Cultural sensitivity in screening adults for a history of childhood abuse: evidence from a community sample.

BACKGROUND: A number of practice guidelines and recommendations call for the assessment of childhood abuse history among adult medical patients. The cultural sensitivity of screening questions, however, has not been examined.
OBJECTIVE: To assess whether questions that inquire about childhood abuse history function differently for black and white patients.
DESIGN: Cross-sectional telephone surveys in 1997 and 2003.
SUBJECTS: Randomly sampled adults from Memphis, Tenn (1997, N = 832; 2003, N = 967). MEASUREMENTS: Physical, emotional, and sexual abuse scales of the Childhood Trauma Questionnaire-Short Form (CTQ-SF). Standardized mean difference technique for differential item functioning to assess for possible bias in CTQ-SF items.
RESULTS: Controlling for total physical abuse scale scores, black respondents were significantly (P < .01) more likely than white respondents to report that they had been punished with a hard object during their childhood, but less likely to report having being hit so hard that it left marks, have been hit so hard that someone noticed, or to believe they had been physically abused.
CONCLUSIONS: Inquiries that do not explicitly differentiate physical punishment from physical abuse may not be useful for black respondents because they tend to identify black respondents who report fewer clearly abusive experiences than comparable white respondents. Although untested in this study, one possible explanation is that physical discipline may be used more frequently and may play a different role among black families than among white families. These results underline the importance of attending to cultural factors in clinical history taking about childhood abuse histories.

BACKGROUND

Racial and ethnic disparities in health care access, experiences, and outcomes are well documented.1 Improving patient-centered communication and the cultural sensitivity of physicians has been proposed as an important step toward improving quality of care for minority populations and eliminating health care disparities.2 Cultural sensitivity may be defined as a set of attitudes, skills, behaviors, and policies that enable individuals to establish effective interpersonal and working relationships that supersede cultural differences.3 For clinicians, culturally competent history taking is especially important when trying to understand a patient’s unique experience and its impact on current health status.

Screening for histories of childhood abuse is potentially one such experience. A history of physical or sexual abuse is reported in as many as 20–50% of patients in adult primary care settings.4 Rates of reported childhood physical and sexual abuse do not appear to differ between black and white respondents in adult primary care practice or community samples,5,6 although differences have been detected in rates of reported emotional abuse and exposure to a wide group of adverse childhood events, including household dysfunction.6,7

A growing body of research shows an association between a history of childhood maltreatment and both psychiatric and nonpsychiatric medical problems in adults. Childhood abuse has been associated with depression and anxiety,8–12 and medical diagnoses such as headache,13 irritable bowel syndrome,14 fibromyalgia,15 and other chronic pain conditions.16 In addition, patients with abuse histories report more health risk behaviors, such as unsafe sexual practices17,18 and alcohol and drug use,12,19 and use more health care resources.20,21

A recent review called for primary and subspecialty care physicians to screen adult patients for a history of childhood abuse as a health risk factor.4 Practice guidelines and recommendations for a number of specific psychiatric and nonpsychiatric medical conditions include an assessment of abuse history.14,22–24 There are no published guidelines, however, for how and under what conditions adults should be screened for childhood abuse histories in primary care settings. Although it is generally accepted that patients will share more information when questioned about specific experiences, instead of using the term “abuse,”25,26 even specific experiences may have different implications and consequences in a cross-cultural context.

Culturally sensitive assessment of childhood abuse history requires that screening questions and interpretation of responses accurately reflect the experiences of patients rather than bias in the assessment process. If questions are unbiased, patients from different racial groups who have similar childhood abuse histories will respond similarly to individual questions about childhood abuse. Methodologically, this means that responses to a given question will be independent of racial group membership among patients who are matched on their responses to a set of related questions about childhood abuse. Otherwise, a reasonable conclusion would be that the question is assessing something related to racial group membership, but not necessarily abuse per se.27,28

It is possible, for instance, that differences in the role of physical punishment between black and white families could influence responses to screening questions about childhood physical abuse. Numerous studies have reported that black families tend to use corporal punishment with children more than other racial and ethnic groups in the United States.29–31 The role of corporal punishment in black families, however, may be different than in white families. Studies have found, for instance, that the association between physical discipline and disruptive behaviors found in white children does not necessarily generalize to black children.30,32,33

The objective of this study was to assess, using a community telephone sample, whether commonly used queries about abuse history may function differently with black and white patients. Although the research literature does not document how primary care physicians in practice typically assess childhood trauma, the Childhood Trauma Questionnaire–Short Form (CTQ-SF)34,35 is the most commonly used retrospective screening tool for childhood maltreatment in medical research. A January 2006 MEDLINE search found that the CTQ-SF had been used in 70 different studies, whereas no other self-report instrument for retrospectively assessing childhood abuse had been used in more than 10 studies. We used questions about childhood abuse from the CTQ-SF to compare responses across black and white respondent groups. Because of differences between black and white families in the use of physical punishment and its relationship with negative behaviors among children, we hypothesized that one question on the CTQ-SF about punishment with a hard object would be endorsed by black respondents at comparatively higher rates than would be expected based on their responses to other questions about childhood physical abuse. To test this hypothesis and to explore other potential differences in responses to questions about childhood physical, sexual, and emotional abuse, we used data from 2 large population samples from the greater Memphis area.

METHODS

Sample Selection

The CTQ-SF was administered in 2 large population samples, the 2003 Mid-South Social Survey Research Program (MSSSRP) survey and the 1997 Memphis Area Study (MAS). In each survey, households were randomly sampled from telephone numbers in the Memphis and Shelby County Telephone System Coles directory. Eligible respondents were English-speaking residents ages 18 to 75 for the 2003 survey and ages 18 to 65 for the 1997 survey. Households were randomly designated as male or female before telephone contact. If the person answering the phone was of the specified sex, only that person could be interviewed. If the person was not of the specified sex, the interviewer asked the person to choose a household member of the specified sex. If a person of the specified sex did not live in the household, the person answering the phone became the selected respondent.

Instrument

The CTQ-SF34,35 is a 28-item retrospective self-report questionnaire designed to assess 5 dimensions of childhood maltreatment: physical abuse, emotional abuse, sexual abuse, physical neglect, and emotional neglect. In this study, we analyzed data from the 15 items of the 3 abuse scales of the CTQ-SF: physical abuse, emotional abuse, and sexual abuse. Each of the 3 abuse scales has 5 items, and the item response options reflect the frequency of maltreatment experiences (never, rarely, sometimes, often, and very often). Bernstein et al.34 reported good internal consistency of the CTQ-SF for each of the abuse scales across 4 heterogeneous samples: physical abuse = 0.83 to 0.86, emotional abuse = 0.84 to 0.89, and sexual abuse = 0.92 to 0.95. Table 1 shows the CTQ-SF items from the 3 scales used in this study with mean scores and standard deviations.

Table 1

Items from the Childhood Trauma Questionnaire–Short Form abuse scales with means and standard deviations

Scale	Item	2003	1997
		Mean (SD)	Mean (SD)
Physical abuse	I got hit so hard by someone in my family that I had to see a doctor or go to the hospital.	1.08 (0.44)	1.10 (0.48)
	People in my family hit me so hard that it left me with bruises or marks.	1.21 (0.65)	1.25 (0.80)
	I was punished with a belt, a board, a cord, or some other hard object.	2.16 (1.16)	2.00 (1.12)
	I got hit or beaten so badly that it was noticed by someone like a teacher, neighbor, or doctor.	1.06 (0.31)	1.07 (0.39)
	I believe that I was physically abused.	1.16 (0.62)	1.18 (069)
Emotional abuse	People in my family called me things like “stupid,” “lazy,” or “ugly.”	1.41 (0.87)	1.44 (0.91)
	I thought that my parents wished I had never been born.	1.17 (0.60)	1.18 (0.66)
	People in my family said hurtful or insulting things to me.	1.50 (0.90)	1.54 (0.97)
	I felt that someone in my family hated me.	1.24 (0.74)	1.25 (0.80)
	I believe that I was emotionally abused.	1.32 (0.89)	1.34 (0.95)
Sexual abuse	Someone tried to touch me in a sexual way, or tried to make me touch them.	1.16 (0.57)	1.15 (0.57)
	Someone threatened to hurt me or tell lies about me unless I did something sexual with them.	1.09 (0.44)	1.05 (0.37)
	Someone tried to make me do sexual things or watch sexual things.	1.12 (0.50)	1.09 (0.43)
	Someone molested me.	1.11 (0.48)	1.10 (0.48)
	I believe that I was sexually abused.	1.13 (0.60)	1.11 (0.53)
Item responses	Never, rarely, sometimes, often, very often

Administration

All interviews were done over the telephone between March 27 and May 23, 2003 for MSSSRP and between February 12 and April 25, 1997 for MAS. Each interviewer received a training manual and at least 6 hours of training, as well as 3–4 practice interviews with supervision. Computer-assisted telephone interviewing files were downloaded automatically at the completion of each interview, and files were checked for reliability in data entry. Interviewers attempted to reach each selected household at least 10–12 times before listing it as a noncontact.

Analyses

Bivariate analyses comparing black and white respondents on demographic variables were conducted by means of χ2 statistics for categorical variables and two-tailed t tests for continuous variables. We assessed potential bias across race on abuse questions by using a standardized mean difference technique for differential item functioning (DIF), which compares the item means of two groups after adjusting for differences on a matching variable.36 DIF is considered to be present when responses to an item depend on a factor (e.g., race) other than the construct that the item is designed to measure (e.g., physical abuse). DIF analysis in this study was done by comparing each item mean score for black and white respondents, adjusted for the total score on the appropriate abuse scale and demographic factors, including age, sex, marital status, and level of education (SPSS General Linear Model, Univariate, Main Effects). Demographic factors were included in the item mean adjustments to assess the independent effects of race above and beyond these factors. Adjusted means for each item were computed for black and white respondents, and an item was classified as functioning differently across black and white respondents if the mean item score was significantly different across racial groups after controlling for the appropriate total scale score and demographic factors. In addition, Cohen’s effect size d37 was also estimated for each item based on pooled standard deviations to estimate the magnitude of the difference. All analyses were conducted using SPSS version 13.0 (Chicago, Ill), and all statistical tests were two-sided with a P < .05 significance level.

RESULTS

Sample Characteristics

In the 2003 and 1997 surveys, interviews with CTQ-SF data were completed in 880 of 1,266 eligible households (69.5% response rate) and 998 of 1,303 eligible households (76.6%), respectively. Only black and white respondents were included in the analyses because of the small number of respondents who identified themselves as belonging to any other racial/ethnic group. Of 832 black and white respondents in the 2003 survey, 812 provided complete data for all abuse items, 828 for all physical abuse items, 823 for all emotional abuse items, and 821 for all sexual abuse items. Of 967 black and white respondents in the 1997 survey, there were complete data on 945 for all abuse items, 956 for all physical abuse items, 953 for all emotional abuse items, and 955 for all sexual abuse items.

As shown in Table 2, for both the 2003 and 1997 surveys, white respondents were significantly older than black respondents (P < .01) and more likely to be married or living with a partner (P < .01), have more education (P < .01), and earn more (P < .01, data for 2003 only). Black respondents tended to be more likely to be female, albeit not significantly.

Table 2

Demographic data by race for 2003 and 1997 samples

Variables	MSSSRP 2003 (N = 832)			MAS 1997 (N = 967)
	Blacks (n = 373)	Whites (n = 459)	P	Blacks (n = 428)	Whites (n = 539)	P
Female, n (%)	261 (70.0)	296 (64.5)	.09	288 (67.3)	330 (61.2)	.05
Age, mean (SD)	38.6 (12.2)	43.0 (12.5)	<.01	37.4 (13.2)	41.2 (12.6)	<.01
Marital status, n (%)
Single	115 (30.1)	77 (16.8)	<.01	178 (41.6)	112 (20.8)	<.01
Married/Live Part.	127 (34.0)	278 (60.6)		149 (34.5)	341 (63.3)
Sep./Div./Widowed	123 (33.0)	102 (22.2)		96 (22.4)	80 (14.8)
Not Reported	8 (2.1)	2 (0.4)		5 (1.2)	6 (1.1)
Education, n (%)
Less than high school	43 (13.1)	18 (3.9)	<.01	72 (16.8)	39 (7.2)	<.01
At least high school or equivalent	250 (67.0)	250 (54.5)		253 (59.1)	285 (52.9)
Completed college	71 (19.0)	190 (41.4)		87 (20.3)	205 (38.0)
Not reported	3 (0.8)	1 (0.2)		16 (3.7)	10 (1.9)
Total household income, n (%)
Less than *$25,000	76 (20.4)	49 (10.7)	<.01	Not available	Not available	N/A
*$25,000 to *$49,999	95 (25.5)	79 (17.2)
*$50,000 to *$74,999	47 (12.6)	80 (17.4)
*$75,000 and above	33 (8.8)	132 (28.8)
Not reported	122 (32.7)	119 (25.9)

MSSSRP = Mid-South Social Survey Research Program, MAS = Memphis Area Study.

Assessment of Racial Differences in Responses to Questions About Childhood Abuse

Adjusted physical, emotional, and sexual abuse scale item means by race are shown in Table 3. In both the 2003 and 1997 surveys, the adjusted item mean for the item, “I was punished with a belt, a board, a cord, or some other hard object,” as hypothesized, was significantly higher for black respondents than for white respondents (P < .01, d = 0.18 in both surveys). In the 2003 survey, white respondents were significantly more likely to indicate that “People in my family hit me so hard that it left me with bruises or marks” (P < .01, d = −0.23) and “I believe that I was physically abused” (P < .01, d = −0.19) compared to black respondents after controlling for overall physical abuse scores and demographics. For the 1997 survey, the adjusted item means were significantly higher for white respondents on 3 items: “People in my family hit me so hard that it left me with bruises or marks” (P = .02, d = −0.13), “I got hit or beaten so badly that it was noticed by someone like a teacher, neighbor, or doctor” (P = .01, d = −0.15), and “I believe that I was physically abused” (P = .03, d = −0.13). Thus, for respondents with a given overall physical abuse scale score, black respondents were more likely to endorse having been punished with a hard object than whites, but were less likely to endorse items that reflect being hit hard enough to leave marks, being hit hard enough that it was noticed, or to state that they believe they were physically abused. For both the 2003 and 1997 data, when the item “I was punished with a belt, a board, a cord, or some other hard object” is removed from the physical abuse scale, no remaining items demonstrated significant differences across race. This was not the case when the other items identified as displaying DIF were removed from the physical abuse total score calculation. No items from the sexual abuse or emotional abuse scales from either survey demonstrated significant differences across race, controlling for the total abuse score on the appropriate scale and demographics.

Table 3

Physical abuse scale item means for black and white respondents adjusted for total physical abuse scale score, age, sex, education level, and marital status

	MSSSRP 2003 (N = 832)				MAS 1997 (N = 967)
	Adjusted mean and standard error		Effect sized		Adjusted mean and standard error		Effect sized
	Blacks (n = 373)	Whites (n = 459)	d	P	Blacks (n = 428)	Whites (n = 539)	d	P
Physical abuse
Hit hard enough...doctor	1.06 (0.02)	1.07 (0.03)	−0.01	.90	1.09 (0.02)	1.11 (0.02)	−0.04	.45
Hit hard enough...bruises	1.14 (0.03)	1.28 (0.03)	−0.23	<.01	1.19 (0.03)	1.27 (0.03)	−0.13	.02
Punished with hard objects	2.23 (0.04)	2.09 (0.04)	0.18	<.01	2.09 (0.04)	1.94 (0.04)	0.18	<.01
Was physically abused	1.05 (0.03)	1.17 (0.03)	−0.19	<.01	1.15 (0.03)	1.23 (0.03)	−0.13	.03
Hit badly enough...noticed	1.06 (0.02)	1.07 (0.02)	−0.02	.69	1.05 (0.02)	1.10 (0.02)	−0.15	.01
Emotional abuse
Called...“stupid,” “lazy,” or “ugly”	1.45 (0.04)	1.42 (0.04)	0.03	.56	1.47 (0.03)	1.44 (0.03)	0.05	.39
Parents wished never born	1.20 (0.03)	1.18 (0.03)	0.03	.58	1.19 (0.03)	1.19 (0.03)	0.00	.98
Said hurtful or insulting things	1.42 (0.03)	1.47 (0.03)	−0.08	.19	1.48 (0.03)	1.53 (0.03)	−0.09	.14
Someone in my family hated me	1.26 (0.03)	1.25 (0.03)	0.01	.83	1.24 (0.03)	1.22 (0.03)	0.03	.60
Was emotionally abused	1.23 (0.04)	1.31 (0.05)	−0.09	.12	1.29 (0.03)	1.36 (0.03)	−0.10	.13
Sexual abuse
Tried to touch me in a sexual way	1.13 (0.01)	1.13 (0.01)	−0.01	.89	1.13 (0.01)	1.14 (0.01)	−0.02	.73
Threatened to hurt me or tell lies	1.11 (0.02)	1.08 (0.02)	0.08	.18	1.07 (0.01)	1.06 (0.01)	0.04	.45
Tried to make me do sexual things	1.11 (0.01)	1.12 (0.02)	−0.04	.54	1.08 (0.01)	1.08 (0.01)	−0.02	.80
Someone molested me	1.12 (0.02)	1.13 (0.02)	−0.04	.46	1.10 (0.01)	1.10 (0.01)	−0.01	.89
Was sexually abused	1.10 (0.02)	1.14 (0.02)	−0.09	.12	1.08 (0.02)	1.10 (0.02)	−0.04	.14

MSSSRP = Mid-South Social Survey Research Program, MAS = Memphis Area Study.

As shown in Table 4, black respondents at higher income and education levels were more likely to have been punished with hard objects than black respondents at lower levels. There was no difference in responses to this question based on educational or income levels for white participants. Overall, black respondents tended to be more likely to endorse having been punished with hard objects across levels of educational attainment and total household income than white respondents.

Table 4

Endorsement of having been punished with hard objects by race and level of education, total household income

	MSSSRP 2003		MAS 1997
	Respondents endorsing/respondents in category		Respondents endorsing/respondents in category
	Black respondents	White respondents	Black respondents	White respondents
Education (%)
Less than high school	25/49 (51.0)	7/18 (38.9)	32/72 (44.4)	19/39 (48.7)
High school degree	66/107 (61.7)	47/80 (58.8)	59/120 (49.2)	59/114 (51.8)
Some college	99/143 (69.2)	96/170 (56.5)	83/132 (62.9)	94/171 (55.0)
College degree	54/71 (76.1)	97/190 (51.1)	49/87 (56.3)	104/204 (51.0)
Total household income (%)
Less than $25,000	49/76 (64.5)	25/49 (51.0)	Not available	Not available
$25,000 to $49,999	54/95 (56.8)	38/79 (48.1)
$50,000 to $74,999	35/47 (74.5)	53/80 (66.3)
$75,000 and above	24/33 (72.7)	69/132 (52.3)

MSSSRP = Mid-South Social Survey Research Program, MAS = Memphis Area Study.

DISCUSSION

This study employed a frequently used childhood abuse questionnaire in 2 population-based samples to compare responses between black and white respondents to inform culturally competent history taking in clinical settings. Several items of the physical abuse scale were identified through DIF analysis to function differently across groups. After controlling for demographic differences and overall levels of physical abuse based on CTQ-SF responses, black respondents were significantly more likely to report having been “punished with a belt, a board, a cord, or some other hard object,” than white respondents, but less likely to report experiences like having been “hit...so hard that it left...bruises or marks,” “hit...so badly that it was noticed,” or “physically abused.” The differences in physical abuse responses between black and white respondents did not appear to reflect differences in socioeconomic indicators because these were controlled for in the analyses.

When individual items are found to have DIF, it is typically because they are measuring something different across groups. In this study, several items of the physical abuse scale initially exhibited DIF. When the item inquiring about having been punished with hard objects was removed from the scale, however, none of the other items functioned differently for black and white respondents. This suggests that the “punished with hard objects” item of the CTQ-SF likely measures both the physical abuse construct being measured by the other physical abuse items, as well as an additional construct that is more prevalent among black respondents. One possibility is that the item measures exposure to both physical abuse and physical discipline and that the relationship between these two constructs is not the same for black and white respondents. This would be the case, for instance, if black respondents were punished more often with hard objects than white respondents even when they did not report other clearly physically abusive experiences.

Physical discipline is common in the United States, with 94% of parents in a national phone survey reporting hand-slapping or spanking a toddler.31 It is more prevalent, however, among black compared with white families,31 and there is evidence to suggest racial and ethnic differences in associations between physical discipline and outcomes.33,38 Deater-Deckard et al.,33 for instance, found higher levels of aggressive behaviors among white adolescents who experienced physical discipline compared to white adolescents who were not physically disciplined, but comparatively lower levels of behavior problems among black adolescents who had been physically disciplined.

Primary care visits present an opportunity to screen patients for a history of childhood abuse.4 This study highlights possible differences between black and white patients’ responses to questions about potentially physically abusive experiences. The results from this study emphasize the need for culturally sensitive assessment with a patient-in-context approach to the evaluation of behaviors and life experiences.39 This requires moving beyond assumptions and providing patients with the time to provide more detailed explanations of their experiences in a safe and confidential setting. For instance, if a patient indicates that he or she was punished with hard objects, it is important to inquire about the circumstances of the punishment, including how it was carried out; what the patient thought of the punishment at the time and now; if the patient would like his/her own children to have a different or similar experience; and the patient’s understanding of childhood abuse, including examples of what does and does not constitute abuse. Because of the known adverse effects of physical abuse on both black and white patients,8–12,14,16,17 it is important for clinicians to screen for physical abuse, but to also understand that different questions about physical abuse may have somewhat different implications for black and white patients.

There are limitations that should be taken into consideration in interpreting results from this study. First, as is the case in most retrospective studies of childhood maltreatment, data in this study was limited to self-report. Thus, the actual relative prevalence of a history of childhood physical abuse among black and white adults is unknown because the majority of cases are never reported to authorities.40 Second, characteristics of nonresponders are not known, creating the possibility of bias. Third, we have not examined the relationship between report of physical punishment or physical abuse in childhood and adult outcomes. Thus, we cannot comment on the relationship between each of these constructs and long-term prognosis. Fourth, it is possible that changes in public discourses on abuse over the course of the sample’s maturation from childhood may have influenced responses. An additional possible limitation involves socioeconomic differences between black and white respondents, as shown in Table 4. However, because all DIF analyses controlled for demographic and socioeconomic differences, it would appear that the study’s results were because of differences in culturally related practices instead of socioeconomic differences.

In summary, this study with a large sample of both black and white respondents employed typical abuse screening questions from the CTQ-SF and found that blacks endorsed being punished with hard objects as children at much greater levels than would be expected by their responses to other questions about physical abuse. Although evidence in this study came from a community telephone sample, the results underline the importance of attending to cultural factors in clinical history taking about childhood abuse histories. They also serve, more broadly, to demonstrate empirically that specific communication and assessment strategies may function differently depending on the cultural background of patients.