Psychological Reports, 1991, 68, 339-347. O Psychological Reports 1991
DEPRESSION A N D DECISION-MAKING A M O N G INTRAVENOUS DRUG USERS ' GEORGE W. JOE
Institute of Behavioral Research Texos Christian University LAVERNE KNEZEK AND DEENA WATSON
DARCO Drz~gServices, Inc Dallus, Texas D. DWAYNE SIMPSON
Institztte of Behavioral Research Texas Christian University Summary.-Two short scales, the TCU Depression Scale and the TCU Decisionmaking Scale, were psychometrically evaluated in a sample of 145 intravenous drug users. Coefficient alpha reliabilities were .78 for the 6-item TCU Depression Scale and .77 for the 9-item TCU Decision-making Scale. Concurrent validity of the former scale was assessed by correlating scores with those on the Beck Depression Inventory, r = .75. Based on the Beck Depression Inventory Clinical cutoff scores, 83% of the sample showed some depression, with 23% severely depressed, 39% moderately depressed, and 21% mildly depressed. Individuals scoring higher on depression on both tests tended to score lower on decision-making. Significant demographic associations of age, gender, education, and race-ethnicity were found for the depression and decision-making scales. More depression was noted for women, those younger, white, and having less education. Older and more educated intravenous drug users tended to score higher on decision-making. Validity for the depression and decision-making scales was assessed by examining correlations with behaviors. Significant positive correlations were found between depression scores and intravenous use of cocaine only, heroin and cocaine combined, and heroin only. Also, intravenous use of cocaine only and of cocaine and heroin combined were negatively related to decision-making. AIDS sex-risky behavior was positively correlated with depression and negatively correlated with decision-making.
According to the Health Belief Model (Rosenstock, 1966), sociopsychological variables might have considerable influence as modifying factors for individuals' perceptions regarding the threat posed by a disease as well as their likelihood of taking preventive action. Two of the psychological status constructs of major interest are depression and decision-making. Depression should affect attention, and readiness for making choices is a basic component of any model of behavioral change. 'This study was supported by the National Institute on Drug Abuse Grant 1 R18 DA05176-02. The interpretations and conclusions, however, do not necessarily represent the position of NIDA or the Department of Health and Human Services. Address correspondence to George W. Joe, Institute of Behavioral Research, Texas Christian University, Fort Worth, TX 76129.
340
G. W. JOE., ETAL
While depression has been recognized as a very important problem among intravenous drug users (e.g., Rounsaville, Weissman, Wilber, CritsChristoph, & Kleber, 1982; Rounsaville, Kosten, Weissman, & Kleber, 1986) and has received considerable study, a similar effort to study decision-making capacity and readiness among intravenous drug users has not been mounted. The importance of examining decision-making in relation to depression is particularly relevant in view of the cognitive-based therapies developed for depression (Hollon & Beck, 1986) and related counseling approaches used for addicts in drug-abuse treatment and other types of interventions to mohfy behavior (Platt & Hermalin, 1989). I n addition to these psychological constructs of depression and decision-making, a number of other constructs are also important in the evaluation of treatments and interventions for addicts. These include self-esteem, anxiety, risk-taking, antisocial personality, and motivation to change. For some of these constructs, instruments are available, but many of these were developed using items intended for a middle-class population. Some of the wordings of items may not: be as appropriate or understandable to an intravenous (IV-drug users) population. This paper describes preliminary findings based on brief scales for depression and decision-making; these were developed as part of an assessment battery for a project on treatment-evaluation focused on AIDS-risk reduction (Simpson, 1990) in IV-drug users. The description includes the psychometric properties of the scales and the predictive (concurrent) validity of rhe short depression scale in relation to the established Beck Depression Inventory. I n addition, scores on these scales are examined in relation to indicators of AIDS-risky behaviors, surrogate health indicators of AIDS risk, and self-reported frequency of drug use.
Sample A sequential sample of 154 IV-drug users who participated in the National AIDS Demonstration Research Project (NADR) in Dallas were asked to complete short self-rating scales immediately before their AIDS Initial Assessment (AIA) interview. These included individuals who had been recruited through community outreach efforts to complete a personal interview and confidential testing for HIV, with the stipulation they had not been in treatment for drug use in the past six months. However, nine of the IV-drug users who had agreed to participate did not complete the psychological scales and were deleted from the analyses. Distributions on demographic characteristics for the sample of 145 are presented in Table 1 . Approximately half (49%) were black, 42% were white, and 10% Hispanic. More than half ( 5 3 % ) were over 34 years old, and 72% were men. A majority also indicated being unmarried (57%), being unemployed at the time of the interview
INTRAVENOUS DRUG USERS: DEPRESSION, DECISIONS
341
(54%), and having graduated from h g h school or having a GED (76%). Almost half (48%) received income from dlegal sources during the past six months, and 25% claimed illegal activities provided them their major source of income. TABLE 1 FREQUENCY DISTRIBUTIONS: A N D DECISION-MAKING, A N D BEHAVIOR ( N = 145) DEMOGRAPHICS, DEPRESSION Measure Age (years) 18-34 35-39 G E 40 Male Race Black White Hispanic
Measure
To
90
Demographics Marital Status 47 Single 32 Married 21 Not currently married 72 Education (grades) 1-11 49 12 42 Over 12 10 Employed Source of Income (illegal) Behavior
NOVA Needle Risk 0 1 2
Hepatitis
3 18 79 25
NOVA Sex Risk 0 1 2 Health-Sex
Depression and Decision-making Beck Clinical (MFPt) 17 Not depressed (0-9) 12 Mild depression (10-18) 30 Moderate depression (19-29) 42 Severe depression (30-63) TCU Decision-making (9) C 7 0- 4 19 10-16 37 17-24 38 25-28 29-36 *MFN, minimizes false negatives. tMFP, minimizes false positives. Beck Clinical (MFN*) Not depressed (0-9) Mild depression (10-15) Moderate depression (16-23) Severe depression (24-63) TCU Depression (6) C 0- 9 5- 8 9- 12 13-24
Scales Depression was measured by two separate scales. One was the 21-item Beck Depression Inventory (Beck & Steer, 1987), and the other was a short 7-item scale developed by the Institute of Behavioral Research at Texas Christian University (TCU). Decision-making was measured by a 10-item
342
G. W. JOE., E T A L .
scale that was also developed. AIDS-risky behaviors were assessed by the two AIA-based NOVA indices (developed by NOVA Research Corporation, 1989) measuring needle-risk behavior and sex-risk behavior. Two surrogate health indicators for AIDS risks, also based on data from the AIA, were (1) whether or not the individuals were told by a physician or nurse that they had ever had hepatitis, and (2) whether or not the persons had been told that they had ever had a sexually transmissible disease (syphilis, gonorrhea, herpes, or chlamydia). The distributions of the AIA-based indices are shown in Table 1. Five drug-use frequency variables taken from the AIA included (1) crack use, (2) IV-amphetamine use, (3) IV-heroin-only use, (4) IV-cocaine use, and (5) the combination of IV-heroin and cocaine use (speedball).
The distributions of the items and the item-total correlations for the TCU Depression Scale and the TCU Decision-making Scale are presented in Table 2. This table shows that on four of the seven depression items, at least 40% of the sample rated themselves as "often" or "almost always" feeling TABLE 2 TCU DEPRESSION AND TCU DECISION-MAKING SCALES: AND ITEM-TOTNCORRELATIONS ( N = 145) ITEM DISTRLBU~ONS Scale Items
Depression 3. You feel interested in life. (Rt) 6. You feel extra tired oc run down. 8. You feel hopeful about the future. (R)* 12. You worry or brood a lot. 13. You have thoughts of committing suicide. 14. You feel sad or depressed. 17. You feel lonely.
Response ItemPercen~aaest- total r a b c d e
-
14 7 17 3 15 15
.SO .50 .19 .54 .41 .70 .49
Decision-maku~g 1 16 50 27 6 1. You make good decisions. 2. You chink of several different ways to solve a problem. 0 1 1 43 28 18 4. You make decisions without thmking about consequences. (R) 12 2 1 42 18 7 3 12 3 1 34 20 5. You think about probable results of your actions. 5 7 22 40 26 7. Your problems are caused by someone else. (R)* 1 3 22 44 3 1 9. You think about what causes your current problems. 7 2 1 37 19 16 10. You plan ahead. 11. You consider how your actions will affect others. 4 14 38 29 16 3 20 5 1 20 7 15. You have trouble malung decisions. (R) 16. You analyze problems by looking at all the choices. 3 14 44 23 15
.55 .60 .21 .42 .OO .23 .56 .44 .47 .49
*Deleted in reduced scale. t R means scoring reversed. $a: never, b: rarely, c: sometimes, d: often, e: almost always.
38 24 7 12 18 24 5 9 57 18 3 13 8 10
27 37 38 34 20 42 42
11 30 13 34 2 27 25
1
343
INTRAVENOUS DRUG USERS: DEPRESSION, DECISIONS
extra tired or run down, worrying or brooding a lot, feeling sad or depressed, and feeling lonely. One item had an item-total correlation below .20 (Item 8-You feel hopeful about the future). Among the 10 decision-making items, over 40% of the sample reported that they "often" or "almost always" thought about solving their or causes of their problems. O n the four items asking about decision-making in terms of consequences or planning ahead, 25% to 35% of the sample rated themselves as "often" or "almost always." One item had an item-total correlation below .2O (Item 7-Your problems are caused by someone else), for which two-thirds of the sample responded "often" or "almost always." I n Table 3, the means, standard deviations, and coefficient alpha internal reliabilities are presented for the depression and decision-making scales. For the TCU Depression Scale the coefficient alpha reliabhties are given for both the original 7-item scale and a reduced 6-item scale (from which Item 8 was deleted). Similarly, the coefficient alpha reliabllities are presented for the original 10-item TCU Decision-making Scale and a reduced 9-item scale (from which Item 7 was deleted). TABLE 3 R E L I A E ~COEFFICIENTS Y FOR DEPRESSIONA N D DECISION-MAKING SCALES(N= 145) Scales
No. Items
M
SD
Coefficient a Raw Standardized
Beck Depression Inventory TCU Depression TCU Depression TCU Decision-making TCU Decision-making
21 7 6 10
21.7 13.0 11.4 24.0 21.3
11.4 4.7 4.4 5.3
.90 .75 .78 .73 .77
9
5.1
.91 .76 .78 .74 .77
The psychometric analysis for these brief scales indicated that their reliability coefficients were acceptable. This included both the original and reduced scales. However, because the reduced scales are psychometrically adequate for this sample and the population of IV-drug users it represents, the remainder of the paper addresses the correlates of the reduced scales. The coefficient alpha reliabilities were .78 for the 6-item TCU Depression Scale, and .77 for the 9-item TCU-Decision-making Scale. Reliability for the longer 21-item Beck Depression Inventory was .90. Principal components analysis (using SAS PROC FACTOR; SAS, 1988) of the 6-item TCU Depression Scale yielded one eigenvalue above one (2.93), and it accounted for 49% of the variance. The analysis of the 9-item TCU Decision-making Scale yielded two eigenvalues above one (3.30 and 1.38), with the first accounting for 37% of the variance. For the Beck Depression Inventory, there were five eigenvalues above one (7.91, 1.50,
344
G. W. JOE., E T A L .
1.19, 1.07, and 1.01), with the first accounting for 38% of the variance. Examination of the eigenvalues suggests only one factor in the TCU Depression Scale, but potentially two factors in the TCU Decision-making Scale. However, the second eigenvalue was small compared to the first. With respect to the clinical cut-off scores for the Beck scale to classify levels of depression (Beck & Steer, 1987), 83% of the sample showed some depression. Based on their established cut-offs to minimize the number of false negatives (MFN), depression was classified as severe for 42%, moderate for 30%, and mild for 12%. In comparison, classifications for minimizing the number of false positives (MFP) showed that 23% were severely depressed, 39% moderately depressed, and 21% mildly depressed. The two Beck clinical classification systems correlated .91. TABLE 4
Scales 1. 2. 3. 4. 5. 6. 7.
Beck Beck Beck TCU TCU TCU TCU
Depression Inventory Clinical', MFN ~ l i n i c a lMFP ~, Depression (6) Depression (6) C' Decision-making (9) Decision-makine (9)
cd
1
2
3
4
5
6
.90* .88t .93t .75t .69t -.45t -.42t
.91t .72t .73t -.37t -.35t
.72t .73t -.43t -.41t
.78' .92t -.44t -.42t
-.43t -.42t
.77* .93t
"Reliabihty coefficient.
t p < .0001. 'Not depressed = 0-9, Mild = 10-15, Moderate = 16-23, 'Not depressed = 0-9, Mild = 10-18, Moderate = 19-29, 'Categorized (0-4, 5-8, 9-12,13-24). dCategorized (0-9, 10-16, 17-24, 25-28, 29-36).
Severe = 24-63. Severe = 30-63.
Table 4 presents Pearson product-moment correlations between the depression and decision-making scales. Scores on the Beck inventory and the two Beck clinical scales, denoted as Beck (MFN) and Beck (MFP), were highly correlated with the TCU Depression Scale and a categorization of the TCU Depression Scale; the latter is referred to as TCU Depression-C and used the cut-offs: 0-4, 5-8, 9-12, 13-24. The correlations of the Beck scale and its clinical classifications with the TCU Depression Scales (uncategorized and categorized) ranged from .69 to .75. These results demonstrate the concurrent (predictive) validity for the short TCU scale. One of the most interesting substantive findings was that the Decisionmaking Scale was significantly correlated with the depression scales (Beck and TCU); persons scoring higher on depression tended to score lower on decision-making (r = -.35 to -.45). The direction of the relationship fits what
345
INTRAVENOUS DRUG USERS: DEPRESSION. DECISIONS
one would expect, namely, that being depressed impairs one's confidence in decision-making. Correlations of demographic variables with the depression and decision-making scales are given in Table 5 . In this sample, women were more depressed than men ( r = .17 to .21 on the Beck scales and .21 for TCU Depression-C); older IV-drug users tended to be less depressed than younger ones ( r = -.23 to -.28), and they also scored higher on decision-making readiness ( r = .29) than younger ones. White respondents were more depressed than black subjects ( r = .28 for TCU Depression-C), and those with more education tended to be less depressed ( r = -.I4 to -.18) and to score hgher on decision-making ( r = .19). No significant relationships were found for marital status and employment status. TABLE 5 PEARSON CORRELA~ONS OF DEMOGRAPHICS WITH DEPRESSION AND DECISION-MAKING SCALES(N= 145) Scales'
Age Sex Race Black White Hispanic Education
Beck Depression Inientory
CfinjcaJ, MFN '
Beck Clinical. MFP
-.28$ ,177
-.23$ .2lt
-.24$ .19t
-.08 .05 .06 -.18t
-.I3 .05 .14 -.I4
-.08 .04 .08 -.I87
TCU Depression (6)
TCU Depression (6) C
TCU Decision-making (9)
TCU Decision-making (9) C
Age -.25$ -.25$ Sex .16 .21t Race -.28$ -.27$ Black Whice .28$ .28$ Hispanic .OO -.02 Education -.08 -.I1 "Marital Status and Whether Employed not significantly correlated with Depression and Decision-making scales.
In terms of the behavioral correlates shown in Table 6, there were significant positive correlations of depression scores with IV-drug use variables measuring cocaine only, heroin and cocaine combined, and heroin only. Depression was not significantly related to IV-amphetamine use or to crack use. In addition, use of IV-cocaine only and heroin and cocaine combined were
346
G. W. JOE., ETAL.
negatively related to decision-making, that is, those who injected more frequently tended to score lower on decision-making. The NOVA sex-risk index for AIDS risks was also negatively correlated with decision-making (r = -.21). O n the other hand, the NOVA sex-risk index was positively correlated with the TCU Depression Scale (Y = ,201. I n other words, persons with higher sex risks had lower scores on decision-making and higher scores in depression. TABLE 6 PEARSONCORRELATIONS OF BEHAVIOR W I - IDEPRESSION AND DECISION-MAKING SCALES(N = 145) Scales
NOVA Needle Risk NOVA Sex Risk Hepatitis Health-Sex Cocaine Only (IV) Amphetamines (IV) Heroin (IV) Heroin and Cocaine (IV) Crack
Beck Depression Inventory .ll .ll .10 -.I2 .20' -.06 .25t .19* .02 TCU Depression (6)
Beck Clinical,
MFN .09 .ll .16 -.I1 .17' -.lo .35t .20* -.03
TCU Depression (6) C
Clinical,
MFP .15 .I0 .12 -.08 .16* -.07 .28t .18* .01
TCU DeTCU Decision-making cision-making (9) (7) c
Nova Needle Risk NOVA Sex Risk Hepatitis Health-Sex Cocaine Only (IV) Amphetamines (IV) Heroin (IV) Heroin and Cocaine (IV) Crack
DISCUSSION Depression and decision-making readiness are proposed as important factors both in development and evaluation of AIDS intervention programs and for evaluation of cognitively-based counseling strategies. The two constructs received empirical support in this preliminary study. Specifically, it showed that a large percentage-either 62% or 72%, depending upon the strictness of the classification system-of this outreach sample was clinically depressed and that depression and decision-making are significantly related to risky behaviors such as frequency of IV-drug use. Methodologically, the
INTRAVENOUS DRUG USERS: DEPRESSION, DECISIONS
347
findings support the validity of the two short TCU scales measuring depression and decision-making. The results for the TCU Depression Scale paralleled the results for the longer and well-established Beck Depression Intheoventory, and the results for the TCU Decision-making- Scale supported .. retical expectations. Significant demographic relationships for age, sex, education, and race-ethnicity were found for the depression and decision-making scales. The preliminary results concerning depression and decision-making in this preliminary study suggest that the interventions need to address these constructs, as well as possibly other psychosocial constructs such as risk-taking and antisocial personality. REFERENCES BECK,A. T., & STEER,R. A. (1987) Beck Depression Inventory manual. San Antonio, T X : The Psycholog~calCorp./Harcourt Brace Jovanovich. HOLLON, S , & BECK, A. T. (1986) Research in cognitive therapies. In S. Garfield & A. E. Bergen (Eds.), Handbook of psychotherapy and behavioral change. New York: Wiley. Pp. 433-482.
SAS INSTITUTE.(1988) The factor procedure. In SASISTAT user's guide, release 6.03 addition. Cary, NC: Author. PLAT^, J. J., & HERMALIN, J. (1989) Social skdl deficit interventions for substance abusers. Psychology of Addictive Behaviors, 3, 114-133. ROSENSTOCK, I. M. (1966) Why people use health services. Milbank Memorial Fund Quarterly, 44, 94-127. ROUNSAVILLE, B. J., KOSTEN,T. R., WEISSMAN,M. M., & KLEBER,H. D. (1986) Prognostic significance of psychiatric disorders in treated opiate addicts. Archives of General Psychiatry, 43, 739-745. ROLINSAVILLE, B. J., WEISSMAN,M. M., WILBER,C. H., CRITS-CHRISTOPH, K., & KLEBER,H. D. (1982) Diagnosis and symptoms of depression in opiate addicts: course and relationship to treatment outcome. Archives of General Psychiatry, 39, 151-156. NOVA RESEARCHCO. (1989) Nefwork. Bethesda, MD: Author, 4600 East-West Highway, Suite 700. SIMPSON,D. D. (1990) Forms manual: drug abuse treatment for AIDS-risks reduction (DATAR). Fort Worth, TX: Texas Christian Univer., Institute of Behavioral Research. Accepted January 31, 1991.
DEPRESSION A N D DECISION-MAKING A M O N G INTRAVENOUS DRUG USERS ' GEORGE W. JOE
Institute of Behavioral Research Texos Christian University LAVERNE KNEZEK AND DEENA WATSON
DARCO Drz~gServices, Inc Dallus, Texas D. DWAYNE SIMPSON
Institztte of Behavioral Research Texas Christian University Summary.-Two short scales, the TCU Depression Scale and the TCU Decisionmaking Scale, were psychometrically evaluated in a sample of 145 intravenous drug users. Coefficient alpha reliabilities were .78 for the 6-item TCU Depression Scale and .77 for the 9-item TCU Decision-making Scale. Concurrent validity of the former scale was assessed by correlating scores with those on the Beck Depression Inventory, r = .75. Based on the Beck Depression Inventory Clinical cutoff scores, 83% of the sample showed some depression, with 23% severely depressed, 39% moderately depressed, and 21% mildly depressed. Individuals scoring higher on depression on both tests tended to score lower on decision-making. Significant demographic associations of age, gender, education, and race-ethnicity were found for the depression and decision-making scales. More depression was noted for women, those younger, white, and having less education. Older and more educated intravenous drug users tended to score higher on decision-making. Validity for the depression and decision-making scales was assessed by examining correlations with behaviors. Significant positive correlations were found between depression scores and intravenous use of cocaine only, heroin and cocaine combined, and heroin only. Also, intravenous use of cocaine only and of cocaine and heroin combined were negatively related to decision-making. AIDS sex-risky behavior was positively correlated with depression and negatively correlated with decision-making.
According to the Health Belief Model (Rosenstock, 1966), sociopsychological variables might have considerable influence as modifying factors for individuals' perceptions regarding the threat posed by a disease as well as their likelihood of taking preventive action. Two of the psychological status constructs of major interest are depression and decision-making. Depression should affect attention, and readiness for making choices is a basic component of any model of behavioral change. 'This study was supported by the National Institute on Drug Abuse Grant 1 R18 DA05176-02. The interpretations and conclusions, however, do not necessarily represent the position of NIDA or the Department of Health and Human Services. Address correspondence to George W. Joe, Institute of Behavioral Research, Texas Christian University, Fort Worth, TX 76129.
340
G. W. JOE., ETAL
While depression has been recognized as a very important problem among intravenous drug users (e.g., Rounsaville, Weissman, Wilber, CritsChristoph, & Kleber, 1982; Rounsaville, Kosten, Weissman, & Kleber, 1986) and has received considerable study, a similar effort to study decision-making capacity and readiness among intravenous drug users has not been mounted. The importance of examining decision-making in relation to depression is particularly relevant in view of the cognitive-based therapies developed for depression (Hollon & Beck, 1986) and related counseling approaches used for addicts in drug-abuse treatment and other types of interventions to mohfy behavior (Platt & Hermalin, 1989). I n addition to these psychological constructs of depression and decision-making, a number of other constructs are also important in the evaluation of treatments and interventions for addicts. These include self-esteem, anxiety, risk-taking, antisocial personality, and motivation to change. For some of these constructs, instruments are available, but many of these were developed using items intended for a middle-class population. Some of the wordings of items may not: be as appropriate or understandable to an intravenous (IV-drug users) population. This paper describes preliminary findings based on brief scales for depression and decision-making; these were developed as part of an assessment battery for a project on treatment-evaluation focused on AIDS-risk reduction (Simpson, 1990) in IV-drug users. The description includes the psychometric properties of the scales and the predictive (concurrent) validity of rhe short depression scale in relation to the established Beck Depression Inventory. I n addition, scores on these scales are examined in relation to indicators of AIDS-risky behaviors, surrogate health indicators of AIDS risk, and self-reported frequency of drug use.
Sample A sequential sample of 154 IV-drug users who participated in the National AIDS Demonstration Research Project (NADR) in Dallas were asked to complete short self-rating scales immediately before their AIDS Initial Assessment (AIA) interview. These included individuals who had been recruited through community outreach efforts to complete a personal interview and confidential testing for HIV, with the stipulation they had not been in treatment for drug use in the past six months. However, nine of the IV-drug users who had agreed to participate did not complete the psychological scales and were deleted from the analyses. Distributions on demographic characteristics for the sample of 145 are presented in Table 1 . Approximately half (49%) were black, 42% were white, and 10% Hispanic. More than half ( 5 3 % ) were over 34 years old, and 72% were men. A majority also indicated being unmarried (57%), being unemployed at the time of the interview
INTRAVENOUS DRUG USERS: DEPRESSION, DECISIONS
341
(54%), and having graduated from h g h school or having a GED (76%). Almost half (48%) received income from dlegal sources during the past six months, and 25% claimed illegal activities provided them their major source of income. TABLE 1 FREQUENCY DISTRIBUTIONS: A N D DECISION-MAKING, A N D BEHAVIOR ( N = 145) DEMOGRAPHICS, DEPRESSION Measure Age (years) 18-34 35-39 G E 40 Male Race Black White Hispanic
Measure
To
90
Demographics Marital Status 47 Single 32 Married 21 Not currently married 72 Education (grades) 1-11 49 12 42 Over 12 10 Employed Source of Income (illegal) Behavior
NOVA Needle Risk 0 1 2
Hepatitis
3 18 79 25
NOVA Sex Risk 0 1 2 Health-Sex
Depression and Decision-making Beck Clinical (MFPt) 17 Not depressed (0-9) 12 Mild depression (10-18) 30 Moderate depression (19-29) 42 Severe depression (30-63) TCU Decision-making (9) C 7 0- 4 19 10-16 37 17-24 38 25-28 29-36 *MFN, minimizes false negatives. tMFP, minimizes false positives. Beck Clinical (MFN*) Not depressed (0-9) Mild depression (10-15) Moderate depression (16-23) Severe depression (24-63) TCU Depression (6) C 0- 9 5- 8 9- 12 13-24
Scales Depression was measured by two separate scales. One was the 21-item Beck Depression Inventory (Beck & Steer, 1987), and the other was a short 7-item scale developed by the Institute of Behavioral Research at Texas Christian University (TCU). Decision-making was measured by a 10-item
342
G. W. JOE., E T A L .
scale that was also developed. AIDS-risky behaviors were assessed by the two AIA-based NOVA indices (developed by NOVA Research Corporation, 1989) measuring needle-risk behavior and sex-risk behavior. Two surrogate health indicators for AIDS risks, also based on data from the AIA, were (1) whether or not the individuals were told by a physician or nurse that they had ever had hepatitis, and (2) whether or not the persons had been told that they had ever had a sexually transmissible disease (syphilis, gonorrhea, herpes, or chlamydia). The distributions of the AIA-based indices are shown in Table 1. Five drug-use frequency variables taken from the AIA included (1) crack use, (2) IV-amphetamine use, (3) IV-heroin-only use, (4) IV-cocaine use, and (5) the combination of IV-heroin and cocaine use (speedball).
The distributions of the items and the item-total correlations for the TCU Depression Scale and the TCU Decision-making Scale are presented in Table 2. This table shows that on four of the seven depression items, at least 40% of the sample rated themselves as "often" or "almost always" feeling TABLE 2 TCU DEPRESSION AND TCU DECISION-MAKING SCALES: AND ITEM-TOTNCORRELATIONS ( N = 145) ITEM DISTRLBU~ONS Scale Items
Depression 3. You feel interested in life. (Rt) 6. You feel extra tired oc run down. 8. You feel hopeful about the future. (R)* 12. You worry or brood a lot. 13. You have thoughts of committing suicide. 14. You feel sad or depressed. 17. You feel lonely.
Response ItemPercen~aaest- total r a b c d e
-
14 7 17 3 15 15
.SO .50 .19 .54 .41 .70 .49
Decision-maku~g 1 16 50 27 6 1. You make good decisions. 2. You chink of several different ways to solve a problem. 0 1 1 43 28 18 4. You make decisions without thmking about consequences. (R) 12 2 1 42 18 7 3 12 3 1 34 20 5. You think about probable results of your actions. 5 7 22 40 26 7. Your problems are caused by someone else. (R)* 1 3 22 44 3 1 9. You think about what causes your current problems. 7 2 1 37 19 16 10. You plan ahead. 11. You consider how your actions will affect others. 4 14 38 29 16 3 20 5 1 20 7 15. You have trouble malung decisions. (R) 16. You analyze problems by looking at all the choices. 3 14 44 23 15
.55 .60 .21 .42 .OO .23 .56 .44 .47 .49
*Deleted in reduced scale. t R means scoring reversed. $a: never, b: rarely, c: sometimes, d: often, e: almost always.
38 24 7 12 18 24 5 9 57 18 3 13 8 10
27 37 38 34 20 42 42
11 30 13 34 2 27 25
1
343
INTRAVENOUS DRUG USERS: DEPRESSION, DECISIONS
extra tired or run down, worrying or brooding a lot, feeling sad or depressed, and feeling lonely. One item had an item-total correlation below .20 (Item 8-You feel hopeful about the future). Among the 10 decision-making items, over 40% of the sample reported that they "often" or "almost always" thought about solving their or causes of their problems. O n the four items asking about decision-making in terms of consequences or planning ahead, 25% to 35% of the sample rated themselves as "often" or "almost always." One item had an item-total correlation below .2O (Item 7-Your problems are caused by someone else), for which two-thirds of the sample responded "often" or "almost always." I n Table 3, the means, standard deviations, and coefficient alpha internal reliabilities are presented for the depression and decision-making scales. For the TCU Depression Scale the coefficient alpha reliabhties are given for both the original 7-item scale and a reduced 6-item scale (from which Item 8 was deleted). Similarly, the coefficient alpha reliabllities are presented for the original 10-item TCU Decision-making Scale and a reduced 9-item scale (from which Item 7 was deleted). TABLE 3 R E L I A E ~COEFFICIENTS Y FOR DEPRESSIONA N D DECISION-MAKING SCALES(N= 145) Scales
No. Items
M
SD
Coefficient a Raw Standardized
Beck Depression Inventory TCU Depression TCU Depression TCU Decision-making TCU Decision-making
21 7 6 10
21.7 13.0 11.4 24.0 21.3
11.4 4.7 4.4 5.3
.90 .75 .78 .73 .77
9
5.1
.91 .76 .78 .74 .77
The psychometric analysis for these brief scales indicated that their reliability coefficients were acceptable. This included both the original and reduced scales. However, because the reduced scales are psychometrically adequate for this sample and the population of IV-drug users it represents, the remainder of the paper addresses the correlates of the reduced scales. The coefficient alpha reliabilities were .78 for the 6-item TCU Depression Scale, and .77 for the 9-item TCU-Decision-making Scale. Reliability for the longer 21-item Beck Depression Inventory was .90. Principal components analysis (using SAS PROC FACTOR; SAS, 1988) of the 6-item TCU Depression Scale yielded one eigenvalue above one (2.93), and it accounted for 49% of the variance. The analysis of the 9-item TCU Decision-making Scale yielded two eigenvalues above one (3.30 and 1.38), with the first accounting for 37% of the variance. For the Beck Depression Inventory, there were five eigenvalues above one (7.91, 1.50,
344
G. W. JOE., E T A L .
1.19, 1.07, and 1.01), with the first accounting for 38% of the variance. Examination of the eigenvalues suggests only one factor in the TCU Depression Scale, but potentially two factors in the TCU Decision-making Scale. However, the second eigenvalue was small compared to the first. With respect to the clinical cut-off scores for the Beck scale to classify levels of depression (Beck & Steer, 1987), 83% of the sample showed some depression. Based on their established cut-offs to minimize the number of false negatives (MFN), depression was classified as severe for 42%, moderate for 30%, and mild for 12%. In comparison, classifications for minimizing the number of false positives (MFP) showed that 23% were severely depressed, 39% moderately depressed, and 21% mildly depressed. The two Beck clinical classification systems correlated .91. TABLE 4
Scales 1. 2. 3. 4. 5. 6. 7.
Beck Beck Beck TCU TCU TCU TCU
Depression Inventory Clinical', MFN ~ l i n i c a lMFP ~, Depression (6) Depression (6) C' Decision-making (9) Decision-makine (9)
cd
1
2
3
4
5
6
.90* .88t .93t .75t .69t -.45t -.42t
.91t .72t .73t -.37t -.35t
.72t .73t -.43t -.41t
.78' .92t -.44t -.42t
-.43t -.42t
.77* .93t
"Reliabihty coefficient.
t p < .0001. 'Not depressed = 0-9, Mild = 10-15, Moderate = 16-23, 'Not depressed = 0-9, Mild = 10-18, Moderate = 19-29, 'Categorized (0-4, 5-8, 9-12,13-24). dCategorized (0-9, 10-16, 17-24, 25-28, 29-36).
Severe = 24-63. Severe = 30-63.
Table 4 presents Pearson product-moment correlations between the depression and decision-making scales. Scores on the Beck inventory and the two Beck clinical scales, denoted as Beck (MFN) and Beck (MFP), were highly correlated with the TCU Depression Scale and a categorization of the TCU Depression Scale; the latter is referred to as TCU Depression-C and used the cut-offs: 0-4, 5-8, 9-12, 13-24. The correlations of the Beck scale and its clinical classifications with the TCU Depression Scales (uncategorized and categorized) ranged from .69 to .75. These results demonstrate the concurrent (predictive) validity for the short TCU scale. One of the most interesting substantive findings was that the Decisionmaking Scale was significantly correlated with the depression scales (Beck and TCU); persons scoring higher on depression tended to score lower on decision-making (r = -.35 to -.45). The direction of the relationship fits what
345
INTRAVENOUS DRUG USERS: DEPRESSION. DECISIONS
one would expect, namely, that being depressed impairs one's confidence in decision-making. Correlations of demographic variables with the depression and decision-making scales are given in Table 5 . In this sample, women were more depressed than men ( r = .17 to .21 on the Beck scales and .21 for TCU Depression-C); older IV-drug users tended to be less depressed than younger ones ( r = -.23 to -.28), and they also scored higher on decision-making readiness ( r = .29) than younger ones. White respondents were more depressed than black subjects ( r = .28 for TCU Depression-C), and those with more education tended to be less depressed ( r = -.I4 to -.18) and to score hgher on decision-making ( r = .19). No significant relationships were found for marital status and employment status. TABLE 5 PEARSON CORRELA~ONS OF DEMOGRAPHICS WITH DEPRESSION AND DECISION-MAKING SCALES(N= 145) Scales'
Age Sex Race Black White Hispanic Education
Beck Depression Inientory
CfinjcaJ, MFN '
Beck Clinical. MFP
-.28$ ,177
-.23$ .2lt
-.24$ .19t
-.08 .05 .06 -.18t
-.I3 .05 .14 -.I4
-.08 .04 .08 -.I87
TCU Depression (6)
TCU Depression (6) C
TCU Decision-making (9)
TCU Decision-making (9) C
Age -.25$ -.25$ Sex .16 .21t Race -.28$ -.27$ Black Whice .28$ .28$ Hispanic .OO -.02 Education -.08 -.I1 "Marital Status and Whether Employed not significantly correlated with Depression and Decision-making scales.
In terms of the behavioral correlates shown in Table 6, there were significant positive correlations of depression scores with IV-drug use variables measuring cocaine only, heroin and cocaine combined, and heroin only. Depression was not significantly related to IV-amphetamine use or to crack use. In addition, use of IV-cocaine only and heroin and cocaine combined were
346
G. W. JOE., ETAL.
negatively related to decision-making, that is, those who injected more frequently tended to score lower on decision-making. The NOVA sex-risk index for AIDS risks was also negatively correlated with decision-making (r = -.21). O n the other hand, the NOVA sex-risk index was positively correlated with the TCU Depression Scale (Y = ,201. I n other words, persons with higher sex risks had lower scores on decision-making and higher scores in depression. TABLE 6 PEARSONCORRELATIONS OF BEHAVIOR W I - IDEPRESSION AND DECISION-MAKING SCALES(N = 145) Scales
NOVA Needle Risk NOVA Sex Risk Hepatitis Health-Sex Cocaine Only (IV) Amphetamines (IV) Heroin (IV) Heroin and Cocaine (IV) Crack
Beck Depression Inventory .ll .ll .10 -.I2 .20' -.06 .25t .19* .02 TCU Depression (6)
Beck Clinical,
MFN .09 .ll .16 -.I1 .17' -.lo .35t .20* -.03
TCU Depression (6) C
Clinical,
MFP .15 .I0 .12 -.08 .16* -.07 .28t .18* .01
TCU DeTCU Decision-making cision-making (9) (7) c
Nova Needle Risk NOVA Sex Risk Hepatitis Health-Sex Cocaine Only (IV) Amphetamines (IV) Heroin (IV) Heroin and Cocaine (IV) Crack
DISCUSSION Depression and decision-making readiness are proposed as important factors both in development and evaluation of AIDS intervention programs and for evaluation of cognitively-based counseling strategies. The two constructs received empirical support in this preliminary study. Specifically, it showed that a large percentage-either 62% or 72%, depending upon the strictness of the classification system-of this outreach sample was clinically depressed and that depression and decision-making are significantly related to risky behaviors such as frequency of IV-drug use. Methodologically, the
INTRAVENOUS DRUG USERS: DEPRESSION, DECISIONS
347
findings support the validity of the two short TCU scales measuring depression and decision-making. The results for the TCU Depression Scale paralleled the results for the longer and well-established Beck Depression Intheoventory, and the results for the TCU Decision-making- Scale supported .. retical expectations. Significant demographic relationships for age, sex, education, and race-ethnicity were found for the depression and decision-making scales. The preliminary results concerning depression and decision-making in this preliminary study suggest that the interventions need to address these constructs, as well as possibly other psychosocial constructs such as risk-taking and antisocial personality. REFERENCES BECK,A. T., & STEER,R. A. (1987) Beck Depression Inventory manual. San Antonio, T X : The Psycholog~calCorp./Harcourt Brace Jovanovich. HOLLON, S , & BECK, A. T. (1986) Research in cognitive therapies. In S. Garfield & A. E. Bergen (Eds.), Handbook of psychotherapy and behavioral change. New York: Wiley. Pp. 433-482.
SAS INSTITUTE.(1988) The factor procedure. In SASISTAT user's guide, release 6.03 addition. Cary, NC: Author. PLAT^, J. J., & HERMALIN, J. (1989) Social skdl deficit interventions for substance abusers. Psychology of Addictive Behaviors, 3, 114-133. ROSENSTOCK, I. M. (1966) Why people use health services. Milbank Memorial Fund Quarterly, 44, 94-127. ROUNSAVILLE, B. J., KOSTEN,T. R., WEISSMAN,M. M., & KLEBER,H. D. (1986) Prognostic significance of psychiatric disorders in treated opiate addicts. Archives of General Psychiatry, 43, 739-745. ROLINSAVILLE, B. J., WEISSMAN,M. M., WILBER,C. H., CRITS-CHRISTOPH, K., & KLEBER,H. D. (1982) Diagnosis and symptoms of depression in opiate addicts: course and relationship to treatment outcome. Archives of General Psychiatry, 39, 151-156. NOVA RESEARCHCO. (1989) Nefwork. Bethesda, MD: Author, 4600 East-West Highway, Suite 700. SIMPSON,D. D. (1990) Forms manual: drug abuse treatment for AIDS-risks reduction (DATAR). Fort Worth, TX: Texas Christian Univer., Institute of Behavioral Research. Accepted January 31, 1991.
