Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Feb 1.
Published in final edited form as: Psychiatry Res. 2020 Dec 21;296:113667. doi: 10.1016/j.psychres.2020.113667

A Psychometric Investigation on the Diagnostic Utility of the Posttrauma Risky Behaviors Questionnaire.

Ateka A Contractor 1, Ling Jin 1, Nicole H Weiss 2, Seanne O’Hara 3
PMCID: PMC7855932  NIHMSID: NIHMS1658661  PMID: 33360968

Abstract

The Posttrauma Risky Behaviors Questionnaire (PRBQ) is a screening measure for posttrauma reckless and self-destructive behaviors (RSDBs). We examined (1) PRBQ’s predictive relations with clinical (vs. not) endorsements of distinct RSDBs, and (2) PRBQ’s optimal cutoff score yielding the most appropriate balance of sensitivity and specificity statistics. The sample included 354 adult trauma-exposed community participants (Mage=35.76 years; 57.90% female). Logistic regression analyses indicated that the PRBQ significantly differentiated individuals endorsing (vs. not) clinical levels of alcohol/drug misuse, disordered eating, problematic gambling, and compulsive buying. Receiver operating characteristic (ROC) curve analyses indicated that the 14-item PRBQ total score had moderate accuracy in differentiating individuals endorsing clinical vs. non-clinical levels of drug misuse, disordered eating, problematic gambling, compulsive buying, and engagement in RSDBs (PTSD’s E2 Criterion); and low accuracy for alcohol misuse. ROC curve analyses indicated 3.5–6.5 as the optimal range of PRBQ cutoff scores. Thus, the PRBQ has good ability to discriminate participants endorsing (vs. not) clinical levels of alcohol/drug misuse, disordered eating, problematic gambling, and compulsive buying (lowest accuracy in identifying participants with clinical levels of alcohol misuse), and a minimum cutoff score between 3.5 and 6.5 may suggest using additional diagnostic assessments and clinical interventions.

Keywords: Posttrauma Risky Behaviors Questionnaire, reckless and self-destructive behaviors, trauma, diagnostic accuracy, receiver operating characteristic curves

1. Introduction

Trauma and posttraumatic stress disorder (PTSD) are linked to several reckless and self-destructive behaviors (RSDBs) such as aggressive behaviors (Armour et al., 2020; Lusk et al., 2017), alcohol/drug misuse (Jacobson et al., 2001), gambling (Ledgerwood and Petry, 2006), disordered eating (Brewerton, 2007), problematic technology use (Contractor et al., 2017a), and self-injury/suicidal behaviors (Pat-Horenczyk et al., 2007). Indeed, theoretical viewpoints support this empirical evidence. For instance, the disinhibition viewpoint indicates that individuals with PTSD symptoms may find it difficult to inhibit impulsive RSDBs (Braquehais et al., 2010) when perceiving distress-reducing situations (Casada and Roache, 2005). The emotion dysregulation perspective indicates that individuals with PTSD symptoms may engage in RSDBs to improve affect (Ben-Zur and Zeidner, 2009; Cooper et al., 2000; Marshall-Berenz et al., 2011). The cognitive framework indicates that individuals with PTSD symptoms may experience limited attentional and informational processing resources, impacting effective decision-making and contributing to RSDBs (Ben-Zur and Zeidner, 2009). Notably, engagement in RSDBs is related to physical and mental health impairment (Contractor et al., 2017b; Drescher et al., 2003), trauma severity (Contractor and Weiss, 2019), and greater engagement in RSBDs with a detrimental cumulative effect (Cooper, 2002; White et al., 1993). Thus, adding the RSDB symptom (Criterion E2) to DSM-5’s PTSD diagnostic criteria was important (American Psychiatric Association, 2013).

Given that there was no brief and validated measure to examine post-trauma RSDBs, Contractor et al. (2020) developed the Posttrauma Risky Behaviors Questionnaire (PRBQ) for these purposes. The PRBQ as a 16-item self-report measure of the extent of engagement in post-trauma RSDBs (PTSD’s E2 Criterion). PRBQ’s development followed recommended guidelines (Furr, 2011; Germain, 2006; Hinkin et al., 1997). PRBQ’s initial psychometric investigations indicate that it measures a unitary RSDB construct and demonstrates sound psychometrics, including internal consistency, construct validity, convergent validity, and incremental validity (Contractor et al., 2020). Additionally, a separate study indicated PRBQ’s strongest relation with PTSD’s externalizing behaviors cluster comprised of items assessing aggression/irritability and RSDB criteria; such results support PRBQ’s construct validity (Contractor et al., 2019). Thus, the PRBQ has support for its utility and validity in the assessment of posttrauma RSDBs. See Appendix A for the PRBQ measure, published in the study by Contractor et al. (2020).

Extending psychometric research on the PRBQ (index test), we examined its clinical and diagnostic utility in comparison to other RSDB measures (reference tests). We examined (1) PRBQ’s predictive relations with clinically significant (vs. not) endorsements of distinct RSDBs, and (2) PRBQ’s optimal cutoff score range yielding the most appropriate balance of sensitivity and specificity statistics as diagnostic accuracy estimates (Bossuyt et al., 2003; Holmbeck and Devine, 2009; Streiner, 2003). We hypothesized that the PRBQ would significantly differentiate participants endorsing (vs. not endorsing) clinically significant levels of alcohol and drug misuse, disordered eating, problematic gambling, and compulsive buying. The hypothesis regarding PRBQ’s optimal cutoff scores is considered exploratory. Obtained results will shed light on PRBQ’s ability to discriminate participants endorsing vs. not endorsing clinical levels of specific RSDBs; and on PRBQ’s optimal cutoff score range to screen for potentially clinical levels of RSDBs (Bossuyt et al., 2003; Holmbeck and Devine, 2009; Streiner, 2003). Contingent on such psychometric investigations, the PRBQ may replace the “black box” practice of administering several measures to examine different RSDBs (Knottnerus and Muris, 2003).

2. Method

2.1. Procedure and Participants

We recruited adult participants from Amazon’s Mechanical Turk (MTurk) platform. We described the current study as a 45 to 60-minute survey to develop a measure of RSDBs endorsed by individuals reporting stressful experiences. Inclusion criteria were (1) aged ≥ 18 years, (2) residing in North America, (3) endorsing English fluency, and (4) endorsing traumatic experience(s) screened with the Primary Care PTSD Screen for DSM-5 (Prins et al., 2015). Eligible participants who provided informed consent and completed the Qualtrics survey validly received $1.25. The Institutional Review Board of University of North Texas approved the current study.

2.2. Exclusions, Missing Data, and Sample Characteristics

Of the obtained 891 responses, we excluded 47 responses of 18 participants who attempted the survey 2+ times, 150 participants who did not meet all inclusion criteria, 122 participants who did not pass all four attention and comprehension validity checks (Meade and Craig, 2012; Thomas and Clifford, 2017), 97 participants who missed data on all measures, 11 participants who did not endorse a trauma/most distressing trauma on the Life Event Checklist for DSM-5 (LEC-5; Weathers et al., 2013a), and 110 participants who missed >30% item-level data on the primary measures. The final sample included 354 participants who averaged 35.76 years (SD = 11.12); 206 participants identified as female (57.90%). Nearly one-third of participants reported symptoms consistent with probable PTSD (n = 122; 34.50%) based on a cutoff score of ≥ 33 (Bovin et al., 2016). In this sample, missing data was minimal (ranging from two items on Drug Abuse Screening Test-10 to one item on Problem Gambling Severity Index). See Table 1 for detailed demographic information.

Table 1.

Demographic and Traumatic Events Data (n = 354)

M (SD)
Age 35.76 (11.12)
Years of Education 15.31 (2.48)
Posttrauma Risky Behaviors Questionnaire Total Score 6.35 (8.91)
Alcohol Use and Disorders Identification Test Alcohol 3.05 (2.53)
Consumption Questions Total Score
Drug Abuse Screening Test-10 Total Score 1.25 (1.98)
Eating Attitudes Test-26 Total Score 9.99 (10.89)
Problematic Gambling Severity Index Total Score 2.20 (4.83)
Compulsive Buying Scale Total Score −2.20 (2.45)
PTSD Checklist for DSM-5 E2 Criterion Score (RSDB) .71 (1.14)
n (%)*
Gender
 Female 205 (57.90)
 Male 144 (40.70)
 Male to Female Transgender 1 (.30)
 Female to Male Transgender 2 (.60)
 Other 2 (.60)
Ethnicity
 Hispanic 46 (13)
 Non-Hispanic 302 (85.30)
 Unknown 6 (1.70)
Race
 White 252 (71.20)
 African American 30 (8.50)
 Asian 37 (10.50)
 American Indian or Alaska Native 23 (6.50)
 Native Hawaiian or Other Pacific Islander 12 (3.40)
Employment
 Full-time 254 (71.80)
 Part-time 55 (15.50)
 Unemployed 45 (12.70)
Income
 < $15,000 29 (8.20)
 $15,000 to $24,999 50 (14.10)
 $25,000 to $34,999 55 (15.50)
 $35,000 to $49,999 47 (13.30)
 $50,000 to $64,999 68 (19.20)
 $65,000 to $79,999 31 (8.80)
 ≥ $80,000 74 (20.90)
Relationship Status
 Single 55 (15.50)
 Dating 107 (30.20)
 Married 164 (46.30)
 Divorced/Separated/Widowed 28 (7.90)
Employment Status
 Part Time 55 (15.50)
 Full Time 254 (71.80)
 Retired 12 (3.40)
 Unemployed 27 (7.60)
 Unemployed Student 6 (1.70)
Index Traumatic Events Endorsed on the LEC-5
 Transportation Accident 65 (18.40)
 Natural Disaster 48 (13.60)
 Sexual Assault 48 (13.60)
 Physical Assault 30 (8.50)
 Life Threatening Illness or Injury 30 (8.50)
 Sudden Violent Death 23 (6.50)
 Sudden Accidental Death 23 (6.50)
 Any Other Very Stressful Event or Experience 20 (5.60)
 Fire or Explosion 18 (5.10)
 Assault with a Weapon 12 (3.40)
 Other Unwanted or Uncomfortable Sexual Experience 9 (2.50)
 Serious Accident at Work, Home, Recreational Activity 7 (2)
 Serious Injury, Harm, or Death You Caused Someone Else 7 (2)
 Severe Human Suffering 4 (1.10)
 Combat or Exposure to a War-zone 3 (.80)
 Prefer Not to Respond 3 (.80)
 Captivity 2 (.60)
 Exposure to a Toxic Substance 2 (.60)
Open in a new tab

Note. PTSD = posttraumatic stress disorder; LEC-5 = Life Events Checklist for DSM-5; RSDB = reckless and self-destructive behaviors;

*

percentages are reported accounting for missing data

2.3. Measures

2.3.1. Demographic information.

Information regarding age, gender, ethnicity, race, income, educational level, employment status, ethnicity, and relationship status was obtained.

2.3.2. Life Event Checklist for DSM-5 (LEC-5; Weathers et al., 2013a).

The LEC-5 is a 17-item self-report measure assessing exposure to lifetime traumatic events. Each item is followed by six response options: happened to me, witnessed it, learned about it, part of my job, not sure, or doesn’t apply. In the current study, participants who selected one of the first four response options were considered to have endorsed a traumatic event consistent with PTSD’s DSM-5 Criterion A (American Psychiatric Association, 2013).

2.3.3. Posttrauma Risky Behaviors Questionnaire (PRBQ; Contractor et al., 2020).

The PRBQ is a 16-item self-report measure developed to evaluate past month engagement in RSDBs (PTSD’s E2 Criterion). The first 14 PRBQ items assess engagement in specific RSDBs; response options range from 0 (never) to 4 (very frequently). The final two items measure functional impairment and the association between RSDB frequency and onset of the worst trauma (start or getting worse). Higher PRBQ summed scores (first 14 items) indicate greater engagement in RSDBs. The PRBQ has shown excellent internal consistency (α = .92 in the current study) and good construct, convergent, and incremental validity (Contractor et al., 2020).

2.3.4. Alcohol Use and Disorders Identification Test Alcohol Consumption Questions (AUDIT-C; Bush et al., 1998).

The AUDIT-C is a 3-item self-report measure assessing heavy drinking and alcohol use disorders with a 5-point Likert scale (Item 1: 0 [never] to 4 [daily or more times a week]; Item 2: 0 [1 or 2] to 4 [10 or more]; Item 3: 0 [Never] to 4 [daily/almost daily]). A score of ≥ 4 on the AUDIT-C indicates probable alcohol use disorder. The AUDIT-C has good psychometrics (Bush et al., 1998); Cronbach’s α in the current study was .74.

2.3.5. Drug Abuse Screening Test-10 (DAST-10; Skinner, 1982).

The DAST-10 is a 10-item self-report measure of drug misuse, including occupational/relational problems, illegal activities, or regret using 1 (yes) and 0 (no) response options. A DAST-10 score of ≥3 indicates probable drug use disorder (Yudko et al., 2007). The DAST has good psychometric properties (Yudko et al., 2007); Cronbach’s α in the current study was .84.

2.3.6. Eating Attitude Test-26 (EAT-26; Garner et al., 1982).

The EAT-26 is a 26-item self-report measure that assesses cognitions, emotions, and behaviors related to eating disorders. Participants rate the frequency of engaging in eating-related thoughts and behaviors on a 6-point Likert scale (0 = never, 5 = always). The EAT-26 yields three subscales: Dieting assesses food avoidance and engagement in dieting behaviors; Bulimia and Food Preoccupation assesses bulimic behavior and concerns related to body image; and Oral Control measures self-control related to food intake (Orbitello et al., 2006). An overall score ≥ 20 indicates a high level of concern about dieting, body weight, or problematic eating behaviors (Dotti and Lazzari, 1998). The EAT-26 has good psychometric properties (Doninger et al., 2005; Garner et al., 1982); Cronbach’s α for the total scale score was .90 in the current study.

2.3.7. Problem Gambling Severity Index (PGSI; Ferris and Wynne, 2001).

The PGSI is a 9-item self-report measure that assesses problem gambling behavior and consequences over the past 12 months with a 4-point Likert scale (0 = never, 3 = almost always). Scores of ≥ 8 indicate probable problem gamblers (Ferris and Wynne, 2001). The PGSI has good psychometric properties (Holtgraves, 2009); Cronbach’s α in the current study was .96.

2.3.8. Compulsive Buying Scale (CBS; Faber and O’guinn, 1992).

The CBS is an 8-item measure of behaviors, motivations, and feelings associated with compulsive buying. Items are rated on a 5-point Likert scale (0 = never/strongly disagree, 4 = very often/strongly agree). Participants with scores ≤ −1.34 are classified as compulsive buyers (Faber and O’guinn, 1992). The CBS has good psychometrics (Faber and O’guinn, 1992); Cronbach’s α in this study was .87.

2.3.9. PTSD Checklist for DSM-5 (PCL-5; Weathers et al., 2013b).

The PCL-5 is a 20-item self-report measure assessing PTSD severity referencing the past month. Response options range from 0 (not at all) to 4 (extremely). Participants completed the PCL-5 item in response to the most distressing trauma endorsed on the LEC-5 (Weathers et al., 2013a). The PCL-5 has excellent psychometric properties (Bovin et al., 2016; Wortmann et al., 2016). For the current study, we used the E2 Criterion assessing RSDBs; an item-level endorsement of ≥ 2 on E2 indicates clinical level of symptomology (Weathers et al., 2013b).

2.4. Statistical Analyses

Preliminary analyses indicated that the PRBQ total score was normally distributed (skewness < 3 and kurtosis < 8; Kline, 2011). As primary analyses, we first conducted logistic regressions using SPSS v.25 to examine PRBQ’s predictive relations with clinical vs. non-clinical levels of distinct RSDBs measured by other assessments (dichotomous variables for alcohol misuse, drug misuse, disordered eating, problematic gambling, and compulsive buying). Next, we conducted receiver operating characteristic (ROC) curve analyses to determine the extent of PRBQ’s accuracy in differentiating positive (e.g., participants endorsing clinical levels of assessed RSDBs) vs. negative (i.e., participants not endorsing clinical levels of assessed RSDBs) cases, and to determine PRBQ’s optimal cutoff score for diagnostic accuracy (Streiner, 2003; Streiner and Cairney, 2007). We used the AUDIT-C (alcohol misuse), DAST-10 (drug misuse), EAT-26 (disordered eating), PGSI (problematic gambling), CBS (compulsive buying), and PCL-5 E2 Criterion (engagement in RSDBs) as the reference tests. Specifically, we computed sensitivity (proportion of individuals with the attribute/disorder accurately identified by the test; true positives) and specificity (proportion of individuals without the attribute/disorder accurately identified by the test; true negatives) statistics to estimate the area under the curve (AUC) and to determine PRBQ’s optimal cutoff score. AUC indicates how well the PRBQ scores distinguished between positive vs. negative cases (Streiner, 2003; Streiner and Cairney, 2007); .50–.70 indicates low accuracy, .70–.90 indicates moderate accuracy, and >.90 indicates high accuracy (Fischer et al., 2003; Swets, 1988). Further, when selecting the optimal cutoff score, we emphasized (1) sensitivity (vs. specificity) statistics with the idea that accurately identifying true positive was more important given functional and health impairment associated with RSDBs; and (2) best balance between sensitivity and specificity values (Habibzadeh et al., 2016). Overall, the ROC curve is a plot of sensitivity against specificity for all possible PRBQ cutoff values (Fischer et al., 2003).

3. Results

3.1. Alcohol misuse.

The PRBQ score accounted for a significant amount of variance in alcohol misuse, Nagelkerke R2 = .085, χ2 (1) = 22.70, p < .001. Each 1 unit increase in the PRBQ total score related to a 6% greater likelihood of endorsing clinical levels of alcohol misuse (OR = 1.06; 95% Confidence Interval [CI: 1.03, 1.09]). ROC curves indicated that the PRBQ had low accuracy in differentiating participants who endorsed vs. did not endorse clinical levels of alcohol misuse (AUC = .69, SE = .03, p < .001, 95% CI [.63, .75]; Supplemental Figure 1). A cutoff score of 3.5 provided the best balance between sensitivity and specificity in classifying participants who did vs. did not endorse clinical levels of alcohol misuse.

3.2. Drug misuse.

The PRBQ total score accounted for a significant amount of variance in drug misuse, Nagelkerke R2 = .258, χ2 (1) = 53.53, p < .001. Each 1 unit increase in the PRBQ total score related to a 13% greater likelihood of endorsing clinical levels of drug misuse (OR = 1.13; 95% CI [1.09, 1.17]). ROC curves indicated that the PRBQ had moderate accuracy in differentiating participants who endorsed vs. did not endorse clinical levels of drug misuse (AUC = .82, SE = .03, p < .001, 95% CI [.76, .88]; Supplemental Figure 2). A cutoff score of 4.5 provided the best balance between sensitivity and specificity in classifying participants who did vs. did not endorse clinical levels of drug misuse.

3.3. Disordered eating.

The PRBQ total score accounted for a significant amount of variance in disordered eating, Nagelkerke R2 = .165, χ2 (1) = 42.89, p < .001. Each 1 unit increase in the PRBQ total score related to a 17% greater likelihood of endorsing clinical levels of disordered eating (OR = 1.17; 95% CI [1.09, 1.24]). ROC curves indicated that the PRBQ had moderate accuracy in differentiating participants who endorsed vs. did not endorse clinical levels of disordered eating (AUC = .78, SE = .04, p < .001, 95% CI [.71, .86]; Supplemental Figure 3). A cutoff score of 5.5 provided the best balance between sensitivity and specificity in classifying participants who did vs did not endorse clinical levels of disordered eating.

3.4. Problematic gambling.

The PRBQ total score accounted for a significant amount of variance in problematic gambling, Nagelkerke R2 = .441, χ2 (1) = 97.78, p < .001. Each 1 unit increase in the PRBQ total score related to a 17% greater likelihood of endorsing clinical levels problematic gambling (OR = 1.17; 95% CI [1.13, 1.22]). ROC curves indicated that the PRBQ had moderate accuracy in differentiating participants who endorsed vs. did not endorse clinical levels of problematic gambling (AUC = .83, SE = .04, p < .001, 95% CI [.75, .92]; Supplemental Figure 4). A cutoff score of 6.5 provided the best balance between sensitivity and specificity in classifying participants who did vs. did not endorse clinical levels of problematic gambling.

3.5. Compulsive buying.

The PRBQ total score accounted for a significant amount of variance in compulsive buying, Nagelkerke R2 = .266, χ2 (1) = 78.49, p < .001. Each 1 unit increase in the PRBQ total score related to a 19% greater likelihood of endorsing clinical levels of compulsive buying (OR = 1.19; 95% CI [1.13, 1.26]). ROC curves indicated that the PRBQ had moderate accuracy in differentiating participants who endorsed vs. did not endorse clinical levels of compulsive buying (AUC = .76, SE = .03, p < .001, 95% CI [.71, .81]; Supplemental Figure 5). A cutoff score of 3.5 provided the best balance between sensitivity and specificity in classifying participants who did vs. did not endorse clinical levels of compulsive buying.

3.6. Engagement in RSDBs (PTSD’s E2 Criterion).

ROC curves indicated that the PRBQ had moderate accuracy in differentiating participants who endorsed vs. did not endorse clinical levels of RSDBs (AUC = .83, SE = .03, p < .001, 95% CI [.77, .89]; Supplemental Figure 6). A cutoff score of 5.5 provided the best balance between sensitivity and specificity in classifying participants who did vs. did not endorse clinical levels of RSDBs.

4. Discussion

Using a sample of individuals reporting traumatic experiences, we examined diagnostic accuracy parameters for the recently developed Posttrauma Risky Behaviors Questionnaire (PRBQ; Contractor et al., 2020). We specifically examined PRBQ’s predictive relations with clinically significant (vs. not) endorsements of distinct RSDBs (alcohol misuse, drug misuse, disordered eating, problematic gambling, and compulsive buying), and PRBQ’s optimal cutoff score yielding the most appropriate balance of sensitivity and specificity statistics as diagnostic accuracy estimates. Results provide support for PRBQ’s discriminative ability; and a cutoff score range of 3.5–6.5 yielded an optimal balance of specificity and sensitivity statistics.

The PRBQ was found to have good ability to discriminate participants endorsing (vs. not) clinical levels of RSDBs such as alcohol misuse, drug misuse, disordered eating, problematic gambling, and compulsive buying. These findings underscore the potential clinical utility of the PRBQ in identifying patients engaging in clinical levels of RSDBs for remedial intervention. Notably, we used only a selectively available set of measures assessing distinct RSDBs and were additionally restricted by the availability of cutoff scores for these measures; additional research would benefit from investigating PRBQ’s discriminative abilities using other self-report as well as more interview-based measures of distinct RSDBs (e.g., Structured Clinical Interview for DSM-5 to assess substance use disorders; First et al., 2015).

There was more variability in the results on PRBQ’s optimal cutoff score, which was expected. In fact, we strategically focused on an optimal range of cutoff scores rather than a specific cutoff score that could indicate clinical levels of RSDB engagement. The rationale is that, although the PRBQ examines distinct RSDBs, it assesses a unidimensional construct of RSDBs represented by one score on this measure (Contractor et al., 2020). A focus on an optimal range of cutoff scores vs. one cutoff score was to account for the heterogeneity in RSDB types and in what defines their specific clinical severity levels; this also is the rationale for why item-level cutoff scores may not be as meaningful as a total cutoff score for clinical/research purposes. Relatedly, even a greater endorsement of a single RSDB on the PRBQ is clinically relevant; thus, we expected a meaningful and optimal cutoff score to be a lower value of the PRBQ total score.

Overall, results indicated that specificity vs. sensitivity statistics had a range with values closer/equal to 1 (ranging from .35 to 1), indicating that the PRBQ may be better at detecting negative cases. Additionally, the optimal cutoff score of 5.5 related to RSBD endorsement as assessed by the PCL-5 E2 Criterion had the best balance of sensitivity and specificity statistics; this was expected as the PRBQ was intended to assess post-trauma RSDBs corresponding to PTSD’s E2 Criterion. These results, in fact, provide additional support for the construct validity of the PRBQ. Further, we could consider a range of cutoff scores from 3.5 to 6.5 as having the highest quality of efficiency in predicting clinical levels of RSDB engagement. In other words, clinicians and researchers may want to probe more and pay attention to scores beyond 3.5–6.5 on the PRBQ; these participants may benefit from further diagnostic assessments and clinical interventions.

Relatedly, results indicated that while the 14-item PRBQ total score had moderate accuracy in differentiating individuals endorsing vs. not endorsing clinical levels of drug misuse, disordered eating, problematic gambling, compulsive buying, and RSDB engagement (PTSD’s E2 Criterion); it demonstrated low accuracy in differentiating clinical vs. non-clinical levels of alcohol misuse. One explanation for this finding is that our measure of alcohol misuse only assessed consumption. Low to moderate consumption of alcohol is quite common in the United States, with over half of all individuals in a national survey indicating alcohol use in the past month (Substance Abuse and Mental Health Services Administration (SAMHSA), 2015). Yet, only a relatively small number of individuals meet criteria for alcohol use disorder (Grant et al., 2015), and it is these individuals who are responsible for most alcohol-related costs (>60%; Moher et al., 2009). These findings suggest that the majority of individuals who consume alcohol encounter no problems associated with its use. Thus, low to moderate consumption of alcohol may not confer the same risk for problematic outcomes as other forms of RSDBs. Additional research is needed to explore PRBQ’s accuracy in differentiating clinical vs. non-clinical levels of alcohol misuse using measures that assess problematic and disordered alcohol use.

Study results need to be interpreted in the light of some limitations. First, we did not have a comprehensive clinician-administered gold-standard measure of RSDBs to obtain diagnostic and clinical cutoff scores. This being said, the PRBQ is conceptualized as a screening test; thus, an initial determination of diagnostic utility against self-report measures is valid and provides indicators/parameters for future research and assessments (Bossuyt et al., 2003; Streiner, 2003). Second, generalizability of study results is restricted by the characteristics of this study’s sample; validation in more diverse samples is needed to determine external validity and applicability of study findings (Bossuyt et al., 2003; Streiner, 2003). Lastly, we used the PRBQ in a research environment to place individuals into broad categories; assessing PRBQ’s sensitivity to track RSDBs in a treatment setting would add to its applications (Bossuyt et al., 2003; Streiner, 2003).

Despite these limitations, this is the first study to our knowledge to examine and support the ability of the 14-item PRBQ to discriminate between individuals endorsing vs. not endorsing clinical levels of specific and distinct RSDBs. Results provide researchers and clinicians a range of scores (3.5 onwards) which clinically may suggest (1) an extent of engagement in RSDBs with impairment, and/or (2) the need to obtain more information about the RSDBs. Notably, as a strength, the current study had minimal sources of bias according to the second version of the Quality Assessment of Diagnostic Accuracy Studies guidelines (QUADAS-2; Whiting et al., 2012) such as using appropriate exclusions, good psychometric properties of the reference tests in terms of classifying their specific conditions/attributes, data from the index and reference tests being collected at the same day, and all participants receiving the same reference test.

Supplementary Material

1
appendix

Table 3.

Diagnostic Accuracy Statistics for Different Posttraumatic Risky Behaviors Questionnaire (PRBQ) Cutoff Scores

AUDIT-C DAST-10 EAT-26 PGSI CBS RSDB
PRBQ Sen Spe Sen Spe Sen Spe Sen Spe Sen Spe Sen Spe
0.5 .85 .42 .96 .40 .90 .36 .88 .35 .82 .50 .88 .38
1.5 .80 .52 .94 .50 .85 .45 .81 .44 .75 .59 .88 .48
2.5 .73 .57 .91 .56 .85 .51 .81 .50 .71 .68 .88 .56
3.5 .67 .62 .85 .61 .81 .57 .81 .56 .67 .74 .83 .61
4.5 .57 .69 .74 .69 .79 .66 .81 .66 .59 .83 .82 .71
5.5 .52 .74 .70 .74 .75 .71 .81 .72 .53 .87 .81 .77
6.5 .43 .78 .65 .79 .73 .78 .79 .78 .46 .91 .75 .83
7.5 .39 .80 .63 .82 .69 .80 .79 .81 .44 .94 .73 .86
8.5 .35 .84 .61 .87 .64 .84 .77 .86 .38 .96 .66 .89
9.5 .34 .85 .57 .88 .62 .85 .77 .87 .36 .97 .65 .91
10.5 .31 .87 .57 .90 .56 .87 .77 .90 .33 .97 .61 .92
11.5 .30 .89 .50 .91 .52 .88 .75 .91 .31 .97 .57 .93
12.5 .27 .90 .44 .92 .48 .88 .71 .92 .28 .97 .53 .94
13.5 .25 .92 .41 .94 .40 .80 .69 .94 .24 .97 .49 .95
14.5 .22 .92 .39 .94 .39 .91 .65 .95 .22 .98 .44 .96
15.5 .21 .93 .37 .95 .38 .92 .60 .95 .20 .98 .42 .96
16.5 .20 .93 .33 .95 .36 .92 .58 .96 .19 .98 .39 .96
17.5 .19 .94 .31 .96 .32 .93 .58 .97 .18 .98 .36 .96
19.5 .17 .94 .30 .96 .29 .93 .56 .97 .17 .98 .36 .97
21.5 .17 .94 .30 .96 .29 .93 .56 .97 .16 .98 .35 .97
22.5 .16 .95 .26 .96 .27 .94 .54 .98 .15 .98 .34 .97
23.5 .14 .95 .22 .97 .25 .94 .52 .98 .14 .99 .31 .98
24.5 .13 .95 .22 .97 .23 .94 .52 .99 .14 .99 .30 .98
25.5 .10 .95 .22 .98 .23 .96 .44 .99 .12 .99 .27 .99
26.5 .09 .96 .19 .99 .21 .97 .33 .99 .09 .99 .23 .99
28.0 .09 .97 .15 .99 .21 .98 .31 .99 .08 .99 .22 1.00
29.5 .08 .97 .13 .99 .21 .98 .27 .99 .08 .99 .21 1.00
30.5 .06 .98 .11 .99 .17 .99 .19 .99 .06 .99 .16 1.00
31.5 .04 .98 .07 .99 .12 .99 .13 .99 .04 .99 .12 1.00
32.5 .03 .98 .06 .99 .12 .99 .13 .99 .04 1.00 .10 1.00
33.5 .02 .98 .06 .99 .10 .99 .10 .99 .04 1.00 .09 1.00
34.5 .02 .98 .06 1.00 .08 .99 .08 .99 .03 1.00 .08 1.00
35.5 .01 .98 .04 1.00 .08 1.00 .06 .99 .03 1.00 .06 1.00
38.5 .01 .99 .02 1.00 .04 1.00 .04 1.00 .02 1.00 .04 1.00
43.0 .00 1.00 .00 1.00 .02 1.00 .02 1.00 .01 1.00 .01 1.00
46.0 .00 1.00 .00 1.00 .00 1.00 .00 1.00 .00 1.00 .00 1.00
Open in a new tab

Note. Sen = Sensitivity; Spe = Specificity; AUDIT-C= Alcohol Use and Disorders Identification Test Alcohol Consumption Questions Total Score; DAST-10 = Drug Abuse Screening Test-10 Total Score; EAT-26 = Eating Attitude Test-26 Total Score; PGSI = Problem Gambling Severity Index Total Score; CBS = Compulsive Buying Scale Total Score; RSDB = Reckless and Self-Destructive Behaviors assessed by the PTSD Checklist for DSM-5 E2 Criterion Score.

Highlights.

  • We examined Posttrauma Risky Behaviors Questionnaire’s (PRBQ) predictive validity and diagnostic accuracy.

  • Sample included 354 adult trauma-exposed community participants.

  • PRBQ differentiated endorsement (vs. not) of clinical levels of reckless behaviors.

  • With moderate accuracy for other behaviors, PRBQ differentiated clinical vs. non-clinical alcohol misuse with low accuracy.

  • Optimal range of PRBQ cutoff scores was between 3.5 and 6.5.

Acknowledgements

We thank Ms. Fallon Keegan for her help in creating the tabulated results.

Funding

The research described here was supported, in part, by National Institutes of Health grants K23DA039327 and P20GM125507, awarded to the third author.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of Interest

None of the authors have a conflict of interest.

Contributor Information

Nicole H. Weiss, Department of Psychology, University of Rhode Island, Kingston, RI, USA

Seanne O’Hara, Department of Psychology, University of North Texas, Denton, TX, USA.

References

  1. American Psychiatric Association, 2013. Diagnostic and Statistical Manual of Mental Disorders, 5th ed. American Psychiatric Association, Washington, DC. [Google Scholar]
  2. Armour C, Greene T, Contractor AA, Weiss NH, Dixon-Gordon K, Ross J, 2020. Posttraumatic stress disorder symptoms and risky behaviors: A network analysis approach. Journal of Traumatic Stress 33, 29–40. [DOI] [PubMed] [Google Scholar]
  3. Ben-Zur H, Zeidner M, 2009. Threat to life and risk-taking behaviors: A review of empirical findings and explanatory models. Personality and Social Psychology Review 13, 109–128. [DOI] [PubMed] [Google Scholar]
  4. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, Moher D, Rennie D, De Vet HC, Lijmer JG, 2003. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Annals of Internal Medicine 138, W1–12. [DOI] [PubMed] [Google Scholar]
  5. Bovin MJ, Marx BP, Weathers FW, Gallagher MW, Rodriguez P, Schnurr PP, Keane TM, 2016. Psychometric properties of the PTSD Checklist for Diagnostic and Statistical Manual of Mental Disorders-Fifth Edition (PCL-5) in veterans. Psychological Assessment 28, 1379–1391. [DOI] [PubMed] [Google Scholar]
  6. Braquehais MD, Oquendo MA, Baca-García E, Sher L, 2010. Is impulsivity a link between childhood abuse and suicide? Comprehensive Psychiatry 51, 121–129. [DOI] [PubMed] [Google Scholar]
  7. Brewerton TD, 2007. Eating disorders, trauma, and comorbidity: Focus on PTSD. Eating Disorders 15, 285–304. [DOI] [PubMed] [Google Scholar]
  8. Bush K, Kivlahan DR, McDonell MB, Fihn SD, Bradley KA, for the Ambulatory Care Quality Improvement Project (ACQUIP), 1998. The AUDIT Alcohol Consumption Questions (AUDIT-C). An effective brief screening test for problem drinking. Archives of Internal Medicine 158, 1789–1795. [DOI] [PubMed] [Google Scholar]
  9. Casada JH, Roache JD, 2005. Behavioral inhibition and activation in Posttraumatic Stress Disorder. The Journal of Nervous and Mental Disease 193 (2), 102–109. [DOI] [PubMed] [Google Scholar]
  10. Contractor AA, Frankfurt S, Weiss NH, Elhai JD, 2017a. Latent-level relations between DSM-5 PTSD symptom clusters and problematic smartphone use. Computers in Human Behavior 72, 170–177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Contractor AA, Weiss NH, 2019. Typologies of PTSD clusters and reckless/self-destructive behaviors: A latent profile analysis. Psychiatry Research 272, 682–691. [DOI] [PubMed] [Google Scholar]
  12. Contractor AA, Weiss NH, Dolan M, Mota N, 2019. Examination of the structural relations between PTSD symptoms and reckless/self-destructive behaviors. International Journal of Stress Management. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Contractor AA, Weiss NH, Dranger P, Ruggero C, Armour C, 2017b. PTSD’s risky behavior criterion: Relation with DSM-5 PTSD symptom clusters and psychopathology. Psychiatry Research 252, 215–222. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Contractor AA, Weiss NH, Kearns NT, Caldas S, Dixon-Gordon K, 2020. Assessment of posttraumatic stress disorder’s E2 criterion: Development, pilot testing, and validation of the Posttrauma Risky Behaviors Questionnaire. International Journal of Stress Management 27, 292–303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Cooper ML, 2002. Alcohol use and risky sexual behavior among college students and youth: Evaluating the evidence. Journal of Studies on Alcohol Supplement 14, 101–117. [DOI] [PubMed] [Google Scholar]
  16. Cooper ML, Agocha VB, Sheldon MS, 2000. A motivational perspective on risky behaviors: The role of personality and affect regulatory processes. Journal of Personality 68, 1059–1088. [DOI] [PubMed] [Google Scholar]
  17. Doninger GL, Enders CK, Burnett KF, 2005. Validity evidence for Eating Attitudes Test scores in a sample of female college athletes. Measurement in Physical Education and Exercise Science 9, 35–49. [Google Scholar]
  18. Dotti A, Lazzari R, 1998. Validation and reliability of the Italian EAT-26.. Eating and Weight Disorders-Studies on Anorexia, Bulimia and Obesity 3, 188–194. [DOI] [PubMed] [Google Scholar]
  19. Drescher KD, Rosen CS, Burling TA, Foy DW, 2003. Causes of death among male veterans who received residential treatment for PTSD. Journal of Traumatic Stress 16 (6). [DOI] [PubMed] [Google Scholar]
  20. Faber RJ, O’guinn TC, 1992. A clinical screener for compulsive buying. Journal of Consumer Research 19, 459–469. [Google Scholar]
  21. Ferris J, Wynne HJ, 2001. The Canadian Problem Gambling Index final report., in: Canadian Centre on Substance Abuse; (Ed.), Ottawa. ON. [Google Scholar]
  22. First MB, Williams JBW, Karg RS, Spitzer RL, 2015. Structured Clinical Interview for DSM-5—Research Version (SCID-5 for DSM-5, Research Version; SCID-5-RV). American Psychiatric Association, Arlington, VA. [Google Scholar]
  23. Fischer JE, Bachmann LM, Jaeschke R, 2003. A readers’ guide to the interpretation of diagnostic test properties: clinical example of sepsis. Intensive Care Medicine 29, 1043–1051. [DOI] [PubMed] [Google Scholar]
  24. Furr M, 2011. Scale construction and psychometrics for social and personality psychology. Sage Publications Ltd, London, UK. [Google Scholar]
  25. Garner DM, Olmsted MP, Bohr Y, Garfinkel PE, 1982. The eating attitudes test: psychometric features and clinical correlates. Psychological Medicine 12, 871–878. [DOI] [PubMed] [Google Scholar]
  26. Germain ML, 2006. Stages of Psychometric Measure Development: The Example of the Generalized Expertise Measure (GEM). Online Submission.
  27. Grant BF, Goldstein RB, Saha TD, Chou SP, Jung J, Zhang H, Pickering RP, Ruan WJ, Smith SM, Huang B, Hasin DS, 2015. Epidemiology of DSM-5 alcohol use disorder: Results from the National Epidemiologic Survey on Alcohol and Related Conditions III. JAMA Psychiatry 72, 757–766. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Habibzadeh F, Habibzadeh P, Yadollahie M, 2016. On determining the most appropriate test cut-off value: The case of tests with continuous results.. Biochemia Medica 26, 297–307. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Hinkin TR, Tracey JB, Enz CA, 1997. Scale construction: Developing reliable and valid measurement instruments. ournal of Hospitality & Tourism Research 2, 100–120. [Google Scholar]
  30. Holmbeck GN, Devine KA, 2009. An author’s checklist for measure development and validation manuscripts. Journal of Pediatric Psychology 34, 691–696. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Holtgraves T, 2009. Evaluating the problem gambling severity index. Journal of Gambling Studies 25, 105–120. [DOI] [PubMed] [Google Scholar]
  32. Jacobson IG, Southwick SM, Kosten TR, 2001. Substance use disorders in patients with posttraumatic stress disorder: A review of the literature. American Journal of Psychiatry 158, 1184–1190. [DOI] [PubMed] [Google Scholar]
  33. Kline RB, 2011. Principles and practice of structural equation modeling, 3rd ed. The Guilford Press, New York, NY. [Google Scholar]
  34. Knottnerus JA, Muris JW, 2003. Assessment of the accuracy of diagnostic tests: The cross-sectional study. Journal of Clinical Epidemiology 56, 1118–1128. [DOI] [PubMed] [Google Scholar]
  35. Ledgerwood DM, Petry NM, 2006. Posttraumatic Stress Disorder symptoms in treatment-seeking pathological gamblers. Journal of Traumatic Stress 19 (3), 411–416. [DOI] [PubMed] [Google Scholar]
  36. Lusk JD, Sadeh N, Wolf EJ, Miller MW, 2017. Reckless self-destructive behavior and PTSD in veterans: The mediating role of new adverse events. Journal of Traumatic Stress 30, 270–278. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Marshall-Berenz EC, Vujanovic AA, MacPherson L, 2011. Impulsivity and alcohol use coping motives in a trauma-exposed sample: The mediating role of distress tolerance. Personality and Individual Differences 50, 588–592. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Meade AW, Craig SB, 2012. Identifying careless responses in survey data. Psychological Methods 17, 437–455. [DOI] [PubMed] [Google Scholar]
  39. Moher D, Liberati A, Tetzlaff J, Altman DG, Prisma Group, 2009. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS med 6 (7), e1000097. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Orbitello B, Ciano R, Corsaro M, Rocco PL, Taboga C, Tonutti L, Armellini M, Balestrieri M, 2006. The EAT-26 as screening instrument for clinical nutrition unit attenders. International Journal of Obesity 30, 977–981. [DOI] [PubMed] [Google Scholar]
  41. Pat-Horenczyk R, Peled O, Miron T, Brom D, Villa Y, Chemtob CM, 2007. Risk-taking behaviors among Israeli adolescents exposed to recurrent terrorism: Provoking danger under continuous threat? American Journal of Psychiatry 164, 66–72. [DOI] [PubMed] [Google Scholar]
  42. Prins A, Bovin MJ, Kimerling R, Kaloupek DG, Marx BP, Pless Kaiser A, Schnurr PP, 2015. The Primary Care PTSD Screen for DSM-5 (PC-PTSD-5). Instrument available from the National Center for PTSD at www.ptsd.va.gov. [DOI] [PMC free article] [PubMed]
  43. Skinner HA, 1982. The Drug Abuse Screening Test. Addictive Behaviors 7, 363–371. [DOI] [PubMed] [Google Scholar]
  44. Streiner DL, 2003. Diagnosing tests: Using and misunsing diagnostic and screening tests. Journal of Personality Assessment 8 (3), 209–219. [DOI] [PubMed] [Google Scholar]
  45. Streiner DL, Cairney J, 2007. What’s under the ROC? An introduction to Receiver Operating Characteristics Curves. The Canadian Journal of Psychiatry 52 (2), 121–128. [DOI] [PubMed] [Google Scholar]
  46. Substance Abuse and Mental Health Services Administration (SAMHSA), 2015. National Survey on Drug Use and Health (NSDUH). Table 2.41B—Alcohol Use in Lifetime, Past Year, and Past Month among Persons Aged 12 or Older, by Demographic Characteristics: Percentages, 2014 and 2015. [Google Scholar]
  47. Swets JA, 1988. Measuring the accuracy of diagnostic systems. Science 240, 1285–1293. [DOI] [PubMed] [Google Scholar]
  48. Thomas KA, Clifford S, 2017. Validity and mechanical turk: An assessment of exclusion methods and interactive experiments. Computers in Human Behavior 77, 184–197. [Google Scholar]
  49. Weathers FW, Blake DD, Schnurr PP, Kaloupek DG, Marx BP, Keane TM, 2013a. The Life Events Checklist for DSM-5 (LEC-5). Instrument available from the National Center for PTSD at www.ptsd.va.gov.
  50. Weathers FW, Litz BT, Keane TM, Palmieri PA, Marx BP, Schnurr PP, 2013b. The PTSD Checklist for DSM-5 (PCL-5). Scale available from the National Center for PTSD at www.ptsd.va.gov.
  51. White HR, Brick J, Hansell S, 1993. A longitudinal investigation of alcohol use and aggression in adolescence. Journal of Studies on Alcohol Supplement 11, 62–77. [DOI] [PubMed] [Google Scholar]
  52. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, Leeflang MM, Sterne JA, Bossuyt PM, 2012. QUADAS-2: A revised tool for the quality assessment of diagnostic accuracy studies.. Annals of Internal Medicine 155, 529–536. [DOI] [PubMed] [Google Scholar]
  53. Wortmann JH, Jordan AH, Weathers FW, Resick PA, Dondanville KA, Hall-Clark B, Foa EB, Young-McCaughan S, Yarvis J, Hembree EA, Mintz J, Peterson AL, Litz BT, 2016. Psychometric analysis of the PTSD Checklist-5 (PCL-5) among treatment-seeking military service members. Psychological Assessment 28, 1392–1403. [DOI] [PubMed] [Google Scholar]
  54. Yudko E, Lozhkina O, Fouts A, 2007. A comprehensive review of the psychometric properties of the Drug Abuse Screening Test. Journal of Substance Abuse Treatment 32, 189–198. [DOI] [PubMed] [Google Scholar]

Table 2.

Logistic Regressions using Posttraumatic Risky Behaviors Questionnaire (PRBQ) Total Score to Predict Reckless and Self-Destructive Behaviors

95% C.I.
B S.E. p OR Lower Upper
Alcohol Use and Disorders .06 .01 < .001 1.06 1.03 1.09
Identification Test Alcohol
Consumption Questions Total Score
Drug Abuse Screening-10 Test Total Score .12 .02 < .001 1.13 1.09 1.17
Eating Attitudes Test-26 Total Score .15 .03 < .001 1.17 1.09 1.24
Problematic Gambling Severity Index .16 .02 < .001 1.17 1.13 1.22
Total Score
Compulsive Buying Scale Total Score .18 .03 < .001 1.19 1.13 1.26
Open in a new tab

Note. OR= Odds Ratio; C.I. = Confidence Interval.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1658661-supplement-1.docx (10.3MB, docx)
appendix
NIHMS1658661-supplement-appendix.jpg (104.4KB, jpg)

RESOURCES