Marijuana

MARIJUANA-SPECIFIC SCREENING IS NEEDED

Marijuana-specific screening merits more clinical attention because of marijuana's prevalence, adverse effects, and clinical assessment dilemmas. Assessing client marijuana use is a clinically challenging task in many settings because of the current social, legal, and scientific debates surrounding marijuana use; inadequate professional training in substance use disorders; DSM IV diagnostic dilemmas with cannabis; and limitations with current screening methods (1-3). A marijuana-specific screening inventory that is not body fluid dependent, but reliable, valid, easily administered, and quickly scored with clinically useful cutoffs, could assist clinicians in assessing marijuana use. This study describes the psychometric properties of the Marijuana Screening Inventory--Experimental Version (MSI-X) with a clinical sample referred to a substance abuse agency for assessment. The study builds upon a previous examination of the MSI-X psychometric properties with a community sample, and pilot use of the MSI-X with mental health clients for screening, promoting assessment discussion, and brief intervention (4,5).

Marijuana's Prevalence and Adverse Effects Support Specific Screening

Clinical screening focused specifically on marijuana use appears necessary based upon prevalence. Marijuana is the most widely used illicit substance. Lifetime use ranges from 32% to 46%, and as high as 53% to 73% among 19- to 40-year-olds, with lifetime cannabis dependence and abuse estimates at 5% to 9% (6-9). Admissions for marijuana abuse increased at public drug and alcohol treatment facilities in 41 of 50 states between 1993 and 1999, totaling 14% in 1999 (10). Marijuana use increased significantly between 2000 and 2001 with 5.4% of the population reporting past month and 9.3% past year use (6). The highest use was among 18- to 25-year-olds, with 16% indicating past month and 26.7% past year use (11). Further highlighting the need for marijuana-specific screening are findings that the majority of all those using illicit drugs in a past year report marijuana use only. In 2001, 56% used only marijuana, 20% used marijuana and another drug, while 24% used illicit drugs other than marijuana (6).

Marijuana specific screening is also needed due to the multiple adverse effects associated with its use. While marijuana's risks may be minimized or confused by culturally or scientifically mixed messages about marijuana, recent reviews and new research continues to document marijuana's adverse psychobiosocial effects (1,8,12-14). New evidence describes marijuana's negative effects on driving, psychosocial development, cognitive ability, educational attainment, work adjustment, depression, health, pregnancy, and as an increased risk factor for mental illness, substance disorders, risky sexual behavior, and aggressive behavior during withdrawal (15-36). While findings sometimes conflict, the multiple adverse effects associated with marijuana use justify developing marijuana-specific screening methods.

The Marijuana Screening Inventory Addresses Clinical Needs

The MSI-X is compatible with recommendations made by the Substance Abuse and Mental Health Service Administration (SAMHSA) 2002 Report to Congress and the President's 2003 New Freedom Commission on Mental Health Report, which emphasize the important role screening should play for effective recognition and treatment of co-occurring mental and substance use disorders (37,38). The SAMHSA report states that "effective treatment ... begins with accurate screening," and its goal is to train providers to "screen all individuals" presenting to multiple settings for co-occurring disorders (37). Others simultaneously emphasize the need to shift from a benign neglect attitude about marijuana toward more focused attention on recognizing and treating cannabis-use disorders (39). Others point to a growing recognition of comorbidity between psychiatric disorders and cannabis dependence (29,35,40).

To improve early recognition of cannabis use, marijuana-specific screening has been recommended for general mental health and primary care settings (1,5,41). Fleming suggests clinicians could detect 80% of drug users by asking just about marijuana use during intake interviews (41). However, relying on initial interviews to fully screen and assess client marijuana use contains dilemmas (1-3,5,42). One dilemma is that, unlike alcohol at risk standards for men and women, no empirically accepted marijuana at risk quantity and frequency standards exist (43). Still, Fleming recommends setting the cutoff for a positive screen at client disclosure of any illicit drug (marijuana) use in the past six months (41). But Plant criticizes "drug use = drug problems" assumptions as "oversimplified or factually incorrect," indicating that research casting "light on patterns of possible consequences" with drug use is what "matters" (44). Traditional DSM-IV guidelines specify criteria for a maladaptive pattern of substance use, progressing in diagnostic severity from abuse (one of four criteria) to dependence (three of seven separate criteria) (9). While the DSM IV-TR offers 11 criteria to fit all substance-use disorders, some question this assumption in relation to cannabis, especially whether physiological dependence or withdrawal apply (1,8,12). Others question the validity of two separate diagnostic constructs for cannabis abuse or cannabis dependence (3,42,45). Authors of an Australian national household study conclude the 11 combined DSM-IV criteria for abuse and dependence represent one general cannabis use disorder diagnosis (3). They also found those reporting any three of the 11 DSM-IV criteria symptoms met the threshold for a cannabis-use disorder, which included 23.7% of those using cannabis [greater than or equal to] 5 times in the past year (3). Thus, exploring DSM IV-TR criterion thresholds with MSI-X scores may be relevant. The MSI-X provides multi-item screening of a client's overall marijuana pattern of risk, to help detect if it is maladaptive and deserving more extensive clinical evaluation for cannabis disorders using DSM IV-TR criteria. Pilot use of the MSI-X within a general screening battery for other disorders found it useful for identifying co-occurring cannabis abuse and dependence with clients presenting with mood and anxiety disorders (5).

The MSI-X Compensates for Limitations with Existing Screening Methods

The MSI-X compensates for limitations with existing screening methods and was developed because of the unavailability of other marijuana-specific screening inventories. These limitations have been described previously and are reviewed here (1,4). Biochemical laboratory screening and general all-purpose drug abuse inventories have major limitations for marijuana screening (1,46,47). Laboratory tests, the predominant method of marijuana screening, consist of four urine tests and a sweat and saliva test (48,49). While generally accurate, lab tests only determine past marijuana use, and marijuana requires 12 to 30 days to leave the body (12,50). However, screening is not just to establish past use, but assess if present marijuana use patterns constitutes a risk. Most clinicians in outpatient mental health or social service settings interested in assessing marijuana use do not have access to laboratory facilities and tests, nor the knowledge necessary to interpret lab values. Clients presenting voluntarily may not agree to a lab referral or the extra costs and requirements for providing urine or blood (51). Interestingly, differences between cannabis self-reports and lab screens have been reported within five substance abuse clinics, with self-reports higher and more accurate than two urine screen methods (52). Thus self-report information, such as the MSI-X, has value even with lab tests.

General all-purpose drug use inventories are not designed to specifically screen for marijuana risk patterns. Existing inventories attempt to screen for all illicit and prescription drugs, but may only mention marijuana among many drugs listed in the instructions, or at best contain one to three items about marijuana (53-63). For example, the 93-item SASSI-3 mentions marijuana (pot) only once (62). Longer screening inventories are seldom used outside of substance abuse specialty clinics (61-63). While some general drug screening inventories have been developed with cannabis user subsamples, they are not specifically tailored for marijuana screening or as suitable as the MSI-X for follow-up assessment questions specific to marijuana use, consequences, and diagnosis (42,58,61). The MSI-X attempts to improve upon or supplement existing screening methods.

Although numerous alcohol-specific screening inventories exist, marijuana specific screens appear rare (64,65). The five marijuana questionnaires located are designed for research with already identified marijuana abusers and not designed for general screening purposes (66,67). (Gorelick D. The Marijuana Quit Questionnaire. Unpublished manuscript and poster session by Boyd, S., Gorelick, D., Huestis, M., Heishman, S., Dermand, J., Simmons, M., Tashkin, D. Quitting and Relapse Among Non-Treatment Seeking Marijuana Smokers. Confirmation MQQ was not designed for screening. Personal communication, 2002) (68,69). Thus the MSI-X appears to be the only marijuana-specific inventory available to screen the clinical study sample. However, the MSI-X requires further evaluation of its psychometric qualities and scoring criteria with clinical samples (4,5,70).

PURPOSE

The purpose of this investigation is to determine: 1) the MSI-X reliability and factor structure with a clinical sample undergoing substance use evaluation at a community agency; 2) the MSI-X scoring cutoff sensitivity and specificity using receiver operator characteristic curves in association with a DSM IV-TR diagnostic criteria measure; and 3) the clinical sample responses to MSI-X items and collective at risk marijuana use based upon the MSI-X empirically derived cutoff scores by total sample, gender, and age.

METHODS Subjects

Clinical sample subjects are adults referred to the Community Outreach Referral and Evaluation (CORE) section of the Houston, Texas, Council on Alcohol and Drugs (Council) agency, completing substance abuse evaluations between September 2001 and June 2002. The Council, is a comprehensive urban alcohol and drug program funded by state, county, private, and program fees. Many CORE referrals are court mandated. Approximately 59% of adults scheduled during the study year kept appointments and completed assessments. Of clients completing assessments, 81% were mandatory court referrals to CORE from Child Protective Services (48.3%); family courts (17.1%); adult probation (4.1%); and DWI offenses (12.1%). Another 19% of CORE clients were referred by family, friends, or others (7.7%); self-referrals (4.9%); employee assistance programs (4.7%); and professional organizations (1.8%) [Shireman, L. CORE section referral information. Personal communication (November 15, 2002)].

A total of 109 individuals voluntarily completed a three-part Marijuana Inventory, representing 14% of those (n = 797) completing CORE assessments during the 10-month data collection period. Agency staff turnover periodically interrupted data collection. Two subjects were deleted for not providing complete MSI-X data, leaving a final study sample of 107 adults. The clinical sample is 60% men (n = 64) and 40% women (n = 43). Age ranged from 18 to 54, with a mean age of 33.2. One subject did not provide age information. Racially, the sample is 63% white, 20% black, and 17% Hispanic. Educationally, 68.9% completed high school, 18.4% one to four years of college, and 9.1% formal education or technical training beyond college. Most are employed, 67% full time for three years and 14.7% part time, but 10% reported frequent unemployment. Half are employed in business, clerical, or blue-collar occupations (54.13%). About 5% are students or retired. Within the sample, 37.6% are currently separated or divorced, 28.4% married, and 29.4% never married.

Procedure

A Marijuana Inventory was administered containing three sections: 11 demographics and lifetime-to-recent-use questions, the embedded 31 item MSI-X, followed by a separate 30 item DSM IV-TR criteria set of questions specific to marijuana. The Marijuana Inventory was administered during CORE's routine three-hour assessment process. Court-mandated clients are aware a CORE report is sent to the court for judicial review. When clients arrived, a receptionist provided them with an agency questionnaire packet, which took one hour to complete, prior to a clinician interview. The agency packet contains six substance-use inventories that are not part of the present study. After completing and returning agency questionnaires, clients wait while clinicians review and score the inventories. The Marijuana Inventory was administered during this waiting period. After returning the required assessment battery, receptionists ask clients if they would complete an additional Marijuana Inventory on a confidential basis while waiting to talk with a clinician. Clients were informed: completion of the Marijuana Inventory was voluntary; for research purposes only; would not be included with in the agency questionnaires just completed; never forwarded to the clinician interviewing them; and placed in a separate research file for future data entry and statistical analyses by an outside investigator. Clients willing to participate were given an agency research consent form to read, sign, and return. They then received a Marijuana Inventory with a numerical code, and they were told not to use their names to ensure confidentiality. Instructions for completing the Marijuana Inventory and additional confidentiality reassurances are printed on a cover sheet. Clients completing the Marijuana Inventory returned it to the receptionist, who placed it in the separate research file. Reasons for client refusal to complete the Marijuana Inventory ranged from: not wishing to fill out an additional inventory; fatigue from already completing multiple inventories; sensitivity about disclosing marijuana information; and distrust and suspicion that marijuana information could be used against them in the evaluation process. In reality, no clinicians conducting interviews had access to client marijuana inventories either during or after evaluation or information on scoring. Agency, state, and university human subject protocols were followed to ensure subject confidentiality. Completed marijuana inventories were retained in a confidential file by the CORE director and released later for data entry purposes.

Description of the Marijuana Screening Inventory (MSI-X)

The 31-item MSI-X embedded within the Marijuana Inventory has been previously described as a marijuana specific, self-report inventory taking clients 10 minutes to complete (4). The MSI-X is patterned after standardized alcohol screens with questions derived from cannabis research, some DSM IV criteria, potential adverse consequences, and possible marijuana use patterns. The MSI-X consists of 31 yes or no scored questions described in Table I. Nonscored questions ask subjects about lifetime, past year, past month and past two-week marijuana use, along with five embedded frequency requests if items 27 to 31 receive Yes responses. Scoring the MSI-X involves adding the yes responses for 31 scored items to arrive at an individual total score. Previous investigation of the MSI-X with a community sample (n = 408) found it to be internally reliable (alpha = .89), with nine factors explaining 65.8% of the variance (4). Preliminary clinical scoring cutoffs were suggested for high, moderate, and low risk patterns of marijuana use, but these scoring thresholds will undergo more rigorous empirical analysis in this study. Pilot use of the MSI-X found it to be clinically useful with clients, but needing additional psychometric evaluation, which is the purpose of this study (5).

Description of the DSM IV-TR Guided Marijuana Inventory (DSM-G-MI)

The final section of the Marijuana Inventory consisted of a DSM IVTR criterion guided set of 30 questions keyed directly to the four abuse and seven dependence criteria, but specifically worded for marijuana use in the last year. This DSM IV-TR guided marijuana inventory (DSM-G-MI) was created to assess four diagnostic classifications of cannabis disorders by level of severity. The DSM-G-MI was developed as a gold standard diagnostic alternative, since lengthy, direct diagnostic interviews with each subject were not possible due to agency confidentiality policies and the research protocol. The DSM-G-MI is patterned after the 11 universal criteria for abuse and dependence described in the DSM IV-TR, which are assumed to be equally applicable to alcohol and all classes of psychoactive substances, including marijuana (9). Each criterion and each of the two diagnostic outcomes are assumed to be independent and to have equal weight, but an abuse diagnosis is considered less severe than dependence (9). Abuse is measured by four criteria, and a diagnosis made if one criterion is endorsed (and a diagnosis of dependence is absent). Dependence is measured by seven criteria and a diagnosis made if at least three criteria are endorsed.

The DSM-G-MI reworded the 11 DSM IV-TR criteria specifically for marijuana as single, double, or multiple items, to avoid professional jargon and improve clarity by separating symptom strings. For example, the first DSM IV-TR abuse criterion for major role disruption was restated as three separate items: While using marijuana in the last year, 1) have you ever skipped school or work for all or part of the day? 2) sometimes neglected taking care of stuff around the house, your kids, household chores? 3) contributed to not performing well on school or work tasks? Twenty questions cover 10 criteria and require circling yes or no responses. The withdrawal criterion was presented as 10 brief symptoms, previously described by Budney, with four Likert scale choices for each, ranging from not at all to severe (19). Four possible diagnostic categories are derived from the DSM-G-MI: 1) Cannabis Abuse Only, based upon a yes response to any one of four abuse criteria, with no dependency criteria reported; 2) Cannabis Use Disorders, based upon an individual's yes responses to any three of 11 abuse or dependency criteria, reflecting the Australian findings of one overall diagnosis and threshold (3); 3) Cannabis Dependency Only, based upon yes responses to any three of the seven dependence only criteria; and 4) Both Cannabis Dependence and Abuse, based on combined yes responses to at least one abuse and three dependence criteria. The withdrawal scale qualified as one dependency criterion only if a subject reported experiencing two symptoms mildly, or one symptom moderately or severely.

The internal consistency reliability of the 20 yes/no items for the DSMG-MI is .84, while the withdrawal criterion scale of 10 symptoms achieved an internal reliability of .94. These reliabilities support the use of the DSMG-MI as a diagnostic comparison measure for the MSI-X scores. For this investigation, the DSM-G-MI is used only as a diagnostic criterion reference measure for assessing the psychometric accuracy of the MSI-X and possible MSI-X cutoff thresholds with each of the four cannabis diagnostic categories (70).

RESULTS

The Statistical Package for the Social Sciences (SPSS) version 11 generated all data analyses. Descriptive statistics for the 107 clinical sample adults indicates 90% report lifetime marijuana use, 48% past year, 18% past month, and 14% marijuana use in the past two weeks. Table 1 summarizes the yes or no responses for 107 clients completing the MSI-X. Individual accumulated MSI-X yes responses ranged from 0 to 23. Ten percent or more of the sample responded yes to 16 of 31 items. Responses to item 14 and 16 reveal 15.9% of this sample report using marijuana every day, while 14% use marijuana most all day, every day. The highest percentage of yes responses are on items 5 and 11, indicating 24.3% feel bad about marijuana use, with 21.7% acknowledging others pointed out marijuana was affecting (them) negatively. The lowest percentage of yes responses are on items 29 and 13, indicating only 2.8% report accidents while using marijuana and 4.8% a cough attributed to marijuana use.

The MSI-X internal consistency reliability, assessed by Chronbach's alpha, is .90. Exploratory factor analysis of the MSI-X item structure using principal components reveals interrelationship patterns among all 31 variables within a nine-factor structure. The final decision on the number of factors to extract was based upon the Kiser-Guttman eigenvalue one criteria and visual inspection of the scree plot test (71). Both support retention of all nine factors, which explained 72.2% of the total variance. Varimax rotation clarifies the maximum variance of squared loading for each factor and allows factor interpretation (71). Table 2 presents the respective eigenvalues, percent of variance accounted for by each factor, and varimax factor-loading matrix for all 31 items comprising the nine factors detected. Items loading [greater than or equal to] .39 are considered significant and highlighted, with the highest loadings on a given factor also underlined. All 31 MSI-X items obtained primary or secondary loadings of .39 or higher. Examination of the varimax factor-loading matrix in Table 2 provides an empirical basis to derive a meaningful set of nine factor-based scales for the MSI-X based upon [greater than or equal to] .39 loadings. In deriving and refining the factor-based scales, factor analytic interpretive conventions suggested by Kim and Mueller (71,72) were followed, providing latitude for inclusion or exclusion of secondary factor loadings below .90 (72). Thus, MSI-X factor-based scale descriptions and related items were derived by including the highest primary loading items per factor (from Table 2), then determining if they clustered into meaningful descriptive sets for each factor. Based upon primary loadings of .39 to .85, all 31 MSI-X items are suitable for inclusion on one of the nine factor-based scales. Items with the highest primary loadings per factor cluster as common construct themes. Twenty MSI-X items obtain one primary loading per factor, while 11 items load on two different factors. Based upon Kim and Mueller's (71,72) interpretation guidance, secondary items loadings of [greater than or equal to] .39 are not included on factor-based scales when their loadings are higher on another scale.

There were two exceptions: although items 22 and 23 load almost equally on two different factors they were placed on factor five, as they clinically fit best on that scale construct. Clinical experience and judgment were involved in deciding what descriptive titles to give each factor, based upon the predominant conceptual theme expressed by the highest loading items. The nine factor titles and their associated items are described below.

Factor I, entitled Job and Interpersonal Interference, accounts for 27.6% of the variance and includes five variables with item summary descriptions as follows: 1) marijuana-created problems with significant others (Item 3); 2) marijuana interferes with work (Item 4); 3) marijuana interferes with education or learning your job (Item 10); 4) losing friends because of marijuana use (Item 19); and 5) difficulty getting a job done (Item 20). Factor II, entitled Frequent Pattern of Use, accounts for 10.4% of the variance and includes five variables: 1) difficulty stopping or controlling marijuana use (Item 2); 2) frequently use marijuana during sexual encounters (Item 9); 3) use marijuana every day (Item 14); 4) use marijuana before noon (Item 15); and 5) use marijuana throughout the day (Item 16). Factor III, entitled Internal Consequences, accounts for 6.5% of the variance and includes five variables: 1) feel bad about marijuana use (Item 5); 2) others pointing out marijuana's affecting you negatively (Item 11); 3) don't feel rested after use (Item27); 4) use marijuana to feel better when nervous or depressed (Item 28); and 5) felt anxious or paranoid when using marijuana (Item 30). Factor IV, entitled External Consequences, accounts for 5.5% of the variance with three variables: 1) arrested while using marijuana, but not for possession or sales (Item 6); 2) trouble at school or job involving marijuana (Item 25); and 3) had an accident while using marijuana (Item 29). Factor V, entitled Memory and Physical Effects, accounts for 5.4% of the variance and includes five variables: 1) memory difficulty while using marijuana (Item 1); 2) cough attributed to marijuana (Item 13); and 3) marijuana withdrawal nervousness, headaches or irritability (Item 17); 4) marijuana tolerance in needing larger amounts to get high (Item 22); and 5) marijuana interferes with getting things done (Item 23). Factor VI, entitled Under the Influence, accounts for 4.9% of the variance with two variables: 1) arrested for drunk driving but marijuana involved (Item 7); 2) difficulty getting schoolwork or job done because high from marijuana (Item 32). Factor VII, entitled Use to Feel Normal with Interpersonal Costs, accounts for 4.4% of the variance with three variables: 1) marijuana interferes in talking about and resolving issues (Item 18); 2) lying about marijuana use to significant others (Item 21); and 3) using marijuana to feel normal Item 31). Factor VIII, entitled Sought Help for Use, accounts for 4.1% of the variance with two variables: 1) consulted helping professional about marijuana-associated problems (Item 12); and 2) attended self-help group because of marijuana (Item 24). Factor IX, entitled, Marijuana Arrest, accounts for 3.4% of the variance with one variable: 1) arrested for possession or sales of marijuana (Item 8).

The DSM-G-MI, based upon responses by 107 subjects (96 reporting lifetime marijuana use), categorized 39 individuals (36%) as meeting DSM IV-TR criteria for cannabis abuse only; 30 (28%) met criteria for the cannabis use disorders construct; 22 (21%) met criteria for cannabis dependence only; and 18 (17%) met the combined criteria for both cannabis dependence and abuse. Subject numbers and diagnostic classifications reflect progression from less to more severe symptoms. The DSM-G-MI was used as a diagnostic criteria reference measure for plotting the MSI-X receiver operating characteristic (ROC) curves with each of the four cannabis diagnostic categories (70). Correlation analysis between the MSIX total score and the DSM-G-MI total number of symptoms reported, indicates strong concurrent validity between the MSI-X and DSM-G-MI 20 yes/no items (r = .804, p = .000), and moderate concurrent validity with the 10 withdrawal scale items (r = .618, p = .000).

Receiver operating characteristic (ROC) curves are used to assess the psychometric accuracy of the MSI-X in two ways: 1) by plotting the summary index of the ROC called the area under the curve (AUC), which computes a probability regarding the accuracy of the MSI-X to correctly identify subjects more likely to have one of the four cannabis disorder classifications; and 2) by determining the MSI-X score significance level at various cutoff values, by computing the sensitivity (true positive rate of cases) and specificity (true negative rate of cases) (70). Receiver operator characteristic curves are presented in Fig. 1 for Methods A, B, C, and D representing the four cannabis diagnostic classifications. These ROC curves plot the sensitivity against 1 - (minus) specificity so that the curve area is an overall measure of MSI-X accuracy. A receiver operating characteristic area of 1 (upper left corner of the graph) theoretically indicates that the test is always correct, and the area of 0.5 (the diagonal line bisecting the plot area) indicates that the accuracy is no better than chance alone. The ROC summary index probabilities for MSI-X accuracy computed by the area the under curve (AUC) are: .91 for both cannabis dependence and abuse; .90 for cannabis use disorder; .90 for abuse only; and .86 for dependence only.

MSI-X optimal cutoffs were computed for each ROC method plotted in Fig. 1 for the four DSM IV-TR diagnostic classifications and different degrees of severity they convey. Each MSI-X score cutoff is associated with the maximal product of sensitivity and specificity reported in the Fig. 1 table below each ROC method. Each method's optimal cutoff point is a best estimate decision based upon obtaining the maximum sensitivity and specificity possible, given their inverse relationship. Examination of the sensitivity and specificity tables reveals two optimal MSI-X screening cutoffs associated with the four cannabis diagnostic classifications methods. The consensus optimal screening cutoff derived from three ROC methods (B, C, and D) is 6, based upon their maximum products of sensitivity and specificity. Thus, the MSI-X score of [greater than or equal to] 6 is the optimal cutoff associated with high risk for cannabis use disorders, cannabis dependence only, and both cannabis dependence and abuse. However, the abuse only optimal cutoff score of 3 suggests consideration of an associated 3 to 5 moderate risk score. Examination of the MSI-X sensitivity tables under each ROC method suggest the best estimate of maximum sensitivity (true positive cases) and specificity (true negative cases) is the 6 cutoff for Method D, both cannabis abuse and dependence, which yields a sensitivity of .833 and specificity of .888. The method A, abuse only, MSI-X cutoff score of 3 reveals a sensitivity of .821 and a specificity of .809.

Table 3 presents the total clinical sample response percentages for lifetime, past year, and past month marijuana use, gender, and age, falling within the MSI-X cutoff scores based upon ROC analysis with the DSM IV-TR diagnostic categories. The MSI-X scoring ranges are: Low Risk = 1 to 2; Moderate Risk = 3 to 5; and High Risk = 6 or more yes responses. Table 3 indicates 23.4% of the clinical sample responded yes to six or more items placing them in the high risk range, 17.8% scored in the moderate risk range, 21.5% in the low risk range, and 26.2% in the no problem range. Among past-year and past-month marijuana users, 28.5% and 12.2% respectively obtained moderate and high risk scores. No significant chi square differences were found between these four risk levels and six age groups, three collapsed age groups, or males and females. Table 3 reveals a higher percentage of males (15.9%) than females (7.5%) scored in the high risk range, but the same percentage (9.3%) of females and males scored in the moderate risk range. Although age range and subject size varies, those reflecting greater percentages of moderate and high risk marijuana use are 18 to 25 (15.1%) and 41 to 54 years of age (13.2%).

DISCUSSION

Clinical Sample Support for MSI-X Psychometrics

Clinical sample support for the psychometric usefulness of the MSI-X comes from the inventory's reliability, item loadings, nine-factor structure, amount of variance explained by the factors, and empirically derived marijuana risk screening score cutoffs with a DSM IV-TR criterion comparison. The MSI-X achieved good reliability (alpha = .90), supporting its item homogeneity and accuracy as a marijuana-specific measure with a clinical sample of mixed alcohol and polysubstance users. The exploratory factor analysis, amount of variance explained, and nine-factor structure, offers further support for the construct validity of the MSI-X as a screening tool. While MSI-X items were expected to cluster in meaningful ways based upon clinical theory, it was not hypothesized in advance what factors should emerge. The nine factor-based scales appear conceptually coherent with descriptive labels conveying reasonable constructs. All 31 MSI-X items are retained based upon the [greater than or equal to] .39 factor-loading criterion, with 26 items achieving higher loadings of [greater than or equal to] .54 and contributing the most correlation value. While Factor II frequency pattern items and the Factor IX arrest item obtained generally higher loadings, they explain less variance (13.8%), than Factor I items regarding adverse interpersonal and job interference consequences (27.6% variance). The five Factor II frequency pattern items offer more precision than ill-defined terms like heavy or regular marijuana user, and may compensate for the lack of an empirically standardized at risk marijuana quantity and frequency measure (1).

The DSM-G-MI internal consistency reliabilities (.84 and .94) and concurrent validity with the MSI-X supports its use as an alternative gold standard method for deriving DSM IV-TR cannabis diagnostic classifications. The DSM-G-MI marijuana-specific criteria and four possible diagnostic classifications also compensate for controversies regarding the generic applicability of DSM IV-TR criteria to cannabis (1,3,8,42,45). The receiver operator characteristic (ROC) curve procedures, comparing the MSI-X with the DSM-G-MI cannabis diagnostic classifications, supports the MSI-X overall accuracy and empirically derived scoring cutoffs for moderate to high risk marijuana use. The ROC summary index (AUC) probabilities for MSI-X accuracy suggest it can correctly identify 91% of those with both cannabis abuse and dependence, 90% of those with cannabis use disorder or cannabis abuse only, and 86% of those with cannabis dependence only. Another measure of MSI-X psychometric accuracy supported by ROC analysis, stems from the sensitivity and specificity calculations for three DSM IV-TR diagnostic methods yielding the same MSI-X optimal cutoffs of 6. The best (maximum) sensitivity and specificity is the 6 cutoff for both cannabis abuse and dependence, suggesting this cutoff will at best detect 83% of true positive cases and 89% of true negative cases for both cannabis dependence and abuse. Thus, for clinical screening purposes any one scoring [greater than or equal to] 6 is considered high risk and merits more intensive follow-up evaluation for the presence of both cannabis dependence and abuse symptoms. The abuse only diagnostic classification method yielded a lower MSI-X cutoff threshold, suggesting MSI-X scores of 3 may accurately detect 82% (sensitivity) of those true positive cases and 81% (specificity) of those true negative cases with cannabis abuse only. For clinical screening purposes this provides a rationale for a less severe moderate risk range, suggesting individuals scoring between 3 and 5 also merit more careful follow-up assessment for cannabis abuse. The moderate risk range may allow clinicians to detect marijuana risks sooner and provide brief intervention earlier.

The ROC analysis cutoffs correspond closely with the previously reported theoretical-clinical scoring cutoffs, which set slightly higher (more conservative) cutoffs (moderate = 4 to 6; high = [greater than or equal to] 7), than the ROC derived cutoffs (4,5). Empirically determined ROC cutoffs are an improvement over theoretical-clinical cutoffs, and support the criterion validity of the MSI-X revised scoring system.

MSI-X CLinical Screening Usefulness

MSI-X psychometric properties and the empirically revised scoring system support its use for clinical pilot screening purposes. The 31 items are not excessive, nor time consuming to complete, or score in the format administered. Others point to the advantages of a direct, nonsubtle screening inventory, like the MSI-X, for facilitating accurate self-reports about sensitive drug use information (73). The MSI-X multiproblem items and scoring cutoffs provide a more extensive and accurate screening of client marijuana patterns, than does a single criterion measure of any use in the past six months (41). The MSI-X is also potentially more accurate than other general drug screens, since it has the advantage of being exclusively marijuana specific in question focus. General drug abuse screens do not focus exclusively on marijuana use, limiting their usefulness for screening marijuana at risk patterns. General drug abuse screens only allow clinicians to assume a problem exists with an unspecified substance. This necessitates additional and wider ranging exploration about multiple substances to determine substance specific DSM IV-TR diagnostic coding and intervention planning. The MSI-X provides a marijuana specific assessment focus for directing assessment questions and cannabis specific diagnostic decisions (5).

In clinical situations, MSI-X screening cutoff scores of [greater than or equal to] 6 and 3 to 5 are suggestive of high and moderate risk marijuana patterns respectively, and merit a more comprehensive follow-up assessment interview. High risk scores of [greater than or equal to] 6 provide the most discriminating cutoff for detecting those most in need of comprehensive follow-up assessment, those most likely at risk for both cannabis abuse and dependence, and those in need of possible intervention. Support for this likelihood comes from the pilot study confirming those with MSI-X high risk [greater than or equal to] 6 scores were diagnosed with co-occurring cannabis disorders after intensive assessment interviews (5). Screening scores in the moderate risk range are associated with risk of cannabis abuse only and may require the most careful follow-up assessment. While screening scores of 3 to 5 reflect less severity and fewer adverse consequences, assessing these individuals for self-report accuracy and readiness to change may pose more challenges, and be deserving of more effort and intervention at this earlier point of risk. Recommended clinical MSI-X administration guidelines could not be followed for this research (5). Normally the MSI-X is also used as a follow-up interview guide to explore and clarify subjects' initial reports for the time flame (lifetime to past two weeks) of their marijuana use. Even when a high or moderate risk marijuana pattern is based only upon lifetime past use, this still has clinical value for assessing relapse potential or prevention strategies.

Marijuana Risk Within This Clinical Sample

Although the MSI-X is designed for individual use, the clinical sample aggregate scores provide screening clues about marijuana risks and assessment priorities with this sample. Within the 90% of the total sample admitting past or present lifetime marijuana use, Table 3 suggests the 43% reporting moderate and high risk use deserve follow-up assessment. Clinically, the 28.5% scoring moderate and high-risk within the 48% reporting past year marijuana use are most at risk and most in need of follow-up assessment for their marijuana use patterns, as they are more likely to meet criteria for marijuana abuse or both dependence and abuse. The 12.2% past month moderate and high-risk users merit this same assessment priority. Moderate and high-risk scores account for 59% of all past year and 69% of all past month users. Past year and past month low risk users need assessment to ensure their self-reported marijuana patterns are accurate. Clinical sample aggregate responses to specific items (Table 1), especially those associated with Factors I, II and III offer additional clues regarding the patterns of use or consequences experienced by this sample.

Finding 15.1% of 18 to 25 year olds with moderate to high risk MSI-X scores reflects national trends indicating higher prevalence among this age group (11). But the 13.2% of 41 to 54 year olds obtaining moderate and high risk MSI-X scores highlights the need for continued marijuana screening among the older baby boomer generation (74). Finding more males (15.9%) obtaining high risk MSI-X scores than females (7.5%) also reflects national marijuana gender trends (11). Although females and males obtained similar moderate risk scores (9.3%), a question remains. Since at risk alcohol quantity and frequency levels are set lower for women due to body differences, would a similar gender threshold difference apply for marijuana use and require a lower MSI-X cutoff score for women? Whether a lower MSI-X cutoff score needs to be set for women remains to be determined.

One would expect a clinical sample referred for substance abuse evaluation to have a larger percentage of high risk marijuana users than a community sample. That is the case when comparing this clinical sample with a previously described community sample (4). Percentage wise, the lifetime high risk clinical sample group (23.4%) is twice as large as the community sample high risk group (9.56%) using similar [greater than or equal to] 6 cutoffs. More statistically sophisticated comparisons between this clinical sample and the community sample are planned for the future.

Limitations and Future Research Directions

Limitations of this MSI-X clinical sample study include: convenience rather than random sampling methods; limited sample size and demographics; limitations inherent with self-report measures (73,75); lower sample return rate than desired; exploratory rather than confirmatory factor analyses; some item loading weakness and a need for statistical item analysis; dependence upon the DSM-G-MI as a gold standard measure without a concurrent structured diagnostic assessment interview or instrument; and no concurrent validity assessment between the MSI-X and other standardized instruments. Some limitations merit more discussion.

Client honesty is frequently questioned on substance use self-report measures, with drug sensitive behaviors like marijuana often underestimated (75). Client self-reports from this sample are open to question since many were court mandated to undergo evaluation and 14.3% report marijuana arrests. Still, MSI-X responses and risk levels are more likely underestimates than overestimates. Administering the MSI-X as an add on rather than with the routine agency assessment battery may have lowered sample return rate. However, separate MSI-X administration on a confidential, research only basis may have lessened concerns about marijuana disclosures. Normally the MSI-X is used as a follow-up interview guide, which also serves as an informal check for item comprehension, importance, discomfort, social desirability tendency, accuracy, or bias. That was not possible in this study. For research administration more formal qualitative and quantitative assessment of these dimensions may be necessary. For example, asking MSI-X respondents to estimate how comfortable and honest they think others would be in answering MSI-X items may quantitatively assess any underreporting bias.

Given client access limitations it was reasonable to use the DSM-G-MI as an alternative gold standard measure, but this may not be an adequate replacement for a lengthy, direct diagnostic assessment interview or the use of a standardized structured interview assessment instrument (63,76). However, diagnostic interviews and structured interview instruments are also self-report dependent, with limitations stemming from inadequate DSM IV diagnostic training for some professionals and reliability and validity issues with structured DSM IV interview tools (77,78). The agency uses a structured interview method, the Addiction Severity Index (ASI), which will be analyzed with the MSI-X in a future study (63). Concurrent and discriminant validity between the MSI-X and standardized drug and alcohol screening instruments routinely used by the agency will also be analyzed. Use of the DSM-G-MI as an independent screening tool will also be explored in the future as it may offer a more direct method for arriving at a cannabis use diagnosis.

Future research is needed to determine the generalizability of the MSIX beyond the present clinical sample and the community sample previously described (4). Replication studies are needed with a new randomly drawn community sample and additional clinical samples from substance abuse, general mental health, and primary medical care clinics. Future research focused on demographic comparisons will ensure the MSI-X is free of gender or race bias, and age appropriate for adolescents.

MSI-X comparisons between this clinical sample and the previous community sample are underway to assess the discriminant validity of the MSI-X and determine if a similar, different, or common factor structure emerges. Factor scales already derived by exploratory factor analyses will serve as hypothesized subscales for future confirmatory factor analyses, which may provide subscale reliabilities and allow subscales to potentially be used as independent variables.

Achieving a more robust MSI-X is a future goal. This may be accomplished by item revisions, deletions, or additions. Adding replacement items based upon new marijuana research is under consideration. For example, evidence that marijuana use at age 14 or younger significantly increases risk factors for future substance abuse, suggests including an age of first use item (23). A shorter MSI-X may be desirable for primary health care settings. A 24-item MSI-X could be evaluated using only those items achieving primary loadings above [greater than or equal to] .54 in this study. Or a 15 item MSI-X could be considered by combining the items from Factors I, II, and III accounting for 44% of the total 72.2% variance explained. However, before proceeding, more statistically sophisticated item analysis needs to be conducted to determine which MSI-X items are most effective at differentiating high risk from low risk subjects.

CONCLUSION

MSI-X psychometrics from this sample built upon those from a previous community sample to support the usefulness of the MSI-X as a screening tool. The MSI-X multiquestion content is internally consistent with a direct and specific focus on one drug, marijuana. The MSI-X empirically derived scoring system, supported by ROC analysis, appears to have sufficient accuracy, sensitivity, and specificity for clinical pilot screening purposes. The MSI-X appears useful for identifying clients with moderate and high risk patterns of marijuana use who require additional assessment. Administering the MSI-X may improve clinician and client awareness of marijuana use risk patterns. The MSI-X is recommended for continued pilot use with other clinical samples While the MSI-X needs psychometric strengthening, it is a useful starting tool for clinicians needing an alternative or adjunct to urine testing. The MSI-X can help clinicians gauge whether clients' marijuana use patterns are potentially problematic or not, and serves as an alert indicator when more assessment is warranted. Research with the MSI-X will continue toward the goal of achieving a robust, valid, marijuana-specific screening tool for use in multiple settings. This MSI-X study with a clinical sample from a substance abuse specialty clinic is a step in that direction.

ACKNOWLEDGMENTS

Support for this research was provided by a University of Houston Faculty Grant and a SAMHSA/AMERSA Project Mainstream Fellowship Grant to the first author. The authors thank Les Shireman, Vicki Longwill, and Mel Taylor of the Houston COUNCIL on Alcohol and Drugs for host agency support with this study, and Lon Sherritt and John Knight of Harvard Medical School for enlightenment regarding ROC analysis.

REFERENCES

(1.) Alexander D. A review of marijuana assessment dilemmas: time for marijuana specific screening methods. J. Social Work Practice in Addict 2003; 3(4):5-28.

(2.) Haack M, Adger H. Executive summary: strategic plan for interdisciplinary faculty development: arming the nation's health professional workforce for a new approach to substance use disorders. Substance Abuse 2002; 23(3:supplement b): 1-21.

(3.) Teesson M, Lynskey M, Manor B, Baillie A. The structure of cannabis dependence in the community. Drug Alcohol Depend 2002; 68:255-262.

(4.) Alexander D. A marijuana screening inventory (experimental version): description and preliminary psychometric properties. Amer J Drug Alcohol Abuse 2003; 29(3):619-645.

(5.) Alexander D. Clinical pilot experiences using the marijuana screening inventory (MSI-X): screening guidelines and case illustrations. J. Social Work Practice in Addict 2003; 3(4):29-51.

(6.) Substance Abuse and Mental Health Services Administration. Results from the 2001 National Household Survey on Drug Abuse: Volume I. Summary of National Findings. Rockville, MD 2002. Office of Applied Studies, NHSDA Series H-17, DHHS Publication No. SMA 02-3758.

(7.) Johnston L, O'Malley P, Bachman J. Monitoring the Future National Survey Results on Drug Use, 1975-2000: College Students and Adults Ages 19 to 40. Bethesda, MD: National Institute on Drug Abuse, 2001. NIH publication No. 01-4925.

(8.) Marijuana and Medicine: Assessing the Science Base. Joy J, Watson S, Benson J, eds. Washington: Institute of Medicine, National Academy Press, 1999.

(9.) American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. Fourth edition. Text Revision, (DSM IV-TR). Washington, DC: American Psychiatric Association, 2000.

(10.) Substance Abuse and Mental Health Services Administration. The DASIS Report. Marijuana Admissions Increase: 1993-1999. SAMHSA Drug and Alcohol Services Information System. Office of Applied Studies, 2002. www.drugabusestatistics.samhsa.gov (accessed Feb 2003).

(11.) Substance Abuse and Mental Health Services Administration. Results from the 2001 National Household Survey on Drug Abuse: Volume II. Technical Appendices and Selected Data Tables. NH SDA Series H-17. Rockville, MD: Office of Applied Studies, 2002. DHHS, Publication No. SMA 02-3759.

(12.) Earlywine M. Understanding Marijuana: A New Look at the Scientific Evidence. New York: Oxford University Press, 2002.

(13.) Gerdes L. Marijuana. San Diego: Greenhaven Press, Inc, 2002.

(14.) National Institute of Drug Abuse. Research Report Series: Marijuana Abuse. Rockville, MD: National Institute of Health, 2002. NIH Publication No. 02-3859.

(15.) Arseneault L, Cannon M, Poulton R, Murray R, Caspi A, Moffitt T. Cannabis use in adolescence and risk for adult psychosis: longitudinal prospective study. BMJ 2002; 325:1212-1213.

(16.) Bovaso G. Cannabis abuse as a risk factor for depressive symptoms. Am J Psychiatry 2001; 158:2033-2037.

(17.) Boyer C, Shafer M, Teitle E, Wibbelsman C, Seeberg D, Schachter J. Sexually transmitted diseases in a health maintenance organization teen clinic: association of race, partner's age and marijuana use. Arch Pediatr Adolesc Med 1999; 153(8):838-844.

(18.) Brook D, Brook J, Zhang C, Cohen P, Whiteman M. Drug use and the risk of major depressive disorder, alcohol dependence, and substance use disorders. Arch Gen Psychiatry 2002; 59:943-955.

(19.) Budney A, Novy P, Hughes J. Marijuana withdrawal among adults seeking treatment for marijuana dependence. Addiction 1999; 94:1311-1322.

(20.) Fergusson D, Horwood J, Swain-Campbell N. Cannabis use and psychological adjustment in adolescence and young adulthood. Addiction 2002: 97:1123-1135.

(21.) Fried P. The consequences of marijuana use during pregnancy: a review of the human literature. J Cannabis Therapeutics 2002; 2:85-104.

(22.) Galif E, Newcomb M, Carmona J. Prospective relationship between drug problems and work adjustment in a community sample. J Appl Psychol 2001; 86(2):337-350.

(23.) Gfroerer JC, Wu L-T., Penne MA. Initiation of Marijuana Use: Trends, Patterns, and Implications. Analytic Series: A-17. Rockville, MD: Substance Abuse and Mental Health Services Administration. Office of Applied Studies, 2002. DHHS Publication No. SMA 02-3711.

(24.) Gorman D, Derzon J. Behavioral traits and marijuana use and abuse: a meta-analysis of longitudinal studies. Addict Behav 2002; 27:193-206.

(25.) Grady M. Cognitive deficits associated with heavy marijuana use appear to be reversible. Nida Notes 2002; 17. Available at www.drugabuse.gov/NIDA_notes/NNindex.

(26.) Green B, Ritter C. Marijuana use and depression. J Health Social Behav 2000; 41:40-49.

(27.) Kouri E, Pope H, Lukas S. Changes in aggressive behavior during withdrawal from long-term marijuana use. Psychopharmacology 1999; 143:302-308.

(28.) Lynskey M, Hall W. The effects of adolescent cannabis use on educational attainment: a review. Addiction 2000; 95:162 l-1630.

(29.) Patton G, Coffey C, Carlin J, Degenhardt L, Lynskey M, Hall W. Cannabis use and mental health in young people: cohort study. BMJ 2002; 325:1195-1198.

(30.) Taylor DR, Fergusson D, Milne B, Horwood LJ, Moffitt T, Sears M, Poulton R. A longitudinal study of the effects of tobacco and cannabis exposure on lung function in young adults. Addiction 2002; 97:1055-1061.

(31.) Schwenk C. Marijuana and job performance: comparing the major streams of research. J Drug Issues 1998; 28:941-970.

(32.) Solowij N, Stephens R, Roffman R, Barbor T, Kadden R, Miller M, Christiansen K, McRee B, Vendetti J. Cognitive functioning of long term heavy cannabis users seeking treatment. JAMA 2002; 287:1123-1131.

(33.) Staton M, Leukefeld C, Logan T, Zimmerman R, Lynam D, Milich R. Risky sex behavior and substance use among young adults. Health Soc Work 1999; 24:147-154.

(34.) Ramaekers J, Robbe H, O'Hanlon J. Marijuana, alcohol and actual driving performance. Hum Psychopharmacol Clin Exp 2000; 15:551-558.

(35.) Zammit S, Allebeck P, Andreasson S, Lundberg I, Lewis G. Self-reported cannabis use as a risk factor for schizophrenia in Swedish conscripts of 1969: historical cohort study. BMJ 2002; 25:1199-2011.

(36.) Zickler P. Study demonstrates that marijuana smokers experience significant withdrawal. Nida Notes 2002; 17, Available at: http://www.nida.gov/NIDA-Notes/NNVol17.

(37.) SAMHSA. Report to Congress on Prevention and Treatment of Co-Occurring Disorders 2002. Rockville, MD: US Dept of HHS, November 2002.

(38.) Presidents New Freedom Commission on Mental Health Report. July 22, 2003. Available at: http://www.mentalhealthcommission.gov (accessed July 24, 2003).

(39.) Dennis M, Babor T, Roebuck M, Donaldson J. Changing the focus: the case for recognizing and treating cannabis use disorders. Addiction 2002; 97(s.1):4-15.

(40.) Agosti V, Nunes E, Levin F. Rates of psychiatric comorbidity among US residents with lifetime cannabis dependence. Am J Drug Alcohol Abuse 2002; 28(4):643-652.

(41.) Fleming M. Chapter 2: screening, assessment, and intervention for substance use disorders in general health care settings. In: Haack M, Hoover A, eds. Vol. 23. Strategic Plan for Interdisciplinary Faculty Development: Arming the Nation's Health Professional Workforce for a New Approach to Substance Use Disorders. Vol. 23. Substance Abuse, 2002; (3: supplement a):319-340.

(42.) Swift W, Copland J, Hall W. Choosing a diagnostic cut-off for cannabis dependence. Addiction 1998; 93:1681-1692.

(43.) National Institute on Alcohol Abuse and Alcoholism. The Physicians Guide to Helping Patients with Alcohol Problems. Rockville, MD, 1995. NIH publication No. 95-3756.

(44.) Plant M. Editorial: patterns and consequences matter. Addict Res 1999; 7:463-468.

(45.) Ridenour T, Cottler L, Compton W, Spitznagel R, Williams R. Is there a progression from abuse disorders to dependence disorders? Addiction 2003; 98:635-644.

(46.) Richter L, Johnson P. Current methods of assessing substance use: a review of strengths, problems, and developments. J Drug Issues 2001; 31:809-832.

(47.) Riley J, Lu N, Taylor B. Drug screening: a comparison of urinalysis results from two independent laboratories. J Drug Issues 2000; 30:171-186.

(48.) James W, Moore D. Examining the relationship between gender and drug-using behaviors in adolescents: the use of diagnostic assessments and biochemical analyses of urine samples. J Drug Educ 1999; 29:235-249.

(49.) Mura P, Klintz P, Paper Y, Ruesch G, Piriou A. Evaluation of six rapid tests for screening of cannabis in sweat, saliva, and tears. Acta Clin Belg 1999; 1:35-38.

(50.) Cone E, Johnson R. Contact highs and urinary cannabinoid excretion after passive exposure to marijuana smoke. Clin Pharmacol Ther 1986; 40:247-256.

(51.) Fendrich M, Kim J. The experience and acceptability of drug testing: poll trends. J Drug Issues 2002; 32:81-96.

(52.) Buchan B, Dennis M, Tims F, Diamond G. Cannabis use: consistency and validity of self report, on-site urine testing and laboratory testing. Addiction 2002; 97(s1):98-108.

(53.) Bastiaens L, Riccardi K, Sakhrani D. The RAFFT as a screening tool for adult substance use disorders. Amer J Drug Alcohol Abuse 2002; 28(4):681-691.

(54.) Brown R, Rounds L. Conjoint screening questionnaires for alcohol and drug abuse. Wis Med J 1995; 94:135-140.

(55.) Brown R, Leonard T, Saunders L, Papasouliotis O. Two item screening test for alcohol and other drug problems. J Fam Prac 1997; 44:151-160.

(56.) Maisto S, Carey M, Carey K, Gordon C, Gleason J. Use of the AUDIT and DAST10 to identify alcohol and drug use disorders among adults with severe and persistent mental illness. Psych Assess 2000; 12:186-192.

(57.) Knight J, Shrier L, Bravender T, Farrel M, Vander-Bilt J, Shaffer H. A new brief screen for adolescent substance abuse. Arch Pediatr Med 1999; 153:591-596.

(58.) Group, WHO ASSIST working. The alcohol, smoking and substance involvement screening test (ASSIST): development, reliability and feasibility. Addiction 2002; 97:1183-1194.

(59.) Winters K. A new multiscale measure of adult substance abuse. J Subst Abuse Treat 1999; 16:237-246.

(60.) Santisteban D, Tejeda M, Dominicis C, Szapocznik J. An efficient tool for screening for maladaptive family functioning in adolescent drug abusers: the problem oriented screening instrument for teenagers. Amer J Drug Alcohol Abuse 1999; 25:197-206.

(61.) Skinner H. The drug abuse screening test (DAST). Addict Behav 1982; 7:363-371.

(62.) Miller G, Lazowski L. The SASSI Manual, Substance Abuse Measures, Substance Abuse Subtle Screening Inventory (SASSI-3) Manual. 2nd ed. Springville, IN: SASSI Institute, 1999.

(63.) McLellan A, Kushner H, Metzger D, Peters R, Smith I, Grissom G, Pettinati H, Argeriou M. Fifth edition of the addiction severity index. J Subst Abuse Treat 1992; 9:199-213.

(64.) Kazmierczak D, Smyth N, Wodarski J. Screening and assessment instruments for alcohol and other drugs. Fam Ther 1999; 26(2):103-119.

(65.) Cooney N, Zweben A, Fleming M. Chapter 3: screening for alcohol problems and at-risk drinking in health care settings. In: Hester R, Miller W, eds. Handbook of Alcoholism Treatment Approaches. 2nd ed. Boston: Allyn & Bacon, 1995:45-60.

(66.) Aarons G, Brown S, Stice E, Coe M. Psychometric evaluation of the marijuana and stimulant effect expectancy questionnaires for adolescents. Addict Behav 2001; 26:219-236.

(67.) Budney A, Novy P, Hughes J. Marijuana withdrawal among adults seeking treatment for marijuana dependence. Addiction 1999; 94: 1311-1322.

(68.) Heishman S, Singleton E, Liguori A. Marijuana craving questionnaire: development and initial validation of a self-report instrument. Addiction 2001;96:1023-1034.

(69.) Stephens R, Roffman R, Curtin L. Comparison of extended versus brief treatments for marijuana use. J Clin Psychol 2000; 68:898-908.

(70.) Edreich L, Lee E. Use of relative operating characteristic analysis in epidemiology: a method for dealing with subjective judgment. Am J Epidemiol 1981; 114:649-662.

(71.) Kim J, Mueller C. Introduction to Factor Analysis: What it is and How to do it. Beverly Hills, CA: Sage Publications, 1981.

(72.) Kim J, Mueller C. Factor Analysis: Statistical Methods and Practical Issues. Beverly Hills, CA: Sage Publications, 1981.

(73.) Svanum S, McGrew J. Prospective screening of substance dependence: the advantages of directness. Addict Behav 1995; 20:205-213.

(74.) Gfroerer J, Penne M, Pemberton M, Folsom R. Substance abuse treatment need among older adults in 2020: impact of the aging baby-boom cohort. Drug Alcohol Depend 2003; 69:127-135.

(75.) Richter L, Johnson P. Current methods of assessing substance use: a review of strengths, problems, and developments. J Drug Issues 2001; 31:809-832.

(76.) Compton W, Cottler L, Dorsey K, Spitznagel E, Mager D. Comparing assessments of DSM-IV substance dependence disorders using the CIDI-SAM and SCAN. Drug Alcohol Depend 1996; 41:179-187.

(77.) Dziegielewski S, Johnson A, Webb E. DSM-IV and social work professionals: a continuing education evaluation. Social Work in Mental Health 2002; 1:27-41.

(78.) Meyer G. Chapter 3: implications of information-gathering methods for a refined taxonomy of psychopathology. In: Beutler L, Malik M, eds. Rethinking the DSM: A Psychological Perspective. Washington: American Psychological Press, 2002:69-106.

Dale E. Alexander, Ph.D., * and Patrick Leung, Ph.D.

University of Houston Graduate School of Social Work, Houston, Texas, USA

* Correspondence: Dale E. Alexander, Ph.D., Associate Professor, University of Houston, Graduate School of Social Work, Houston, TX 77204-4013, USA; E-mail: dalexander@uh.edu.

COPYRIGHT 2004 Taylor & Francis Ltd.
COPYRIGHT 2004 Gale Group