Test–Retest reliability of a lifetime drug use questionnaire

Test–Retest reliability of a lifetime drug use questionnaire

Addictive Behaviors 30 (2005) 361 – 368 Short communication Test–Retest reliability of a lifetime drug use questionnaire Christoph Czermaka,b,c,*, M...

137KB Sizes 1 Downloads 82 Views

Addictive Behaviors 30 (2005) 361 – 368

Short communication

Test–Retest reliability of a lifetime drug use questionnaire Christoph Czermaka,b,c,*, Michael Lehoferc, Peter Gasser-Steinerd, Sandra Ettingere, Leonidas Lemonisb,c, Alfred Rohrhoferb,c, Thomas Leglb,c, Peter M. Liebmanna,c a

Institute of Pathophysiology, University of Graz, Heinrichstrasse 31a, Graz A-8010, Austria b Drug Rehabilitation Center ‘‘Gruener Kreis’’, 2872 Moenichkirchen 25, Austria c Addiction Research Society, Hans-Resel-Gasse 18, Graz 8020, Austria d Institute of Sociology, University of Graz, Universitaetsplatz 15, Graz 8010, Austria e Wagner-Jauregg Clinics, Linz 4020, Austria

Abstract By adopting the investigational principle of the Lifetime Drinking History (LDH) interview, we developed a Lifetime Drug Use (LDU) Questionnaire to assess the amount and frequency of lifetime drug consumption. The Pearson Correlation Coefficients for test – retest reliability, investigated in a sample of N = 47 residents of a drug rehabilitation center and averaged over the investigated seven drug categories, were r =.95 for the abstinence-corrected total duration of regular use, r =.89 for the mean number of consumption days per month of regular use, r =.89 for the mean daily amount, r =.83 for the variability in consumption days per month of regular use, and r =.75 for month variability in daily amount. Applicability and further research needs are discussed. D 2004 Elsevier Ltd. All rights reserved. Keywords: Substance-related disorders; Human; Questionnaires; Psychometrics; Reproducibility of results; Therapeutic community

1. Background Retrospective recording of details on drug consumption over an individual’s lifetime period for the main, presently available drugs is valuable for treatment planning as well as for * Corresponding author. Institute of Pathophysiology, University of Graz, Heinrichstrasse 31a, Graz A-8010, Austria. Tel.: +43-676-708-6475; fax: +43-316-380-9640. E-mail address: [email protected] (C. Czermak). 0306-4603/$ – see front matter D 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.addbeh.2004.05.005

362

C. Czermak et al. / Addictive Behaviors 30 (2005) 361–368

research on addictive behaviors. However, limited attempts have yet been made to develop a respective interview instrument. The Lifetime Drinking History (LDH) yields highly reliable respective data, but only for alcohol (Skinner, 1984; Skinner & Sheu, 1982). Therefore, we adopted its interview principle to assess lifetime consumption data for a variety of different psychoactive substances and developed a Lifetime Drug Use (LDU) questionnaire.

2. Method 2.1. LDU questionnaire The LDU can be administered as a standardized interview or as a self-report form (a form unit is shown in Appendix A). According to the LDH interview principle, the respondent is asked to first think of the time when he/she started to regularly1 consume a certain psychoactive substance, to go then forward in time until the consumption pattern for this substance had changed. Thereby, a consumption phase is obtained, for which additional data on details of consumption is requested. Data are thus recorded from the beginning of regular consumption up to the present day, in various phases. Questions on ‘‘maximal amount’’ and ‘‘important life events and/or reasons for consumption of this drug’’ should serve for improving memory. Substance units for the amount of consumption are suggested to follow the local standard. For each psychoactive substance, five indices (LDU1–LDU5) are calculated from the raw data, with LDU1 describing the abstinence-corrected total duration of regular use and LDU2–LDU5 summarizing the frequency and amount of use, calculated in analogy to arithmetic mean and standard deviation (Appendix B). 2.2. Participants and procedure In this study, approved by the local ethics committee (LKH Graz, Austria, Nr. 12-147 ex 01/02), the test–retest reliability for the LDU was investigated in a sample of 50 (42 males, 8 females; mean F S.D. age: 29.0 F 10.7) ex-addicts, who were permanent residents at the Austrian drug rehabilitation center Gruener Kreis and had been drug abstinent for 10.3 F 7.0 (mean F S.D.) months. The LDU was applied as a self-report form, the test–retest time interval was 30 days, and the following substance categories were investigated: alcohol, cigarettes, cannabinoids, heroin, cocaine, sedatives, and hallucinogens. Test–retest reliability was investigated for each LDU index of each psychoactive substance with indicated regular use at one of the two sessions. If the consumption of a specific substance was indicated at only one of the two sessions, the parameters of nonreported use were set to zero and included in the study. Both Pearson Correlation Coefficients and intraclass correlation coefficients (ICCs, computed for consistency of random raters and estimates on single rating) were used to describe reliability.

1

The definition of ‘‘regular’’ consumption is left to the respondent.

C. Czermak et al. / Addictive Behaviors 30 (2005) 361–368

363

3. Results Three individuals dropped out from the rehabilitation program during the test–retest period due to relapse. As they did not differ in any of the demographic or LDU substance consumption variables from the rest of the group, they were completely excluded from the data analysis. It took participants 25 F 10 (mean F S.D.) min to complete the LDU at the first application (test) and 15 F 6 min at the retest. At the first application, the mean F S.D. number of regularly consumed substances from the investigated seven categories was 4.0 F 2.1 per person. Two persons reported regular use of a substance, that is, cocaine, respectively, hallucinogens, at only one test application but not on the other. Data on LDU indices obtained from the test and retest sessions (median and 10th–90th percentile range), together with the respective ICCs and Pearson correlation coefficients—all statistically significant at P < .05—are shown in Table 1. Among the LDU indices, the highest test – retest correlation was obtained for the abstinence-corrected total duration of regular use (LDU1), with an average Pearson coefficient of r=.95 (averaged over the different drugs and corrected by Fischer z transformation), followed by the mean number of consumption days per month of regular use (LDU2; average r=.89), mean daily amount (LDU4; average r=.89), variability in consumption days per month of regular use (LDU3; average r=.83), and month variability in daily amount (LDU5; r=.75). Among the different psychoactive substances, the highest test– retest correlation was obtained for cigarette consumption, with an average test–retest Pearson coefficient of r=.96 (averaged over the different LDU indices and corrected by Fischer z transformation), followed by hallucinogens (average r =.89), alcohol (average r =.87), sedatives (average r=.83), heroin (average r=.82), cocaine (average r=.82), and cannabinoids (average r=.78).

4. Discussion The high reliability coefficients obtained in this study may have been partly due to the administration of the LDU in the familiar atmosphere of a therapeutic community, with extensive (psycho)therapeutic focus on the individuals’ history of substance abuse. The reliability of drug use self-reports has been postulated to be promoted by the context of trusting relationship (Skinner, 1984) and may substantially differ in, for example, outpatients without care. Although reliability is a prerequisite for validity, a direct assessment of the validity of the LDU by presently used methods appears as limited. Biochemical assays, such as urine and hair analyses, can detect only recent drug use (Martin, Wilkinson, & Kapur, 1988), and individuals providing collateral reports may often not be in the position to accurately observe an individual’s drug consumption (Cordingley, Wilkinson, & Martin, 1990), in particular, over a prolonged time period. Possible applications of the LDU comprise the clinical field, as well as research on addictive behaviors. In the clinical context, LDU-obtained data can serve as basis for

364

C. Czermak et al. / Addictive Behaviors 30 (2005) 361–368

Table 1 LDU data given at the test and retest sessions (median and 10th – 90th percentile range), together with respective Pearson correlation (R) and intraclass correlation (ICC) coefficients Test session

Retest session

R

ICC

LDU1: Abstinence-corrected total duration of regular use (years) Alcohol (n = 37) 7.0 (2.2 – 25.1) 8.0 (1.8 – 25.9) Cigarettes (n = 38) 10.8 (6.8 – 25.5) 11.8 (6.1 – 25.7) Cannabinoids (n = 30) 6.8 (1.3 – 14.0) 7.3 (1.9 – 14.9) Heroin (n = 25) 3.0 (0.6 – 12.7) 4.4 (0.9 – 11.7) Cocaine (n = 26) 4.8 (0.4 – 12.2) 5.1 (1.6 – 13.3) Sedatives (n = 15) 6.0 (0.9 – 10.8) 5.0 (0.8 – 13.8) Hallucinogens (n = 20) 3.5 (0.5 – 12.8) 3.7 (0.9 – 14.1)

0.97** 0.98** 0.95** 0.94** 0.87** 0.85** 0.97**

0.97** 0.98** 0.95** 0.94** 0.87** 0.84** 0.97**

LDU2: Mean number of consumption days per Alcohol (n = 37) 18.6 Cigarettes (n = 38) 30.0 Cannabinoids (n = 30) 19.5 Heroin (n = 25) 21.4 Cocaine (n = 26) 13.0 Sedatives (n = 15) 10.6 Hallucinogens (n = 20) 4.3

(4.9 – 30.0) (30.0 – 30.0) (4.4 – 30.0) (10.5 – 30.0) (6.6 – 30.0) (1.3 – 30.0) (1.0 – 28.5)

0.89** 0.98** 0.79** 0.82** 0.79** 0.81** 0.89**

0.89** 0.98** 0.79** 0.82** 0.79** 0.80** 0.88**

(0 – 12.2) (0 – 0) (0 – 11.8) (0 – 11.8) (0 – 8.9) (0 – 3.3) (0 – 3.5)

0.72** 0.98** 0.44* 0.62** 0.72** 0.95** 0.77**

0.72** 0.98** 0.44** 0.62** 0.72** 0.83** 0.76**

(2.6 – 62.0) (6 – 100) (0 – 38) (0 – 73) (11.1 – 51.0) (0.1 – 4.9) (0.2 – 1.7) (0.1 – 3.5) (0.3 – 20.0) (0 – 1.5)

0.89** 0.76** 0.87** 0.79** 0.95** 0.75** 0.89** 0.88** 0.80** 0.93**

0.85** 0.76** 0.85** 0.79** 0.95** 0.74** 0.89** 0.87** 0.79** 0.86**

(0 – 13.5) (0 – 12.6) (0 – 1.7) (0 – 0.5)

0.67** 0.88** 0.70** 0.69**

0.62** 0.87** 0.70** 0.61**

month of regular (5.7 – 30.0) (30.0 – 30.0) (4.2 – 30.0) (8.4 – 30.0) (1.7 – 30.0) (1.0 – 30.0) (1.0 – 16.9)

use 18.0 30.0 18.8 21.1 17.1 15.7 5.3

LDU3: Variability in consumption days per month of regular use Alcohol (n = 37) 4.0 (0 – 12.4) Cigarettes (n = 38) 0 (0 – 0) Cannabinoids (n = 30) 2.5 (0 – 12.8) Heroin (n = 25) 6.6 (0 – 13.0) Cocaine (n = 26) 0 (0 – 9.1) Sedatives (n = 15) 0 (0 – 4.8) Hallucinogens (n = 20) 0 (0 – 1.7)

4.8 0 2.8 7.6 0 0 0

LDU4: Mean daily amount Alcohol (gram alcohol; n = 37) % Beer % Wine % Hard drinks Cigarettes (n = 38) Cannabinoids (g; n = 30) Heroin (g; n = 25) Cocaine (g; n = 26) Sedatives (tablets; n = 15) Hallucinogens (daily intake; n = 20)

15.9 45 14 35 20.0 1.8 0.6 0.8 4.5 0.2

(3.7 – 76.0) (8 – 100) (0 – 50) (0 – 73) (9.9 – 44.3) (0.1 – 4.0) (0.2 – 1.6) (0.1 – 4.6) (0.2 – 20.0) (0 – 1.5)

LDU5: Month variability in daily amount Alcohol (gram alcohol; n = 37) 4.7 (0 – 36.7) Cigarettes (n = 38) 4.1 (0 – 13.5) Cannabinoids (g; n = 30) 0.2 (0 – 2.0) Heroin (g; n = 25) 0.2 (0 – 0.6)

13.9 42 12 29 20.0 1.5 0.6 0.8 5.1 0.3

5.8 3.7 0.1 0.3

C. Czermak et al. / Addictive Behaviors 30 (2005) 361–368

365

Table 1 (continued ) Test session LDU5: Month variability in daily amount Cocaine (g; n = 26) Sedatives (tablets; n = 15) Hallucinogens (daily intake; n = 20) * P < .05. ** P < .01. g: grams.

0 (0 – 0.5) 0 (0 – 0.8) 0 (0 – 0.2)

Retest session 0 (0 – 0.8) 0 (0 – 2.2) 0 (0 – 0.4)

R

ICC

0.79** 0.59* 0.68**

0.73** 0.45* 0.68**

individually adapted therapy planning. LDU raw data can be easily transformed to diagrams showing timelines of amount/frequency of consumption and providing a quick overview of an individual’s drug career (Fig. 1 shows an example). The distinction of different consumption phases can help to identify, individually, the relevant factors for drug consumption. The LDU can be also used for research on psychoactive substance consumption behaviors. For example in the field of polydrug use, LDU-obtained data can be used to

Fig. 1. Example for a graphical display directly obtained from LDU raw data on lifetime cannabis and heroin consumption of a 29-year-old patient. Consumption phases, as well as possible mutual relationships between consumption of the different drugs, can be further analyzed together with the client.

366

C. Czermak et al. / Addictive Behaviors 30 (2005) 361–368

investigate mutual relationships between the consumption of different psychoactive substances, for example, whether certain drugs are preferably taken simultaneously or alternatively, as well as whether certain drugs generally serve as a predecessor of others. Cluster analysis may be used to identify different types of drug users according to their consumption behaviors. In conclusion, the present study has corroborated previous findings (Fals-Stewart, O’Farrell, Freitas, McFarlin, & Rutigliano, 2000; Hasin, Carpenter, McCloud, Smith, & Grant, 1997; Martin, Pearlman, & Li, 1998; Sobell, Kwan, & Sobell, 1995) that a systematic assessment of details on drug use of a further past can be reliably obtained from abstinent drug users in treatment. Further studies are needed for, for example, outpatients, different treatment settings and time points of rehabilitation, as well as for assessing the validity of LDU obtained data by the use of adjusted methods.

Appendix A. LDU form unit Phase*

Frequency

From: To:

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5

From: To:

From: To:

From: To:

From: To:

From: To:

Sporadic: on_____ days per month Weekly: on_____ days per week Daily No consumption Sporadic: on_____ days per month Weekly: on_____ days per week Daily No consumption Sporadic: on_____ days per month Weekly: on_____ days per week Daily No consumption Sporadic: on_____ days per month Weekly: on_____ days per week Daily No consumption Sporadic: on_____ days per month Weekly: on_____ days per week Daily No consumption Sporadic: on_____ days per month Weekly: on_____ days per week Daily

Important life events and/or reasons for consumption of this drug

Amount consumed on a typical day**

Abstinence, if less than one year*

Maximum: Average:

From: To:

Maximum: Average:

From: To:

Maximum: Average:

From: To:

Maximum: Average:

From: To:

Maximum: Average:

From: To:

Maximum: Average:

From: To:

*Data is given in terms of age. **Substance units used in the study: alcohol: half-liter beer (Austrian consumption standard unit), liter wine, liter hard drinks; cigarette consumption: number of cigarettes; cannabinoids, heroin, and cocaine: gram; sedatives: number of tablets consumed; hallucinogens: number of intake per day.

Index description

Calculation formula

LDU1 = abstinence-corrected total duration of regular use LDU2 = mean number of consumption days per month of regular use LDU3 = variability in consumption days per month of regular use

= Sum of all phase lengths, with indicated regular use minus indicated abstinence periods.

LDU4 = mean daily amount during regular use time LDU5 = month variability in daily amount during regular use time

= Total number of consumption days/LDU1 (in months) Note: The total number of consumption days follows directly from the raw data. sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi P 2 ðphase length ðin monthsÞðfrequency per monthðLDU2ÞÞ Þ

¼

phases

LDU1ðin monthsÞ

Note: Only phases with indicated regular use are taken into account. = Total amount of the substance consumed during the total duration of regular use/LDU1 (in days). Note: The total amount of the substance consumed follows directly from the raw data+ sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi P 2 2 ðnumber of consumption daysphase ðdaily amountphase ðLDU4ÞÞ Þþðnumber of abstinent daysphase ðLDU4Þ Þ

¼

phases

LDU1ðin daysÞ

+ For alcohol, the investigated liter-amounts for consumption of beer, wine, and hard drinks are summarized into the respective gram-amount of pure alcohol consumed, using conversion coefficients according to the local standard.

C. Czermak et al. / Addictive Behaviors 30 (2005) 361–368

Appendix B. Index description and calculation formulas for LDU indices

367

368

C. Czermak et al. / Addictive Behaviors 30 (2005) 361–368

References Cordingley, J., Wilkinson, A., & Martin, G. W. (1990). Corroborating multiple drug users’ posttreatment selfreports by collaterals. Behavioral Assessment, 12, 253 – 264. Fals-Stewart, W., O’Farrell, T. J., Freitas, T. T., McFarlin, S. K., & Rutigliano, P. (2000). The timeline followback reports of psychoactive substance use by drug- abusing patients: Psychometric properties. Journal of Consulting and Clinical Psychology, 68, 134 – 144. Hasin, D., Carpenter, K. M., McCloud, S., Smith, M., & Grant, B. F. (1997). The alcohol use disorder and associated disabilities interview schedule (AUDADIS): Reliability of alcohol and drug modules in a clinical sample. Drug and Alcohol Dependence, 44, 133 – 141. Martin, G. W., Pearlman, S., & Li, S. (1998). The test – retest reliability of the frequency of multiple drug use in young drug users entering treatment. Journal of Substance Abuse, 10, 275 – 290. Martin, G. W., Wilkinson, D. A., & Kapur, B. M. (1988). Validation of self-reported cannabis use by urine analysis. Addictive Behaviors, 13, 147 – 150. Skinner, H. A. (1984). Instruments for assessing alcohol and drug problems. Bulletin of the Society of Psychologists in Addictive Behaviors, 3, 21 – 33. Skinner, H. A., & Sheu, W. J. (1982). Reliability of alcohol use indices. The Lifetime Drinking History and the MAST. Journal of Studies on Alcohol, 43, 1157 – 1170. Sobell, L. C., Kwan, E., & Sobell, M. B. (1995). Reliability of a drug history questionnaire (DHQ). Addictive Behaviors, 20, 233 – 241.