral agents in either year. To further compare the effect of product group prescription share on HEDIS score, a multivariate regression was conducted for each year with analog insulin share, human insulin share, and branded oral share as independent variables (Table 3). Since the four categories total 100 percent, the generic oral value is completely dependent on the values of the other three shares and was dropped from the model. The multivariate regression clearly discriminated between the effect of analog and human insulin prescription share, with higher analog insulin utilization continuing to be statistically significantly associated in both 2005 and 2006 (p = 0.004 and p = 0.013, respectively) with better HEDIS scores. Higher human insulin utilization was not associated with better HEDIS scores. Branded oral share continued to be associated with poorer HEDIS scores at a statistically significant level (p = 0.011 and p = 0.002 in 2005 and 2006, respectively). Evaluation of plan enrollment and location in the multiple regression analysis revealed that only two U.S. regions reached statistical significance (South Atlantic/Central and Mountain) with both regions unfavorably correlated to HEDIS scores (Table 4). However the effects were small compared to the product prescription shares. The inclusion of plan enroll- TABLE 3 Multivariate regression analysis of HEDIS score with analog insulin, human insulin, and branded oral prescription share HEDIS 2005 model coefficient (N = 194)* −0.302 0.010 0.363 HEDIS 2006 model coefficient (N = 211)* −0.258 0.156 0.492 Drug class Analog insulin share Human insulin share OAD share (brand ) P value† 0.004 0.961 0.011 P value† 0.013 0.427 0.002 *A negative model coefficient indicates a favorable correlation between HbA1c control and prescription share. Model statistics: r2=0.091, p<0.001, F-test (year 2005); r2=0.141, p<0.001, F-test (year 2006). † t test OAD = oral antidiabetic drug 42 MANAGED CARE / FEBRUARY 2009
For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.