ELSEVIER

Appointment Attendance, Pill Counts, and Achievement of Goal Blood Pressure in the African American Study of Kidney Disease and Hypertension Pilot Study

Jeannette Y. Lee, PhD, Paul G. Greene, PhD,

Margaret Douglas, MPH, Clarence Grim, MD,

Katharine A. Kirk, PhD, John W. Kusek, PhD,

Sharon Mill&an, PhD, Delia E. Smith, PhD, and

Paul K. Whelton, MD, MSc, for the AASK Pilot

Study Investigators

ABSTRACT: The African American Study of Kidney Disease and Hypertension (AASK) Pilot Study evaluated the feasibility of conducting a ‘/-year clinical trial to assess the effect of two levels of blood pressure control based on mean arterial pressure (MAP) (low goal G 92 mm Hg or usual goal of 102-107 mm Hg) and three antihypertensive drug regimens (atenolol, amlodi- pine, or enalapril) as initial therapy in slowing the decline of renal function in African Americans with clinically diagnosed hypertensive nephrosclerosis. Ninety-four African American men and women between 18 and 70 years of age were randomized and followed for an average of 4.6 months.

On average, participants attended 87.5% of the scheduled monthly follow-up visits and achieved an acceptable level of medication adherence (80%-100% of prescribed doses by pill count) at 65.4% of those visits. Blood pressure levels within goal were observed in 17.5% and 25.6% of the participants in the low- and usual MAP goal groups, respectively. Neither attendance nor medication adherence by pill count was associated with attainment of goal blood pressure. Although AASK Pilot Study participants maintained excellent attendance, their pill counts were lower than previously reported among clinical trial participants and goal blood pressure levels were difficult to achieve during the short period of follow-up.

KEY WORDS: AASK Pilot Study, hypertension, kidney disease

INTRODUCTION

Adherence to treatment is an important element in the control of hypertension 175,761. It has been estimated that 50% of hypertensive patients miss their clinic

Address reprint requests to: Dr. John W. Kusek, NIDDK, NIH, Natcher Building, Room 6AS-131, 45 Center Drive, MSC 6600, Bethesda, MD 20892-6600.

Received March 15, 1995; revised October 17, 1995.

Controlled Clinical Trials 16:34%39S (1996) 0 Elsevier Science Inc. 1996 655 Avenue of the Americas, New York, NY 10010

0197-2456/96,‘$15.00 PI1 SO197-2456(96)0080-3

Appointment Attendance, Pill Counts, and Goal Blood Pressure 35s

appointments and that only 60% take their medications as prescribed [771, with minority patients at higher risk for nonadherence [75,78,79]. In long-term clinical trials of hypertension, appointment attendance rates of 83-96% and medication adherence rates by pill count of 79-83% have been reported 176,80-831. In compari- son to clinic populations, high clinic visit attendance and medication adherence rates are reported among clinical trial participants and reflect the use of aggressive strategies to retain participants and to enhance appropriate medication intake. Minority participants in these clinical trials, however, had lower attendance rates for follow-up clinic visits [76,81] and drug adherence rates [80] than their white counterparts.

Control of blood pressure is important in the management of patients with chronic renal disease prior to end-stage ]28-301. Recently, it has been suggested that the level of blood pressure control may be an important factor in the progression of chronic renal disease [5,31-361. In the Modification of Diet in Renal Disease (MDRD) Study 151, participants were randomized to one of two blood pressure goals based on mean arterial pressure (MAP): low-MAP goal < 92 mm Hg and usual MAP goal c 107 mm Hg, with 84.7% and 81.7% of participants randomized to the low- and usual MAP goal groups, respectively, attaining average blood pressure within goal [84]. However, the difference in mean MAP between the low- and usual MAP goal groups (4.6 mm Hg) was only about one third of that expected. Recently, among clinical trial participants with presumed hypertensive nephro- sclerosis, both mean diastolic blood pressure (DBP) and mean arterial pressure were significantly lower in the “strict” blood pressure group (diastolic blood pres- sure 65-80 mm Hg) compared to a conventional group (DBP 85-95 mm Hg) (5.7 and 5.5 mm Hg, respectively) [8].

Nonadherence to experimental treatment protocols can have a significant impact on the design and analysis of a clinical trial. Even modest rates of nonadherence require substantial increases in sample size with a concomitant rise in cost 185-871. If low adherence rates occur within each intervention arm of a randomized clinical trial, the interventions may be rendered ineffective and significantly different out- comes are likely to be achieved. Disparate levels of medication adherence between treatment arms can yield misleading results as differences in outcome measures may reflect differences in adherence rather than therapeutic efficacy.

The African American Study of Kidney Disease and Hypertension (AASK) Pilot Study was conducted to assess the feasibility of a 7-year clinical trial to evaluate the effects of two levels of blood pressure control and three antihypertensive drug regimens in slowing the decline in renal function in African Americans with hypertensive nephrosclerosis [9]. One of the objectives of the AASK Pilot Study was to assess adherence to the study protocol and the ability to achieve blood pressure levels within randomly assigned goals. Appointment attendance and pill count were monitored as measures of participant adherence to the study protocol. Mean arterial pressures were monitored as an intervention check to determine if participants consistently achieved protocol-defined blood pressure goals.

METHODS

Pill counts for all prescribed antihypertensive drugs were taken at each visit by the study coordinator and discussed with the study participant. Procedures were implemented at all clinical centers to promote overall adherence to the protocol.

36s J.Y. Lee et al.

Table 1 Average Percent of Follow-up Visits at Which an Acceptable Level of Medication Adherence Was Achieved

Number Drug of participants Mean (%) SE

Randomized drug 94 65.4 3.1 Furosemide 60 57.1 4.4 Doxazosin 42 61.6 5.8 Clonidine 22 73.5 6.2 Minoxidil 9 40.7 14.5 Hydralazine 11 63.6 11.9

Study coordinators were trained in coinseling techniques regarding medication adherence at a centralized training session held prior to participant enrollment. An adherence interview was conducted at each follow-up visit to elicit information on barriers to antihypertensive drug use and to develop participant-specific strate- gies to enhance appropriate medication intake. Study participants were also in- structed to record the date and time of each drug dose in a diary.

Because of the short duration of follow-up and the staggered entry of participants, the attendance rates and medication adherence are reported for the first six follow- up visits. For each participant, the appointment attendance rate was calculated as the proportion of scheduled monthly follow-up visits that were attended. No adjustment was made for nonattendance due to withdrawal from the study or other reasons in order to assess differential attendance, which may be associated with the assigned blood pressure goal or antihypertensive drug regimen.

An acceptable level of medication adherence was defined as pill count between 80% and 100% of the total pills dispensed which were not returned at a subsequent clinic visit. For each participant, the medication adherence rate was defined as the proportion of monthly follow-up visits at which an acceptable level of medication adherence was achieved. As it was anticipated that a minimum of 2 months would be required to adjust drug dosages to reach MAP goal, the proportion of participants in each blood pressure goal group whose mean MAP for follow-up visits 3-6 was within goal was determined.

STATISTICAL METHODS

Two-way analyses of variance were used to evaluate the effects of blood MAP goal group and assigned antihypertensive drug regimen on appointment attendance rate and medication adherence rate for the randomized drug. Logistic regression analysis was used to relate achievement of MAP goal with blood pressure goal group, antihypertensive drug regimen, appointment attendance rate, and medica- tion adherence rate.

RESULTS

The appointment attendance and medication adherence rates did not vary by MAP goal group or assigned antihypertensive drug regimen. The average atten- dance rate (mean ? SE) was 87.5 + 1.7. The medication adherence rate for the randomized drug averaged 65.4 + 3.1 and ranged from 40.7 to 73.5 for the nonran- domized drugs (Table 1). The mean number of drugs prescribed was 2.5 and did not vary by MAP goal groups or antihypertensive drug regimens.

Appointment Attendance, Pill Counts, and Goal Blood Pressure 37s

Blood pressure levels within goal were observed in 17.5% participants in the low-MAP goal group and in 25.6% of the participants in the usual MAP goal group. Achievement of blood pressure goal was not associated with MAP goal group, antihypertensive drug regimen, attendance rate, or medication adherence rate.

DISCUSSION

The low educational level and economic status among the AASK Pilot Study participants posed potential barriers to attendance and medical adherence. The high appointment attendance rate in AASK is similar to that of other studies in which several strategies were employed to retain participants [80,88,891. The attendance rate reported here may be somewhat underestimated because no adjust- ment was made for follow-up visits missed due to participant withdrawal or other reasons. The motivation to attend clinic visits may be related to the participants’ reasons for joining the study and the demands of the screening period. In the AASK Exit Survey which was administered at the conclusion of the study, the most frequently cited reasons for joining the study were to improve health (94%); to learn more about high blood pressure and renal disease (89%); and close, frequent monitoring (82%) 1141. During the 3-month screening period prior to randomization, participants had their renal function measured twice by a low-level radioisotope method and about half of the participants underwent a renal biopsy. Thus, random- ized participants made a substantial commitment to the study prior to treatment assignment and might be expected to demonstrate a high degree of adherence. Work-related problems were cited as a barrier to participation by only 8% of the employed AASK participants. Furthermore, reimbursement for travel expenses to and from the clinic was made available to all participants.

There is little recent information on medication adherence among African Ameri- cans. The medication adherence rates observed in the AASK Pilot Study are lower than observed in clinical trials of antihypertensive therapy ]76,80,821. In these studies, medication adherence by pill count was lower among black participants than among their white counterparts [76,80,82]. Although medication adherence based on pill count may be overestimated because pill counts can be recorded only when medication bottles are returned, pill containers were not returned at only 10.2% of the AASK follow-up visits. In contrast, 36% and 29% of participants in the Hypertension Detection and Follow-Up Program (HDFP) on singIe and multiple drug therapy, respectively, failed to return pill containers 1761.

The lower medication adherence rate among AASK Pilot Study participants occurred despite implementing multiple adherence promotion strategies, recruiting primarily among patients receiving ongoing medical care from physicians, and providing the study medications free of charge to the participant. As most AASK participants were under a physician’s care for hypertension prior to joining the study, they might have been expected to be more likely to follow the trial’s pre- scribed medication regimen. Participation in the study did, however, require most participants to change their blood pressure medications, which may have required changing longstanding habits [14]. The prescriptions and diaries were reviewed frequently with the AASK participants as part of the interview conducted at each follow-up visit. While the medication diaries were cited as being helpful by 65% of the respondents to the AASK Exit Survey 1141, they reflect self-reported medica- tion use, which often overestimates actual medication intake [9&94]. Thus, more

38s J.Y. Lee et al.

aggressive measures must be instituted to increase the medication adherence rate to a more satisfactory level.

A relationship between medication adherence and attainment of the target blood pressure was observed in the HDFP [76]. In the HDFP, blood pressure goal was met by 79.4%, 70.5%, and 62.5% of the participants whose pill counts were 3 80% of prescribed, < 80% of prescribed, and who did not return pills, respectively [761. The lack of association between medication adherence and achievement of target blood pressure in the AASK Pilot Study may reflect the short duration of follow- up and the limitations in the use of pill counts in assessing medication adherence. Pill counts can be recorded only when the medication bottle is returned and will be overestimated if pills that are removed from the bottle are not taken by the participant. Furthermore, pill counts provide no information on the time intervals between medication doses 1951. Electronic monitoring was used in addition to pill count in the AASK Pilot Study. Participants who were adherent by electronic monitoring and pill count were more likely to achieve goal blood pressure than those who were nonadherent by one or both methods 1431.

In AASK, blood pressure levels in goal were observed among only 17.5% and 25.6% of the low- and usual MAP goal participants, respectively. Among partici- pants assigned to the usual MAP goal, 33% achieved blood pressure levels lower than 102 mm Hg. In comparison to the MDRD Study, blood pressure goals were achieved less frequently in AASK. Factors other than the short duration of follow- up in AASK may have influenced achievement of goal MAP. For example, there were several differences between the AASK and MDRD study populations. The MDRD Study participants were heterogeneous with respect to their underlying renal disease 151. In the MDRD Study, only 8% of the study population were African Americans, 13.8% were normotensive, and 30% had renal function that was more severely compromised (13-24 ml/min/1.73 m2) [5] than in the AASK population.

MDRD Study participants were not randomly assigned to specific classes of antihypertensive drug therapy, a high proportion of MDRD Study participants had blood pressures in goal at baseline, and there was no lower limit to the usual MAP goal in the MDRD Study 1841. Because there was substantial overlap in MDRD between the two blood pressure goal groups with respect to achieved blood pres- sure, a lower bound was added to the usual goal group in AASK to assess the feasibility of attaining separation between the two goal groups. In AASK, the mean (2 SE) difference is achieved MAP levels was 11.4 mm Hg (+ 3.9) at 6 months, which reflected a significant difference between the two goal groups 1171. Among black MDRD Study participants, an average difference of 6.8 mm Hg between the two goal groups was achieved over 12 months [lo]. Recently, Toto et al. [S] observed a mean difference in MAP of 5.5 mm Hg between patients with presumed hyper- nephrosclerosis (75% black) randomized to strict (DBP 65-80 mmHg1 compared to conventional (85-95 mm Hg) blood pressure control.

Both blood pressure goal groups in AASK might be expected to have difficulty achieving an acceptable level of medication adherence. Since the low-MAP goal is lower than the currently recommended level of blood pressure control, participants may have been prescribed higher dosages and more antihypertensive drugs than they had taken prior to joining the AASK Pilot Study. Participants in the low-MAP goal group experienced more symptoms than those assigned to the usual MAP goal group 1151. To achieve blood pressure levels within the narrow window required to achieved goal in the usual MAP goal group may have required frequent

Appointment Attendance, Pill Counts, and Goal Blood Pressure 39s

dosage changes, which may have posed an additional barrier to medication ad- herence.

CONCLUSIONS

Although multiple adherence promotion strategies were implemented in the AASK Pilot Study and participants maintained a high attendance rate, medication adherence as measured by pill count was relatively low and blood pressure goals were infrequently achieved. The lack of correlation between pill count and achieve- ment of target blood pressure suggests that the pill count may be an inadequate measure of adherence. In order to enhance adherence in the full-scale AASK trial, more innovative strategies need to be implemented and better measures for as- sessing adherence should be considered.