AIDS PATIENT CARE and STDs

Volume 21, Number 3, 2007

© Mary Ann Liebert, Inc.
DOI: 10.1089/apc.2006.0077

Improvement of Psychosocial Adjustment to HIV-1

Infection through a Cognitive-Behavioral Oriented
Group Psychotherapy Program: A Pilot Study

ARACELI ROUSAUD, Psych.,1 JORDI BLANCH, M.D., Ph.D.,2

MARTIN HAUTZINGER, Ph.D.,3 ELISA DE LAZZARI, Ph.D.,4 JOSEP MARIA PERI, Ph.D.,2
OLGA PUIG, Psych.,2 ESTEBAN MARTINEZ, M.D., Ph.D.,5
GUILLEM MASANA, M.D., Ph.D.,2 JOAN DE PABLO, M.D., Ph.D.,2
and JOSEP MARIA GATELL, M.D., Ph.D.5

ABSTRACT

The present study evaluated the efficacy of a group therapy program in improving psychosocial
adjustment to HIV infection, and tried to identify variables predictive of greater improvement.
The outcome of 47 completing patients was analyzed, comparing the measures between T1 (1
month before therapy), and T2 (first session), and between T2 and T3 (last session) using the
Wilcoxon matched-pairs signed-ranks test for each dimension of the Psychosocial Adjustment
to Illness Scale (PAIS). The therapy consisted of 16 weekly 2-hour sessions following a struc-
tured time-limited cognitive-behavioral group psychotherapy program. During the intervention
(between T2 and T3) a significant improvement was observed in health care orientation, voca-
tional environment, domestic environment, sexual relation, extended family relationships, social
environment, and total PAIS. There were no changes during baseline (between T1 and T2) in
any of the PAIS subscales, or in the total PAIS score. Sexual route of transmission was inde-
pendently associated with an improvement in health care orientation (  2.525). Time since HIV
diagnosis (  0.022) and being employed (  2.548) were independently associated with an
improvement in adjustment to vocational environment. Men who have sex with men showed a
poorer improvement in adjusting to family relations after the intervention (  2.548). Finally,
a lower CD4 count (  0.005) and being employed (  3.054) were independently associated
with an improvement in adjustment to social environment. Our psychotherapy program im-
proved psychosocial functioning in a heterogeneous sample of HIV-1–infected patients referred
to a consultation-liaison psychiatry unit.

INTRODUCTION life expectancies for persons living with HIV in

industrialized countries,1 quality of life and

S INCE RECENT PHARMACOLOGIC ADVANCES, par-

ticularly the advent of highly active anti-
retroviral therapy (HAART), have increased
psychosocial functioning have become impor-
tant concerns.2,3 Psychosocial adjustment is de-
fined as the capacity of an individual to per-

1Fundació Clínic per la Recerca Biomèdica-IDIBAPS, Institut d’Investigacions Biomèdiques Agustí Pi i Sunyer, Hos-

pital Clínic de Barcelona, Barcelona, Spain.

2Clinical Institute of Neurosciences, Hospital Clínic de Barcelona, Barcelona, Spain.
3Clinical and Physiological Psychology Department, Eberhard-Karls-University of Tübingen, Tübingen, Germany
4Epidemiology and Biostatistics Unit, Hospital Clínic de Barcelona, Barcelona, Spain.
5Infectious Diseases and Microbiology Department, Hospital Clínic de Barcelona, Barcelona, Spain.

212
IMPROVING ADJUSTMENT TO HIV THROUGH GROUP THERAPY 213

form social and domestic roles so as to meet the Subjects

challenges of every day living without emo-
Patients were recruited from August 1998 to
tional distress or physical disability.4 The con-
July 2003. Inclusion criteria were proven HIV-
cept includes interactions between the individ-
positive status for at least 3 months and a clin-
ual and other individuals and the institutions
ical judgment of significant psychosocial dis-
that make up his or her sociocultural environ-
tress due to HIV-infection established by both
ment. Such interactions are usually achieved
the psychiatrist (J.B.) and the psychologist
through loosely prescribed behavioral patterns
(A.R.), and based on their clinical experience.
termed “roles.” The functional efficiencies of an
Exclusion criteria were nonstabilized axis-I
individual’s role behavior (e.g., as spouse, par-
psychiatric disorders where patients had to be
ent, professional) tend to be highly correlated
ascribed mainly to psychopharmacologic treat-
with judgments concerning his or her level of
ment (for example, mood disorders with risk
psychosocial adjustment. Furthermore, it has
of suicide, presence of psychotic symptoms,
been consistently observed in many major dis-
current frequent panic attacks), organic mental
eases that the nature of the patient’s psychoso-
disorders, substance abuse during the last 6
cial adjustment can be just as important as the
months, symptomatic non-HIV medical dis-
status of his or her physical disease in deter-
ease, significant cognitive impairment, and any
mining the quality of an illness experience.5
medical conditions that could interfere with at-
Although there is substantial evidence that
tendance at the therapy sessions.
psychotherapy interventions improve mood,6
adherence to treatment,7 and quality of life,8
few studies have specifically focused on the as- Intervention
sessment of the improvement of psychosocial Each group was closed and consisted of 8 to
adjustment through a psychotherapy interven- 12 participants, a leader or therapist (A.R.) who
tion.9–10 This is the case in spite of the consid- was a psychologist, and a coleader or cothera-
erable demand from HIV-infected patients for pist who was a psychiatrist (J.B.) or a psychol-
public mental health assistance, especially from ogist (O.P., J.M.P.). The therapy consisted of 16
those referred to a consultation-liaison psychi- weekly 2-hour sessions following a structured
atry department. To optimize use of public time-limited cognitive-behavioral psychother-
mental health resources and to benefit as many apy program. The intervention program has
patients as possible, short but efficient struc- been described elsewhere, and was based on a
tured interventions are needed. treatment manual.6,12 The main objectives of
As a continuation of a previous study pub- the therapy were to teach the participants to
lished elsewhere by our group,6 the aim of the use cognitive and behavioral skills and strate-
present study is to evaluate the efficacy of a struc- gies. Patients were introduced to exercises for
tured time-limited, cognitive-behavioral oriented training in problem-solving skills focused on
group therapy for HIV-infected patients who specific HIV-related concerns, and shown edu-
were referred to a consultation-liaison psychiatry cational presentations on HIV and mental
unit from the infectious disease department and health. Concerns and problems that are com-
who presented difficulties in the psychosocial ad- mon in the context of the illness13 were han-
justment to the HIV infection. The study also aims dled from a cognitive problem-solving per-
to identify baseline characteristics associated with spective: disclosure of serostatus, safer sex
improvement and successful outcome. practices if sexually active, treatment compli-
ance and adherence to medical controls, and
dealing with adverse effects such as lipodys-
MATERIALS AND METHODS trophy, isolation, lack of future plans, etc.
Subjects’ recruitment, type of intervention,
and study design were described in detail in a Measures
previous publication6 and will be summarized Measures were performed at three time
as follows. points: T1 or baseline (at least 1 to 2 months be-
214 ROUSAUD ET AL.

fore beginning of the therapy), T2 (during the could not implement a control group design. Pa-
first session of the group-therapy), and T3 (dur- tients were referred to our unit from the infec-
ing the last session). tious disease department for psychopharmaco-
At baseline, all recruited patients were inter- logic treatment or general psychotherapy. Most
viewed using the Spanish adaptation of the of the patients who were not permitted to enter
Structured Clinical Interview for DSM-III-R, a group because it was complete or already on-
patient version.14 The psychiatric interview going were not able to remain on a waiting list
also included questions about patient charac- for a long time, so they received interpersonal
teristics, as well as about HIV-related data such psychotherapy or psychopharmacologic treat-
as the date of knowledge of their serostatus, ment. This meant that we were unable to com-
route of transmission, current CD4 cell count pile a control group from a waiting list. For com-
and viral load, and opportunistic diseases. At parisons, we decided to use the patients
all time points (T1, T2, and T3) patients com- themselves as their own controls by performing
pleted the Psychosocial Adjustment to Illness two preintervention measures (T1 and T2).
Scale (PAIS).15
Given the predominance of physical dimen- Statistics
sions in most of the instruments used to measure
Patients who did not finish the study
quality of life in HIV infection, and the need to
(dropouts) were compared to those who did
provide a greater emphasis on psychosocial do-
(completers) using the Wilcoxon rank sum test
mains, we decided to use the PAIS, because it
for continuous variables and the 2 test or
was designed to assess functional, emotional,
Fisher’s exact test for categorical variables. The
and social dimensions of quality of life in patients
outcome of completing patients was analyzed
with medical illnesses.16 It evaluates the psy-
comparing the measures between T1 and T2,
chological and social adjustment to illness dur-
and between T2 and T3 using the Wilcoxon
ing the previous 4 weeks. We used the Spanish
matched-pairs signed-ranks test for each di-
validation of the questionnaire17 replacing the
mension of the PAIS. No outcome data were
original term “illness” with the term “HIV in-
collected on patients who withdrew from the
fection.” The PAIS comprises 46 questions that
study. Univariate and multivariate logistic re-
were divided into seven assessment domains: (1)
gression analysis were performed to identify
health care orientation; (2) vocational environ-
patient characteristics or baseline measures
ment, (3) domestic environment; (4) sexual rela-
able to predict better outcome, in terms of de-
tions; (5) extended family relationships, (6) social
creases on dimension scores between T2 and
environment; and (7) psychological distress. All
T3. Univariate logistic regression analysis was
items were also summed together for the total
used initially to screen for association between
score. This instrument has shown a good degree
each outcome variable and each of the baseline
of reliability for both individuals and groups.15
independent variables considered individu-
The seven domains are relatively independent of
ally. A stepwise multivariate logistic regression
each other in the assessment of the overall score,
analysis was used to create a multivariate
and each gives a measure of the PAIS validity.
model for the baseline factors associated with
Higher scores indicate poorer functioning.
changes on the PAIS dimension scores. All p
values were two-tailed and statistical signifi-
Design
cance was set at 0.05. All calculations were per-
Participants were followed for 20 weeks: 4 formed with the STATA 7.0 software.18
weeks of baseline (between T1 and T2) and 16
weeks of therapy (between T2 and T3). Be-
tween T1 and T2 the patients did not receive
psychotherapy of any kind, and this period RESULTS
served as a control condition in which no
Subjects’ characteristics
changes in the measures were expected. The
group therapy took place between T2 and T3. During the period of recruitment, of 82 out-
For ethical, clinical, and practical reasons we patients sent to our consultation-liaison psychi-
IMPROVING ADJUSTMENT TO HIV THROUGH GROUP THERAPY 215

atry unit from the infectious disease service, had started work 1 week before the first ses-
who were offered to participate in our study, 66 sion of the therapy; 4 patients did not answer
(80.5%) agreed, though only 47 finished the en- the PAIS questionnaire correctly or did not
tire program. Subjects were assigned to partici- complete it inside the established time-limits.
pate in 6 different groups. The mean number of Reasons for dropping out during the therapy
participants per group was 11. Table 1 shows included medical complications (6 patients),
the characteristics of the patients who agreed to beginning a new job (4 patients), not feeling
participate in the study, divided as follows: pa- comfortable with the group (3 patients), and
tients who finished the whole follow-up period imprisonment (1 patient).
(completers) and patients who were recruited Comparison of variables at baseline be-
but abandoned the study (dropouts). tween dropouts and patients who finished the
follow-up period (completers) did not show
significant differences on any of the following
Dropouts
variables at baseline: age, gender, sexual ori-
Of the 66 participants, 47 (31 men and 16 entation, route of transmission of HIV infec-
women) finished the complete follow-up pe- tion, time since HIV diagnosis, CD4 count,
riod. Most dropouts (19 patients, 29%) left the presence or absence of opportunistic diseases,
study during the intervention period. One pa- psychiatric diagnosis and PAIS dimension
tient did not come to the first visit because he scores (p  0.05 for each; Table 1).

TABLE 1. DEMOGRAPHIC AND CLINICAL CHARACTERISTICS OF STUDY PATIENTS:

DIFFERENCES BETWEEN COMPLETERS AND DROPOUTS (N  66)

Completers Dropouts Total

(n  47) (n  19) (n  66) p value

Age (years), median (IQR) 39 (35–46) 37 (33–44) 39 (35–45) n.s.

Gender (male), n (%) 31 (66) 11 (58) 42 (64) n.s.
Sexual orientation, n (%) n.s.
Heterosexual 28 (60) 10 (53) 38 (58)
MSM 19 (40) 9 (47) 24 (36)
With partner, n (%) 13 (28) 8 (42) 21 (32) n.s.
Route of transmission, n (%) n.s.
Sexual 34 (72) 18 (95) 52 (79)
Intravenous drug users 13 (28) 1 (5) 14 (21)
Time since HIV diagnose (months), median (IQR) 64 (18–110) 72 (26–104) 70 (19–104) n.s.
Previous opportunistic diseases, n (%) 8 (17) 6 (32) 14 (21) n.s.
Good compliance with HAART, n (%) 42 (89) 16 (84) 58 (88) n.s.
Psychiatric diagnoses, n (%) n.s.
Major depression 6 (13) 1 (5) 7 (11)
Dysthymia 9 (19) 0 9
Adjustment disorder 28 (60) 17 (89) 45 (68)
Generalized anxiety disorder 4 (9) 1 (5) 5 (8)
PAISa subscales (baseline) n.s.
Health care orientation, median (IQR) 8 (6–11) 8 (6–10) 8 (6–11) n.s.
Vocational environment, median (IQR) 9 (5–12) 7 (5–9) 8 (5–12) n.s.
Domestic environment, median (IQR) 5 (2–10) 9 (6–12) 5 (2–10) n.s.
Sexual relations, median (IQR) 9 (5–12) 11 (5–13) 9.5 (5–12) n.s.
Extended family relations, median (IQR) 2 (0–6) 2 (0–7) 2 (0–6) n.s.
Social environment, median (IQR) 12 (6–15) 13 (3–15) 12 (6–15) n.s.
Psychological distress, median (IQR) 12 (8–14) 10 (8–16) 12 (8–14)
Total, median (IQR) 59 (38–76) 52 (33–72) 57 (37–78) n.s.

Referred patients were divided into those who finished the whole follow-up period (completers), patients who were
recruited to participate, but abandoned during the study (dropouts). Proportions were compared by using Fisher’s
exact test; continuous variables were compared by Wilcoxon rank sum test. Only significant p values (p  0.05) were
presented.
PAIS, Psychosocial Adjustment to Illness Scale; n: number of patients (rate value in brackets); IQR, interquartile
range; MSM, men who have sex with men; n.s., not significant.
216 ROUSAUD ET AL.

Participants’ characteristics 0.001), and total PAIS score (z  4.155, p 

Median scores and interquartile range on the 0.001). There were no changes during baseline
PAIS subscales and the rest of the characteris- (between T1 and T2) in any of the PAIS sub-
tics of the participants (64% men, n  42; 36% scales, or in the total PAIS score. Table 2 shows
women, n  24) are presented in Table 1. The the degree of improvement for each subscale
most frequent psychiatric condition was ad- and the total PAIS score (Table 2).
justment disorder (60%) (with depressed mood
(13%), with anxiety (15%), or mixed (32%), fol- Predictors of improvement
lowed by dysthymia (19%), and major depres-
The impact of the variables of age, gender,
sion (single episode or recurrent) (13%).
sexual orientation, route of transmission, time
The mean number of sessions attended dur-
since HIV diagnosis, CD4 count, viral load,
ing the period of therapy was 14.
presence of opportunistic diseases, employ-
ment, partnership, TARV compliance, and psy-
Effect of intervention
chiatric diagnosis in each PAIS subscale is
Wilcoxon matched-pairs signed-ranks test shown in Tables 3 and 4.
showed significant changes during the inter- For example, compared to heterosexuals,
vention period (between T2 and T3) in health men who have sex with men (MSM) showed
care orientation (z  2.771, p  0.006), voca- more improvement in adjustment to health
tional environment (z  2.062, p  0.039), do- care (p  0.041), but less improvement in fam-
mestic environment (z  2.409, p  0.016), ily relationships (p  0.049). Employed patients
sexual relations (z  3.723, p  0.001), ex- reported more improvement in their vocational
tended family relationships (z  2.087, p  environment (p  0.023) and patients infected
0.037), social environment (z  3.671, p  through sexual intercourse showed a greater

TABLE 2. MEDIANS, INTERQUARTILE RANGES, WILCOXON MATCHED-PAIRS SIGNED-RANKS TEST,

AND DEGREE OF IMPROVEMENT FOR EACH PAIS SUBSCALE AND TOTAL PAIS SCORE
OF PATIENTS COMPLETING THE COGNITIVE-BEHAVIORAL GROUP THERAPY (N  46)

Time-point of measure Statistical data

T1 T2 T3 Differencea Differencea
(1 month (first (last between T1 between T2 Degree of
before) session) session) and T2 and T3 improvement
between T2
Median Median Median p p and T3
PAIS (IQR) (IQR) (IQR) z value z value %

Health care 8 9 7 0.085 n.s. 2.771 0.006 8.3

orientation (6–11) (6–12) (5–10)
Vocational 9 8 5 0.984 n.s. 2.062 0.039 16.6
environment (5–12) (4–12) (2–11)
Domestic 5 5 4 0.215 n.s. 2.409 0.016 4.1
environment (2–10) (2–10) (1–8)
Sexual relations 9 9 7 0.236 n.s. 3.723  0.001 10.1
(5–12) (5–12) (4–9)
Extended family 2 3 1 1.230 n.s. 2.087 0.037 11.1
relations (0–6) (1–7) (0–5)
Social 12 10 6 0.986 n.s. 3.671  0.001 22.3
environment (6–15) (5–15) (2–11)
Psychological 12 11 7 0.495 n.s. 4.690  0.001 19.1
distress (8–14) (9–14) (4–10)
Total 59 54 39 0.657 n.s. 4.155  0.001 10.8
(38–79) (35–76) (25–62)
aWilcoxon matched-pairs signed ranks test.
PAIS, Psychosocial Adjustment to Illness Scale; IQR, interquartile range; n.s., not significant; MSM, men who have
sex with men.
 TABLE 3. UNIVARIATE LOGISTIC REGRESSION ANALYSIS OF THE FACTORS ASSOCIATED WITH IMPROVEMENT
IN EACH SUBSCALE OF THE PSYCHOSOCIAL ADJUSTMENT TO ILLNESS SCALE (N  47)

Items of the Psychosocial Adjustment to Illness Scale

Factors used in
the multivariate Statistical Health care Vocational Domestic Sexual Family Social Psychological
analysis parameters orientation environment environment relations relationships environment distress Total

Agea Coefficient NS NS NS NS NS NS NS NS
95% CI
p-value
Gender (female) Coefficient NS NS NS NS NS NS NS NS
95% CI
p-value
Sexual Coefficient 1.827 NS NS NS 2.814 NS NS NS
orientation 95% CI (0.126–3.528) (0.091–5.537)
(MSM) p-value 0.041 0.049
Employed Coefficient NS 2.407 NS NS NS NS NS NS
95% CI (0.408–4.406) (0.283–4.968)
p-value 0.023 0.087
With partner Coefficient NS NS NS NS NS NS NS NS
95% CI
p-value
HIV Coefficient 2.525 NS NS NS NS NS NS NS
transmission 95% CI (0.714–4.336)
(sexual) p-value 0.009
Time since HIV Coefficient NS 0.020 NS NS NS NS NS NS
diagnosisb 95% CI (0.002–0.038)
p-value 0.032
CD4 counta Coefficient NS NS NS NS NS 0.004 NS NS
95% CI (0.008–0)
p-value 0.062
Viral loada Coefficient NS NS NS NS NS NS NS NS
95% CI
p-value
Opportunistic Coefficient NS NS NS NS NS NS NS NS
diseases 95% CI
p-value
TARV compliance Coefficient NS NS NS NS NS NS NS NS
95% CI
p-value
Psychiatric Coefficient NS NS NS NS NS NS NS NS
diagnosis 95% CI
p-value
aContinuous variable.
CI, confidence interval; NS, not statistically significant; IVDU, intravenous drug users; MSM, men who have sex with men.
218 ROUSAUD ET AL.

TABLE 4. RESULTS OF STEPWISE MULTIVARIATE LOGISTIC REGRESSION: EFFECT OF BASSELINE VARIABLES ON THE RATE OF
IMPROVEMENT BETWEEN T2 (FIRST SESSION) AND T3 (LAST SESSION) IN EACH PAIS SUBSCALE (N  47)

Adjusted
PAIS subscales Variable Categories coefficient 95% confidence interval p value

Health care Route of Sexual 2.525 0.714–4.336 0.009

transmission IVDU
Vocational Time since 0.022 0.005–0.038 0.016
environment HIV 2.548 0.654–4.442 0.012
diagnosis
Employed
Family Sexual MSM 2.814 0.091–5.537 0.049
relations orientation heterosexual
Social Employed 3.054 0.491–5.618 0.024
environment CD4 count 0.005 0.009–0.001 0.022

PAIS, Psychosocial Adjustment to Illness Scale; IVDU, intravenous drug users; MSM, men who have sex with men.

improvement in health care (p  0.009) than needed when patients have problems in ad-
patients infected through intravenous drug justing with the new situations caused by the
use. Time since HIV diagnosis was associated infection. In a previous report we demon-
with a greater improvement in adjustment strated the immediate and long-term efficacy
to the vocational environment (p  0.032), of a cognitive-behavioral oriented group psy-
whereas CD4 count was inversely associated chotherapy program for HIV-infected pa-
with a greater improvement in adjustment to tients with anxiety and depressive symptoms
the social environment (p  0.062) (Table 3). referred to a consultation-liaison psychiatry
The multivariate linear regression analysis of department.6 The present study seeks to show
each subscale of the PAIS is shown in Table 4. the ability of this psychotherapy program to
Sexual route of transmission was indepen- improve the psychosocial functioning of a
dently associated with an improvement in heterogeneous sample of patients with HIV-
health care orientation through the psycho- 1 infection. Many previous studies suggest
therapy intervention (  2.525, p  0.009). that cognitive-behavioral group interventions
Time since HIV diagnosis (  0.022, p  in HIV-positive patients reduce depression,
0.016) and being employed (  2.548, p  anxiety and generalized psychiatric distress
0.012) were independently associated with an indicators,13 although to our knowledge,
improvement in adjustment to vocational en- only two published studies have examined
vironment. MSM showed a poorer improve- the efficacy of psychotherapy in improving
ment in adjusting to family relations after the psychosocial functioning in HIV-1–positive
intervention (  2.548, p  0.049). Finally, a patients.10,11 Molassiotis and coworkers10
lower CD4 count (  0.005, p  0.002) and demonstrated the efficacy of a cognitive-be-
being employed (  3.054, p  0.024) were in- havioral group therapy in improving quality
dependently associated with an improvement of life in Chinese patients with symptomatic
in the adjustment to social environment. HIV disease compared to a peer support/
counseling group therapy and to a group re-
ceiving routine treatment with no formal psy-
DISCUSSION chosocial intervention. In contrast, in a ran-
domized 3  3 block design study, Eller and
Although HIV-1 infection in first world is coworkers11 found no significant improve-
no longer regarded as a fatal illness, it is still ment in quality of life in a sample of patients
considered a long-term process with a great receiving guided imagery or progressive
many stressful factors that patients have to muscle relaxation compared to a control
confront. Psychotherapeutic intervention is group.
IMPROVING ADJUSTMENT TO HIV THROUGH GROUP THERAPY 219

PAIS subscale improvement during therapy Factors associated with better outcome
The present study shows that a structured, Analyzing the different domains separately
time-limited, cognitive–behavioral oriented we were able to obtain more information on the
group therapy program may be clinically use- effect of the therapy on the psychosocial func-
ful for improving psychosocial adjustment to tioning of patients with specific characteristics.
HIV-infection in a heterogeneous sample of pa- In the multivariate regression analysis, pa-
tients referred to a consultation-liaison psychi- tients who were infected through sexual inter-
atry unit. All PAIS dimensions improved dur- course showed a greater improvement in the
ing the intervention period, but not during the health care orientation dimension than patients
waiting period (baseline). The dimension that infected through intravenous drug use. The
showed the greatest improvement was social health care orientation dimension of the PAIS
environment, followed by vocational function- addresses the nature of the respondent’s health
ing and psychological distress. care posture, and whether it will function to pro-
Given the lack of scientific evidence for these mote a positive or negative adjustment to the
phenomena, we suggest the following expla- HIV infection. It is generally accepted that per-
nations on the strength of our clinical experi- sons infected through intravenous drug use fol-
ence: low antiretroviral treatment and medical con-
trols worse than patients infected via other
• The Social Environment subscale reflects routes. In a study by Mizuno and coworkers21
the status of the patient’s current social in a sample of patients almost all of whom were
and leisure time activities, and the degree drug users, more than two thirds (63%) of the
to which the patient has suffered impair- participants did not rank HIV as their top life
ment or constriction of these activities as a priority. For the participants of Mizuno’s study,
result of the HIV infection and/or its treat- housing, money, and safety from violence were
ment.15 Therapy allows the improvement particularly salient. It is also noteworthy that
of social functioning via the planned in- HIV was not necessarily ranked as the second
crease in social activities. Patients are also or third most important by significant numbers
trained to use strategies to confront the of participants, in fact, nearly half the subjects
disclosure of serostatus to significant oth- ranked HIV as fourth priority or lower, and
ers. These strategies are complemented by more than 1 in 10 ranked HIV as the least im-
the group effect, which enables subjects to portant life issue.21 Furthermore, it seems that
meet others in the same situation as them- the participants who were drug users did not
selves. benefit from our therapy program in the same
• The Vocational Environment subscale ad- way as the rest of the participants, possibly be-
dresses the impact that a medical disorder cause cognitive–behavioral interventions pres-
may have on vocational adjustment, defin- ent only limited advantages over the standard
ing “vocational” flexibly to indicate work, interventions in terms of both magnitude and
school, or home, whichever is most appro- frequency of HIV risk reduction in drug users.22
priate.15 The therapy promotes the onset of Participants who were employed when they
new projects and the acquisition of problem began the therapy showed a significant im-
solving-strategies. provement both in their functioning at work or
• The Psychological Distress subscale is de- at home and in their social relationships. In a
signed to measure dysphoric thoughts and previous longitudinal study, most HIV-in-
feelings that accompany the patient’s disor- fected men who were unemployed at baseline
der, or are a direct result of the illness and did not return to work after 30 months of fol-
its sequelae.15 Anxiety is improved through low-up.23 Those authors concluded that spe-
the presentation of relaxation exercises,19 cific interventions may be needed to promote
and depression is improved through the ac- return to work for people with HIV/AIDS. Our
quisition of strategies following Beck’s in- results suggest that specific psychotherapy
tervention model.20 programs are also useful for patients who are
220 ROUSAUD ET AL.

still working. HIV-infected patients have many not tend to have as much opportunity as MSM
limitations at work and in their social relations, to speak about their infection with friends. Dur-
especially those who are afraid of disclosing ing therapy, the heterosexual participants are
their serostatus, and who have to take their able to share impressions with other HIV-pos-
medication and attend the control visits with- itive persons. This may reduce their feelings of
out disrupting their everyday professional and guilt and encourages them to confront the fam-
social activities. Through our psychotherapy ily reaction.
program participants were trained to use Similar to our previous report,6 the present
strategies to solve these limitations at work and study avoids the frequent limitations observed
in their social relations. Our results suggest that in other publications. Our patients had all been
these strategies were more useful for those em- referred to a Consultation-Liaison Psychiatry
ployed than for those unemployed. We should Department, unlike most previous studies,28–30
recall at this point that the vocational dimen- in which community announcements or ad-
sion assessed by the PAIS also applies to stu- vertisements were used to recruit patients in
dents and housewives, given the special defi- order to obtain large participant samples. This
nition of the term used by the designer of the type of recruitment produces a selection bias,
instrument.5 because people who volunteer for this type of
Time since HIV infection was also signifi- study may have a special motivation and may
cantly associated with an improvement in vo- be more receptive to intervention of any kind.
cational environment. In former times HIV-in- Our patient sample was heterogeneous,
fected patients stopped working either because whereas in previous studies most of the pa-
they were physically limited or because they tients were MSM.28–30 Public mental health set-
were considered permanently disabled by our tings should offer an efficient psychotherapeu-
National Health Service. With the introduc- tic and supportive intervention program for a
tion of highly active antiretroviral treatments heterogeneous group of patients.
(HAART) many patients improved their phys- Nevertheless, several limitations of this
ical status, although they continued to be con- study should be mentioned. This study fol-
sidered as patients. Through our psychother- lowed an open-label, observational design and
apy program many patients returned to no randomisation or comparison group was in-
activities they had given up many years before cluded. Therefore, the problems inherent in
when their HIV positive status was diagnosed. observational studies may be present in this in-
In our study, MSM showed a lower im- vestigation. The dropout rate during the inter-
provement in family relationships than hetero- vention period was very high (approximately
sexuals. In fact, family relationship is shown to 30%). Finally, we did not use postintervention
be the second most endorsed area of impaired follow-up measures to ascertain treatment ef-
functioning in MSM.24 HIV-seropositive gay fects. An efficient therapeutical intervention
and bisexual men tend to disclose their HIV program should also aim to achieve long-last-
serostatus less frequently to family members ing changes.
than to friends and partners.25,26 This may be The results of this study and a review of the
because disclosing their serostatus often im- most recent literature suggest that efforts
plies revealing their homosexuality to their should be focused on the following areas of en-
family for the first time. Furthermore, HIV-in- hanced intervention development: adapting
fected MSM reported that friends and lovers the intervention to the client’s needs and in-
responded more helpfully than relatives.26 cluding a component in the intervention that
They seem to withhold their serostatus from specifically targets drug users in order to im-
the family both to avoid personal rejection and prove their health care, MSM to improve their
to avoid worrying them.27 Although the ther- family relationships, and unemployed patients,
apy may help MSM to deal with their HIV con- who seem to have more difficulties than em-
dition in a social environment, it does not seem ployed patients in adjusting to HIV infection.
to help to improve family relations, as is the Our study seems to demonstrate that our
case in heterosexual patients. Heterosexuals do psychotherapy program for HIV infected pa-
IMPROVING ADJUSTMENT TO HIV THROUGH GROUP THERAPY 221

tients improves psychosocial functioning. A 10. Molassiotis A, Callaghan P, Twinn SF, Lam SW,
therapeutic intervention program should not Chung WY, Li CK. A pilot study of cognitive-
behavioral group therapy and peer support/counsel-
only reduce distress and improve adjustment
ing in decreasing psychologic distress and improving
to illness, but should also aim to achieve long- quality of life in Chinese patients with symptomatic
lasting changes. Further analysis of more de- HIV disease. AIDS Patient Care STDs 2002;16:83–96.
finitive data from this ongoing study will be 11. Eller LS. Effects of cognitive-behavioral interventions
presented in forthcoming studies to determine on quality of life in persons with HIV. Int J Nurs Stud
whether this kind of psychotherapeutic inter- 1999;36:223–233.
12. Riemer D, Bock J, Escobar LC, Hautzinger M. Grup-
vention is also effective at long-term follow-up, pen-Psychotherapie für Menschen mit HIV/AIDS.
and able to improve other variables such as Psychotherapeutics 1998;43:198–204.
antiviral treatment compliance. 13. Kelly JA. Group psychotherapy for persons with HIV
and AIDS-related illnesses. International J Group Psy-
chother 1998;48:143–1626.
14. Spitzer RL, Williams JBW. Structured Clinical Inter-
ACKNOWLEDGMENTS
view for DSM-III-R, patient version with psychotic
symptoms. New York, NY: New York Psychiatric In-
The study was partially supported by the Eu- stitute, 1987.
ropean Commission Grant SOC 97 201135 15. Derogatis LR. The psychosocial adjustment to illness
05F02. scale (PAIS). J Psychom Research 1986;30:77–91.
16. Wright C, Feurer ID, Payne JL, Wise PE, Shockley S,
Speroff T. Health-related quality of life after different
types of solid organ transplantation. Ann Surg 2000;
REFERENCES 232:597–607.
17. Crespo MD, Ferre F. Versión española del PAIS-SR:
1. Holtgrave DR. Causes of the decline in AIDS deaths, Un instrumento de medida para la adaptación psi-
United States, 1995–2002: Prevention, treatment or cosocial a la enfermedad física. Psiquis 1992;13:38–44.
both? Int J STD AIDS 2005;16:777–781. 18. StataCorp. 2001 Stata Statistical Software: Release 7.0.
2. Eller L S. Effects of cognitive-behavioral interventions College Station, TX: Stata Corporation.
on quality of life in persons with HIV. Int J Nurs Stud 19. Jacobson E. Progressive Relaxation. Chicago: Univer-
1999;36:223–233. sity of Chicago Press, Chicago, 1938.
3. Worthington C, Myers T, O’Brien K, Nixon S, Cock- 20. Beck AT. Cognitive models of depression. J Cogn Psy-
erill R. Rehabilitation in HIV/AIDS: Development of cother 1987;1:5–37.
an expanded conceptual framework. AIDS Patient 21. Mizuno Y, Purcell D, Borkowski M, Knight K, and the
Care STDs 2005;19:258–271. SUDIS team. The life prorities of HIV-seropositive in-
4. Tarter RE, Switala J, Arria A. Plail J, Van Thiel D. jection drug users: Findings from a community-based
Quality of life before and after orthotopic hepatic sample. AIDS Behav 2003;7:395–403.
transplantation. Arch Intern Med 1991;151:1521–1526. 22. Hershberger SL, Wood M, Fisher G. A cognitive-behav-
5. Derogatis LR, Derogatis MF. The Psychosocial Ad- ioral intervention to reduce HIV risk behaviors in crack
justment to Illness Scale (PAIS & PAIS-SR). Adminis- and injection drug users. AIDS Behav 2003;7:229–243.
tration, Scoring, and Procedures Manual II. Townson, 23. Rabkin JG, McElhiney M, Ferrando SJ, Van Gorp W,
MD: Clinical Psychometric Research, 1990. Lin SH. Predictors of employement of men with
6. Blanch J, Rousaud A, Hautzinger M, et al. Assessment HIV/AIDS: a longitudinal study. Psychosom Med
of the efficacy of a cognitive-behavioural oriented group 2004;66:72–78.
psychotherapy program for HIV-infected patients re- 24. Berg M, Mimiaga M, Safren S. Mental health concerns
ferred to a consultation-liaison psychiatry department. of HIV-infected gay and bisexual men seeking men-
Psychotherapy and Psychosomatics 2002;71:77–84. tal health services: An observational study. AIDS Pa-
7. Weber R, Christen L, Christen S, et al. Effect of indi- tient Care STDs 2004;18:635–643.
vuidual cognitive behaviour intervention on adher- 25. Wolitski RJ, Rietmeijer CA, Goldbaum GM, Wilson RM.
ence to antiretroviral therapy: prospective, random- HIV serostatus disclosure among gay and bisexual man
ized trial. Antivir Ther 2004;9:85–95. in four American cities: General patterns and relation to
8. Lechner SC, Antoni MH, Lydston D, et al. Cognitive- sexual practices. AIDS Care 1998;10:599–610.
behavioral interventions improve quality of life in 26. Hays RB, McKusick L, Pollack L, Hilliard R, Hoff C,
women with AIDS. J Psychosom Res 2003;54:253–261. Coates TJ. Disclosing HIV seropositivity to significant
9. Sherman AC, Mosier J, Leszcz M, et al. Group inter- others. AIDS 1993;7:425–431.
ventions for patients with cancer and HIV disease. 27. Mason HR, Marks G, Simoni JM, Ruiz MS, Richard-
Part I: Effects on psychosocial and functional out- son JL. Culturally sanctioned secrets? Latino men’s
comes at different phases of illness. Int J Group Psy- nondisclosure of HIV infection to family, friends,
chother 2004;54:29–82. lovers. Health Psychol 1995;14:4–5.
222 ROUSAUD ET AL.

28. Weiss JJ, Mulder CL, Antoni MH, De Vroome, Address reprint requests to:
Garssen B, Goodkin K. Effects of supportive-expres- Araceli Rousaud
sive group intervention on long-term psychosocial
Fundació Clínic per la Recerca
adjustment in HIV-infected gay men. Psychother Psy-
chosom 2003; 72:132–140. Biomèdica-IDIBAPS
29. Lee MR, Cohen L, Hadley SW, Goodwin FK. Cogni- Institut d’Investigacions Biomèdiques Agustí
tive-behavioural group therapy with medication for Pi i Sunyer
depressed gay men with AIDS or symptomatic HIV Hospital Clínic de Barcelona
infection. Psychiatr Serv 1999;50:948–952. Rosellon 140
30. Goodkin K, Blaney NT, Feaster DJ, Baldewicz T,
Burkhalter JE, Leeds B. A randomised controlled clin-
Barcelona 08036
ical trial of a bereavement support group intervention Spain
in Human Immunodeficiency Virus type 1-serposi-
tive and –seronegative homosexual men. Arch Gen
Psychiatry 1999;56:52–59. E-mail: [email protected]