Assessing the Validity of the Ask Suicide-Screening Questions (ASQ) in Black Youth

Abstract

Purpose:

Given the increasing rates of suicide and non-fatal suicide attempts among Black youth in the United States, it is crucial that screening tools are valid in identifying Black youth at risk for suicide. This study assessed the validity of the Ask Suicide-Screening Questions (ASQ) among Black youth.

Methods:

This analysis used pooled data from three ASQ validation studies of pediatric medical patients ages 10-21 years. All participants completed the ASQ and the gold standard Suicidal Ideation Questionnaire.

Results:

Of the 1,083 participants, 330 (30.5%) were non-Hispanic Black and 753 (69.5%) were non-Hispanic White. ASQ psychometric properties for Black and White participants were equivalent (sensitivity = 94% vs. 90.9%; specificity = 91.4% vs. 91.8%, respectively).

Conclusions:

There were no significant differences in ASQ psychometric properties between Black and White youth, indicating that the ASQ is valid for screening Black youth at risk for suicide.

Introduction

Screening for suicide risk is a prevention strategy aimed at early detection of at-risk individuals who may benefit from prompt assessment and intervention. Screening tools, such as the 4-item Ask Suicide-Screening Questions (ASQ),¹ have been developed to rapidly identify youth presenting to healthcare settings with suicide risk. Equitable performance of screening tools across racial groups is vital to ensure prevention strategies reach everyone.² In the original ASQ development study,¹ the sensitivity of the ASQ was reported to be significantly lower in Black participants compared with non-Black participants. Given the increasing rates of suicide and non-fatal suicide attempts among Black youth in the United States,^3-5 it is imperative to ensure that screening tools are valid in identifying Black youth at risk for suicide. Therefore, in this study, we assessed the validity of the ASQ among Black youth.

Methods

Sample

This analysis uses data from three ASQ validation studies of pediatric medical patients ages 10-21 years conducted in emergency departments (ED),¹ inpatient medical/surgical units,⁶ and outpatient specialty/primary care clinics⁷ of four U.S. children’s hospitals. Institutional Review Board approval was obtained from the National Institutes of Health and all participating study sites.

Measures

Ask Suicide-Screening Questions (ASQ):

All participants completed the ASQ, a brief 4-item tool used to identify suicide risk. The ASQ was originally developed in the pediatric ED, with a sensitivity of 96.9% and specificity of 87.6%,¹ and has since been validated for use in both pediatric and adult medical inpatient medical units and outpatient primary care and specialty settings with equally strong psychometric properties.^6-8 While the ASQ is intended to identify clinically significant suicidal ideation/past behavior, it has also been shown to demonstrate predictive validity for future suicide attempts.⁹

Suicidal Ideation Questionnaire (SIQ):

The SIQ/SIQ-JR was used as the criterion standard comparison for this analysis. The SIQ was developed using a sample of adolescents aged 14-19 and was found to have strong reliability and validity.¹⁰ The SIQ-JR was developed using a sample that was 71.4% Black youth aged 12-14 and demonstrated strong validity.¹¹ The SIQ validation sample contained a sufficient number of White and Black youth to establish that no significant differences were found between White and Black participants; however, there was insufficient data on other racial and ethnic groups.¹² Recent research has found the SIQ/SIQ-JR to be valid among various racial and ethnic groups,^13,14 however, this remains a limited area of research. Therefore, this analysis was confined to a comparison of non-Hispanic Black (hereafter, Black) and non-Hispanic White (hereafter, White) youth, self-reported by participants.

Data Analysis

The percentages of positive risk by ASQ and by SIQ/SIQ-JR were compared by racial group (Black and White) and by study setting. Validity statistics included sensitivity, specificity, likelihood ratio positive (LR+), likelihood ratio negative (LR−), and area under receiver operating characteristic curves (AUC) were calculated for each racial group. Pairwise comparisons across racial groups using Chi-Square tests at 2-sided significance level (p-value) of 0.05 were completed, with no adjustment for multiple comparisons.

Results

Of the 1,083 participants included in this analysis, 330 (30.5%) were Black and 753 (69.5%) were White (Table 1). The proportion of positive screens was not significantly different across the three ASQ validation studies (Table 1); therefore, the samples were pooled across the three sites. When compared to the criterion standard SIQ/SIQ-JR in the combined samples, the ASQ had a sensitivity of 92.0% and a specificity of 91.7%. Among Black participants, the ASQ had a sensitivity of 94.1% and a specificity of 91.4%. Comparatively, among White participants, the ASQ had a sensitivity of 90.9% and a specificity of 91.8% (Table 2). No significant differences in ASQ sensitivity, LR+/−, and AUC were observed for Black and White participants (Table 2). ASQ screen positive rates were significantly higher in females than in males for both Black and White participants (see Table 1).

Table 1.

Racial/ethnic Identity and Suicide Risk Breakdown

	Sample		Screening Test: ASQ		Criterion Standard: SIQ/SIQ Jr
	N	Percent %	Positive n	Percent %	Positive n	Percent %
Total	1083	100.0	132	12.2	50	4.6
Racial/Ethnic Group
Non-Hispanic White	753	51.8	89	11.8	33	4.4
Non-Hispanic Black	330	22.7	43	13.0	17	5.2
			χ2=0.3, p-value =.575		χ2= 0.3, p-value =.579
Sex and Racial/Ethnic Group
Non-Hispanic Whitea
Female	401	53.8	57	14.2	18	4.5
Male	344	46.2	30	8.7	12	3.5
			χ2=5.4, p-value =.020		χ2=0.5, p-value =.489
Non-Hispanic Black
Female	204	61.8	34	16.7	14	6.9
Male	126	38.2	9	7.1	3	2.4
			χ2=6.2, p-value =.012		χ2=3.2 , p-value =.073
Medical Setting
Emergency Department	266	24.6	26	9.8	9	3.4
Inpatient	465	42.9	58	12.5	22	4.7
Outpatient	352	32.5	48	13.6	19	5.4
			χ2=2.2, p-value =.337		χ2 =1.4, p-value =.492

^aEliminated 8 participants with unknown sex

Table 2.

Psychometric Properties of the ASQ across racial/ethnic subgroup

	Total			Non-Hispanic White			Non-Hispanic Black			NH-White vs. NH- Black comparison
	Estimate	95% CI		Estimate	95% CI		Estimate	95% CI		p-value
Sensitivity	0.920	0.845	0.995	0.909	0.811	1.000	0.941	0.829	1.000	0.613
Specificity	0.917	0.900	0.934	0.918	0.898	0.938	0.914	0.883	0.945	0.739
LR+	11.051	8.639	13.462	11.094	8.129	14.059	10.911	6.769	15.053	0.944
LR−	0.087	0.005	0.169	0.099	−0.008	0.206	0.064	−0.058	0.187	0.922
AUC	0.918	0.880	0.957	0.914	0.863	0.964	0.928	0.868	0.987	0.121

Note: LR+: The likelihood ratio of a positive test result; LR−: The likelihood ratio of a negative test result; AUC: Area under receiver operating characteristic curve

Discussion

Pooling data from three ASQ validation studies revealed no significant differences in ASQ psychometric properties between Black and White youth, providing initial evidence that the ASQ is valid tool for screening Black youth at risk for suicide.

Limitations

Study limitations include: 1) both our sample and the SIQ/SIQ-JR validation study¹² used a predominantly English-speaking sample, and race categories under analysis were limited to Black and White participants due to small sample sizes of other races; 2) the validation studies were cross-sectional without follow-up of participant outcome; 3) while the ASQ has been studied in youth starting at age 10, the SIQ/SIQ-JR is only validated for youth 12 and older; and 4) the number of Black (30.5%) participants was smaller compared to the White sample (69.5%). Given the increasing suicide rates among Black, American Indian/Alaska Native, Asian and Pacific Islander, and Latino populations, it is important future studies ensure the validity of suicide risk instruments across all racial and ethnic groups.

Conclusions

The Ask Suicide-Screening Questions (ASQ) psychometric properties for Black and White participants were equivalent (sensitivity = 94% vs. 90.9%; specificity = 91.4% vs. 91.8%, respectively), providing initial evidence that the ASQ is valid for screening Black youth at risk for suicide. Further research is needed to confirm the validity of the ASQ among Black youth and other racial and ethnic groups.