Cohen JS. Adverse Drug Effects, Compliance, and Initial Doses of Antihypertensive Drugs Recommended by the Joint National Committee vs the Physicians' Desk Reference. Arch Intern Med. 2001;161(6):880–885. doi:10.1001/archinte.161.6.880
Compliance problems are common causes of the inadequate treatment of hypertension, with 16% to 50% of patients quitting treatment within 1 year. Dose-related adverse drug events (ADEs) frequently cause compliance problems, and many ADEs occur with the initial doses of antihypertensive drugs. Thus, it is an established tenet to initiate antihypertensive therapy at low doses to avoid ADEs that diminish patients' quality of life and reduce compliance. However, what are the lowest effective doses of antihypertensive drugs?
To compare the initial doses recommended in the Physicians' Desk Reference (PDR) with those recommended by the Sixth Report of the Joint National Committee on the Detection, Evaluation, and Treatment of High Blood Pressure (JNC VI).
Review of the latest JNC VI report (1997) and the 1999 and 2000 editions of the PDR and the medical literature.
The JNC VI recommends substantially lower initial doses for 23 (58%) of 40 drugs, compared with the PDR. In addition, for 37 (82%) of 45 drugs, PDR guidelines do not suggest lower initial doses for old or frail patients than for younger adults.
Although the PDR is the drug reference most used by physicians, it does not reflect the lowest initial doses that are recommended by the JNC VI for many of the most prescribed antihypertensive drugs. Because avoidance of ADEs is essential to maintaining compliance with antihypertensive therapy, and because many antihypertensive ADEs are dose related, physicians must know the very lowest, effective, least ADE-prone doses. Patients and physicians would benefit by establishing mechanisms to make this information readily available to all practicing physicians.
DURING THE past quarter century, the medical profession has made a major effort to improve awareness among people who have hypertension and to start their treatment.1 The result has been a significant reduction in the long-term sequelae of hypertension, including myocardial infarctions and cerebrovascular accidents. However, during the 1990s, improvements in these measures slowed,2,3 and only 29% of the 50 million Americans with hypertension had attained a blood pressure below 140/90 mm Hg.4,5 Therefore, efforts continue to be directed at (1) reaching unaware or untreated hypertensive persons and facilitating treatment; (2) improving the percentage of patients achieving treatment goals by optimizing therapeutics; and (3) improving compliance by reducing the percentage of patients quitting treatment. Optimizing therapeutics and improving compliance are the focus of this article.
Compliance may be defined as the extent to which a patient's behavior conforms to medical advice.6 Long-term compliance with antihypertensive medication regimens has been poor. In one study, only 49% of patients took more than 80% of their prescribed dosages during the first year of treatment.7 Other studies indicate that 16% to 50% of hypertensive patients quit taking their medications within the first year of treatment.7- 10
These numbers improve considerably once patients have become established in treatment regimens. A survey of 79 000 hypertensive patients in Saskatchewan found a dropout rate among new patients of 22% in the first year and 54% after 4.5 years, leaving a persistence rate that the authors deemed "remarkably poor."11 Although patients with newly diagnosed hypertension frequently quit treatment, the authors noted that "patients who have successfully initiated antihypertensive treatment tend to remain in treatment." Thus, the successful initiation of treatment itself is a goal, particularly in terms of ensuring the continuation of treatment. The authors underscored this point: "Our analyses emphasize the importance of maximizing the likelihood of successful early treatment of hypertension and thus increasing the proportion of patients taking medication over the long term."11
There are many factors that affect compliance, such as the cost of medications and the inadequacy of physicians' explanations to patients of the importance of treating hypertension, which often is asymptomatic. A primary reason for poor compliance among patients receiving antihypertensive medications is adverse drug events (ADEs), many of which are dose related.12- 14 In one study, 34% of patients reported unacceptable ADEs.15 Because of the high incidence of ADEs, physicians must consider quality-of-life issues in selecting antihypertensive drugs and doses. One difficulty is that many antihypertensive medications cause ADEs at therapeutic doses. Although physicians may consider ADEs such as dizziness, headaches, constipation, low energy, or sedation as minor, these can greatly interfere with normal functioning, which many patients find unacceptable. Not infrequently, sexual functioning is impaired, another difficult ADE for patients to accept.
These problems are compounded by the nature of hypertension, which may cause few symptoms. Treatment that provokes an ADE may make some patients feel subjectively worse than before treatment initiation. Patients who believe that they must choose between a comfortable but shortened life vs an ADE-afflicted but prolonged life often choose the former. Because compliance is intimately related to the avoidance of troublesome ADEs, physicians are advised to start antihypertensive therapy at the very lowest effective drug dosages. However, what are the lowest effective doses of antihypertensive drugs? And are the data on the lowest effective doses readily available to physicians? This analysis attempts to answer these questions.
The Physicians' Desk Reference (PDR)16,17 was selected because it contains the dosages that are recommended by the drugs' manufacturers and approved by the Food and Drug Administration, and the dosage recommendations of other drug references are usually very similar to those of the PDR. The PDR is the drug reference used most often among physicians18- 20; approximately 90% of physicians rely on the PDR for dosage information.18 It is also used extensively by consumers, who purchase half a million copies annually. The 1999 and 2000 editions were consulted for this analysis.
The Sixth Report of the Joint National Committee on the Detection, Evaluation, and Treatment of High Blood Pressure (JNC VI)1 was selected because the JNC is a respected panel of experts on hypertension that issues regular reports on the status of antihypertensive therapy in America. The JNC VI, the most recent report, was published in 1997 and contains a comprehensive list of antihypertensive drugs with the panel's recommended dosages. These dosages were compared with the dosage recommendations in the PDR.
The JNC VI lists 57 antihypertensive drugs, but 12 are no longer listed in the 1999 or 2000 PDR or are listed only briefly without dosage guidelines. Another 5 drugs (guanfacine hydrochloride, nifedipine, prazosin hydrochloride, terazosin hydrochloride, and valsartan) are not produced in a form that would allow the use of lower dosages. Of the remaining 40 drugs, the JNC VI recommended initial doses for 23 (58%) that were substantially lower than those recommended by the PDR (Table 1). All 23 dose disparities occurred among the 34 drugs (68%) in 5 frequently prescribed groups: angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, β-blockers, calcium antagonists, and diuretics. Except for chlorthalidone and one brand of metoprolol succinate, the PDR initial doses were at least 100% higher than the JNC VI doses. These disparities occurred with some of the most prescribed antihypertensive drugs, such as amlodipine besylate, atenolol, bisoprolol fumarate, diltiazem hydrochloride, lisinopril, losartan potassium, metoprolol, propranolol hydrochloride, and ramipril.21 There were no drugs for which the PDR recommended lower initial doses than JNC VI. Regarding older patients, for whom experts often recommend lower initial doses than for younger adults, the PDR guidelines recommended lower doses with only 8 (18%) of 45 drugs (Table 2).
Even when the JNC VI and PDR agree on the initial doses of 17 drugs, even lower doses of these drugs might be effective. For example, the American Hospital Formulary Service recommends an initial enalapril dose of 2.5 mg vs 5 mg by JNC VI and the PDR.22 The American Hospital Formulary Service also recommends 0.05 mg of clonidine hydrochloride twice daily initially for some patients vs 0.1 mg twice daily by the JNC VI and the PDR.
The range of manufactured doses varies considerably among antihypertensive drugs. Some drugs are produced with 16- or 20-fold dosage ranges (eg, doxazosin mesylate dosage range, 1-16 mg/d), whereas others are produced with 2- or 3-fold ranges (Table 3). The most common dosage ranges are 4- and 8-fold. Individual variation in drug response due to differences in age, weight, sex, ethnic background, state of health, concomitant medication use, and genetic polymorphisms in drug metabolism is a long-accepted pharmacological principle. Therefore, drugs offering wider ranges of dosages that allow maximum flexibility in titrating treatment may be generally preferable.
Because of the importance of avoiding ADEs at the beginning of treatment, the JNC VI cautions physicians against trying to bring mild-to-moderate hypertension under control too quickly: "Therapy for most patients (uncomplicated hypertension, stages 1 and 2) should begin with the lowest dosage . . . to prevent adverse effects of too great or too abrupt a reduction in blood pressure."1 Other authorities agree.23,24 Sorrentino23 states aptly, "Unless acute target organ involvement is already present, there are no compelling reasons to lower blood pressure quickly."
Thus, initial interventions are usually aimed at lifestyle modifications such as weight loss, stress reduction, exercise, dietary changes, reduction of alcohol use, and cessation of smoking. Medication interventions are usually considered secondary, because lifestyle changes alone represent major interventions and adjustments for many patients and can have substantial impact on blood pressure.25- 27 Because more than 75% of all ADEs are dose related,28,29 starting with the lowest effective doses that minimize ADEs is recommended. MacConnachie and Maclean12 link compliance directly to dosage: "Therefore, any measure [that] reduces the dosage requirement of an antihypertensive while maintaining therapeutic efficacy, or otherwise limits the possibility of commonly encountered ADEs, will encourage good patient compliance and so improve long-term control of hypertension." For these reasons, some physicians subscribe to starting with a dose that may be subtherapeutic rather than a dose that is "correct" but may provoke troublesome ADEs.
First-dose reactions are ADEs that occur with the initial dose of a drug or when the dosage is increased. First-dose reactions have been described with the use of α-receptor blockers,16 calcium channel blockers,16 ACE inhibitors,30,31 and β-blockers,32 but clinical experience suggests that first-dose reactions may occur with any antihypertensive drug. First-dose reactions due to antihypertensive drugs are frequently dose related and may result from an abrupt lowering of blood pressure, causing postural hypotension, dizziness, syncope, headaches, lethargy, or other symptoms. First-dose reactivity explains why a large proportion of ADEs due to antihypertensive drugs occur at the beginning of treatment9 and may indicate that some patients are sensitive to the pharmacological effects of the standard initial doses of antihypertensive drugs that physicians are prescribing.
Hypertension is most prevalent among people older than 60 years,1,33,34 and treatment can be especially challenging because of altered pharmacokinetics (eg, reduced liver and kidney function, and increased receptor sensitivity), which can produce even greater extremes in individual drug response than in younger adults.35,36 In addition, treatment is further complicated because two thirds of people older than 65 years take at least 1 medication daily, with the average being 3 prescription and/or nonprescription drugs daily.29,37,38 Thus, the overall incidence of ADEs is 2 to 3 times greater in elderly than in young adults. These ADEs in older patients are typically dose related38,39 and are the leading cause of older patients' discontinuing antihypertensive therapy.9
For these reasons, some experts recommend that older patients, especially those who are frail, ill, or taking other medications, should be prescribed initial doses that are lower than those for younger adults.29,36,38,39 However, 37 (82%) of the 45 antihypertensive drugs examined herein are not recommended at a lower dose for seniors by the PDR(Table 3). Some of these are drugs for which the JNC VI recommends lower initial doses not only for seniors, but for all patients. Moreover, the pharmacokinetic data of some drugs suggest the appropriateness of lower doses for seniors. For example, the plasma levels of acebutolol and bisoprolol are doubled in seniors compared with young adults,16 yet the PDR-recommended initial doses are the same. With other drugs (eg, nifedipine and diltiazem), the PDR lacks pharmacological data on the response of older patients, so the physician's ability to make educated judgments about dosage is limited. With some drugs, physicians wanting to use half doses are stymied because the pills do not facilitate this (capsules or irregular or coated tablets).
Hypertension is the most common indication for visits to US physicians,40 and it is a leading cardiovascular risk factor.11,41 Because of considerable variation among patients, the optimal pharmacological treatment of hypertension requires a flexible approach.42 Such flexibility involves the use of the lowest effective initial doses to facilitate a positive therapeutic alliance and to avoid ADEs that may affect compliance. Low-dose therapy allows patients time to adjust psychologically to the fact they have hypertension and to begin making lifestyle changes recommended by the physician—no small undertaking for most patients. Some patients experience distress about having hypertension and possibly requiring lifelong drug therapy, and they develop anxiety symptoms that may be mistaken for ADEs, which may lead to skipping doses or quitting treatment. Drug therapy commencing with minimal doses is generally more acceptable to patients, allays fears about unpleasant side effects, and minimizes possible confusion between drug-related ADEs vs anxiety-related symptoms or coincidental factors.
The danger in commencing treatment with the lowest recommended doses is, of course, undermedication, which also is a problem in antihypertensive management today.1 However, undermedication is usually a result of inadequate follow-up rather than a low initial dose. The stepwise approach to treating hypertension presupposes that proper titration will lead to higher doses in some patients or to the addition of 1 or 2 more antihypertensive drugs, when necessary. The "start low, go slow" approach, which is intended to minimize dose-related ADEs that hinder compliance, is effective if proper follow-up and dosage titration are provided.
Some patients experience first-dose reactions when their dosages are increased. This may sometimes occur because the recommended dosage range is too narrow or a 100% increase in dosage exceeds the patient's tolerance. Smaller dosage increases, which may require splitting pills, may sometimes prevent this complication. Drugs produced as scored pills in wide ranges of doses provide the most dosage flexibility.
Overall, optimal antihypertensive pharmacotherapy in mild-to-moderate hypertension is most often accomplished when initiated at the lowest effective doses, to maximize compliance by minimizing ADEs. To provide optimal therapy, physicians must have a readily available source of current information that defines the very lowest effective doses of antihypertensive drugs. The PDR, which was initiated 54 years ago as a promotional device, is now the leading source of drug information among physicians, mainly because of its easy-to-use format and excellent indexes, and because it is distributed free to physicians each year.
However, the information in the PDR consists mainly of the limited prerelease data that the manufacturer and the Food and Drug Administration deemed necessary for the safe and effective use of medications at the time of their approvals. This information may not be adequate for making therapeutic decisions in the much wider range of patients seen in clinical practice compared with patients undergoing evaluation in prerelease studies. Ray et al43 have stated: "Although these studies generally ensure that a drug is efficacious and does not cause unacceptable harm, pre-marketing studies often fail to provide much of the information needed to make therapeutic decisions." However, there is no requirement and little incentive for manufacturers to update the PDR information regularly to reflect postrelease studies or reports. Thus, the point of this article is not that the producers of the PDR are failing to fulfill an expected responsibility, but rather that a situation has developed in which there is no readily available source where physicians can obtain current information about the best methods of initiating antihypertensive drug therapy.
This situation requires solutions. Because of the wide popularity of the PDR, the ideal solution might be to establish mechanisms by which the information in the PDR can be kept current to reflect the evolving standards of care and the full range of proven effective drug dosages. If such mechanisms cannot be established, then perhaps the free dissemination of the PDR should be discouraged, and an alternative, objective source of accurate, current information should be created in its place.
The findings of this article raise many questions. What initial doses of antihypertensive drugs are physicians actually prescribing? Studies have shown that the recommendations of the JNC VI have had little impact on the types of antihypertensive drugs that physicians prescribe,44,45 but this has not been examined in regard to doses. Furthermore, the implementation of optimal pharmacotherapeutic methods ultimately depends on physicians. If physicians were better informed about the JNC VI– recommended initial doses of antihypertensive drugs, would they alter their methods? What would best motivate them to do so? What would motivate drug manufacturers to define the lowest, safest doses of new drugs in their prerelease research, to provide pills that allow for flexible dosing, to provide rational guidelines for older patients, and to reverse the trend that seems apparent in drug advertising toward 1-size-fits-all and other simplistic methods of dosing?
With a perennial high incidence of ADEs, most of which are dose related, and with well-defined problems with ADEs and compliance in treating hypertensive patients, these questions need to be answered and solutions need to be found and implemented.
The optimal pharmacotherapy of hypertension depends on the availability of information on the full range of effective dosages of antihypertensive drugs. This includes effective dosages that may be lower than those recommended by drug manufacturers. The PDR is the most used source of drug information among physicians, and it is heavily relied on by hospital staff who may lack the clinical experience to know that PDR dosage recommendations are general guidelines based on limited prerelease research, not on hard and fast rules of optimal therapeutics. This analysis reveals that, compared with the PDR, the initial doses of antihypertensive drugs recommended by JNC VI are substantially lower for 23 (58%) of 40 drugs for which the use of lower doses was possible. Moreover, although experts generally suggest reduced initial doses for older patients, the PDR does not recommend such reductions for old or even very old patients with 37 of 45 drugs. If the JNC VI recommendations, which represent prerelease and postrelease data, are considered the state of the art, mechanisms need to be implemented by which these recommendations are incorporated into the PDR and the corresponding package inserts, where physicians are most likely to see and use them. If such mechanisms cannot be established for the PDR, another source of current, readily available drug information should be created so that physicians' methods will keep pace with evolving standards of optimal pharmacotherapy for patients with hypertension.
Accepted for publication September 14, 2000.
Corresponding author: Jay S. Cohen, MD, 2658 Del Mar Heights Rd, #120, Del Mar, CA 92014 (e-mail: email@example.com).