Adherence to Statin Therapy and LDL Cholesterol Goal Attainment by Patients With Diabetes and Dyslipidemia

This study used information from the ∼6,000 patients enrolled in the diabetes disease management program of a 188,000-member independent practice association model HMO. All statins prescribed were identified from the pharmacy claims database for the 2 years between 1 January 2001 and 31 December 2002. All LDL cholesterol measurements available from these patients over the same period were also included.

Patients without a LDL laboratory value or whose most recent LDL cholesterol measurement was before or within 30 days after the first statin prescription in the database were excluded. The remaining records were used to compute a 9-month MPR, defined as the percentage of days when medication was available, beginning with the first prescription in the database. Adherence calculations were performed with Standardized Therapy Adherence Research Tool (START, Pfizer, New York, NY) software (16). Patients were also excluded if they were first prescribed a statin after 31 March 2002 (to have a complete 9 months of therapy available for analysis). The LDL cholesterol goal for all patients was <100 mg/dl, as specified by NCEP ATPIII (5).

Student’s t tests, χ² tests, and calculation of linear correlation coefficients and 95% CIs were used to evaluate results. A P value <0.05 was required for statistical significance in all analyses.

Among the 653 patients whose records met the inclusion criteria and were analyzed, men had a significantly higher average MPR than women (Table 1). Overall, 44% (n = 290) of patients achieved their LDL cholesterol goal. Once again, the rate of achievement was higher among men.

As the MPR increased, the LDL cholesterol level declined (Fig. 1). A Pearson correlation showed a significant positive association between MPR and LDL (r = −0.393, P < 0.001).

In this study, the probability of achieving the LDL cholesterol goal rose progressively with MPR (Fig. 2). The mean MPR for the 290 patients who reached their LDL cholesterol target (0.82 ± 0.25 [95% CI 0.79−0.85]) was significantly higher (P < 0.05) than that (0.61 ± 0.30 [0.58−0.64]) for the 363 patients who were not at the goal by study end. Further, patients whose MPR was >0.8 had a 56–78% chance of therapeutic success; this probability decreased to 20–42% when the MPR was <0.8.

The choice of a statin appeared to have an influence on the LDL cholesterol level but not on the MPR (Table 2). Atorvastatin, the most frequently prescribed statin in this database (n = 470), produced a mean LDL cholesterol level of 109.07 mg/dl, the lowest level save for 108 mg/dl achieved with lovastatin, which was taken by only two patients. Other mean LDL cholesterol levels ranged from 113.26 mg/dl with simvastatin to 125.05 mg/dl with fluvastatin.

Medical care for the patient with diabetes, especially medication regimens, has become more complex over time, producing a barrier to achieving evidence-based goals of treatment. The number of primary care visits that include the use of oral hypoglycemic, antihypertensive, and lipid-lowering medications in the same patient has significantly increased over the past decade (17). Such polypharmacy is a growing barrier to proper medication adherence and the attainment of treatment goals for all conditions being managed. Increasingly stringent and complicated lipid treatment goals (18,19) and the limited amount of time available during office and clinic visits to discuss the multiple aspects of this care (17) are other significant factors that currently contribute to the low implementation of published guidelines for managing dyslipidemia in diabetic patients. The consequences of poor adherence in the diabetic population are becoming clearer. A recent evaluation showed that an oral hypoglycemic medication MPR of <80% correlated with an increased risk of hospitalization in the following year (22).

Our study using a diabetes disease management database demonstrates that adherence to statin therapy by patients with diabetes and dyslipidemia is an important determinant of LDL cholesterol reduction and goal attainment. Highlighting the relationship between MPR and LDL cholesterol goal attainment was a marked rise in the probability of successful treatment when the MPR was >0.80. Based on these findings, we have taken steps within our disease management program to have case managers assess medication adherence and potential barriers when clinical goals are not being reached. The barriers now monitored and addressed include patient knowledge and perception of therapy benefit (26,29).

The relationship between adherence to therapy and LDL cholesterol levels achieved in this study is consistent with findings from other recent reports. Among 207 consecutive patients with CHD and dyslipidemia, total cholesterol and LDL cholesterol levels were lower and HDL cholesterol levels higher in compliant than in noncompliant patients treated with statins (20).

Poor adherence to statin therapy and failure to achieve LDL cholesterol goals amplify the significant morbidity and mortality already present in patients with dyslipidemia. In one evaluation of 5,590 patients who had an incident myocardial infarction and at least one additional attack, the risks for recurrence and mortality were lower in patients who adhered to statin therapy than in those who did not (23). The proven effectiveness of statins in bringing adherent patients to the LDL cholesterol goal suggests that nonadherence should be strongly suspected when goals are not achieved and that barriers to adherence should be immediately investigated. Among the potential obstacles to adherence are nonwhite race, low income, old age, psychiatric illness (depression, dementia), the occurrence of CHD events during therapy, smoking, comorbid hypertension, and substance abuse, although these have been determined to be inconsistently correlated with adherence (12,24,25). One of the limitations of our study, it should be pointed out, was the absence of more detailed demographic information, which may or may not have provided greater insight into why poor adherence occurred; however, our goal was focused on the relationship between adherence and goal attainment in this population.

The 44% of patients who reached their LDL cholesterol goal in the present analysis is somewhat higher than that reported in previous surveys. Overall, only 38% of patients achieved NCEP ATPIII–specified LDL cholesterol target levels in the Lipid Treatment Assessment Project (L-TAP)—68% of low-risk patients, 37% of high-risk patients, and only 18% of patients with CHD (21). In an analysis of 6,586 patients with diabetes who were members of a single managed care organization, only 15.8% demonstrated LDL cholesterol levels <100 mg/dl (11). The present results must be compared cautiously with those from other studies because of the relatively short (9-month) period in this evaluation and the known decline in adherence to statin therapy over time (12). Possible explanations for the higher LDL goal attainment may include the impact of the disease management program and differences in adherence rates.

Looking forward, attempts to enhance adherence will require a multifaceted, comprehensive, and empathetic approach (25,26). Central to any such efforts are education about the importance of controlling dyslipidemia, clear instructions on dosing, continued interaction between patient and health care professionals, implementation of strategies to overcome high drug costs, adverse effect evaluation and management, and integration of statins with other medications, including use of a daily pill organizer (27). The mutual establishment of clinical goals between the health care provider and the patient, along with frequent follow-up on these goals, are also important for improving adherence (28). We also believe that a stronger message about the benefits of medicine is needed and should be incorporated into patient counseling, because patients who believe in the necessity of a particular medication are more likely to be adherent (29). For example, the patient perception that statins primarily “lower cholesterol” should be replaced with an emphasis on the “prevention of heart attack and stroke.”

Adherence to statin therapy, as reflected by the MPR, is closely related to LDL cholesterol goal attainment by patients with diabetes and dyslipidemia, with the probability of success appearing to increase substantially when the MPR is >0.80. In this database, adherence to therapy and goal achievement is higher among men than among women. This study also demonstrates the usefulness of pharmacy records in identifying patients at increased risk of failure to attain LDL cholesterol goals with statins (by calculation of the MPR). Because outcomes are directly related to patients’ medication-taking behavior, adherence should be the first item assessed when clinical goals (such as serum cholesterol levels) are not being reached.