Antihypertensive Drug Effects on Renal Function and Myocardial Infarction and Implications of the ALLHAT Study Results

In this issue of the Journal, Grissinger, Globus, and Fricker, from the Institute of Safe Medication Practices (ISMP), focus on the patient-practitioner interaction as a primary opportunity to reduce medication errors. This is also the principal focus of a campaign launched by the U.S. Food and Drug Administration (FDA) in August 2002 that had the theme: “Think through the risks and benefits of medicines.” This consumer-oriented education campaign urged patients to ask questions of their physicians and pharmacists and to become more active in the process of assessing the benefits and risk of prescription drugs: “before using any medicine—as with many things that you do every day—you should think through the benefits and the risks in order to make the best choice for you.” Managed care pharmacy can effectively use the same metaphor: when driving a car, you wear your seat belt; when taking medications, you talk to your pharmacist and physician. Managed care pharmacy can also have a measurable effect on the medication errors and ADEs through reliance on the principles of continuous quality improvement, a fundamental, core area of the Academy of Managed Care Pharmacy’s “Pharmacy’s Framework for Drug Therapy Management in the 21st Century.” The framework’s self-assessment tool contains specific tasks and components within key functional areas that permit individual, organization, and system analysis of opportunities for quality improvement in drug therapy management. Key functional areas in patient safety and reduction of medication errors and ADEs are interspersed throughout the framework, with emphasis on patient-practitioner interaction in interpersonal communication (area 1.1), patient education (1.3, 4.4, 5.3, and 6.3), patient and worker safety (1.4), drug selection (3.2), coordination of care (3.5), etc. Many of the solutions necessary to prevent and reduce medical errors and medication errors will involve changes in processes and systems that “make it easy to do it right.”

improvement in drug therapy management. Key functional areas in patient safety and reduction of medication errors and ADEs are interspersed throughout the framework, with emphasis on patient-practitioner interaction in interpersonal communication (area 1.1), patient education (1. 3, 4.4, 5.3, and 6.3), patient and worker safety (1.4), drug selection (3.2), coordination of care (3.5), etc. 43 Many of the solutions necessary to prevent and reduce medical errors and medication errors will involve changes in processes and systems that "make it easy to do it right." 44 ■■ Quality of Health Economic Studies (QHES)-Tool or Mask? Managed care pharmacists face a mountain of data when making decisions about the relative value of alternate drug therapies in individual patients and in the selection of preferred agents in prescription drug formularies. The objective is to apply rules of evidence-based medicine to derive the information that will be important to develop and refine clinical practice guidelines (CPGs) and clinical practice models (CPMs) that will make it possible to achieve effective disease management. This paradigm might be made more clear by thinking of this continuum in terms of structure, process, and outcome in which the "structure" derives from evidence-based medicine, the "process" from application of CPGs and CPMs, and the outcome as successful attainment of disease management.
The U.S. National Library of Medicine reported an average 10,000 new lines (articles) referenced in MEDLINE each week at year-end 2001. 45 The amount of data and information in the medical literature is growing further and is now quite easily overwhelming, setting aside the additional data and information disseminated in the lay press and on the Internet. It is now more important than ever to find tools to help filter and interpret enormous amounts of data and thousands of medical literature references. The Quality of Health Economics (QHES) instrument may be such a tool. On the other hand, this tool, as any tool, can be misapplied. In addition to possible inherent flaws in the instrument, some of which will only be discovered upon repeated use and scrutiny of the results, users of the QHES have the potential to distort the findings of the studies that they are measuring.
Ofman, Sullivan, Neumann, et al. in this issue of the Journal, take the bold step of introducing a new instrument, the QHES. 46 The true value of this instrument and method will not be determined immediately, and readers have reason to be critical. Shaya and Lyles in an accompanying editorial suggest that managed care pharmacy should evaluate critically this new instrument and method. 47 Motheral argues for caution and even rejection of the instrument and method. 48 Science advances through scholarly debate. By articulating and applying the QHES, Ofman et al. permit others to critique the instrument and method and to propose changes that will enhance value by increasing its validity, reliability, and usefulness. We are certainly in need of better tools to evaluate published data.
The QHES has value, perhaps not so much for its final numeric "score," but in its qualitative analysis of the results of assessment of individual items in the 16-item instrument. Some researchers will no doubt want to change the weight of individual items to improve the utility of the QHES in application to particular uses. The architects of the QHES will need to explain for other researchers that the 3 compound items (numbers 5, 8 and 11) in the QHES require affirmation of both questions. Item number 5, "Was uncertainty handled by: 1) statistical analysis to address random events; 2) sensitivity analysis to cover a range of assumptions?" could have one "yes" and one "no" answer, yet the weight for the item is "9." Item number 8 asks 2 questions and has a weight of "7": "Did the analytic horizon allow time for all relevant and important outcomes?" "Were benefits and costs that went beyond 1 year discounted (3% to 5%) and justification given for the discount rate?" Item no. 11 also has 2 questions, with one score of "7": "Were the health outcome measures/scales valid and reliable? If previously tested valid and reliable measures were not available, was justification given for the measures/scales used?" At the least, the QHES instrument and method present us with a useful platform for scholarly debate on attempts to bring more science to pharmacoeconomics, a relatively young discipline still in search of credibility in the scientific community.

■■ Antihypertensive Drug Effects on Renal Function and Myocardial Infarction and Implications of the ALLHAT Study Results
In managed care we must study the effects of drugs outside the realm of randomized clinical trials (RCTs) to ascertain their real value in uncontrolled, real-world settings. This is particularly true in the ubiquitous treatment of hypertension. One area of particular interest is in the growing body of evidence that some drugs have renal protective effects in excess of their hemodynamic effects in blood pressure reduction. A randomized trial involving 1,094 African Americans aged 18 to 70 years followed for a minimum period of 3 years and up to 6.4 years found that ramipril, an angiotensin converting enzyme inhibitor (ACEI), in a dose range of 2.5 mg up to 10 mg per day, appeared to be more effective than beta-blockers (metoprolol dosed between 50 mg and 200 mg per day) or dihydropyridine calcium channel blockers (amlodipine dosed between 5 mg and 10 mg per day) in slowing the decline in glomerular filtration rate (GFR), an important indicator of kidney function. 49 The ACEI was associated with risk reduction of 22% in the composite outcome (reduction in GFR by 50% or more [or greater than or equal to 5 mL/min per 1.73 m 2 ]) from baseline, endstage renal disease or death), compared to metoprolol or amlodipine; there was no significant difference in the clinical composite outcome between the amlodipine and metoprolol groups; and there was no apparent additional benefit in slowing progression of hypertensive nephrosclerosis associated with a lower blood pressure goal of 92 mm Hg or less, compared to the usual blood pressure goal of 102 mm Hg to 107 mm Hg.
Nephropathy, defined as proteinuria greater than 300 mg in 24 hours, will develop in 35% of patients with type 1 diabetes, usually manifested first as persistent microalbuminuria that appears 5 to 10 years after the onset of diabetes. Nephropathy will progress to end-stage renal failure. Drug therapy that lowers blood pressure and protects diabetics from development of nephropathy is obviously important in reducing morbidity and mortality. Ten years ago, captopril 25 mg 3 times a day was found to reduce by 50% the risk of the combined endpoint of death, dialysis, and transplantation compared to placebo in patients with type 1 diabetes. 50 Since then, lisinopril dosed at 10 mg to 20 mg per day was found to significantly reduce albumin excretion and microalbuminuria in normotensive type 1 diabetics, 51 thereby demonstrating the utility of ACEIs in the prevention of diabetic nephropathy as well as in its treatment. Meta-regression analysis has shown that ACEIs can decrease proteinuria and preserve GFR in patients with diabetes mellitus. 52 The MICRO-HOPE trial of more than 3,500 diabetic patients from the HOPE trial (97% of whom had type 2 diabetes) showed a 24% reduction in risk of nephropathy in an average 4.5 years of follow-up in patients who received the ACEI, ramipril. 53 There was a significantly lower albumin-creatinine ratio in the ramipril group, and these effects were greater than could be attributed to reduction in blood pressure alone.
The dihydropyridine (DHP) calcium channel blockers (CCBs) have not been shown to have comparable effects in protection from nephropathy in either diabetic or non-diabetic patients. In the Irbesartan Diabetes Nephropathy Trial, the DHP CCB, amlodipine, at 10 mg per day appeared to fare worse than placebo in the composite endpoint of time to doubling of baseline serum creatinine, development of ESRD or death from any cause, in type 2 diabetics with hypertension. 54 The effects of CCBs on risk of MI in patients with or without diabetes warrant further study. While short-acting and intermediate-acting DHPs may be associated with an increased risk of myocardial infarction (MI) 55-56 this relationship has not been found with the long-acting DHPs, 57 but experts have requested caution in the use of CCBs in treating hypertension, 58 particularly in diabetics. 59 The researchers at Wake Forest University School of Medicine who found in 1995 that short-acting calcium channel blockers may cause more harm than benefit, presented at the European Society of Cardiology in Amsterdam in August 2000 the results of a meta-analysis of 9 RCTs that compared outcomes of the use of calcium channel blockers versus diuretics, ACE inhibitors or beta-blockers for hypertension. 60 The pooled data showed a 27% higher risk of heart attack and a 26% higher risk of heart failure in patients on calcium channel blockers versus alternative therapies for hypertension: diuretics, ACEIs, or beta-blockers. 61 Overall survival was not significantly different among the alternative therapies.
Anderson, Alabi, Kelly, Diseker, and Roblin, in this issue of the Journal, failed to find a higher risk of MI in high-risk diabetic patients prescribed a combination antihypertensive drug regimen that included a DHP CCB or a nondihydropyridine CCB. 62 The well-conceived study design was undermined by a small sample size and missing data for variables such as ethnicity, smoking status, vital signs, and laboratory tests. Nevertheless, this research is useful in articulating a study design that may be used by others in managed care pharmacy to examine retrospective databases to ferret out the relationships of combination drug therapies in the development of adverse myocardial and renal outcomes in high-risk patients. Future research should also address the important question regarding dose-related effects of CCBs in combination with ACEIs or diuretics.
The impetus to perform more clinical studies of the relative value of ACEIs and CCBs in the treatment of hypertension increased in light of the results of the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALL-HAT), released in December 2002. It is hard to overstate the importance and potential impact of the results of the ALLHAT study. These results suggest that most of the more than 40 million Americans with hypertension 63 could be treated more effectively and more safely with low-cost diuretics such as chlorthalidone and hydrochlorothiazide, at a fraction of the cost of ACEIs such as ramipril or quinapril and CCBs such as amlodipine, felodipine and long-acting nifedipine and diltiazem. 64 The 8-year study, with a mean follow-up of 4.9 years, produced unequivocal evidence that amlodipine and lisinopril were associated with the same incidence of the primary outcome of combined fatal coronary heart disease or nonfatal MI as chlorthalidone; all-cause mortality was also the same among the three treatment groups. However, chlorthalidone was superior to amlodipine in the 6-year rate of heart failure (HF), 7.7% vs. 10.2%, relative risk 1.38, and chlorthalidone was superior to lisinopril in the 6-year rates of combined cardiovascular disease, 30.9% versus 33.3%, RR 1.10, stroke (5.6% versus 6.3%, RR 1.15) and HF (8.7% versus 7.7%, RR 1.19). Robert Anderson, one of the ALLHAT investigators, has additional observations regarding the ALLHAT study findings in this issue of the Journal. 65

■■ Prior Authorization to Manage Drug Utilization and Costs
In 1998, state Medicaid programs were responding in various ways to double-digit increases in prescription drug benefit costs. In some states such as Massachusetts, the cost increases threatened to push drug benefit costs ahead of the costs of hospital acute care services by the year 2003. Massachusetts Medicaid planned to implement "more aggressive utilization practices," Kentucky required prescribers to use lower-cost first-generation antihistamines and Florida used prescriber profiling, focusing on the use of "fourth-generation" antibiotics for upper respiratory infection in adults. 66 Florida also planned "provider education forums" to focus on drug prices. Drug formularies were not an option since state Medicaid programs have essentially open formularies because nearly all drug manufacturers pay rebates to obtain formulary status, leaving only prior authorization (PA) programs to discourage use of certain high-cost drugs. Medicaid rebates for single source and innovator multiple-source brand drugs in 1998 was the greater of 15.1% of average manufacturer price or AMP less best price. For noninnovator drugs, the rebate was 11%.
By 1999 and 2000, state Medicaid officials began to complain more openly about requirement of state programs to operate open drug formularies as part of the OBRA 1990 statute on mandatory drug manufacturer rebates. Oklahoma Health Care Authority CEO, Michael Fogarty, testified at a 29 March 2000 Senate Finance Committee hearing that two factors in the Medicaid "best price" approach contributed to the "evaporation" of savings: (a) open formularies, and (b) price adjustments by manufacturers to compensate for the mandatory rebates. 67 Fogarty recommended that state Medicaid programs be permitted to institute closed formularies to make drug manufacturers compete with lower prices.
OBRA 1993 amended the OBRA 1990 language and no longer required state Medicaid programs to reimburse for new drugs approved by the FDA, for the first 6 months after introduction. However, federal law and regulations prohibit states from denying access to drugs by Medicaid recipients, and 43 states and the District of Columbia had PA programs in place in 1996 to limit the use of nonformulary and nonpreferred drugs. 68 States have reported limited success with PA programs, and New York adopted a mandatory generic drug substitution program in November 2002 in which brand-name drugs with a generic equivalent would require a PA to be covered by the Medicaid program. The PA program could be utilized by physi-cians using voice recognition or a keypad and required the physicians to answer a "brief set of questions about why the patient required the brand product." For a multiple-source brand product to be dispensed, the prescription must include the PA number obtained by the physician, and must indicate "DAW" (dispense as written) and "brand (medically) necessary." 69 New York hoped to push its generic substitution rate to 95% from 88% of multiple-source brand drug prescriptions with the mandatory generic/PA program, a modest goal for managed care organizations working in the private employer sector.
The absence of reliable data on the cost-effectiveness of prescription drug benefit prior PA programs was highlighted in a previous issue of the Journal. 70 Any reasonable assessment of the cost-effectiveness of PA programs in prescription drug benefits would include consideration of the potential, and predictable, adverse service outcomes, including physician, pharmacy, and beneficiary, satisfaction as well as wait time and additional service costs. Such considerations in the public arena appear to be outweighed by budget concerns. Beginning with the Florida PA program for nonpreferred drugs that was launched in July 2001, several other states implemented similar programs by mid-2002. 71 The principal feature of all of these state Medicaid PA programs, including Michigan, Illinois, Louisiana, and North Carolina, was to extract additional rebates from pharmaceutical manufacturers to bring their drugs down to the "reference price" within therapeutic classes. Failure to match the reference price through supplemental rebates subjects the drug to PA. Cox-2 drugs were the first of 2 drug classes implemented in the Louisiana Medicaid preferred drug program in June 2002.
In this issue of the Journal, LaPensee describes the experience of a drug PA program in a Medicaid managed care organization (MCO) in the northeast in early 2002. 72 While Medicaid HMOs operate under a different set of rules than private MCOs and have fewer tools to manage drug benefit costs, particularly the absence of tier-copay benefit designs, the description of this Medicaid PA program has some useful information for all managed care observers. Nearly 4 out of 5 PA requests were for formulary drugs. Second, the PA rejection rate was low: only 4.4% of the more than 22,000 PA requests received each month were denied. Clearly, this is a large administrative burden, nearly 1,000 PA requests per day, of which an average of 44 were denied. The 93% acceptance rate for PAs for nonformulary drugs compares with an acceptance rate of about 75% for drug benefit PAs for commercial health plans in 1999 and 2000. 73-74

■■ HIPAA Effects on Health Research and PBM Functions in Drug Utilization Review
The Health Insurance Portability and Accountability Act (HIPAA) of 1996 was crafted originally with the intent to better protect health insurance coverage for employees and their families when employees changed employers; i.e., to ensure portability of health insurance. 75 By the time the legislative process was complete, HIPAA had 2 primary impacts beyond portabil-