چگونه روشهای خوب رهنمودهایی در تجویز فشار خون مواد مخدر به دنبال دارند؟: نقش بیمه سلامت

Background The US Joint National Committee (JNC) on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure issues guidelines on the optimal first-line drug therapy in treating hypertension. Despite broad dissemination of these guidelines, prescribing practices have long remained discrepant with recommendations. The purpose of this study was to examine the role of insurance type in the selection of drugs for hypertension treatment in light of the JNC guidelines. Methods Subjects were derived from the 1996 Medical Expenditures Panel Survey who had a diagnosis of essential hypertension and who were prescribed a diuretic, beta-blocker, calcium channel blocker (CCB), or ACE inhibitor (ACEI) as monotherapy. Using the nationally representative sample, this study presents the first estimates of the impact of insurance policies on the choice of antihypertensive drugs while controlling for predisposing, enabling, and need variables in the context of a logistic health-care utilization model. Results Nationally in 1996, more than twice as many subjects (7.3 million) were taking ACEIs or CCBs compared to diuretics or beta-blockers (3.1 million) as the first-line drug therapy, a sharp contrast to the JNC guidelines. Patients with health maintenance organization (HMO) insurance were much less likely than fee for service (FFS) patients to follow the JNC guidelines in this respect (odds ratio 0.50, P < .01), controlling for all other factors. Individuals with all other public insurance and no insurance were not statistically different from the FFS group in the use of the study drugs. Other significant factors in the regression model were being of African American descent, being unmarried, having higher out-of-pocket payment, being in excellent physical health, having diabetes, and being diagnosed with essential hypertension after 1988. Each was associated with a decreased likelihood of following the JNC recommendations for the use of diuretics or beta-blockers. Conclusions After controlling for other predisposing, enabling, and need variables, patients who had HMO coverage were significantly more likely than FFS patients to receive ACEIs or CCBs. Given a popular public perception of HMOs being most cost conscious in providing health care, further research is needed to understand why prescribing patterns associated with HMOs have poorly followed the JNC recommendations.

Hypertension is associated with increased risk of stroke, myocardial infarction, atrial fibrillation, heart failure, peripheral vascular disease, and renal disease [1,2]. Hypertension has become a major public health threat to Americans, both medically and economically. About 50 million Americans, or 25% of American adults, have high blood pressure [3]. The total cost of antihypertensive treatment was estimated to be as high as $40 billion in the United States in 1999. The indirect cost of hypertension is even more striking, including the cost of increased mortality and morbidity from coronary heart disease, stroke, and congestive heart failure that occurs in untreated or undertreated hypertensive patients. Drug therapy has played a key role in hypertension control and treatment, with six major classes of antihypertensive drugs available: diuretics, beta-blockers, calcium channel blockers (CCBs), angiotensin-converting enzyme inhibitors (ACEIs), adrenergic inhibitors, and angiotensin II receptor blockers [4,5]. In the United States, the Joint National Committee (JNC) on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure issues a consensus review report every fourto five years that includes stepped therapy guidelines for clinicians in the management and treatment of hypertension. The report, based on the most current clinical evidence at the time, has become a respected standard for the initial evaluation and treatment of hypertension. The history of JNC recommendations for hypertension is presented in Table 1. Since 1977, the JNC reports have consistently recommended diuretics, and since 1984, diuretics or beta-blockers as the first-line drug treatment for uncomplicated hypertension [4,6–10]. Beginning in the early 1980s, when ACEIs and CCBs became available, these two drugs quickly became widely used antihypertensive agents, coupled with diuretics and beta-blockers. Based on the available evidence on similarities in short-term clinical efficacy, the 1988 JNC Report IV added both ACEIs and CCBs to diuretics and beta-blockers to its recommended list of first line therapy drugs [9]. In the 1990s, several studies were carried out that indicated an association of diuretics and betablockers with a decreased risk of cardiovascular morbidity and mortality in clinical trials [11–15]. For example, a prospective, randomized, and double-blind intervention by Dahlof et al. [13] demonstrated a significant effect of three betablockers (atenolol, pindolol, and metoprolol) and two diuretics (hydrochlorothiazide and amiloride) on the reduction of stroke mortality and morbidity among a Swedish population cohort aged 70–84 years, compared to placebo. The Systolic Hypertension in the Elderly Program [11] study demonstrated that the 5-year total incidence of stroke was 5.2 per 100 participants in treatment group with diuretics and beta-blockers compared to 8.2 per 100 for the placebo group, with a relative risk of 0.64 (P = .0003). A meta-analysis of by Psaty and colleagues [15] of 18 long-term randomized trials also showed that diuretics and beta-blockers were more effective compared to the placebo, for major disease end points including the incidence of stroke and congestive heart failure, as well as coronary disease and total mortality. In contrast, there was little clinical trial evidence available indicating comparable longterm outcome effect of ACEIs and CCBs [16]. As a result, the following JNC guidelines in 1993 and 1997 recommended only diuretics and betablockers as the first-line therapy [4,10]. More recently reported trials (since JNC VI) offer affirmative evidence that diuretics and beta-blockers are at least as effective in reducing overall morbidity/ mortality as other agents [17]. In addition, cost data also favored diuretics and beta-blockers over ACEIs and CCBs. A costminimization study by Pearce et al. [18] documented that the cost of medication treatment was much lower with diuretics or beta-blockers than with ACEIs or CCBs. Furthermore, Edelson and colleagues [19] estimated the cost-effectiveness of various initial monotherapies for mild to moderate hypertension for 20 years (1990–2010). Using the Coronary Heart Disease Policy Model, they found that the dollar cost per year of life saved was higher for an ACEI (captopril) and a CCB (nifedipine) than for a diuretic (hydrochlorothiazide) or beta-blocker (propranolol). More specifically, their simulation indicated that the cost per year of life saved was $72,100 for captopril (ACEI) and $31,600 for nifedipine (CCB), compared to $16,400 for hydrochlorothiazide and $10,900 for propranolol hydrochloride. Despite the consistent JNC recommendations and cost advantages of diuretics and beta-blockers, medical practices in the United States have remained a sharp contrast [20–22]. For example, in the most recent study of antihypertensive drugs in retail channels in the United States, the book by Kaplan and Lieberman [5] showed a sustained trend of increased use of ACEIs and CCBs ever since 1986. In the meantime, the use of diuretics and beta-blockers continued to decrease. Beginning in the early 1990s, the total number of prescriptions for CCBs and ACEIs exceeded the number of prescriptions for beta-blockers.

In light of the sharp contrast between prescribing patterns in practice and the JNC guidelines for the treatment of uncomplicated hypertension, this study investigated the insurance impact on the choice of alternative antihypertensive drugs based on a nationally representative sample of population derived from the 1996 MEPS data. After controlling for other observed confounding factors in the context of Anderson health utilization model, this study found that insurance type had a statistically significant impact on the determination of antihypertensive drug choice. Contrary to a popular hypothetical view, patients with HMO insurance had proportionately less use of diuretics and betablockers compared to FFS patients. While further study is needed to explain why it was so, this finding does suggest that insurance coverage is an important determinant of drug choice and use in the treatment of hypertension. A methodologic strength of this study lies in the fact that the MEPS provides representative estimates of health-care use at the national level over time, for a sample of a noninstitutionalized US population, while most previous related studies were restricted to elderly or poor people [30,32]. In addition, we used the multivariate approach to control confounders in the context of Anderson model for antihypertensive drug use. This study has several limitations that should be noted. First, the MEPS design is a stratified sampling based on demographic characteristics; therefore, extrapolation of the results for the institutionalized US adult population such as nursing home patients with essential hypertension should be done with caution. Second, this is a cross-sectional study design. Because the 1996 MEPS data do not contain the time of each prescription, the switch of drug treatment before and in 1996 could not be considered in the study design. The database was not sufficiently large for us to consider initial prescribing decisions occurring during this target year, although we did have available the year of first diagnosis of hypertension. We could not examine switches among the four antihypertensive drugs due to nonresponse or to side effects. Third, there is no information available in the MEPS 1996 about the prescriber (e.g., age, physician type—resident, primary care physician, or specialist), which might influence the prescribing patterns of antihypertensive drugs. In addition, while we tried to control for all important comorbidities related to antihypertensive drug use, some conditions were too few or nonexistent to include such as congestive heart failure and left ventricular hypertrophy. Fourth, the MEPS data set did not provide specifics concerning the nature of the pharmacy benefit, if any. However, our model included four insurance models, including HMO versus FFS. Assuming that HMOs are more likely to have a pharmacy benefit, our model should have captured some part of the pharmacy benefit impact. Finally, owing to the potential complexity of study design, we had to exclude some patients with other medications for essential hypertension while investigating the utilization pattern of the four most commonly used antihypertensive drug classes. Although we did not explicitly examine the differences, we believe that excluded patients might have had more severe hypertension or other related cardiovascular conditions. Further research is needed to discern the underlying reason why an association was found between insurance type and the patterns of antihypertensivedrug use. Another question is whether or not insurance type also affected longer-term costs of care, clinical outcomes, or quality of life of patients with hypertension. Outcomes research based on a national population sample may provide better insight into the underlying reasons for the discrepancy between medical practice and the JNC recommendations. Moreover, such outcomes research data may also offer stronger grounds for the JNC guidelines in an attempt to increase the compliance of both patients and health-care providers, especially in managed care organizations.