You are on page 1of 8

Psychological Assessment Copyright 1997 by the American Psychological Association, Inc.

1997, Vol. 9, No. 3, 269-276 1040-3590/97/S3.00

Factor Structure of the Psychosocial Adjustment to Illness Scale


(Self-Report) for Persons With Cancer
Thomas V. Merluzzi and Mary Ann Martinez Sanchez
University of Notre Dame

The Psychosocial Adjustment to Illness Scale (PAIS-SR) is a frequently used self-report measure,
yet its factor structure, reliability, and validity have not been tested adequately on a sample of persons
with cancer. A group of persons with cancer (N = 502) completed the PAIS-SR and other measures
of adjustment and coping. A principal-axis factor analysis with varimax rotation yielded 7 factors:
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Social and Leisure Activities (.86), Job and Household Duties (.85), Psychological Distress (.87),
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Sexual Relationship (.92), Relationships With Partner and Family (.70), Health Care Orientation
(.61), and Help From Others (.63). Values in parentheses are Cronbach's as for the factors; a for
the entire scale was .93. Correlations with measures of disease impact, adjustment, and coping
support the validity of the PAIS-SR and its use for cancer research.

Advances in medical care and changing attitudes toward ill- The PAIS was devised originally for, and initially validated
nesses such as cancer and renal disease have resulted in increas- on, persons with cancer and their families (Morrow et al.,
ing numbers of people who are living longer with chronic dis- 1978). A recent review of measures of psychological function-
eases. Treatments for many serious and chronic diseases have ing used in.research on adjustment to cancer (Gotay & Stern,
severe side effects that may tax the person's coping capacity. 1995) indicated that the PAIS is still widely used to assess
This increase in longevity and the severity of side effects have psychosocial adjustment in persons with cancer. However, Gotay
provoked a concomitant interest in the psychosocial adjustment and Stern (1995) have strongly endorsed the need for research
of those who survive serious diseases and endure treatments for on the psychometric properties of those measures and their re-
those diseases. That interest has spawned a number of measures finement. Furthermore, they suggest that factor analytic and va-
to assess adjustment (e.g., Derogatis & Derogatis, 1990; Viet & lidity studies need to be conducted to ascertain the structure
Ware, 1983) and quality of life (e.g., Cella et al., 1993). The and quality of measures used in psychosocial oncology research.
relevance of the concept of adjustment to illness is underscored The PAIS-SR is very similar to the PAIS in that both mea-
by the finding from a prospective longitudinal study that for sures contain 46 items that cover seven domains of psychosocial
the same disease the cost of health care for people who are functioning. The subscale Health Care Orientation assesses "pa-
poorly adjusted to chronic illness is approximately $24,000 an- tient attitudes, quality of information, and the nature of the
nually compared with approximately $10,000 for their well- patient's expectations about his/her disorder and its treatment"
adjusted counterparts (Browne et al., 1990). (Derogatis & Derogatis, 1990, p. 8). Vocational Environment
One of the most prevalent measures of adjustment appearing taps "perceived quality of job performance, job satisfaction,
in the health psychology literature is the Psychosocial Adjust- lost time, job interest, and a number of other variables that are
ment to Illness Scale, which was devised by Derogatis and his associated with the nature of vocational adjustment'' (p. 8). The
colleagues (Derogatis & Derogatis, 1990; Morrow, Chiarello, & items that compose the Domestic Environment scale ' 'measure a
Derogatis, 1978). There are two versions of the Psychosocial number of aspects of family living, including financial impact
Adjustment to Illness Scale. One is administered in a structured of the illness, quality of relationships, family communications,
interview format (PAIS); the other is a self-report form and effects of physical disabilities" (p. 9). The Sexual Relation-
(PAIS-SR). ship scale assesses "sexual interest, frequency, quality of perfor-
mance, and level of satisfaction" (p. 10). Extended Family
Relationships contains items that assess "any negative impact
Thomas V. Merluzzi and Mary Ann Martinez Sanchez, Department of the illness upon communication, quality of relationships, in-
of Psychology, University of Notre Dame. terest in interacting with family and other variables reflective of
We wish to acknowledge the support of Rafat Ansari, Thomas Troeger, this domain" (p. 11). The Social Environment scale is used to
David Taber, Juan Garcia, Rhonda Critchlow, and the staff of Michiana "determine the degree to which the patient has suffered incur-
Hematology-Oncology; Marti Verfurth, executive director of Memorial sions due to illness into his/her typical social and leisure activi-
Hospital's Regional Cancer Center, and her staff; and the many dedicated ties" (p. 12). Finally, Psychological Distress covers "indicators
undergraduates in the seminar on psychooncology who served as research of psychological distress such as anxiety, depression and hostil-
assistants. Also, we would like to extend our gratitude to the many persons
ity, as well as reduced self-esteem, body image problems and
with cancer who graciously agreed to participate in this research.
Correspondence concerning this article should be addressed to inappropriate guilt" (p. 13).
Thomas V. Merluzzi, Department of Psychology, University of Notre The PAIS-SR continues to be used in research on adjustment
Dame, Notre Dame, Indiana 46556. Electronic mail may be sent via the to cancer (e.g., Gotcher, 1992; Howes, Hoke, Winterbottom, &
Internet to merluzzi.l@nd.edu. Delafield, 1994; Lowery, Jacobsen, & DuCette, 1993; Mer-

269
270 MERLUZZI AND MARTINEZ SANCHEZ

luzzi & Martinez Sanchez, 1997), including bone marrow trans- based regional cancer center and a large private-practice oncology clinic.
plantation (Andrykowski et al., 1992; Jenkins, Linington, & Participants ranged in age from 21 to 90 years, with a median of approxi-
Whittaker, 1991); choice of breast surgery (Wolberg, Tanner, mately 60. Sixty-nine percent of the participants were married, 13%
Romsaas, Trump, & Malec, 1987); and family functioning widowed, and 7% divorced. Equal numbers (36%) were employed or
retired, and 14% were full-time homemakers. Approximately 55% made
(Friedman et al., 1988). It also has been used to study adjust-
$25,000 per year or less, and 45% made more than $25,000 per year.
ment to lupus (Engle, Callahan, Pincus, & Hochberg, 1990); Most participants had completed high school (44%), some had attended
renal disease (Fricchione et al., 1992); heart surgery (Langelud- college (18%), and still fewer had earned college degrees (8%). Most
decke, Tennant, Fulcher, Baird, & Hughes, 1989); pain (Crook, were Protestant (55%) and Caucasian. Members of minority ethnic
Tunks, Kalaher, & Roberts, 1988); and hypertension (DeVon & groups constituted 8% of the sample, but it had a broad range of age
Powers, 1984). and socioeconomic status (SES).
Although the PAIS-SR has been used extensively in research A variety of cancer diagnoses were represented: breast (37%),
on adjusting to cancer, its psychometric properties, in particular lymphoma (13%), lung (13%), colorectal (8%), prostate (5%), uter-
its internal consistency and factor structure, have not been re- ine/ovarian/cervical (8%), leukemia (3%), brain (3%), organ/stom-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

ported. Internal consistency data are available for a small sample ach (2%), testicular (2%), and other (6%). Most had been treated
This document is copyrighted by the American Psychological Association or one of its allied publishers.

with chemotherapy (72%), surgery (66%), radiation (49%), or some


of cardiac patients but no such data on persons with cancer have
combination of these treatments. Forty-nine percent had been diagnosed
been provided in the PAIS manual. Also, internal consistency less than a year and 51% more than a year before participating in this
data are provided for the PAIS (i.e., interview format) based research. For 124 participants in this sample, the following information
on lung cancer patients; however, the sample size is quite small on stage of cancer (at diagnosis) was obtained: Stage 1, 24%; Stage 2,
(N = 89), and there have been some revisions of the PAIS since 28%; Stage 3, 10%; Stage 4, 26%; not staged, 12%.
those internal consistency values were reported (Derogatis &
Derogatis, 1990). Finally, a review of many studies that used Materials
the PAIS-SR revealed no additional internal consistency data.
In many research studies authors reported the internal consis- Psychosocial Adjustment to Illness Scale—Self Report. Participants
completed the Psychosocial Adjustment to Illness Scale (PAIS-SR; De-
tency values for the interview format of the PAIS -while using
rogatis & Derogatis, 1990), which correlates with measures of depres-
the self-report version. sion and global adjustment. Other validity studies have demonstrated its
A factor analysis of the PAIS (i.e., interview format) on data ability to discriminate between "good" and "bad" adjusters to renal
derived from 120 lung cancer patients was presented in the dialysis (Derogatis & Derogatis, 1990). The only reported internal con-
manual (Derogatis & Derogatis, 1990). The authors reported a sistency coefficients in the manual were from a sample of 69 cardiac
seven-factor solution. In part, the solution does appear to support patients. Across seven scales those coefficients alpha ranged from .47
the original scales of the PAIS. However, there are several prob- to .85.
lems involved in the interpretation of that factor analysis. First, For each of 46 items in the PAIS-SR, participants chose one of
the sample was quite small (TV = 120). Tabachnick and Fidell four responses, which are graded to represent poor-to-good adjustment.
(1996) suggested that for factor analyses a sample size of 100 Generally, high scores are indicative of poor adjustment. We reversed the
scoring, however, so that a high score was indicative of good adjustment,
is poor, 200 is fair, 300 is good, and 500 is very good. Second,
because correlations with other measures were easier to interpret if, for
the sample consisted of only lung cancer patients, and therefore all measures, high scores indicated positive psychological or physical
the results may not be generalized to another type of cancer. states. For the analyses conducted in this research, raw scores were used
Third, only factor loadings of greater than .40 are reported in instead of standard scores because there was no need for comparisons
the manual. Fourth, although 80% of the items had a factor with other populations or clinical interpretation.
pattern that did correspond to the original scales of the PAIS, The measures presented next were used to assess the validity of the
20% did not. Finally, the factor analysis was performed on data PAIS-SR. These measures were chosen on the basis of their strong
derived from the PAIS, which is conducted in interview format, psychometric properties (i.e., reliability and validity) and their use in
and therefore the reported structure may vary from the PAIS- health psychology research.
SR, which uses the self-report format. Disease impact. The Sickness Impact Profile (SIP; Bergner, Bobbit,
Carter, & Gilson, 1981) was used to determine the functional impact
The purpose of this study was to conduct a factor analysis
of disease on the person's activity. The participants completed the scales
of the PAIS-SR on data obtained from a large sample of patients of this measure, which relate to the disease's physical impact and impair-
who had a variety of cancer diagnoses. The goals were to con- ment of independence (i.e., ambulation, mobility, and body care and
firm the seven conceptual scales of the PAIS-SR, provide infor- movement) but not psychosocial impact. A score was obtained by sum-
mation concerning internal consistency, and investigate its valid- ming the scale values of each of the endorsed items and dividing the
ity by examining correlations with other measures. We antici- sum by the number of scale items within that category. Then a total
pated that the PAIS-SR would (a) correlate negatively with score was obtained by summing the category scores and multiplying
measures of disease impact, functional status, and stage at diag- that score by 100. A higher score is indicative of greater impairment.
nosis, (b) correlate positively with measures of adjustment, so- This measure has been used extensively in research on coping with
cial support, and coping, and (c) have a weak or no relationship cancer.
to demographic variables. The Karnofsky Performance Status scale (KPS; Karnofsky, Abel-
mann, & Graver, 1948) was designed to assess the general physiological
Method and functional impact of cancer. This measure relates the disease to
daily functioning and has found widespread usage in cancer research
Participants (Greico & Long, 1984; Karoly, 1985). Ratings were done by a health
A sample of 502 persons with cancer (331 women, 171 men) was care professional, and scores could range from 0 (dead) to 100 (normal
recruited for participation in this research from two sites: a hospital- with no complaints).
FACTOR STRUCTURE OF THE PAIS-SR 271

Adjustment. The Mental Health Index (MHI), developed by Viet global scores of the PAIS-SR differed as a function of site.
and Ware (1983), is an indicator of psychological distress and psycho- Approximately 85% of the participants were tested at the clinic,
logical well-being in the general population. The scale consists of 38 whereas the remaining were recruited from the hospital. A ran-
items that were rated on a 0-5 scale, with higher scores indicating dom sample of 100 participants from the clinic were compared
greater psychological well-being.
with the 78 from the hospital. No overall difference was found
Social support. The Interview Schedule for Social Interaction (Ber-
geman et al., 1990) contains 24 items for which the participant indicated on the global score of the PAIS-SR; thus, the data from all
(on a 3-point scale) how often a supportive behavior occurred and, on participants were grouped together for further analyses.
the same scale, how satisfied he or she was with that support. A greater
level of support was associated with higher scores on this measure. Factor Analysis
Coping. The COPE Scale (Carver, Scheier, & Weintraub, 1989) is
a 52-item measure that contains scales for problem-focused and emotion- The goals of the factor analysis were threefold. The first goal
focused coping. Participants indicated on a 4-point scale the extent to was to confirm the conceptually derived scales of the PAIS-
which each item applied to them ( 1 = 7 usually don't do this at all to SR, the second was to examine the internal consistency of the
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

4 = 1 usually do this a lot). High scores are indicative of high levels factors, and the third was to explore the relationship of the
This document is copyrighted by the American Psychological Association or one of its allied publishers.

of problem- or emotion-focused coping.


factors of the PAIS-SR with other measures used in research
Religiosity was assessed by asking participants to rate three questions
on 7-point Likert scales: (a) How often do you attend religious services? on coping with cancer. On the basis of those goals and the
(b) How often do you pray? and (c) Are you a spiritual person? A total recommendations of Floyd and Widaman (1995), a principal-
religiosity score was computed by summing the responses with a high axis (common factors) factor extraction was chosen. Also, be-
score, indicative of more religiosity. cause the original scales of the PAIS-SR were developed to tap
The Cancer Behavior Inventory (CBI; Merluzzi & Martinez Sanchez, distinct domains of functioning, an orthogonal rotation (i.e.,
1997) is a 43-item measure of self-efficacy for coping with cancer. Each varimax) was used. That rotation method was also used in the
item describes a coping behavior (e.g., coping with treatment-related factor analysis of the interview format of the PAIS that was
side effects), which the patients rated (on a 9-point scale) in terms of reported in the PAIS manual.
their confidence that they could perform the behavior. The CBI has a Although a seven-factor solution was needed to test the con-
Cronbach's alpha of .96 and correlates with measures of coping with
cancer.
firmation hypothesis, six-, seven-, and eight-factor solutions
Medical and demographic information. For some of the participants, were examined. However, we determined that a seven-factor
information was obtained from charts, such as stage at diagnosis, time solution was not only adequate but also optimal on the basis of
since diagnosis, type of treatments received, and so on. The participants scree curve analysis, variance accounted for (52%, unrelated;
also completed an extensive demographics sheet. 44%, rotated), and relationship to the conceptually derived
scales (see Table 1). The seven factors were labeled as follows
Procedure (with eigen-values in parentheses): Social and Leisure Activities
Data were collected by undergraduate research assistants (RAs) who (11.04), Job and Household Duties (3.48), Psychological Dis-
were enrolled in a seminar on psychooncology and received information tress (2.24), Sexual Relationship (2.02), Relationship With
about the medical and psychological aspects of cancer. They also role- Partner and Family (1.81), Health Care Orientation (1.68), and
played to improve their communication skills with persons with cancer, Help From Others (1.40). The variance accounted for was less
were engaged in a discussion of ethical issues in the conduct of research
than that reported in the PAIS manual (63%) for the factor
(e.g., confidentiality), and toured the treatment facilities. The training
lasted for 10 hr over a 4-week period. The RAs received this training analysis of the interview version of the scale. The difference in
because they had no previous research experience with cancer patients. variance accounted for between the solutions may be due to
RAs approached patients in the waiting rooms of the two research two key distinctions between those analyses. First, the factor
sites and asked them to participate in a study exploring ways that people analysis reported in the manual was performed on the interview
cope with cancer. Informed consent was obtained, and the participant form of the PAIS, which may have produced more reliable
was given a booklet to complete that contained the PAIS-SR and the scores than the self-report version (i.e., the interviewer may
other measures. The RA answered any questions, and on rare occasions have assisted in the clarification of the questions and response
(for less than 1% of the sample), assisted the participant in the comple- options). Second, the analysis reported in the manual was per-
tion of the questionnaires. Because data collection continued over 2 formed on only lung cancer patients, whereas, in this study there
years, with three distinct testing periods, some of the questionnaire
was a mixture of cancer diagnoses. Thus, the solution reported
packets varied depending on the length or number of questionnaires
included. Thus, the participants took from 15 to 45 min to complete the here for the self-report version may contain more unexplained
packet. Some participants were allowed to take the questionnaire packet variance than the solution for the interview version reported in
home. A reminder letter was sent if the packet was not received within the PAIS manual.
a week. Across all samples tested, approximately 60% of people who The factor structure that emerged does overlap with the origi-
had been approached agreed to participate. Approximately 80% of par- nal scales of the PAIS-SR. Correlations of the original concep-
ticipants completed the research materials during their office visits. tually derived scales and the empirically derived factors indi-
About 55% of those who had taken the questionnaires home mailed cated that the original Social Environment scale and the Social
them back to the office. and Leisure Activities factor correlated .98. The Vocational En-
Results vironment scale and the Job and Household Duties factor corre-
lated .94. The Psychological Distress, Sexual Relationship, and
Preliminary Analyses Health Care Orientation scales correlated with the Psychological
Because data were collected from two sites, a clinic and a Distress, Sexual Relationship, and Health Care Orientation fac-
hospital, analyses were conducted to determine whether the tors .99, .96, and .87, respectively. The Domestic Environment
272 MERLUZZI AND MARTINEZ SANCHEZ

Table 1
Factor Analyses of the Psychosocial Adjustment to Illness Scale—Self Report
Factor loadings
Original Item
scale no. Item summary 1 2 3 4 5 6 7

Factor 1: Social and Leisure Activities


SE 38 Interest in social activities .72 .19 .17 .17 .06 .09 .13
SE 36 Interest in leisure activities .72 .18 .16 .17 .07 -.02 .14
SE 37 Participation in leisure activities .70 .27 .16 .23 -.03 .03 .06
SE 39 Participation in social activities .69 .31 .23 .20 .01 .07 .05
SE 35 Actual participation in activities .66 .29 .19 .24 -.01 .00 .08
SE 34 Leisure activities and hobbies .61 .26 .21 .13 .02 .06 .10
EF 32 Socialize with extended family .33 .17 .08 .06 .18 -.01 .09
.02 -.05 .21 .21
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

EF 30 Interest in extended family .26 .01 .20


This document is copyrighted by the American Psychological Association or one of its allied publishers.

Factor 2: Job and Household Duties


VE 10 Job performance .22 .75 .11 .16 .09 .00 .14
VE 9 Interference with job .25 .74 .13 .15 .06 .00 .13
VE 13 Changing job goals .20 .58 .15 .06 .09 .10 .07
DE 17 Interference with household duties .39 .57 .23 .16 -.02 .04 .19
DE 21 Extent of physical disability .36 .56 .21 .18 -.01 .04 .16
VE 11 Lost time at work .14 .53 .10 .13 .06 .04 .13
VE 12 Importance of job .08 .48 .03 .02 .12 .20 .02
HCO 5 Expectations of outcomes .32 .44 .37 .06 .02 .15 .04
HCO 2 Special attention to illness -.10 -.19 - .09 -.08 .09 .08 .04

Factor 3: Psychological Distress


PD 41 Depression .23 .13 .70 .10 .14 .15 .17
PD 40 Anxiety .15 .09 .66 .04 .02 .21 .18
PD 44 Worry .22 .15 .62 .05 .07 .18 .07
PD 45 Self-deprecation .25 .28 .55 .15 .16 .18 .00
PD 42 Anger .03 .02 .54 .02 .22 .12 .19
PD 43 Self-blame .13 .25 .53 .15 .15 .20 .03
PD 46 Illness affects appearance .20 .17 .41 .06 .10 .11 .04

Factor 4: Sexual Relationship


SR 26 Sexual satisfaction .18 .14 .11 .82 .07 -.04 .04
SR 25 Sexual activities .20 .16 .10 .82 .06 -.03 .08
SR 24 Sexual interest .19 .09 .10 .72 .01 .06 .08
SR 27 Sexual performance .21 .20 .05 .67 .04 .08 -.00

Factor 5: Relationship With Partner and Family


DE 15 Relationship with partner -.00 .05 .13 .02 .73 .07 .06
SR 23 Changes in relationship with partner .01 .09 .17 .08 .70 .02 .04
SR 28 Arguments with partner -.00 .04 .14 .08 .66 .03 .12
EF 29 Contact with extended family .23 -.06 - .05 -.03 .39 .11 .26
EF 33 Getting along with extended family .15 -.02 .00 -.02 .38 .23 .33
DE 19 Communication with family .23 .09 .23 -.01 .37 .10 .27
DE 16 General relationship with family -.00 .15 .10 -.00 .23 -.02 .22

Factor 6: Health Care Orientation


HCO 8 Getting complete information .05 .02 .08 .07 -.02 .58 .06
HCO 3 Quality of medical care .01 .02 .14 .06 .07 .57 .05
HCO 6 Information from medical staff .01 .13 .10 -.01 -.02 .53 .05
HCO 7 Trusting treatments .05 .10 .21 -.03 .12 .42 .08
HCO 4 Treatment and medical staff .00 .04 .08 .02 .03 .38 .05
VE 14 Problems with coworkers, etc. .08 .25 .08 .00 .26 .27 .19
HCO 1 Attention to health care needs -.02 -.11 .06 -.11 .07 .19 -.01

Factor 7: Help From Others


EF 31 Help from extended family .28 .08 .09 -.00 .13 .15 .64
DE 20 Needing help from others .08 .14 .14 .06 .09 .05 .58
DE 18 Shifts in family duties .04 .15 .13 .15 .15 .09 .45
DE 22 Financial hardship .08 .17 .24 .04 .16 .11 .32

Note. Boldface type indicates the factor with the highest loading. SE = Social Environment, EF = Extended
Family, VE = Vocational Environment, DE = Domestic Environment, HCO = Health Care Orientation, PD =
Psychological Distress, SR = Sexual Relationship.
FACTOR STRUCTURE OF THE PAIS-SR 273

scale and the Help From Others factor correlated .82. The Ex- chological Distress (Factor 3), which had a very strong correla-
tended Family scale and the Relationship with Partner and Fam- tion with the MHI.
ily factor had no strong counterparts. Social support. The positive correlations of the Psychologi-
cal Distress, Relationships With Partner and Family, and Health
Care Orientation factors with the ISSI indicated that more social
Internal Consistency Analyses support is associated with better adjustment to cancer in those
Coefficient alpha for the entire 46-item PAIS-SR scale was domains. However, it is interesting to note that for Help From
.93 for this sample. Analyses of the reliability of the original Others, a factor for which we anticipated a positive correlation,
conceptually derived scales revealed a range of values from .50 none was obtained.
to .87, and the alpha values for the scales derived from the factor Coping. The correlations of the factors of the PAIS-SR
analysis were somewhat higher than those for the original scales, with the CBI, a measure of self-efficacy for coping with cancer,
ranging from .61 to .92 (see Table 2). reflected a relationship between confidence in one's ability to
perform a variety of coping behaviors (e.g., cope with treat-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

ment-related side effects, ask doctors questions, maintain a posi-


This document is copyrighted by the American Psychological Association or one of its allied publishers.

Correlations of the PAIS With Other Measures tive attitude, etc.) and positive adjustment. Clearly, confidence
in the ability to execute coping behaviors and strategies contri-
The unweighted factor scores from the 46-item PAIS-SR were butes to more positive adjustment to cancer.
correlated with measures of disease impact, adjustment, social The correlations of the factors with the Problem Focused
support, coping, and medical and demographic information. (PF) and Emotion Focused scales of the COPE were mixed and
Disease impact and stage. As anticipated, the factors of the not as anticipated. For example, the negative correlation of the
PAIS-SR were correlated negatively with the SIP and positively Social and Leisure Activities factor with the COPE-PF indicates
with the KPS, indicating that increases in the impact of the that higher levels of problem-focused coping are associated with
disease were associated with decreased adjustment (Table 3). poorer social adjustment. We expected the opposite. Similarly,
That relationship was particularly true of those factors that in- greater emotion-focused coping was associated with poorer ad-
volve social (Factor 1), vocational, household (Factor 2), or justment with respect to Help From Others. That is, managing
sexual activities (Factor 4), as well as receiving help from others emotions was associated with poorer adjustment to receiving
(Factor 7). Also, stage at diagnosis was correlated with adjust- help from others. Other correlations are more interpretable. For
ment across most of the factors of the PAIS-SR. The more example, greater emotion-focused coping was associated with
advanced the stage of the disease at diagnosis, the poorer the better adjustment on the Health Care Orientation factor. Perhaps
adjustment in most domains. However, the strength of those controlling extreme emotions allows better access to medical
correlations was less than that of the SIP. information and better treatment by medical staff. Finally, a
Adjustment. Relatively strong correlations with the MHI single significant correlation with Religiosity on the Health Care
across all factors are indicative of a pervasive mental health Orientation factor may indicate that those who endorsed engag-
component in the PAIS-SR. That was particularly true for Psy- ing in religious practices were better adjusted with respect to
the medical aspects of the disease than those who did not engage
in religious practices.
Table 2 Demographic and medical. Time since diagnosis was not
Cronbach 's Alpha Coefficients for the Original Scales of the related to adjustment. Age, on the other hand, had an interesting
PAIS-SR and the Seven-Factor Solution pattern of correlations with the factors. The correlations of age
with Psychological Distress and Relationship With Partner and
Scale a Family indicate that greater adjustment is associated with in-
creasing age. Thus, older people who have cancer may be better
Original scales adjusted psychologically and in their close relationships than
Social Environment .83
Vocational Environment .84 younger people. The negative correlation of age with the Sexual
Psychological Distress .87 Relationship factor may reflect the effects of both disease
Sexual Relationship .84 and age.
Domestic Environment .74 The correlations of income with several factors of the PAIS-
Health Care Orientation .50
Extended Family .51 SR attest to the notion that SES has an effect on adjustment to
Total scale (46 items) .93 disease. The greater the income of the person with cancer, the
Seven-factor solution greater the adjustment in Social and Leisure Activities, Job and
Social/Leisure Activities .87 Household Duties, Sexual Relationship, and Help From Others.
Job and Household .85 Education was minimally related to adjustment, except that
Psychological Distress .87
Sexual Relationship .92 more highly educated people with cancer adjust better with
Partner/Family Relationships .70 respect to their sexual relationship.
Health Care Orientation .61
Help From Others .63
Total scale (46 items) .93 Discussion

Note. PAIS-SR = Psychosocial Adjustment to Illness Scale—Self- With some exceptions, the original conceptually derived
Report. scales of the PAIS-SR emerged in the factor solution. Two of the
274 MERLUZZI AND MARTINEZ SANCHEZ

Table 3
Correlations of the Factors of the PAIS-SR With Disease-Related Measures
and Demographic Variables
Factor

Measure

Disease impact
SIP _ 49** -.54** -.30** -.40** -.16** -.12** -.31**
KPS" .40** .44** .16 .21** .02 .18* .33**
MHI" .35** .50** .71** .33** .41** .46** .46**
Support
issr .16 .15 .18* .08 .33 .25** .07
Coping
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

CBI" .39** .38** .48** .27** .26** 27** .31**


This document is copyrighted by the American Psychological Association or one of its allied publishers.

COPE PEb -.19* -.18 -.15 -.18 -.23* .05 -.14


COPE EF" -.12 -.16 -.12 -.15 -.19 .25** -.26**
Religiosity0 .13 .13 .07 -.03 .10 .25* -.06
Medical
Stage" -.38** -.39** -.18* -.33** -.14 -.21* -.23**
Time* .07 .00 -.01 .01 -.09 .00 .00
Demographic
Age" -.11* -.07 .20** -.22** .20** .10* .06
Education' .10* .03 .01 .18** .00 -.05 .07
Income" 22** .17** .07 .17** .08 -.01 .32**

Note. PAIS-SR = Psychosocial Adjustment to Illness—Self-Report. Factors: 1 = Social and Leisure


Activities, 2 = Job and Household Duties, 3 = Psychological Distress, 4 = Sexual Relationship, 5 =
Relationships With Partner and Family, 6 = Health Care Orientation, 7 = Help From Others. SIP = Sickness
Impact Profile; KPS = Karnofsky Performance Scale; MHI = Mental Health Index; ISSI = Interview
Schedule for Social Interaction; CBI = Cancer Behavior Inventory; COPE PF = Problem Focused Coping;
COPE EF = Emotion Focused Coping; Time = time since diagnosis.
a
Data compiled across all 502 participants. b Data derived from 124 women with breast cancer. c Data
derived from 137 persons with a variety of cancer diagnoses.
* p < .05. ** p < .01.

original scales, Psychological Distress and Social Environment, from the Domestic Environment Scale loaded on that factor.
emerged intact in the factor analysis. The Social Environment These items focused on the degree to which the person is physi-
items were strongly associated with Factor 1. The two additional cally disabled (Item 21) and how much the illness interferes
items in Factor 1, from the original Extended Family scale, were with household duties and chores (Item 17). These items had
associated with this factor, which may reflect interest in social relatively high loadings on Factor 2 but also had moderate load-
activities with the extended family. Thus, Factor 1 emerged as ings on Factor 1, which may indicate that physical disability
similar to the original Social Environment scale and reflects and interference with routine household activities is also associ-
interest and involvement in social and leisure activities. Mainte- ated with the ability to engage in leisure and social activities.
nance of those activities is tantamount to positive adjustment to We might have expected Item 18 from the Domestic Environ-
illness. ment scale, dealing with shifts in family duties, to load on Factor
The original Sexual Relationship scale did emerge, in part, 2 (Job and Household Duties); however, it was one of four
as Factor 4. However, it is interesting to note that two items items to form Factor 7 (Help From Others). Factor 7 also has
from the original scale appeared to be associated more strongly one item from the Extended Family Scale, which reflects the
with Factor 5 (Relationship With Partner and Family) than with need for help from the extended family. This factor represents
Factor 4 (Sexual Relationship). A close examination of these the support one might get from family and friends to help com-
items revealed that, in the case of one item (Item 23) the content pensate for the impact of the illness. However, the negative
focuses on how an illness can cause problems in a relationship correlations of this factor with measures of disease impact indi-
with a spouse or partner. There is no mention of sexual problems cate that more impaired individuals have a more difficult time
in that item. The other item (Item 28) does mention interference adjusting to the help they receive from others.
with a couple's sexual relationship but tends to emphasize, to The most unique scale to emerge from the factor analysis is
a greater extent, the fact that illness may cause arguments. The Factor 5, Relationship With Partner and Family. It draws items
association of those two items with Factor 5 (Relationships With from the original Domestic Environment, Sexual Relationship,
Partner and Family) does confirm their emphasis on relation- and Extended Family scales. This factor represents the quality
ships rather than on sexual interest or performance. of the significant relationships in the patient's life. The items
Items from the Vocational Environment scale did emerge on cohere around a theme of how the patient's illness has affected
Factor 2, which reflected the degree to which the illness inter- those relationships, which in turn has an impact on adjustment.
feres with the performance of one's job. In addition, two items The correlations with measures of disease impact and stage
FACTOR STRUCTURE OF THE PAIS-SR 275

support the notion that cancer and its treatments may take their not needed before the disease. Perhaps people who have more
toll on patients and that in turn reduces their level of adjustment. material resources are able to take advantage of not only family
That reduction in level of adjustment is found to a greater extent resources (e.g., paying a relative for transportation) but also
in the social, vocational, and sexual domains than in other do- commercial services that make daily living easier (e.g., a house-
mains. Thus, it appears that limitations on physical activity have cleaning service). Thus, having greater financial resources may
direct bearing on adjustment in domains that involve remaining allow for better adjustment because it provides the ability to
active. These correlations are significant and strong but do allow maintain certain aspects of one's lifestyle despite the disease.
for other influences on adjustment. The similarity in the pattern
of correlations between the SIP, which is self-report, and the
Conclusion
KPS, which is "other" report, provide some measure of confi-
dence in the findings. Moreover, there are implications for medi- The PAIS-SR is an internally consistent measure with a fac-
cal and psychological practice. To the extent that medical and tor structure that taps seven domains of adjustment to disease.
psychological interventions either prevent debilitating states or The validation data suggest that it is significantly related to
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

provide for recovery from those states, the level of adjustment variables that have been associated with adjustment to cancer.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

will be maintained or increased. Moreover, its global approach to psychosocial adjustment as


The correlations of the PAIS-SR with the MHI illustrate the well as its component'scales or factors give the PAIS-SR some
strong and pervasive emotional health component in the PAIS- advantage over other measures that tap a single dimension (e.g.,
SR and therefore support its validity as a measure of adjustment. depression, mood, etc.). In fact, Gotay and Stern (1995) sug-
That relationship was particularly evident in the correlation of gested that the PAIS-SR may be more like a quality-of-life
the MHI with the Psychological Distress factor of the PAIS- measure than a measure of psychological state. On the other
SR. The correlations with the CBI also illustrate the psychologi- hand, it is not as broad based as the better quality-of-life mea-
cal nature of the PAIS-SR. Greater self-efficacy for coping with sures (e.g., Cella et al., 1993).
cancer is associated with greater adjustment to the disease. Not Future research should investigate the stability of the PAIS-
only does this support the validity of the PAIS-SR, it also SR over time or changes as a function of interventions, disease
indicates that improvements in efficacy expectations may im- status, and life events. Also, given the amount of medical care
prove adjustment. Thus, interventions designed to improve effi- services required by poorly adjusted patients when compared
cacy may also improve psychosocial adjustment. with well adjusted patients, the PAIS-SR could be used as a
We anticipated that social support would be positively related screening device to target those who may need interventions to
to adjustment. The correlation of social support with Relation- improve their adjustment to cancer. A profile based on the factors
ships With Partner and Family indicates that support from family of the PAIS-SR could help identify areas that need attention.
and friends is tantamount to positive adjustment to disease. Of Considering that research has established that persons with can-
course, the reverse may also be true: Positive adjustment with cer who receive psychological treatment live longer than those
respect to relationships may preserve or strengthen social sup- who do not (Spiegel, Bloom, Kraemer, & Gottheil, 1989), ad-
port. A more interesting finding is that social support is associ- justment may be a key variable in those outcomes.
ated with positive adjustment to the medical system. Support
from one's social network may allow individuals to navigate
through the medical maze of information and obtain high-quality References
medical care. In a similar vein, although many of the correla- Andrykowski, M. A., Altmaier, E. M., Harriett, R. L., Otis, M. L., Gin-
tions of the PAIS-SR with religiosity were not significant, the grich, R., & Henslee-Downey, P. J. (1992). The quality of life in adult
one that was significant was with Health Care Orientation. Per- survivors of allogenic bone marrow transplantation. Bone Marrow
haps greater religiosity evokes more trust in and satisfaction Transplantation, 50, 399-406.
with the medical treatment people receive and therefore more Bergeman, C. S., Plomin, R., Pedersen, N. L., McClearn, G. E., & Nes-
selroade, J. R. (1990). Genetic and environmental influences on social
comfort and better adjustment to the medical aspects of the
support: The Swedish adoption/twin study of aging. Journal of Ger-
disease.
ontology: Psychological Sciences, 45, 101-106.
The positive correlations of age with the Psychological Dis- Bergner, M., Bobbit, R. A., Carter, W. B., & Gilson, B. S. (1981). The
tress factor and the negative correlation of age with the Sexual Sickness Impact Profile: Development and final revision of a health
Relationship factor may reveal the small but not insignificant status measure. Medical Care, 19, 787—805.
role that life stage plays in coping with serious illness. With Browne, G. B., Arpin, K., Corey, P., Fitch, M., & Gafni, A. (1990).
respect to Psychological Distress, perhaps with age comes an Individual correlates of health service utilization and the cost of poor
increase in coping capacity based on a lifetime of experiences. adjustment to chronic illness. Medical Care, 28, 43-57.
Also, a disease such as cancer may be more normative in old Carver, C. S., Scheier, M. F., & Weintraub, J. K. (1989). Assessing cop-
age than in youth and therefore more expected. That is, con- ing strategies: A theoretically based approach. Journal of Personality
tracting cancer at a young age may be nonnormative and there- and Social Psychology, 56, 267-283.
Cella, D. K, TUlsky, D. S., Gray, G., Sarafin, B., Linn, E., Bonomi, A.,
fore less easy to adjust to when compared with contracting it Silberman, M., Yellen, S. B., Winicor, P., Brannon, J., Eckberg, K.,
in old age. Lloyd, S., Purl, S., Blendowski, C., Goodman, M., Barnicle, M., Stew-
The relationship between Help From Others and income is art, I., McHale, M., Bonomi, P., Kaplan, E., Taylor, S., Thomas, C., &
an interesting one. With greater income there is more positive Harris, J. (1993). The Functional Assessment of Cancer Therapy
adjustment to the help one might need. The questions that consti- Scale: Development and validation of the general measure. Journal
tute this factor contrast the help one may receive now that was of Clinical Oncology, 11, 570-579.
276 MERLUZZI AND MARTINEZ SANCHEZ

Crook, J., Tunks, E., Kalaher, S., & Roberts, J. (1988). Coping with Jenkins, P. L., Linington, A., & Whittaker, J. A. (1991). A retrospective
persistent pain: A comparison of persistent pain sufferers in a specialty study of psychosocial morbidity in bone marrow transplant recipients.
pain clinic and in a family practice clinic. Pain, 34, 175-184. Psychosomatics, 32, 65—71.
Derogatis, L. R., & Derogatis, M. F. (1990). The Psychosocial Adjust- Karnofsky, D. A., Abelmann, W. H., & Craver, L. F. (1948). The use of
ment to Illness Scale: Administration, scoring, and procedures man- nitrogen mustards in the palliative treatment of carcinoma. Cancer, 1,
ual-H. Towson, MD: Clinical Psychometric Research. 634-656.
DeVon, H. A., & Powers, M. J. (1984). Health beliefs, adjustment to Karoly, P. (1985). Measurement strategies in health psychology. New
illness, and control of hypertension. Research in Nursing and Health, York: Wiley.
7, 10-16. Langeluddecke, P., Tennant, C., Fulcher, G., Baird, D., & Hughes, C.
Engle, E. W., Callahan, L. E, Pincus, T., & Hochberg, M. C. (1990). (1989). Coronary artery bypass surgery: Impact upon the patient's
Learned helplessness in systemic lupus erythematosus: Analysis using spouse. Journal of Psychosomatic Research, 33, 155—159.
the Rheumatology Attitudes Index. Arthritis and Rheumatism, 33, Lowery, B. J., Jacobsen, B. S., & DuCette, J. (1993). Causal attribution,
281-286. control, and adjustment to breast cancer. Journal of Psychosocial
Floyd, F. J., & Widaman, K. F. (1995). Factor analysis in the develop- Oncology, 10, 37-53.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

ment and refinement of clinical assessment instruments. Psychological Merluzzi, T. V., & Martinez Sanchez, M. (1997). Assessment of self-
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Assessment, 7, 286-299. efficacy and coping with cancer: Development and validation of the
Fricchione, G. L., Howanitz, E., Jandorf, L., Krossler, D., Zervas, I., & Cancer Behavior Inventory. Health Psychology, 16, 163-170.
Woznicki, R. M. (1992). Psychological adjustment to end-stage renal Morrow, G. R., Chiarello, R. J., & Derogatis, L. R. (1978). A new scale
disease and the implications of denial. Psychosomatics, 33, 85-91. for assessing patients' psychosocial adjustment to medical illness.
Friedman, L. C., Baer, P. E., Nelson, D. V, Lane, M., Smith, F. E., & Psychological Medicine, 8, 605-610.
Spiegel, D., Bloom, J. R., Kraemer, H. C., & Gottheil, E. (1989). Effect
Dworkin, R. J. (1988). Women with breast cancer: Perception of
of psychosocial treatment on survival of patients with metastatic breast
family functioning and adjustment to illness. Psychosomatic Medicine,
cancer. Lancet, 2, 888-891.
50, 529-540.
Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics
Gotay, C. C., & Stern, J. D. (1995). Assessment of psychological func-
(3rd ed.). New York: Harper Collins.
tioning in cancer patients. Journal of Psychosocial Oncology, 13,
Viet, C. T, & Ware, J. E. (1983). The structure of psychological distress
123-160.
and well-being in general populations. Journal of Consulting and
Gotcher, J. M. (1992). Interpersonal communication and psychosocial Clinical Psychology, 51, 730-742.
adjustment. Journal of Psychosocial Oncology, 10, 21-39. Wolberg, W. H., Tanner, M. A., Romsaas, E. P., Trump, D. L., & Malec,
Greico, A., & Long, C. J. (1984). Investigation of the Karnofsky Perfor- J. F. (1987). Factors influencing options in primary breast cancer
mance Status as a measure of quality of life. Health Psychology, 3, treatment. Journal of Clinical Oncology, 5, 68-74.
129-142.
Howes, M.J., Hoke, L., Winterbottom, M., & Delafield, D. (1994). Received August 1, 1996
Psychosocial effects of breast cancer on the patient's children. Journal Revision received February 27, 1997
of Psychosocial Oncology, 12, 1-21. Accepted March 3, 1997 •

You might also like