Most practicing physicians do not own ambulatory blood pressure recorders and therefore probably obtain this service through a local hospital or clinic. Ambulatory blood pressure referrals to our hypertension unit have increased markedly in recent years. However, the effects of ambulatory blood pressure on clinical hypertension management by the referring physician remain unknown. To our knowledge, no previous study has been done to determine the influence of data from ambulatory blood pressure monitoring on physicians' practice habits regarding management of hypertensive patients. Thus, our study addressed physicians' ordering of ambulatory blood pressure monitoring and the effects of the information on patient management and office blood pressure control.

Methods

Top Methods Results Discussion Author & Article Info References

Patients and Physicians

All patients with complete 24-hour ambulatory blood pressure reports generated by the University of Connecticut's Section of Hypertension and Vascular Diseases between January 1989 and May 1992 were included in the study. Referrals from the hypertension unit and the author's (WBW) practice were excluded. Referrals came from either hospital-based physicians or physicians in private or community practice in central Connecticut (Figure 1). Our service was not advertised; referrals for clinical studies came via word of mouth. During the study period, a standard instrument (Accutracker II, Suntech Medical Instruments, Raleigh, North Carolina) [5] was used for all recordings, and a standardized two-page consultative report was generated and sent to the referring physician. The report, prepared by members of our section, included patient demographic characteristics, summary statistics for average 24-hour, "waking" and "sleeping" blood pressures, blood pressure load [6], percentage change in waking and sleeping blood pressures and heart rates, customized blood pressure and heart rate graphs, and a brief clinical impression of the 24-hour profile. The analysis was derived from data obtained from the Accutracker II, which was programmed to record the blood pressure and heart rate at 15-minute intervals from 0600 h to 2200 h and at 30-minute intervals from 2200 h to 0600 h. Detailed activities, symptoms, and medication dosing times were recorded in journals as previously described [3, 7].

View larger version (13K): [in this window] [in a new window]   Figure 1. Number of referring physicians according to clinical specialty. Community-based physicians were defined as practitioners with private office practices; hospital-based physicians were defined as full-time hospital employees with clinics or office practices in a community or teaching hospital.

Study Questionnaires

We developed two questionnaires for each referring physician. We used a general questionnaire to assess physicians' demographic characteristics, practice specialty, and familiarity with ambulatory blood pressure monitoring; the clinical diagnoses they considered appropriate for monitoring a patient; and their level of confidence in the precision of the test and expected future use of ambulatory blood pressure monitoring.

A second questionnaire focused on the management of physicians' referred patients. This questionnaire was used to evaluate how the results of ambulatory blood pressure monitoring affected patient management in terms of additional diagnostic test ordering and changes in diagnosis and antihypertensive therapy. The office blood pressure readings determined by the referring physician before ambulatory blood pressure monitoring, within 3 months after testing, and within 1 to 2 years after testing, were requested through the data questionnaires and follow-up telephone calls, if necessary.

Statistical Analyses

Comparisons of the responses of hospital-based and community-based physicians on the questionnaires were done using chi-square analysis. Statistical comparison of the office systolic and diastolic blood pressures before and after ambulatory blood pressure monitoring was done using analysis of variance. Tests for normality of blood pressure distributions were done using the Shapiro-Wilk test.

Results

Top Methods Results Discussion Author & Article Info References

Of 65 referring physicians, 47 (72%) completed the questionnaires. Twenty-nine were community-based physicians, and 18 were hospital-based physicians who attended at one of six community or university teaching hospitals in the greater Hartford, Connecticut, area. Incomplete questionnaires were not included in the analysis.

Physicians' Use of Monitoring

Most ordering physicians were internists (57%); clinical cardiologists formed the second largest group (see Figure 1). The physicians indicated awareness of ambulatory blood pressure monitoring for an average of 4.3 years. Sixty-one percent of the community-based physicians felt well informed about ambulatory blood pressure monitoring, whereas only 28% of the hospital-based physicians placed themselves in that same category [P = 0.02]. Seventy-seven percent of the entire physician group indicated they had learned about ambulatory blood pressure monitoring through journal articles, 66% through discussion with colleagues, 58% through lectures and conferences, and 43% through the ambulatory blood pressure monitoring report.

The hospital-based physicians said they would order an ambulatory blood pressure study to aid in the diagnosis of "white-coat" or "office" hypertension, borderline hypertension, severe hypertension, and symptomatic hypertension and to assess antihypertensive therapy and blood pressure control. Similar findings were noted among the community-based physicians, who showed a greater tendency to order an ambulatory blood pressure monitoring study in cases of suspected "white coat" or "office" hypertension (P = 0.03) or borderline hypertension or in cases in which assessment of drug therapy was indicated.

Physicians in both types of practice indicated they felt confident in the precision of the 24-hour ambulatory monitoring devices (80% of total group). After recent experiences with ambulatory blood pressure monitoring, 24% of the physicians indicated they would probably use the test more often, 4% indicated they would use the test less often, and 68% indicated they would order the test with approximately the same frequency as in the recent past.

Of 237 patients who underwent ambulatory blood pressure monitoring, 22% were studied for suspected "white coat" hypertension, 27% for borderline hypertension, 25% for assessment of antihypertensive therapy and control of pressure, and 16% for drug-resistant hypertension (Figure 2). Questionnaires for individual patients were completed (including office blood pressure data) for 122 patients referred by 37 of the hospital- and community-based physicians. The average age of these patients was 56 ± 6 years; 51% were men, and 65% were receiving antihypertensive therapy at the time of the ambulatory blood pressure study. Physicians had referred an average of 3.3 patients to the ambulatory blood pressure monitoring laboratory, and the mean duration that a patient had been under the care of the referring physician was 6.6 ± 3.4 years.

View larger version (19K): [in this window] [in a new window]   Figure 2. Clinical indications for patient referral for ambulatory blood pressure monitoring. No statistical differences were observed between types of referrals from community-based and hospital-based physicians.

Changes in Patient Management

The influence of the ambulatory blood pressure monitoring results on patient management was assessed by evaluating test-ordering behavior and changes in diagnosis and therapy. Only 6 of the 47 physicians (13%) ordered additional tests (exercise testing, 3%; echocardiography, 7%; and angiography, 3%) based on monitoring results. Forty-three percent of the hospital-based physicians and 38% of the community-based physicians altered their diagnosis based on data provided by ambulatory blood pressure monitoring. In patients with variable, borderline hypertension, the most common new diagnoses were "white coat" hypertension (40%) and sustained hypertension (26%). Twenty-two percent had a diagnosis that was categorized as sustained hypertension while receiving drug therapy with a "white-coat" component (that is, patients whose out-of-office blood pressure measurements were substantially lower, albeit abnormal, when compared with their office measurements).

Antihypertensive therapy was changed in 53% of the hospital-based physicians' patients and in 34% of the community-based physicians' patients based on information provided by the ambulatory blood pressure monitoring study. Overall, 88 patients had changes in their treatment regimen. Nearly one third of the 88 patients initially received antihypertensive therapy because of the findings from monitoring (Table 2).

Changes in Blood Pressure

We compared office blood pressure readings obtained before monitoring with office readings obtained within 3 months and within 1 to 2 years after monitoring. Before the ambulatory blood pressure monitoring study, the mean office blood pressure was 161/96 ± 22/12 mm Hg and was normally distributed (Figure 3). Three months after ambulatory blood pressure monitoring, the mean office blood pressure fell 13/10 ± 3/2 mm Hg (P = 0.004 for systolic blood pressure; P < 0.001 for diastolic blood pressure). The distribution of the changes in office blood pressure is shown in Figure 4. The office blood pressure of most patients was lower after the procedure. One to 2 years later, these changes in office blood pressure had stabilized: At that time, blood pressure was 15/10 ± 3/2 mm Hg lower when compared with the prestudy measurement and 2/0 ± 1/1 mm Hg lower when compared with the measurement obtained within 3 months after monitoring (P > 0.3 for both systolic and diastolic pressure).

View larger version (20K): [in this window] [in a new window]   Figure 3. Office blood pressure measurements obtained by referring physicians before ambulatory blood pressure monitoring in 237 patients.

View larger version (17K): [in this window] [in a new window]   Figure 4. Distribution of changes in office blood pressure measurements within 3 months after ambulatory blood pressure monitoring relative to prestudy measurements.

Seventy-two percent of the 122 patients had a lower (defined by a decrease of more than 10 mm Hg in either systolic or diastolic blood pressure) office blood pressure 3 months after the ambulatory blood pressure study. In this subgroup, 41% had a change in their antihypertensive regimen: Seventeen percent were given an antihypertensive agent, 12% received an additional agent or had a dose increase in existing medication, and 12% had a substitution in medication.

Discussion

Top Methods Results Discussion Author & Article Info References

Our study showed that physicians are incorporating ambulatory blood pressure monitoring into their practice for assessment of hypertensive patients. Their use is in accordance with the recommendations of the National High Blood Pressure Education Group Working Party [1]; the main clinical indications for ambulatory blood pressure monitoring by physicians in our area (central Connecticut) included the assessment of antihypertensive therapy (28%), possible "white coat" or "office" hypertension (22%), and borderline hypertension (27%) (see Figure 2). The utility of ambulatory blood pressure monitoring in these instances has been well documented [1, 7-10]. The test-ordering behavior of our physicians should be of interest to health policy experts who have been concerned about the potential for both inappropriate and excessive use of ambulatory blood pressure monitoring [1, 11]. Only 9% of the physicians surveyed indicated that they would use ambulatory blood pressure monitoring as a standard test in the evaluation of hypertension. Thus, it is possible that if ambulatory blood pressure monitoring were used appropriately (see Table 1), it would become cost-effective in the management of mild hypertension [12].

Ambulatory blood pressure monitoring appeared to affect the management of the patients in our study: Forty-one percent of the patients underwent a change in diagnosis and 46% had some sort of change in antihypertensive therapy after ambulatory blood pressure monitoring (see Table 2). As expected, many of the patients had their diagnosis changed to "white coat" hypertension or to sustained hypertension with a "white coat" component. In some instances, these patients may have been spared unnecessary or excessive medication, a change in lifestyle, and the psychological effects of being labeled "hypertensive" [1].

The high percentage of patients receiving antihypertensive therapy who were referred for an ambulatory blood pressure study (65%) was surprising. Waeber and colleagues [13] reported that some patients with symptomatic hypertension receive excessive treatment; however, many patients in our study had inadequate blood pressure control, and office blood pressures were lower after monitoring studies (see Figure 4). This finding suggests the unexpected benefit of ambulatory blood pressure monitoring for detection of persons with poorly controlled hypertension; one might speculate that improved blood pressure control in such patients could yield reduced target organ involvement [4, 6, 14].

Ambulatory blood pressure monitoring may be useful in several clinical situations related to hypertension management (see Table 1). Physicians in practice appear to be adhering to general recommendations of the Working Party Report for these indications and have not demonstrated abuse of this study. However, because the physicians in our study did not receive remuneration for ordering this test, this finding is not necessarily applicable to a scenario in which ambulatory blood pressure monitoring is widely approved by third-party insurance carriers.

Our study had several limitations. First, the study did not include controls who were evaluated and treated without ambulatory blood pressure monitoring. The reductions in office blood pressure (see Figure 4) may be related to the ambulatory blood pressure data, but without a control group, the direct relevance is unknown. Second, although the general outcome in our patient population was positive, these findings cannot be extended to the general population of hypertensive patients. Referring physicians probably were having diagnostic or therapeutic challenges with their patients. Some physicians may have already been poised to make therapeutic changes in their patients, and, although the monitoring helped with that decision, other clinical issues were probably part of the physician's decision-making process.

The many readings provided by ambulatory blood pressure monitoring during activities of daily living are superior to the casual readings taken in the doctor's office for the detection and evaluation of specific types of hypertensive problems [1]. However, recommendations for use of ambulatory blood pressure monitoring have been tempered by the fact that only one prognostic study [15] has shown any relationship between ambulatory blood pressure and cardiovascular risk. The main effects of ambulatory blood pressure monitoring on blood pressure management include more precision in diagnosis and significant, albeit indirect, improvement in hypertension control. Our findings, as well as other published studies, should stimulate investigators to examine whether ambulatory blood pressure monitoring changes the outcome of care and whether its use is cost-effective.