Drug screening and changing marijuana policy: Validation of new single question drug screening tools

Accepted Manuscript Title: Drug screening and changing marijuana policy: Validation of new single question drug screening tools Authors: J. Paul Seale, J. Aaron Johnson, Nicholas Cline, Christopher Buchanan, Chris Kiker, Lindsey Cochran PII: DOI: Reference:

S0376-8716(18)30688-4 https://doi.org/10.1016/j.drugalcdep.2018.08.030 DAD 7150

To appear in:

Drug and Alcohol Dependence

Received date: Revised date: Accepted date:

27-4-2018 14-8-2018 17-8-2018

Please cite this article as: Seale JP, Johnson JA, Cline N, Buchanan C, Kiker C, Cochran L, Drug screening and changing marijuana policy: Validation of new single question drug screening tools, Drug and Alcohol Dependence (2018), https://doi.org/10.1016/j.drugalcdep.2018.08.030 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.

Drug screening and changing marijuana policy: Validation of new single question drug screening tools

SC RI PT

J. Paul Sealea,b, [email protected], J. Aaron Johnsonc, [email protected], Nicholas Clinea, [email protected], Christopher Buchanana,

[email protected], Chris Kikerd, [email protected], Lindsey Cochrane, [email protected]

Mercer University School of Medicine, 1501 Mercer University Dr, Macon, GA 31207, USA

b

Navicent Health, 777 Hemlock St, Macon, GA 31201, USA

c

Augusta University Institute of Public and Preventive Health, 1120 15th St, Augusta, GA 30912,

A

N

U

a

M

USA

Seneca Lakes Family Medicine Residency, 11082 N Radio Station Rd, Seneca, SC 29678, USA

e

University of Texas Southwestern, 5323 Harry Hines Blvd., Dallas, TX 75390, USA

J. Aaron Johnson

EP

*Correspondence:

TE

D

d

Institute of Public and Preventive Health, Augusta University, 1120 15th Street, CJ2300,

A

CC

Augusta, GA 30912, E-mail: [email protected]

Highlights

Sensitivity and specificity were similar across four single question drug screens



All four single question screens showed acceptable sensitivity and high specificity



Saliva testing increased drug use detection by 2 percent, lowering sensitivity



These alternative questions may be useful as laws around marijuana use change

SC RI PT



U

Abstract

N

Background: Illicit drug use is common in U.S. medical settings. A validated Single Drug

A

Screening Question (SDSQ) containing the word “illegal” is widely used. As marijuana policies

M

change, the present wording may not perform as expected in states legalizing use. This study

D

compares the performance of the validated SDSQ with three different SDSQ wordings.

TE

Methods: Patients, 18 years and older, presenting to a U.S. hospital emergency department were asked to complete a survey containing the existing SDSQ, three new SDSQ versions without the

EP

word “illegal,” and the Mini International Neuropsychiatric Interview. Patients were also encouraged to provide a sample for saliva drug testing.

CC

Results: Of 297 eligible patients, 200 (67.3%) completed the survey and 141 (70.5% of interview participants) completed saliva testing. Overlapping confidence intervals on the Area

A

Under the Curve (AUC) analysis confirmed that sensitivities of all SDSQs were statistically similar for detecting self-reported drug use (65-71%). Combining questionnaire and saliva testing increased drug use detection by 2% and resulted in lower SDSQ sensitivity (58-67%). Specificities were 99-100%. The SDSQ with the highest sensitivity was: “In the last twelve

months, did you smoke pot (marijuana), use another street drug, or use a prescription medication “recreationally” (just for the feeling, or using more than prescribed)?” Conclusions: All four SDSQs demonstrated acceptable sensitivity and high specificity.

SC RI PT

Removing the term ‘illegal’ does not result in higher, or lower, reported drug use, though results could vary in states with legalized marijuana. Future research should replicate this study in one or more of these states.

A

CC

EP

TE

D

M

A

N

U

Keywords: Drug Screening, Brief Screening, Drug Use

1. Introduction

Illicit drug use, which includes the use of drugs that are illegal to process, sell and

SC RI PT

consume as well as the use of prescription medications other than as prescribed, is common in the U.S. and has demonstrated progressive increases for more than a decade. In 2016, 10.6

percent of Americans aged 12 or older reported having used an illicit drug in the past 30 days

(Center for Behavioral Health Statistics and Quality, 2017) This percentage has risen every year since 2002, due in large part to the burgeoning use of marijuana (Center for Behavioral Health

U

Statistics and Quality, 2017) A National Drug Intelligence Center report estimated the economic

N

cost of illicit drug use to the U.S. in 2007 was $193 billion (National Drug Intelligence Center,

A

2011), while in 2016, the human toll included over 64,000 deaths from drug overdose (National

M

Institute on Drug Abuse, 2017). Attitudes and policies about drug use have changed significantly in recent years, especially regarding marijuana. Medical marijuana is now legal in 29 states and

D

Washington D.C., while both recreational and medical marijuana use is legal in 9 states and

TE

Washington D.C. (ProCon.org, 2018). The decriminalization of marijuana in some states has led

EP

to problems with its description as an illicit drug. Terminology is now in flux, as some clinicians are instead referring to the use of mood-altering substances or unhealthy drug use.

CC

A concern for healthcare providers in clinical settings is the lack of screening by healthcare providers to detect those who use or misuse mood-altering drugs. Use of mood-

A

altering drugs is common among patients in clinical settings, with greater than monthly drug use reported by 7% of past-year emergency department patients and 3% of past-year primary care patients (Cherpitel and Ye, 2008). More often than not, such use goes undetected. Although the percent of emergency department patients screened for drug misuse is unknown, a consensus

statement from a national conference on SBIRT in emergency departments held in 2009 concluded that routine screening for drug use in emergency departments is almost non-existent, due in part to the lack of validated brief screen for drug use (Cunningham et al., 2009). In 2015,

SC RI PT

only about 10.8% of those aged 12 or older who needed treatment for an illicit drug or alcohol use problem received it (Lipari et al., 2016). Increased screening could help identify patients’ illicit drug use and help them access treatment.

Most current drug screening instruments, such as the Drug Abuse Screening Test (DAST) and the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST), are comprised

U

of multiple questions and are time-consuming to administer (WHO ASSIST Working Group,

N

2002). In recent years, brief Single Drug Screening Questions (SDSQs) have been developed to

A

allow more efficient and timely screening of patients in clinical settings (McNeely et al., 2015a).

M

Thus far, the only validated SDSQ for drugs asks: “How many times in the past year have you used an illegal drug or used a prescription medication for non-medical reasons?” (Smith et al.,

D

2010). Several studies have demonstrated that SDSQs can identify drug use in clinical settings

TE

with sensitivities ranging from 71% to 93% and specificities ranging from 79% to 94%

EP

(McNeely et al., 2015a, 2015b; Saitz et al., 2014; Smith et al., 2010). These same studies also note that the performance of drug screening questionnaires composed of a single item or question

CC

is similar to that of multi-item questionnaires such as the DAST and ASSIST. Despite these promising results, there are still a number of concerns with the current

A

validated SDSQ. One concern is the potential confusion of patients regarding the definition of "non-medical reasons." The developers of the SDSQ suggest that if patients ask for clarification of the meaning of “non-medical reasons,” the person administering the screen should respond “for instance, because of the experience or feeling it causes” (Smith et al., 2010). It is unclear,

however, how many patients may be confused by this terminology yet not ask for clarification, or how many clinicians may administer the screen without clarification. Another concern is the use of the word illegal, which may deter patients from answering honestly about their drug

SC RI PT

history due to embarrassment or fear that law enforcement or other authorities may access this information. The term illegal may also impact responses to the question in states where

marijuana, the most commonly used mood-altering drug, is now legal. Thus, the currently validated SDSQ as worded may be expected to become less and less useful in detecting

marijuana use. The objective of this study was to test the sensitivity and specificity of three

U

alternative SDSQs that avoid the use of the terms "illegal drug" and "non-medical reasons," and

A

N

to compare their sensitivity and specificity with the current validated SDSQ.

M

2. Methods 2.1 Sample

D

Data for this study were collected from patients presenting to the emergency department

TE

of a Level 1 trauma center in the southeastern U.S. as part of a summer scholars research project

EP

for medical students. Data were collected over two summers, July to September 2013 and 2014. The study was conducted in a state is not legal, either medically or recreationally. All patients 18

CC

years and older, fluent in English, cognitively intact, and not pregnant were eligible for participation. During the day and early evening, research assistants systematically approached

A

adult patients receiving care in exam rooms in the medical-surgical area of the emergency department. Of 297 emergency department patients approached about participation, 8 (2.7%) did not meet eligibility criteria, 68 (22.9%) refused to participate, and 21 (7.1%) began but were unable to complete the interview process. A total of 200 (69.2% of eligible) patients consented to

and completed the interview, with 141 (70.5% of interview participants) also consenting to and providing a saliva sample for testing for drug metabolites (Figure 1). The study was approved by

SC RI PT

the Institutional Review Boards of the Medical Center Navicent Health and Mercer University.

2.2 Measures

After obtaining appropriate written informed consent from the participants, the research assistants administered a survey containing the existing validated SDSQ plus three additional SDSQs which had been crafted to address the perceived limitations of the existing question

U

(Table 1). Variations between the questions included using a yes-or-no format (“Did you use…”)

N

vs. a frequency-of-use format (“How many times…), mentioning specifically the use of "pot

A

(marijuana)" vs. not explicitly referencing it, using the term "street drug" rather than "illegal

M

drug," and asking about the use of prescription drugs "recreationally (just for the feeling, or more than prescribed)" vs. "for nonmedical reasons (for example, because of the experience or feeling

D

it caused)" vs. "for nonmedical reasons" without any further definition. The survey also included

TE

the drug use section of the Mini International Neuropsychiatric Interview (MINI), which was

EP

used as the criterion measure for assessment of drug use, abuse, and dependence according to the DSM-IV (Lecrubier et al., 1997; Sheehan et al., 1998, 1997). The MINI includes an extensive

CC

list of drugs from eight different categories. Patients were asked to indicate if, in the past 12months they had used any of the drugs to get high, to feel better or to change their mood. They

A

were then read the list of drugs. For the purpose of this analysis, the criterion measure was considered positive if the patient responded affirmatively to any of the drugs on the list. The most commonly reported drug on the MINI was marijuana.

All interviews were conducted in a private, confidential exam room. To avoid possible ordering effect in patient responses, the four different SDSQs were presented in rotating order using 4-different versions of the assessment questionnaire. SDSQs were also separated in the

SC RI PT

survey so that they were not all asked consecutively. Patients were informed before the survey that some questions may seem redundant but were asked to answer each question honestly based on their understanding of the question. Interviewers were told not to react or probe if answers

across SDSQs were inconsistent. All patients completing the survey received a $20 gift card for their participation.

U

Patients were then offered the opportunity to provide a saliva sample to test for residuals

N

of medications and drugs, and those agreeing to participate signed a second consent form.

A

Patients were given an additional $5 gift card for participating in the saliva screening portion of

M

the study. To collect a saliva sample with the OraSure® Oral Specimen Collection Device, the research assistant placed the device's pad between the cheek and gum for 2-5 minutes, then

D

inserted the device into the provided buffer solution and snapped the wand off. The cap was then

TE

replaced and the sample was carefully marked with the patient's study number and stored in a

EP

double-lock area for up to 7 days; collected samples were shipped on a weekly basis to Clinical Reference Laboratory in Lenexa, Kansas for testing using screening enzyme immunoassay (EIA)

CC

to test for the presence of amphetamines, methamphetamines, opioids, cocaine, PCP, ecstasy, THC, benzodiazepines, and barbiturates. Oral fluid testing has been shown to have an accuracy

A

similar to urine drug screening tests (Drummer, 2006; McNeely et al., 2015a). Patients providing a saliva sample were also asked to report any current prescription medications. Saliva test results matching the reported prescriptions were not considered a positive test. 2.3 Data Analysis

Data analyses were performed using SPSS Version 24 statistical software. In addition to sensitivity and specificity in detecting any drug use, the predictive performance of each SDSQ was evaluated with respect to the area under the curve (AUC) obtained from the receiver

SC RI PT

operating characteristics (ROC) curve. Because men and women may respond differently to question wording, results were also separated by gender. Initial analyses compared the 4 SDSQs to the self-report criterion measure (the MINI). Results from the MINI were then combined with saliva test results (when available), and identical analyses were conducted comparing the

performance of the 4 SDSQs against this combined criterion measure (patients positive on

U

questionnaires screening and/or saliva testing).

N

3. Results

A

3.1 Demographics

M

Participants ranged in age from 18-90, had both a mean and median age of 56, and females (56%) and African Americans (53%) represented a slight majority. Most had a high

D

school education or less (62.0%), while few had completed a bachelor’s degree or higher

TE

(10.5%). 25.5% of participants were currently employed, 40.5% were unemployed due to

EP

disability, and 25.0% were retired. About 1 in 5 males (19.5%) and 1 in 8 females (12.5%) reported past 12 month illicit drug use based on their responses on the MINI (Table 2).

CC

3.2 Sensitivity and Specificity of Single Questions compared to MINI Diagnoses When compared with responses obtained using the MINI diagnostic interview, the

A

sensitivity of the 4-questions ranged from 65% to 71%, while specificity ranged from 99% to 100% (Table 3). Q1, a yes-or-no question that asked about use of “street drugs” and "nonmedical" prescription drug use, had the lowest sensitivity and specificity (65% and 99%, respectively), while Q3, a yes-or-no question referring to street drugs and use of prescription

drugs "recreationally," had the highest sensitivity and high specificity (71% and 99%, respectively). Three additional people (1.5%) screened positive on Q3 than on Q2, the previously validated SDSQ (Smith et al., 2010) which referred to use of an "illegal drug" and “non-medical

SC RI PT

use” of prescription drugs. Q2 had a sensitivity of 68% and a specificity of 100%, while Q4, which asked frequency-of-use, use of “street drugs,” and use of prescription medications

“recreationally,” had a sensitivity of 68% and a specificity of 99%. Area Under the Curve

(AUC), which measure the ability of the test to correctly classify those with and without drug use, ranged from 0.82 to 0.85. Though Q3 again demonstrated the highest value (0.85),

U

overlapping confidence intervals suggest that these differences are not statistically significant. A

N

crosstabulation of the previously validated SDSQ (Q2) with the other three wording revealed an

A

80% to 90% overlap in the patients identified as positive.

M

3.3 Saliva testing results

Of the 141 saliva samples, 13 (9.2%) were positive for drug metabolites, including 5 for

D

benzodiazepines, 4 for barbiturates, 3 for cocaine, and 1 for opiates. When compared with verbal

TE

information provided during the MINI questionnaire interviews, 3 of the 13 positive results were

EP

patients who had denied drug use, capturing an additional 2.1% whose drug use would have remained undetected by question-based screening alone. Of the 23 subjects who disclosed drug

CC

use during the MINI and also submitted a saliva sample, only 4 (17.4%) tested positive, leaving

A

82.6% of self-reported positives undetected by saliva testing.

3.4 Sensitivity and Specificity of Single Questions Compared to MINI Diagnoses or Positive Saliva Test

When saliva results were used to complement self-report data in identifying drug use, the sensitivity of all 4-single questions decreased, with a range of 58% to 64% (see Table 4).

SC RI PT

Discrimination as measured by ROC curves also decreased, ranging from 0.79 to 0.82.

4. Discussion

All 4 SDSQ variations demonstrated acceptable sensitivity and specificity for use in the emergency department, with extremely high levels of specificity (99-100%). Question 3 (Q3),

the highest performing question, was a yes-or-no question that referred to use of “street drugs”

U

rather than illegal drugs and to "recreationally" using prescription drugs. This study’s finding

N

that including the word “illegal” in the SDSQ led to a decrease in sensitivity of only 3% when

A

compared to the other questions suggests that inclusion of the term “illegal” does not have a

M

major impact on patient responses. Likewise, specifically asking patients about marijuana use did not result in a marked increase in drug use detection in this study, though this could in part be

D

due to the study having been conducted in a state where marijuana use is illegal.

TE

The addition of saliva testing to the questionnaire slightly increased the detection of drug

EP

use, while also resulting in a slight decrease in the sensitivity of all four SDSQs. At the same time, the saliva results from most self-reported drug users were actually negative, reflecting the

CC

limitations of oral fluid testing and its ability to detect only very recent drug use (up to 48 hours for most illicit drugs) (Verstraete, 2004). These findings are consistent with previous evidence

A

that drug screening which relies on self-report measures typically demonstrates good accuracy in research studies where patients are guaranteed anonymity (McNeely, 2015a). Though the similarities across the different variations of the SDSQ are encouraging and have practical implications for healthcare settings planning to integrate single-item screening

questions into their patient intake/triage protocol, the study had several limitations that should be considered. First, with four slight changes to terminology (“marijuana or other drugs” vs. “drugs”; frequency of use vs. any use; “street drug” vs. “illegal drug”; and “recreational” vs.

SC RI PT

“non-medical”), there are sixteen possible permutations of the SDSQ. In an effort to avoid participant response set due to too many similarly worded questions, we selected four

permutations including the original SDSQ wording. While similar results across the four suggest that other wordings would have likely yielded similar results, this cannot be determined with the current data. Second, data were collected from a single emergency department, in a mid-size city,

U

in the southeastern U.S. As a result, the population served was likely older and had a lower rate

N

of substance use than emergency departments in large urban areas. This combined with the

A

relatively small sample size (n=200) resulted in a relatively small number of patients reporting

M

substance use (n=31). Third, the refusal rate for saliva samples was relatively high (29%). Those refusing the saliva test may have refused because they were concerned they would test positive

D

which would have resulted in changes to the sensitivity and specificity in the saliva test results.

TE

One strength of the study was the ability to test the performance of these four questions in

EP

a state that has been largely unaffected by changes in marijuana policy. The reality, however, is that the questions could perform very differently in states where marijuana is now legalized. For

CC

example, patients in these states might omit marijuana use from the currently validated SDSQ since it is no longer “illegal.” Future research should test these questions in one or more states

A

whose marijuana laws differ from those in Georgia to determine the potential impact of marijuana policy on their validity.

5. Conclusions

Having several single drug screening questions of comparable validity available to screen for use of marijuana, illicit “street drugs,” and misuse of prescription drugs should be increasingly helpful as legal use of marijuana for medical or recreational reasons continues to

SC RI PT

increase across the U.S. This will allow practitioners to select a screening question with acceptable sensitivity and specificity regardless of the status of marijuana in their state.

Contributors JP Seale – Study design, manuscript preparation

N

Nicholas Cline – Data collection, manuscript preparation

U

JA Johnson – Study design, data analysis, manuscript preparation

M

Chris Kiker – Data collection, data analysis

A

Christopher Buchanan – Data collection, manuscript preparation

Lindsey Cochran – Data collection, data analysis

EP

Author Disclosures

TE

D

All authors have read and approved the final manuscript.

CC

Role of Funding Source SAMHSA grant # provided partial salary support for all authors, covered costs of incentive

A

payments for participants, and lab costs associated with saliva tests.

Conflict of Interest No conflict declared.

Acknowledgements

The authors gratefully acknowledge SAMHSA grant #TI1019545. We are also grateful for the support of Medical Center Navicent Health for facilitating access to the emergency department

A

CC

EP

TE

D

M

A

N

U

SC RI PT

during the data collection phase of the study.

References

Center for Behavioral Health Statistics and Quality, 2017. 2016 National Survey on Drug Use

SC RI PT

and Health: Detailed Tables. Available at https://www.samhsa.gov/data/sites/default/files/NSDUH-DetTabs-2016/NSDUH-DetTabs2016.pdf (accessed 4.26.18).

Cherpitel, C.J., Ye, Y., 2008. Drug use and problem drinking associated with primary care and emergency room utilization in the US general population: Data from the 2005 National

N

https://doi.org/10.1016/j.drugalcdep.2008.03.033

U

Alcohol Survey. Drug Alcohol Depend. 97, 226–230.

A

Cunningham, R.M., Bernstein, S.L., Walton, M., Broderick, K., Vaca, F.E., Woolard, R.,

M

Bernstein, E., Blow, F., D’Onofrio, G., 2009. Alcohol, tobacco, and other drugs: Future directions for screening and intervention in the emergency department. Acad. Emerg. Med.

D

16, 1078–1088. https://doi.org/10.1111/j.1553-2712.2009.00552.x

TE

Drummer, O.H., 2006. Drug testing in oral fluid. Clin. Biochem. Rev. 27, 147–59.

EP

Lecrubier, Y., Sheehan, D. V., Weiller, E., Amorim, P., Bonora, I., Sheehan, K.H., Janavs, J., Dunbar, G.C., 1997. The Mini International Neuropsychiatric Interview (MINI). A short

CC

diagnostic structured interview: Reliability and validity according to the CIDI. Eur. Psychiatry 12, 224–231. https://doi.org/10.1016/S0924-9338(97)83296-8

A

Lipari, R.N., Park-Lee, E., Van Horn, S., 2016. America’s need for and receipt of substance use treatment in 2015. CBHSQ Rep. 1, 10–15.

McNeely, J., Cleland, C.M., Strauss, S.M., Palamar, J.J., Rotrosen, J., Saitz, R., 2015a. Validation of self-administered single-item screening questions (SISQs) for unhealthy

alcohol and drug use in primary care patients. J. Gen. Intern. Med. 30, 1757–1764. https://doi.org/10.1007/s11606-015-3391-6 McNeely, J., Strauss, S.M., Saitz, R., Cleland, C.M., Palamar, J.J., Rotrosen, J., Gourevitch,

SC RI PT

M.N., 2015b. A brief patient self-administered substance use screening tool for primary care: Two-site validation study of the substance use brief screen (SUBS). Am. J. Med. 128, 784.e9-784.e19. https://doi.org/10.1016/j.amjmed.2015.02.007

National Drug Intelligence Center, 2011. The economic impact of illicit drug use on american society. Washington, DC. Available at http://www.justice.gov/archive/ndic/.

U

National Institute on Drug Abuse, 2017. Overdose death rates. Available at

N

https://www.drugabuse.gov/related-topics/trends-statistics/overdose-death-rates. (accessed

A

4.26.18)

M

ProCon.org, 2018. 29 legal medical marijuana states and DC: Laws, fees, and possession limits.

(accessed 4.26.18).

D

Available at https://medicalmarijuana.procon.org/view.resource.php?resourceID=000881

TE

Saitz, R., Cheng, D.M., Allensworth-Davies, D., Winter, M.R., Smith, P.C., 2014. The ability of

EP

single screening questions for unhealthy alcohol and other drug use to identify substance dependence in primary care. J. Stud. Alcohol Drugs 75, 153–7.

CC

https://doi.org/10.15288/JSAD.2014.75.153 Sheehan, D. V., Lecrubier, Y., Sheehan, K.H., Amorim, P., Janavs, J., Weiller, E., Hergueta, T.,

A

Baker, R., Dunbar, G.C., 1998. The Mini-International Neuropsychiatric Interview (M.I.N.I.): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J. Clin. Psychiatry 59, (Suppl 20), 22–33. https://doi.org/10.1016/S0924-9338(99)80239-9

Sheehan, D. V., Lecrubier, Y., Sheehan, K.H., Janavs, J., Weiller, E., Keskiner, A., Schinka, J., Knapp, E., Sheehan, M.F., Dunbar, G.C., 1997. The validity of the Mini International Neuropsychiatric Interview (MINI) according to the SCID-P and its reliability. Eur.

SC RI PT

Psychiatry 12, 232–241. https://doi.org/10.1016/S0924-9338(97)83297-X Smith, P.C., Schmidt, S.M., Allensworth-Davies, D., Saitz, R., 2010. A single-question screening test for drug use in primary care. Arch. Intern. Med. 170, 1155–1160. https://doi.org/10.1001/archinternmed.2010.140

Verstraete, A.G., 2004. Detection times of drugs of abuse in blood, urine, and oral fluid. Ther.

U

Drug Monit. 26, 200–205. https://doi.org/10.1097/00007691-200404000-00020

N

WHO ASSIST Working Group, 2002. The Alcohol, Smoking and Substance Involvement

A

Screening Test (ASSIST): Development, reliability and feasibility. Addiction 97, 1183–

A

CC

EP

TE

D

M

1194. https://doi.org/10.1046/j.1360-0443.2002.00185.x

Figure Legend

A

CC

EP

TE

D

M

A

N

U

SC RI PT

Figure 1. Consort Diagram of the patient Recruitment Process

Figure 1. Consort Diagram of the Patient Recruitment Process 297 Approached

8 Ineligible for study 3 Unable to communicate 2 Pregnant 2 Lacked English fluency 1 under 18

SC RI PT

   

289 Eligible

N

26 Too ill 12 Sleeping or too tired 10 Timing interfered with medical care 8 Not interested 5 Questions too personal 7 Other

D

M

A

     

U

68 Declined participation

CC

EP

TE

221 Enrolled

21 Did not complete interview

A

200 Completed interview

141 Completed both saliva testing and interview

59 Refused saliva testing

Table Table 1. Single Drug Screening Questions

A

CC

EP

TE

D

M

A

N

U

SC RI PT

*Q1. How many times in the past year have you used an illegal drug or used a prescription medication for nonmedical reasons (for example, because of the experience or feeling it caused)? Q2. In the last twelve months, did you smoke pot (marijuana), use another street drug, or use a prescription painkiller, stimulant, or sedative for a non-medical reason? Q3. In the last twelve months, did you smoke pot (marijuana), use another street drug, or use a prescription medication ‘recreationally’ (just for the feeling, or using more than prescribed)? Q4. In the last twelve months, on how many days did you smoke pot (marijuana), use another street drug, or use a prescription medication ‘recreationally’ (just for the feeling, or more than prescribed)? *Original wording of SDSQ as validated by Smith et al. (2010).

Table 2. Characteristics of patient study sample All Subjects (n=200) 55.7

Variable

Male (n=88) 57.9 (16.3)

p-value .11 .01*

71 (63.4) 38 (33.9) 3 (2.7)

A

CC

EP

TE

D

M

A

N

U

SC RI PT

Age in years, mean (SD) Race/ethnicity, n (%) African American 106 35 (39.8) White 83 45 (51.1) Other 11 8 (9.1) Drug Measures, n (%) Illicit Drug Use from MINI 31 (15.5%) 17 (19.5%) DSM IV Abuse or Dependence 14 (7.0%) 8 (9.1%) SD, standard deviation; any drug use as determined by the MINI

Female (n=112) 54.0 (16.8)

14 (12.5%) 6 (5.4%)

.23 .30

A

CC

EP

TE

D

M

A

N

U

SC RI PT

Table 3. Performance of SDSQs in detecting illicit drug use as measured by the MINI Q1: In the last twelve months, did you smoke pot (marijuana), use another street drug, or use a prescription painkiller, stimulant, or sedative for a non-medical reason? Se Sp PPV NPV AUC (95% CI) Test % % % % 65 99 91 94 0.82 (0.71-0.92) All subjects 71 100 100 93 0.85 (0.72-0.99) Males 57 98 80 94 0.78 (0.61-0.94) Females Q2: How many times in the past year have you used an illegal drug or used a prescription medication for nonmedical reasons (for example, because of the experience or feeling it caused)? 68 100 100 94 0.83 (0.74-0.94) All subjects 77 100 100 95 0.88 (0.76-1.00) Males 57 100 100 94 0.79 (0.62-0.95) Females Q3: In the last twelve months, did you smoke pot (marijuana), use another street drug, or use a prescription medication ‘recreationally’ (just for the feeling, or using more than prescribed)? 71 99 92 95 0.85 (0.75-0.95) All subjects 77 100 100 95 0.88 (0.76-1.00) Males 64 98 82 95 0.81 (0.66-0.97) Females Q4: In the last twelve months, on how many days did you smoke pot (marijuana), use another street drug, or use a prescription medication ‘recreationally’ (just for the feeling, or more than prescribed)? 68 99 91 94 0.83 (0.73-0.93) All subjects 77 100 100 95 0.88 (0.76-1.00) Males 57 98 80 94 0.77 (0.61-0.94) Females Se, sensitivity; Sp, specificity; PPV, positive predictive value; NPV, negative predictive value; AUC, area under the curve; CI, confidence interval; risky drug use determined by the MINI

Table 4. Performance of SDSQs in detecting illicit drug use as measured by the MINI and saliva test

M

A

N

U

SC RI PT

Q1: In the last twelve months, did you smoke pot (marijuana), use another street drug, or use a prescription painkiller, stimulant, or sedative for a non-medical reason? Se Sp PPV NPV AUC Test % % % % 95% (CI) 58 99 96 92 .79 (0.69-0.89) All subjects 63 100 100 91 0.82 (0.68-0.95) Males 52 99 90 92 0.76 (0.61-0.91) Females Q2: How many times in the past year have you used an illegal drug or used a prescription medication for nonmedical reasons (for example, because of the experience or feeling it caused)? 58 100 100 92 .79 (0.69-0.89) All subjects 68 100 100 92 0.84 (0.71-0.97) Males 47 100 100 91 0.74 (0.58-0.89) Females Q3: In the last twelve months, did you smoke pot (marijuana), use another street drug, or use a prescription medication ‘recreationally’ (just for the feeling, or using more than prescribed)? 67 100 100 93 0.83 (0.74-0.93) All subjects 68 100 100 92 0.84 (0.71-0.97) Males 65 100 100 94 0.82 (0.68-0.97) Females Q4: In the last twelve months, on how many days did you smoke pot (marijuana), use another street drug, or use a prescription medication ‘recreationally’ (just for the feeling, or more than prescribed)? 64 100 100 93 0.82 (0.72-0.92) All subjects 68 100 100 92 0.84 (0.71-0.97) Males 59 100 100 93 0.79 (0.65-0.94) Females

D

Se, sensitivity; Sp, specificity; PPV, positive predictive value; NPV, negative predictive value; AUC, area

A

CC

EP

TE

under the curve; CI, confidence interval; risky drug use determined by the MINI and saliva test results