824
Adherence of Human Immunodeficiency Virus–Infected Patients to Antiretroviral Therapy Nina Singh, Stephen M. Berman, Susan Swindells, Janice C. Justis, Jeffrey A. Mohr, Cheryl Squier, and Marilyn M. Wagener
From the Veterans Affairs Medical Center, Pittsburgh, Pennsylvania; Veterans Affairs Medical Center, Long Beach, California; and University of Nebraska Medical Center, Omaha, Nebraska
The impact of demographic, psychosocial, and medical regimen–related variables on adherence of 123 human immunodeficiency virus (HIV)–infected patients to antiretroviral therapy was assessed by means of refill methodology. Satisfaction with social support (P 5 .029), problem-focused coping (P 5 .027), and active-behavioral coping (P 5 .011) correlated significantly with adherence, whereas loss of motivation (P 5 .006), hopelessness (P 5 .16), and avoidant coping (p 5 .015) correlated with nonadherence. At the 6-month follow-up, the mean CD4 cell count differed significantly among adherent versus nonadherent patients (a mean increase of 78/mm3 vs. a mean decrease of 5/mm3; P 5 .018). Adherence did not correlate with the number of antiretroviral medications consumed per day (mean, 3.0 vs. 2.5). Non-Caucasian patients were more likely to be nonadherent than Caucasian patients (relative risk, 2.5; 95% confidence interval, 1.2–5.3; P 5 .013); this difference was not explained by age, education, employment, income, history of intravenous drug use, or medical regimen. Non-Caucasian patients, however, were less satisfied with their social support (P 5 .04) and informational support (P 5 .016) and were more likely to utilize emotion-focused coping (P 5 .01). Thus, satisfaction with social support and coping style significantly impacted adherence and likely accounted for the observed racial difference in adherence among HIV-infected patients.
The remarkable success of the newer antiretroviral therapeutic regimens, with their ability to achieve durable suppression of HIV replication, have transformed HIV infection into a chronic manageable disease [1]. Adherence to antiretroviral therapy, however, is critically important for the success of therapy [1, 2]. In an acute illness, the rewards or benefits of complying with therapy are immediately apparent to the patient, thereby creating a sense of accomplishment that reinforces adherent behavior [3]. Such a reinforcement may be lacking or may diminish over time in chronic diseases, as patients are required to remain adherent for prolonged or indefinite periods of time. In HIV infection, the immense complexity and demands posed by the difficult antiretroviral regimens may further undermine adherence. Suboptimal adherence ultimately has implications not only for significant restriction on future drug regimens but also for the potential transmission of drugresistant virus [4]. Thus, imperfect adherence in HIV infection, as compared to other chronic medical illnesses, has unparalleled and unprecedented relevance. A growing body of evidence suggests that social and psy-
Received 26 January 1999; revised 20 May 1999. Presented in part at the 36th Annual Meeting of the Infectious Diseases Society of America, held in Denver in November 1998. Reprints or correspondence: Dr. Nina Singh, Infectious Disease Section, Veterans Affairs Medical Center, University Drive C, Pittsburgh, Pennsylvania 15240. Clinical Infectious Diseases 1999;29:824 –30 © 1999 by the Infectious Diseases Society of America. All rights reserved. 1058 – 4838/99/2904 – 0017$03.00
Downloaded from https://academic.oup.com/cid/article-abstract/29/4/824/451590 by guest on 05 June 2018
chological variables, including social support, are amongst the most significant factors that influence adherence to medical therapy [5–7]. In a review of the adherence literature in the non-HIV setting, 64% (16) of the 25 published studies that included social support as a variable showed a positive correlation between social support and adherence, only 4% (1) showed an inverse correlation, and 32% (8) showed no correlation with adherence [8]. HIV-infected patients, because of their physical debilitation and the psychosocial impact of their infection, remain particularly vulnerable to social isolation [9]. Sources of social support, even if existent, may not be perceived as being satisfactory [10, 11]. No study to our knowledge, however, has assessed the impact of social and psychological variables on adherence by HIV-infected patients. In a prospective, multicenter study, we assessed the role of demographic, psychosocial, quality-of-life, medical-regimen, and illness-related characteristics on adherence by patients with HIV infection to antiretroviral therapy.
Methods Consecutive HIV-infected patients followed in an HIV clinic at the three participating medical centers between March 1996 and December 1997 were eligible for inclusion in the study. Patients who were in a moribund state, were unable to read and complete the study questionnaire, or refused to give informed consent were excluded. The study was approved by the institutional review board at each of the participating centers, and all patients provided written informed consent. The data form included demographic characteristics, stage of HIV infection,
CID 1999;29 (October)
Adherence of HIV-Infected Patients to Therapy
CD4 cell counts, antiretroviral drugs prescribed, and numbers of refills per patient. The questionnaire and the psychosocial assessment were completed at baseline and at 6 months. The baseline levels were used as dependent measures for the endpoints and the 6-month levels to assess any change from baseline. Psychosocial measures assessed were as follows. Quality of life was assessed by the Medical Outcome Study Health Survey Short Form–36 (MOS-SF36) [12, 13]. This self-rated 36-item form measures overall quality of life and the domains of physical functioning, role functioning, social functioning, mental health, health perceptions, and pain. Higher scores are indicative of better quality of life. Satisfaction with and the availability and sources of social support were assessed by the Social Support Questionnaire [14, 15]. Three types of social support were assessed: (1) emotional support, defined as receipt of emotional comfort that serves to gratify an individual’s basic social needs for nurturance, approval, esteem, and belonging; (2) tangible, instrumental, or practical support, defined as provision of material aid, physical assistance, or help in crises; and (3) informational support, which included receipt of advice, guidance, feedback, or information on a variety of issues, including the patient’s medical illness. Patients rated their satisfaction with social support on a 4-point scale, from very satisfied to not at all satisfied, and higher scores were indicative of lower satisfaction, as previously reported [15]. For each type of support, the source or sources of the particular social support were also assessed. Coping was assessed by the Billings and Moos Inventory of Coping with Illness Styles [16]. The focus of coping assessed was as follows: problem-focused coping, which included attempts to modify or eliminate the sources of stress through behavior, or emotion-focused coping, which included cognitive responses whose primary function was to manage the emotional consequences of stressors in order to maintain emotional equilibrium. Methods of coping responses were categorized as activecognitive, active-behavioral, or avoidance, as reported previously [16]. Active-cognitive coping included attempts to manage the stressfulness of events by using thoughts and attitudes that defend against concerns or create meaning out of a dismal situation [16, 17]. Active-behavioral coping is overt behavioral attempts to deal directly with the problem and its effects. Avoidance coping is attempts to avoid confronting the problem, denying that the problem exists, or indirectly reducing emotional tension by behaviors such as eating or smoking. Hopelessness was assessed by the Beck Hopelessness Scale [18]. Total hopelessness scores, as well as subscores for factors labeled as feelings about the future, loss of motivation, and future expectations, were assessed. The scale consisted of 20 true-false statements. The possible scores ranged from 0 to 20; higher scores were indicative of greater hopelessness [18].
Downloaded from https://academic.oup.com/cid/article-abstract/29/4/824/451590 by guest on 05 June 2018
825
Adherence was assessed by the refill methodology reported by Monane et al. [19]. This method has been employed previously in the assessment of adherence of HIV-infected patients [20]. All patients filled their prescriptions exclusively through the site pharmacy. Patients were provided a 30-day supply of medications each month by the pharmacy. Prescription refills were picked up in person by the patients each month from the site pharmacy. Mailing out the refill at the request of the patient was allowed, provided the care providers or the pharmacy was notified by the patient. Automatic mailing out of prescriptions was not permitted at any of the participating institutions. There was no other source for acquiring or filling the antiretroviral drug prescriptions for the study patients. Computerized pharmacy refill records were reviewed for each patient on a monthly basis. Patients who obtained ,90% of the prescribed refills for any of the antiretroviral drugs during the preceding 6 months were considered nonadherent. The demographic characteristics and medication and illnessrelated data were recorded by the health care providers (J. A. M., J. C. J., and C. S.), and the psychosocial data were provided by the patients; all self-administered patient forms were checked for completeness and accuracy. Statistical Analysis
Patient demographics (age, sex, underlying liver diseases, etc.) and psychiatric measures (Beck [Hopelessness Scale] scores and coping assessments) were entered into the PROPHET Statistics program (version 5.0; BBN Systems and Technologies, Cambridge, MA). The x2 or Fisher’s exact test was used to compare categorical variables (presence or absence of a condition). Continuous variables (age, MOS-SF36 scores, Beck scores, etc.) were examined with use of the t-test or the Mann-Whitney test. Baseline and follow-up values were compared with the paired t-test or the Wilcoxon signed-rank test. To examine the relationship between two continuous measures, we obtained a best-fit line by the method of least squares. The slope of this line was then tested against the null hypothesis of a slope of zero. A Pearson r correlation coefficient was also calculated. Relative risks (RRs) and 95% confidence intervals were calculated on contingency data with use of Epistat (version 5.3; Epistat Services, Richardson, TX). Results Patients’ Characteristics
Of 147 patients eligible for inclusion in the study, 14 refused to give informed consent; none were excluded because of illiteracy. Of 138 patients enrolled, a 6-month follow-up for the refill-methodology assessment of adherence was completed for 123, of whom the study population was comprised (3 patients died within 6 months and 7 were lost to follow-up). The
826
Singh et al.
patients lost to follow-up did not differ significantly from those in the study sample with regard to age, race, CD4 cell count, employment, or satisfaction with social support. The study patients ranged in age from 24 to 71 years (median, 41 years). Ninety-three percent were male. Seventy-two percent of the patients were Caucasian and 28% were nonCaucasian. The latter included 19% black, 6% Hispanic, and 1% Asian. The predominant HIV-infection risk factors were homosexual/bisexual lifestyle (71%) and intravenous drug use (14%). Forty-seven percent (58) of the 123 patients were employed. The median CD4 cell count at baseline was 269/ mm3 (range, 4 –796/mm3); 39% (48) of the patients had CD4 cell counts of ¶200/mm3. The Karnofsky Performance score ranged from 60 to 100 (median, 90). Patients from one of the sites (Omaha) were more likely to be younger and Caucasian and comprised a greater number of female patients. Only 7% of the patients were antiretroviral therapy–naive and were beginning treatment with an antiretroviral regimen, and 93% were receiving a stable regimen.
Determinants of Adherence
Eighteen percent (22) of the 123 patients were nonadherent on the basis of the criteria outlined in the Methods section. The mean (median) adherence rate among the “nonadherent” patients was 72% (77%) and ranged from 0% to 87%; 77% (12 of 22) had adherence rates of 70% to 90%, 14% (3 of 22) of 50% to 70%, 5% (1 of 22) of 20% to 50%, and 5% (1 of 22) of ,20%. Age, income, employment, education, intravenous drug use, homosexual/bisexual lifestyle, diagnosis of AIDS, or baseline CD4 cell count did not correlate with refill adherence (table 1). Non-Caucasian patients were significantly more likely to be nonadherent than Caucasian patients (31% [11 of 35] vs. 12% [11 of 88]; RR, 2.5; 95% CI, 1.2–5.3; P 5 .013). Neither living alone nor having a significant other was associated with refill adherence. Satisfaction with one’s social support, however, was a significant predictor of adherence; nonadherent patients reported significantly less satisfaction with their overall social support (mean 6 SEM: 22.9 6 3.3 vs. 16.8 6 .75; P 5 .029). Lesser satisfaction with tangible social support (7.7 6 1.1 vs. 5.5 6 .31; P 5 .007) and informational social support (7.9 6 1.1 vs. 6.1 6 .32; P 5 .041) but not emotional social support (7.2 6 1.1 vs. 5.4 6 .3; P 5 .31) was predictive of nonadherence (table 2). Satisfaction with social support correlated with greater adherent behavior, regardless of the source of that support. Nonadherent patients demonstrated a trend toward greater hopelessness, i.e., had higher Beck hopelessness scores (mean 6 SEM: 6.4 6 1.4 vs. 3.8 6 .4; P 5 .16), greater negative feelings about the future (1.9 6 .5 vs. 1.2 6 .2; P 5 .18), and significantly higher loss of motivation (1.75 6 .5 vs. .6 6 .1; P 5 .006). A strong inverse correlation between
Downloaded from https://academic.oup.com/cid/article-abstract/29/4/824/451590 by guest on 05 June 2018
CID 1999;29 (October)
Table 1. Sociodemographic variables predictive of adherence of HIV-infected patients to antiretroviral therapy. Percentage (no.) of patients
Variable Age (mean), y Race Caucasian Non-Caucasian Education Grade school Technical school High school College Postgraduate Risk behavior Male-to-male sex Other(s) IV drug use No iv drug use Income ($/mo) 0–500 500–1,000 1,000–1,500 .1,500 Not stated Employed Yes No Household status Living alone Living with someone Have a significant other (spouse or partner) Yes No
Nonadherent Adherent P (n 5 22) (n 5 101) value RR* (95% CI) 41.9
41.0
50 (11) 50 (11)
76 (77) 24 (24)
0 (0) 9 (2) 36 (8) 55 (12) 0 (0)
5 (5) 4 (4) 43 (43) 41 (41) 8 (8)
64 (14) 36 (8) 18 (4) 82 (18)
27 (27) 73 (74) 13 (13) 86 (87)
23 (5) 41 (9) 5 (1) 23 (5) 9 (2)
17 (17) 33 (33) 26 (26) 22 (22) 3 (3)
41 (9) 59 (13)
49 (49) 51 (52)
27 (6) 73 (16)
23 (23) 77 (78)
.81 .013
... 2.5 (1.2–5.3)
.31
...
.36
.69 (.32–1.5)
.51
1.4 (.53–3.6) ...
41 (9) 59 (13)
.125
.52
1.3 (.59–.28)
.65
1.2 (.52–2.8)
.82
1.1 (.51–2.4)
44 (44) 56 (57)
* Calculated on presence of risk factor in dichotomized variables: nonCaucasian, male-to-male sex, iv drug use, unemployment, living alone, and no significant other.
satisfaction with social support and loss of motivation was documented. As satisfaction with tangible social support (coefficient of determination [r2] 5 0.171; P 5 .05), informational support (r2 5 0.292; P 5 .001), emotional support (r 2 5 0.3369; P 5 .0001), and overall social support (r 2 5 0.3399; P 5 .0003) declined, loss of motivation score on the Beck Hopelessness Scale increased. There was a significant correlation between coping style and refill adherence of the patient. Whereas problem-focused coping (mean 6 SEM: 7.1 6 .2 vs. 6.0 6 .5; P 5 .027) and active-behavioral coping (6.6 6 .2 vs. 5.2 6 .5; P 5 .011) were associated with significantly higher refill adherence, avoidant coping was associated with lower refill adherence (3.3 6 .3 vs. 2.6 6 .2; P 5 .015). There was no association between emotion-focused coping and active-cognitive coping and refill
CID 1999;29 (October)
Adherence of HIV-Infected Patients to Therapy
Table 2. Psychosocial variables nonadherent vs.-adherent patients.
for
antiretroviral
therapy–
Mean (6SEM) score
Variable Satisfaction with social support* Emotional Tangible Informational Total support Beck hopelessness score† Loss of motivation Negative feelings about future Future expectations Total hopelessness score Coping style Problem-focused Emotion-focused Active-cognitive Active-behavioral Avoidant coping Quality of life‡ (MOS-SF36) score Health perceptions Physical functioning Physical limitations Emotional limitations Social functioning Pain Psychological functioning Vitality Total quality of life score
Nonadherent patients
Adherent patients
P value
7.1 (61.1) 7.7 (61.1) 7.9 (61.1) 22.9 (63.3)
5.4 (6.3) 5.5 (6.3) 6.1 (6.3) 16.8 (6.75)
.31 .0065 .04 .029
1.75 (6.5) 1.9 (6.4) 2.3 (6.4) 6.4 (61.4)
.6 (6.1) 1.2 (6.2) 1.8 (6.2) 3.8 (6.4)
.006 .18 .21 .16
6.0 (6.5) 6.8 (6.5) 4.8 (6.4) 5.2 (6.5) 3.3 (6.3)
7.1 (6.2) 6.6 (6.2) 5.4 (6.1) 6.6 (6.2) 2.6 (6.2)
.02 .72 .28 .011 .015
18.5 (61.2) 23.8 (61.1) 6.3 (6.3) 4.6 (6.3) 5.9 (6.2) 7.1 (6.6) 20.2 (61.5) 13.9 (61.1) 100.4 (64.9)
19.8 (6.6) 25.6 (6.4) 6.4 (6.2) 5.1 (6.1) 6.2 (6.1) 7.4 (6.2) 22.4 (6.5) 14.3 (6.4) 107.5 (62.0)
.39 .18 .64 .11 .24 .71 .21 .72 .23
* Higher numbers indicate less satisfaction. † Higher numbers indicate more helplessness. ‡ Higher numbers indicate better quality of life; MOS-SF36 5 Medical Outcome Study Health Survey Short Form (36 items).
adherence (table 2). Quality of life scores did not correlate with refill adherence (table 2). The number of antiretroviral medications consumed by the patient per day did not differ significantly between patients who were nonadherent and those who were adherent with obtaining refills (mean, 2.5 vs. 3.0). Overall, 59% (72) of the 123 patients were receiving treatment with a regimen that included a protease inhibitor; 14% of these patients were nonadherent, as were 24% of those whose regimen did not include a protease inhibitor (P 5 .17). Adverse effects attributable to antiretroviral therapy were commonly documented. However, there was no difference in the rate of reported side effects in adherent patients (56%) as compared with nonadherent patients (58%; P 5 .9). At 6 months, the change in CD4 cell count was significantly different in adherent vs. nonadherent patients. Overall, whereas the CD4 cell count had increased by a mean of 77.7 cells/mm3 in the adherent patients, it declined by a mean of 5.1 cells/mm3 in the nonadherent patients (P 5 .018). Amongst patients
Downloaded from https://academic.oup.com/cid/article-abstract/29/4/824/451590 by guest on 05 June 2018
827
whose regimen included a protease inhibitor, the adherent patients had a mean increase in CD4 cells of 76.1/mm3, and nonadherent patients had a mean decrease of 6.7/mm3 (P 5 .2). Amongst patients not receiving a protease inhibitor, the adherent patients had a mean increase in CD4 cells of 44/mm3 and nonadherent patients had a mean decrease of 14/mm3 (P 5 .08). In order to explain the association between race and refill adherence, racial differences in a number of demographic and psychosocial variables were analyzed (table 3). There were no differences in age, education, employment, income, history of intravenous drug abuse, diagnosis of AIDS, or CD4 cell count at baseline between Caucasian and non-Caucasian patients (table 3). Non-Caucasian patients reported significantly lower satisfaction with their overall social support (mean 6 SEM: 21.5 6 2.2 vs. 16.5 6 .7; P 5 .04) and informational support (7.7 6 .7 vs. 5.9 6 .3; P 5 .016). Non-Caucasian patients were also more likely to utilize emotion-focused coping (7.5 6 .3 vs. 6.3 6 .2; P 5 .01) and cognitive coping (5.8 6 .2 vs. 5.0 6 .2; P 5 .01). There was no difference in Beck hopelessness scores between the Caucasian and non-Caucasian patients (4.5 vs. 4.1). The proportion receiving protease inhibitors (51% vs. 61%) and the number of antiretroviral drugs consumed daily (mean, 2.5 vs. 2.6) did not differ significantly between Caucasian and non-Caucasian patients (table 3). Discussion Although it is intuitively logical to assume that adherence would lead to an improvement in virological surrogate markers, e.g., CD4 cell count, our study provides objective evidence of this association. CD4 cell count differed significantly between the patients who were adherent vs. nonadherent with prescription refills on follow-up. At the 6-month follow-up, adherent patients in our study had increases in their CD4 cell counts, whereas the nonadherent patients were shown to have decreases in their CD4 cell counts. While complexity of the therapeutic regimen and medical illness–related variables can clearly undermine adherence, these represent relatively enduring and, in many cases, unalterable variables. Adherence did not correlate with the number of antiretroviral drugs taken per day in our study. Other studies have reported similar findings [20, 21]. Indeed, Becker and Maiman have shown that despite similar complexity and adverse characteristics of the therapeutic regimens likely to be associated with a high probability of default, some patients are able to follow the recommended therapies, whereas others are not [7]. These authors have instead identified social behavior factors as being critical determinants of adherence and more productive dimensions for intervention than are the therapeutic regimen or illness-related characteristics [7]. Our study therefore focused primarily on psychosocial factors as determinants of adherence among HIV-infected patients.
828
Singh et al.
CID 1999;29 (October)
Table 3. Racial differences in demographic and psychosocial characteristics of the HIV-infected patients whose adherence to antiretroviral therapy was studied.
Variable
Non-Caucasian Caucasian patients patients (n 5 35) (n 5 88) P value RR* (95% CI)
Age, mean (y) Employed Risk factor Male-to-male sex IV drug use Education Less than high school Technical school High school College Postgraduate Income ($/mo) 0–500 500–1,000 1,000–1,500 .1,500 Living alone Having significant other AIDS diagnosis CD4 cells/mm3 (baseline, mean) Satisfaction with social support (mean score 6 SEM) Tangible Informational† Emotional† Overall support† Coping style Problem-focused Emotion-focused‡ Active-cognitive‡ Active-behavioral Avoidance Beck hopelessness score Loss of motivation Negative feelings about future Future expectations Total hopelessness score Adherent to therapy With regimen including protease inhibitors Total no. of antiretroviral drugs taken per d
40.6 51% (18/35)
40.7 45% (40/86)
.307 .549
... 1.1 (.76–1.7)
54% 15%
78% 14%
.007 .88 .74
.69 (.5–.96) 1.1 (.4–2.8) ...
3% 30% 49% 37% 9%
5% 6% 39% 45% 6% .89
...
21% 32% 21% 26% 26% 29% 57% 280
18% 37% 24% 21% 23% 49% 62% 278
.72 .04 .58 .95
1.1 (.6–2.2) .58 (.33–.98) .91 (.65–1.3) ... ...
6.7 7.7 6.6 21.5
(6.8) (6.7) (6.7) (62.3)
5.5 5.9 5.4 16.5
(6.3) (6.3) (6.3) (6.8)
.56 .016 .07 .04
6.9 7.5 5.8 6.3 3.1
(6.3) (6.4) (61.2) (6.3) (6.3)
6.9 6.3 5.0 6.3 2.5
(6.2) (6.2) (6.2) (6.2) (6.2)
.39 .01 .01 .73 .10
...
... .87 (6.3) 1.2 (6.3) 2.1 (6.3) 4.5 (6.8) 69% (24/35) 51% 2.5 (6.1)
.76 (6.2) 1.3 (6.2) 1.8 (6.2) 4.1 (6.5) 885 (77/88) 61% 2.6 (6.1)
.34 .45 .34 .59 .013 .31 .29
2.5 (1.2–5.3) .84 (.58–1.2) ...
* Calculated for dichotomized variables. Nonwhite patients were significantly less satisfied with their emotional, informational, and overall support. ‡ Nonwhite patients were significantly more likely to use emotion-focused coping, cognitive coping, and avoidant coping. †
Social support, coping style, and loss of motivation significantly influenced adherence in our study. Social support may enhance adherence, either directly (through encouragement, reassurance, reinforcement, systematic cues, bolstering of competence, and motivation) or indirectly (by buffering the effect of those variables that might interfere with adherence, e.g., life’s stresses, anxiety, and depression). The three major components of social support are proposed to be tangible, emotional, and informational support [11, 14, 15]. Distinction
Downloaded from https://academic.oup.com/cid/article-abstract/29/4/824/451590 by guest on 05 June 2018
among these is important since their functions are not necessarily interchangeable. Social support must also be conceptualized in terms of not merely availability but also its perceived adequacy [10, 11, 22]. Support may not be considered helpful unless the individual perceives it as being supportive. Our data show that satisfaction with overall support, in particular, satisfaction with tangible and informational support, significantly correlated with adherence. In the non-HIV setting, informational support has been
CID 1999;29 (October)
Adherence of HIV-Infected Patients to Therapy
shown to empower patients with skills that enhanced adherent behavior; those with greater knowledge of their treatment regimen had better adherence, had fewer somatic complaints, and coped more effectively with the demands of their regimen [3]. An important source of informational support can be health care providers. It has been shown that hypertensive patients who perceived greater social support from the physician were significantly more likely to regard the consequences of nonadherence as being serious [3]. Given the uncertainties, complexities, and continual evolution of therapeutic and management strategies, the implications of the adequacy of informational support are even more relevant for HIV-infected patients. Thus, adherence-enhancing measures for HIV-infected patients must ensure that their tangible and informational social support needs are met. Respondents’ “fighting spirit” and faith in the future are reflected in low scores on the Beck Hopelessness Scale. High scores indicate hopelessness, which is considered a core characteristic of depression. Patients nonadherent with their refills in our study tended to have greater hopelessness, i.e., lacked fighting spirit and demonstrated significantly greater loss of motivation. Depression has been shown to be associated with devaluation of self-worth, lack of mastery of one’s environment, and unhealthy self-caring behaviors, including inability to transform information into effective self-care strategies [23]. In a study of hypertensive patients, depression was significantly associated with a failure to promptly refill prescriptions [3]. In a previous study of HIV-infected patients, depression was significantly associated with nonadherence with zidovudine therapy [20]. Therefore, lack of fighting spirit, loss of motivation, and hopelessness were identified as characteristics of the patients at risk for nonadherence with therapy. Coping strategies utilized by the patients were significantly predictive of their adherence with filling prescriptions. Individuals who confronted the stress with problem-solving and behavior-modifying approaches were significantly more likely to be adherent than those who coped by denial or who utilized emotions to create meaning out of dismal situations. Problemsolving and active-behavioral coping have previously been shown to be associated with lower total mood disturbance, higher self-esteem, and better quality of life [17, 24, 25]. Patients using avoidant coping, on the other hand, had the highest levels of depression and lowest self-esteem [24]. Avoidant coping or denial may indeed be an expression of helplessness and despair [24]. Thus, psychological interventions that encourage or enhance problem-solving behavior rather than passive acceptance or submission to the illness may improve adherence. Recognition of denial or avoidant coping would also be relevant in identifying patients at risk for nonadherence or defaulting. Adherence with refills of antiretroviral drugs was significantly lower among nonwhite patients in our study. Similar
Downloaded from https://academic.oup.com/cid/article-abstract/29/4/824/451590 by guest on 05 June 2018
829
observations have been made by other investigators, although the precise variables contributing to the racial disparity in adherence rates have not been well-defined [20, 26, 27]. Our data show that while education, employment, depression, income, stage of HIV infection, and medical regimen were not different, dissatisfaction with overall social, informational, and emotional support and use of unhealthy coping strategies were significantly more likely in non-Caucasian patients than in Caucasian patients. In a previous study group comprising asymptomatic HIVpositive men, African American patients were documented to be more likely to cope by feeling helpless, by denial, and by turning to religion and were less likely to seek emotional support to cope with AIDS [24]. Non-Caucasian patients in our study were more likely to utilize cognitive and emotionfocused coping. Such coping has been shown to lead to ruminations and obsessive thinking and may not contribute substantively to an individual’s psychological well-being [17, 25]. Thus, less satisfaction with their social support and use of unhealthy coping strategies may largely explain the higher refill-nonadherence rate among non-Caucasian patients. Our study, however, has limitations that must be acknowledged. At the time of initiation of this study, employment of HIV-RNA levels had just come to be accepted as a standard of care but was not yet routine. Thus, CD4 cell counts were primarily utilized as the surrogate markers to follow disease progression. The primary objective of our study, however, was not to demonstrate whether nonadherence led to disease progression but rather to discern variables predictive of nonadherence. Second, a “gold standard” for assessment of adherence does not exist. Patients’ self-reports have been consistently shown to overestimate adherence, both in the HIV setting and non-HIV setting [21, 28, 29]. Although pill counts have been utilized extensively in the non-HIV setting, such methodology can have important drawbacks for the assessment of adherence by HIVinfected patients. Patients are often prescribed a number of antiretroviral agents and would be required to bring the container of every medication during every clinic visit. Given the large number of antiretroviral pills prescribed, pill counting for the entire regimen can be cumbersome and time-consuming. Furthermore, many of the patients utilize pill boxes rather than taking pills directly from the bottles. The expense of electronic monitoring devices was considered prohibitive, given that adherence in our study was assessed for the entire antiretroviral regimen and not a particular drug. We therefore utilized the pharmacy refill method as a measure of adherence [2]. This method has been used previously for the assessment of adherence among HIV-infected patients and has been identified as a reliable measure of adherence as long as the prescriptions are consistently filled from the same source [19], as was the case for our patients. Finally, our patient population comprised primarily males
830
Singh et al.
and those with HIV risk factor(s) other than intravenous drug use. Whether the study findings would differ in other subpopulations of HIV-infected patients remains to be determined. In summary, dissatisfaction with one’s social support, lack of fighting spirit, loss of motivation, and lesser use of problemfocused coping were the identified variables that significantly increased the likelihood of nonadherence among HIV-infected patients. Strategies to enhance adherence should include psychological and social support interventions targeted particularly toward patients thus identified. Differences in satisfaction with social support and coping style may also account for the observed racial differences in adherence rates. References 1. Altice FL, Friedland GH. The era of adherence to HIV therapy. Ann Intern Med 1998;129:504 –5. 2. Hecht FM. Measuring HIV treatment adherence in clinical practice. AIDS Clin Care 1998;10:57–9. 3. Caplan RD, Robinson EAR, French JRP, et al. Adhering to medical regimens: pilot experiments in patient education and social support. Ann Arbor, Michigan: University of Michigan, 1976. 4. Hecht FM, Grant RM, Petropoulos CJ, Dillon B, Chesney MA, Tian Heal. Sexual transmission of an HIV-1 variant resistant to multiple reverse transcriptase and protease inhibitors. N Engl J Med 1998;339:307–11. 5. Janis IL. The role of social support in adherence to stressful decisions. American Psychologist 1983;38:143–59. 6. Levy RL. Social support and compliance. J Hypertension 1995;3:45–9. 7. Becker MH, Maiman LA. Social behavioral determinants of compliance with health and medical care recommendations. Med Care 1975;13:10–23. 8. Haynes, RB, Sackett DL. A workshop/symposium: compliance with therapeutic regimens. Annotated bibliography. Hamilton, Ontario: McMaster University Medical Centre, 1974. 9. Namir S, Wolcott DL, Fawzy FI. Social support and HIV spectrum disease: clinical and research perspectives. Psychiatr Med 1989;7:97– 104. 10. Bruhn JG, Philips BU. Measuring social support: a synthesis of current approaches. J Behav Med 1984;2:151– 69. 11. Schaefer C, Coyne JC, Lazarus RS. The health-related functions of social support. J Behav Med 1981;4:381– 407. 12. Ware JE, Sherbourne CD. The MOS 36-item short-form health survey, conceptual frame work and item selection. Med Care 1992;30:473– 83. 13. Wachtel T, Piette J, Mor V, Stein M, Fleishman J, Carpenter C. Quality of life in persons with human immunodeficiency virus infection: measurement by the medical outcomes study instrument. Ann Intern Med 1992;116:129 –37.
Downloaded from https://academic.oup.com/cid/article-abstract/29/4/824/451590 by guest on 05 June 2018
CID 1999;29 (October)
14. Sarason IG, Levine HM, Basham RB, Sarason BR. Assessing social support: the social support questionnaire. J Pers Soc Psychol 1983;44: 127–39. 15. Hays RB, Turner H, Coates TJ. Social support, AIDS-related symptoms, and depression among gay men. J Consult Clin Psychol 1992;60:463–9. 16. Billings AG, Moos RM. The role of coping responses and social resources in attenuating the stress of life events. J Behav Med 1981;4:139 –57. 17. Wolf TM, Balson PM, Morse EV, et al. Relationship of coping style to affective state and perceived social support in asymptomatic and symptomatic HIV-infected persons: implications for clinical management. J Clin Psychiatry 1991;52:171–3. 18. Beck AT, Weissman A, Lester D, Trexler L. The measurement of pessimism: the hopelessness scale. J Consult Clin Psychol 1974;42:861–5. 19. Monane M, Gurwitz JH, Monane S, Avorn J. Compliance issues in medical practice. Hospital Physician 1993;(April):35–9. 20. Singh N, Squier C, Sivek C, Wagener M, Nguyen MH, Yu VL. Determinants of compliance with antiretroviral therapy in patients with human immunodeficiency virus: prospective assessment with implications for enhancing compliance. AIDS Care 1996;8:261–9. 21. Paterson DL, Swindells S, Mohr JA, et al. Adherence with protease inhibitor therapy for human immunodeficiency virus infection [abstract no I-72]. In: Program and abstracts of the 38th International Conference on Antimicrobial Agents and Chemotherapy (San Diego). Washington, DC: American Society for Microbiology, 1998. 22. Cutrona CE. Objective determinants of perceived social support. J Pers Soc Psychol 1986;50:349 –55. 23. Valente SM, Saunders JM, Ulman G. Self-care, psychological distress and HIV disease. J Assoc Nurs AIDS Care 1993;4:15–25. 24. Leserman J, Perkins DO, Evans DL. Coping with the threat of AIDS: the role of social support. Am J Psychiatry 1992;149:1514 –20. 25. Namir S, Wolcott DL, Fawzy FI. Coping with AIDS: psychological and health implications. J Applied Social Psychology 1987;17:309 –28. 26. Easterbrook PJ, Keruly JC, Creagh-Kirk T, et al., and The Zidovudine Epidemiology Study Group. Racial and ethnic differences in outcome in zidouvudine treated patients with advanced HIV disease. JAMA 1991; 266:2713– 8. 27. Muma RD, Ross WW, Parcel GC, Polland RB. Zidouvudine adherence among individuals with HIV infection. AIDS Care 1995;7:439 – 48. 28. Melbourne K, Geletko S, Brown S, et al. Electronic adherence assessment versus self-report in HIV-infected individuals [abstract no I-75]. In: Program and abstracts of the 38th International Conference on Antimicrobial Agents and Chemotherapy (San Diego). Washington, DC: American Society for Microbiology, 1998. 29. Haynes RB, McKibbon KA, Kanani R. Systematic review of randomized trials of interventions to assist patients to follow prescriptions for medications. Lancet 1998;348:383– 6.