PREGNANCY AND DOMESTIC VIOLENCE - A Review of the Literature
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
10.1177/1524838003259322
TRAUMA,
Jasinski / PREGNANCY
VIOLENCE, &AND
ABUSE
DOMESTIC ARTICLE
/ JanuaryVIOLENCE
2004
PREGNANCY AND DOMESTIC VIOLENCE
A Review of the Literature
JANA L. JASINSKI
University of Central Florida
Pregnancy-related violence is a serious public health issue. Although there is a
growing body of research on this subject, there are still many unanswered ques-
tions regarding the prevalence of this type of victimization, the risk factors, and the
consequences. The purpose of this literature review is to organize and synthesize
the interdisciplinary empirical research on pregnancy-related violence and to pro-
vide direction for both researchers and practitioners for future work in this area.
Key words: domestic violence, pregnancy
ESTIMATES OF VIOLENCE against women To incorporate information from the variety
suggest that almost 2 million women are physi- of disciplines in which research on pregnancy-
cally assaulted annually and more than 50 mil- related violence takes place, the following data-
lion are assaulted in their lifetime (Tjaden & bases were searched: Sociological Abstracts,
Thoennes, 2000). Although the sheer magni- PsychINFO, CINAHL, MEDLINE, Social Ser-
tude of this problem has generated a great deal vices Abstracts, and Ageline. The focus of this
of interest on the part of the public health, advo- review is on research published after 1996,
cate, and academic communities, it has only the date of the last comprehensive review of the
been recently that significant attention has been literature on pregnancy-related violence
paid to the intricacies of the relationship be- (Gazmararian et al., 1996); however, occasion-
tween pregnancy and violence. And, although ally research published prior to 1996 is used for
the research literature is growing every day, it is illustrative purposes. In addition, the focus of
often difficult to determine the exact nature of this review is primarily on research conducted
pregnancy-related violence. This has posed dif- with samples from the United States, due to
ficulties for both researchers and practitioners, possible differences in health care systems
who need a clear understanding of the relation- cross-culturally.
ship between intimate partner violence (IPV)
and pregnancy to develop and implement effec- Prevalence Rates of Pregnancy-Related
tive prevention and intervention programs. The Violent Victimization
purpose of this literature review is to provide a
framework for understanding the wide variety One of the first areas of debate regarding the
of research studies on the topic of pregnancy relationship between pregnancy and violence is
and violence. the issue of prevalence. An early review synthe-
TRAUMA, VIOLENCE, & ABUSE, Vol. 5, No. 1, January 2004 47-64
DOI: 10.1177/1524838003259322
© 2004 Sage Publications
4748 TRAUMA, VIOLENCE, & ABUSE / January 2004
KEY POINTS OF THE remains an important factor to consider when
RESEARCH REVIEW interpreting prevalence estimates. Notably,
analysis of population-based data from the
• Estimates of the prevalence of pregnancy-related
violence vary due to differences in research
Centers for Disease Control and Prevention’s
designs, measures used, and populations sam- (1999) Pregnancy Risk Assessment Monitoring
pled. System (PRAMS) 1996 Surveillance Report
• The debate about whether pregnant women are found reported rates of pregnancy-linked abuse
at increased risk for violence continues as to be much lower than studies using hospital-
hospital- and clinic-based studies find pregnancy
based samples, ranging from 2.9% to 5.7%
a time of increased risk for violence, whereas na-
tional studies do not find an association between among several thousand women across 11
pregnancy and intimate partner violence. states participating in PRAMS (also see
• Consequences of pregnancy-related violence in- Gazmararian et al., 1995).
clude later entry into prenatal care, low birth Certainly the type of sample is an important
weight babies, premature labor, fetal trauma, un- factor to consider when examining the preva-
healthy maternal behaviors, and health issues for
the mother.
lence of pregnancy-related violence. For exam-
• Health care providers who have received train- ple, much of the research considering the rela-
ing are more likely to screen for violence; how- tionship between pregnancy and violence uses
ever, very few providers have received training primarily hospital- or clinic-based samples
as part of their medical education. (e.g., samples of either postpartum women or
interviews with women during a prenatal care
visit) (Martin, Mackie, Kupper, Buescher, &
sizing the results of 13 studies found that the Moracco, 2001; Muhajarine & D’Arcy, 1999;
prevalence of violence during pregnancy Rachana, Suraiya, Hisham, Abdulaziz, & Hai,
ranged from 0.9% to 20.1% (Gazmararian et al., 2002). Prevalence estimates using these samples
1996). It was suggested that the wide-ranging are estimates of violence among women who
estimates were likely a result of the use of a vari- are pregnant. In contrast, researchers using na-
ety of violence measures, and differences in tional probability samples are estimating the
both the populations sampled, and methodolo- risk for victimization among either all women
gies used. In fact, the study that found the low- or all women of childbearing age regardless of
est prevalence rate used a sample from a pri- pregnancy status. Researchers using national
vate clinic in which more than one third of the probability samples have found prevalence
women reported incomes greater than $50,000 a rates of 15% (Gelles, 1990), 14.5% (Anglo
year. In contrast, the study that reported the women), 23.6% (Hispanic women) (Jasinski &
highest prevalence rate used a sample of Kaufman Kantor, 2001), and 20.5% (couples in
women from a public clinic in which two thirds which the male partner was persistently
of the women reported incomes of less than violent) (Jasinski, 2001).
$20,000 a year. Additional differences were ap- In addition to physical abuse, several re-
parent in the assessments used to determine searchers have also considered other forms of
victimization, the points during the preg- abuse, including verbal abuse and sexual abuse,
nancy at which the assessment was adminis- as well as different severity levels of physical
tered, and the method in which the assessment violence (Jasinski & Kaufman Kantor, 2001;
was administered. Prevalence estimates re- Parker, McFarlane, & Soeken, 1994; Shumway
ported in research that has taken place since the et al., 1999), each of which could impact preva-
Gazmararian et al. (1996) review have remained lence estimates of pregnancy-related victimiza-
primarily within this range. For example, do- tion. Parker and associates (1994), for example,
mestic violence assaults accounted for 22% of found that 23% of the teenagers in their sample
the cases of pregnant patients seen in several and 28% of the adults were physically or sexu-
North Carolina emergency rooms for trauma ally abused in the year prior to their first pre-
(Connolly, Katz, Bash, McMachon, & Hansen, natal visit. More recently, Shumway et al. (1999)
1997). The issue of research sample, however, reported that 36% of their sample of womenJasinski / PREGNANCY AND DOMESTIC VIOLENCE 49
attending an obstetrical clinic reported verbal most commonly used measure of violence in re-
abuse, 16% reported moderate physical vio- search using clinical samples is the Abuse
lence, and 14% reported severe physical vio- Assessment Scale. Most likely, this is due to the
lence during their pregnancy. Although much ease of administration and established reliabil-
of the research that considers pregnancy-related ity and validity of the scale. In contrast, research
violence focuses primarily on abuse by intimate using nonclinical community samples has more
partners, some researchers have included any commonly used the Conflict Tactics Scales to
experience of violence regardless of the identity measure violence victimization. Terminology
of the offender. A recent study found, for exam- differences regarding marital status, time peri-
ple, that although 11.9% of the adolescents who ods of exposure, and the definition of domestic
experienced some type of physical violence in violence also make comparisons nearly im-
the year prior to the research were assaulted by possible (Peterson, Saltzman, Goodwin, &
the father of the baby, the remaining victims re- Spitz, 1997). As a result, it is extremely difficult
ported experiencing violence at the hands of to provide any reliable information for practi-
other family members or relatives or as a wit- tioners.
ness to a fight in which someone was seriously
hurt (Wiemann, Agurcia, Berenson, Volk, & Are Pregnant Women at
Rickert, 2000). Greater Risk of IPV?
Although the increasing body of research ex-
amining pregnancy-related violence provides From a public health perspective, an impor-
much-needed information, it remains difficult tant question is that of risk. If being pregnant in-
to compare individual studies. For example, the creases the risk for violent victimization, then
PRAMS 1996 Surveillance Report found re- certain interventions are warranted. For the
ported rates of pregnancy-linked abuse to be most part, however, the majority of the re-
consistent with only the lower-bound estimates searchers examining pregnancy-related vio-
of several other studies noted above. One expla- lence use small samples
nation of the lower-than-anticipated rates may of either postpartum Comparisons of
be that the PRAMS asks only a few limited ques- women or women at- studies examining the
t io n s o n a bus e a n d que s t io n s a re n o t tending a prenatal clinic pregnancy-violence
behaviorally specific. For example, women without a comparison relationship are
were asked only whether they were “physically group of women who are difficult to make due
abused by a husband or partner during the 12 not pregnant (Bullock & to inconsistent
months preceding their most recent pregnancy McFarlane, 1989; Camp- terminology, a wide
or during their most recent pregnancy.” Both bell, Poland, Waller, & variety of instruments,
the limited number of items and the use of Ager, 1992; Gelles, 1974; and different
the term “abuse” (subject to varying interpreta- Stark & Flitcraft, 1995; methodological
tions) may lead to underestimates of assaults Stewart, 1994). Although techniques (Petersen,
preceding or coinciding with pregnancy. In ad- much of this research sug- Gazmararian et al.,
dition, comparisons of studies examining the gests that pregnancy may 1997) . . . Terminology
pregnancy-violence relationship are difficult to be a time of increased risk differences regarding
make due to inconsistent terminology, a wide for violence, at least for marital status, time
variety of instruments, and different method- some women (Berenson, periods of exposure,
ological techniques (Petersen, Gazmararian, Stiglich, Wilkinson, & and the definition of
et al., 1997). For example, violence has been as- Anderson, 1991; Gelles, domestic violence
sessed with a number of different instruments 1974; Smikle, Sorem, also make
including the Conflict Tactics Scales (Straus, Stain, & Hankins, 1996; comparisons nearly
1979, 1990a), the Danger Assessment Scale Webster, Sweett, & Stolz, impossible (Peterson,
(Campbell, 1986), and the Index of Spouse 1994), the reliance on an- Saltzman, Goodwin, &
Abuse (Hudson & McIntosh, 1981), among oth- e cd o t a l re p o rt s f ro m Spitz, 1997).
ers (See Table 1 for a list of recent studies). The pregnant women or hos-50
TABLE 1: Prevalence of Pregnancy-Related Violence
Author, Year Sample Description Violence Measure Prevalence
Cokkinides & Coker, 1998 6,718 women from South Carolina PRAMS pro- Asked if woman was involved in physical fight in 10.9% violence during pregnancy (physical
ject, includes women with live single births last 12 months before delivery? Was the fights and hurt by partner), 5.1% physical vio-
woman physically hurt by male partner in last lence only
12 months before delivery?
Goodwin et al., 2000 PRAMS: 39,348 women in 14 states who had Asked women if they were physically abused 3.2% abused only in year prior to pregnancy,
delivered a live infant, sampled using birth (pushed, hit, kicked, slapped) in the 12 1.5% abused only during pregnancy, 4.1%
certificates months before they became pregnant and/or abused both before and after pregnancy
during their current pregnancy
Glander et al., 1998 486 women seeking outpatient abortion Five questions on abuse asking about any vic- 39.5% had any history of abuse; no information
timization ever, within past year, and during given on percentage abused during preg-
current pregnancy? Also asked if pregnancy nancy or within last year
was result of forced sex and if first inter-
course was forced?
Hedin, 2000 207 Swedish-born women attending antenatal Severity of Violence Against Women Scale 24% reported threats, physical or sexual abuse
clinics, mailed survey 8 weeks postpartum modified to refer to pregnancy period as well during pregnancy
as past year
Jasinski, 2001 2,484 couples in which the female partner was Two questions: In serious disagreements, how 11.5% of male partners were violent in first
younger than 50 years old, part of the Na- often did R end up hitting or throwing things wave, 8.9% were violent in second wave
tional Survey of Families and Households at partner? During the past year, how many
fights with partner resulted in you/your part-
ner hitting, shoving, throwing things at him/
her?
Jasinski & Kaufman Kantor, 1,336 Anglo and Hispanic families in which the Conflict Tactics Scale Anglo: Violence reported in 14.5% of pregnant
2001 female partner was 50 years of age or youn- couples vs. 12.8% in nonpregnant couples.
ger, part of National Alcohol and Family Vio- Hispanic: 23.9% for pregnant vs. 15.3 for not
lence Survey pregnant
Martin, Clark, Lynch, Kupper, 703 Women 12 to 19 years old, Prenatal inter- Abuse Assessment Screen 29% victims of some type of violence
& Cilenti, 1999 view at health department
Martin, English, Clark, Cilenti, 2,092 prenatal patients in health department Abuse Assessment Screen 26% reported violence in lifetime, 23% only be-
& Kupper, 1996 fore current pregnancy, 2% before and during
pregnancy
Martin, Mackie, Kupper, 2,648 women who recently delivered live in- Asked if had been pushed, hit, slapped, kicked, 6.9% abused in 12 months before pregnancy,
Buescher, & Moracco, 2001 fants: Part of North Carolina PRAMS or physically hurt in some other way during 6.1% abused during pregnancy, 3.2% abused
three periods: 12 months before becoming postpartum
pregnant, during pregnancy, and after
delivery
Muhajarine & D’Arcy, 1999 543 women receiving prenatal services inter- Abuse Assessment Screen 8.5% abused in year preceding interview, 5.7%
viewed in third trimester abused during pregnancy
Norton, Peipert, Zierler, Lima, 334 new registrants for prenatal care Abuse Assessment Screen 15% in past year, 10% during current
& Hume, 1995 pregnancyRachana, Suraiya, Hisham, 7,105 women with single live birth between Case file review. Physical violence: physically 21% reported physical violence
Abdulaziz, & Hai, 2002 1996 and 1999 attending clinics of two uni- hurt by husband or involvement in physical
versity teaching hospitals fight 10 months before delivery
Sable, Fieberg, Martin, & 80 pregnant prisoners and 1,623 matching Asked if ever had been hit, slapped, kicked, or 37% of prisoners and 31% of patients experi-
Kupper, 1999 women admitted to prenatal care at health otherwise physically hurt and if ever had enced some type of violence
department, 17 to 37 years old been forced to have sexual activity with
someone
Shumway et al., 1999 401 women attending hospital clinic Conflict Tactics Scales 66% experienced physical or verbal abuse,
16% moderate physical abuse, 14% severe
violence
Smikle et al., 1996 563 questionnaires given to pregnant patients Adapted Abuse Assessment Screen (AAS) 18% history of physical or sexual abuse, 1.2%
at prenatal orientation classes at medical physically assaulted by partner during current
center serving military personnel and their pregnancy; lifetime prevalence of physical
dependents abuse 12%
Torres et al., 2000 1,004 Puerto Rican, Cuban American, Central Index of Spouse Abuse (ISA) and AAS 5.6% report abuse during pregnancy, 5.6% us-
American, African American, and Anglo ing ISA and 5.2% using AAS
Women from Florida and Massachusetts
Hospital sample
Wiemann, Agurcia, Berenson, 724 adolescents from hospital planning on Modified Abuse Assessment Screen 29% experienced violence in past year, 11.9%
Volk, & Rickert, 2000 keeping baby by the father of the baby
NOTE: Articles were included in this table if they met two criteria: (a) Part of their major focus was investigating the prevalence of pregnancy-related violence and (b) They were not included in
the extensive review conducted by Gazmararian et al., 1996.
5152 TRAUMA, VIOLENCE, & ABUSE / January 2004 pital samples of pregnant women does not Sweet, Bumpass, & Call, 1988). The results indi- make it possible to empirically test whether or cated that pregnant women were no more likely not pregnancy per se increases the risk for vio- to be victims of IPV than women who were not lence. Furthermore, this body of research is pri- pregnant. However, persistent violence was marily focused on examining the consequences more likely to occur among couples in which the of violent behavior for the infant (e.g., preterm male partner perceived that the pregnancy of labor, fetal death, low birth weight) (Parker his female partner occurred sooner than et al., 1994; Webster, Chandler, & Battistutta, intended. 1996) as well as improving assessment tech- Although studies using probability samples niques among physicians (McFarlane, Parker, seem to agree that pregnancy does not increase Soeken, & Bullock, 1992; Norton, Peipert, the risk for violent victimization, they were not Zierler, Lima, & Hume, 1995). Although this is designed to specifically look at this issue and likely to be beneficial in improving health care therefore have not included the necessary ques- outcomes for women and children, it still does tions to create a complete picture of the violence- not address the following question: Are pregnancy relationship. Both the Gelles (1990) pregnant women at a greater risk of assault by and Jasinski and Kaufman Kantor (2001) stud- their male partners compared to women who ies, for example, used reference periods of the are not pregnant? past year when asking about pregnancy and Research using national probability samples victimization status. No other pregnancy-related can address the question of risk as the sample questions were asked. Jasinski’s (2001) study studied includes both women who are pregnant used a data set that was not designed to address and women who are not, and studies using violence, and as a result, those questions are these types of samples have consistently re- weak. In none of these studies is it possible to ported no difference in risk due to pregnancy. determine the causal order between pregnancy For example, Gelles’s 1990 analysis of data from and violence. Furthermore, it should be noted the 1985 National Family Violence Survey that these studies, although finding no in- found, after controlling for age, that pregnant creased risk for victimization, also have not women were not significantly more likely to found a decreased risk. be victims of assaults by their male partners compared to women who were not pregnant. Ethnicity and Pregnancy-Linked Abuse Unfortunately, this study did not consider other pregnancy-related factors (either pregnancy- or In addition to the examination of pregnancy non-pregnancy-related) or demographic fac- as a risk factor for violence, it is also important tors (e.g., ethnicity) as potential risk markers for to consider the presence of other factors that violence. Similarly, researchers analyzing the may increase the possibility of violence. As with 1992 National Alcohol and Family Violence Sur- research on IPV in general, researchers are just vey (Kaufman Kantor, Jasinski, & Aldarondo, now beginning to consider racial/ethnic differ- 1994) found that for both Anglo and Hispanic ences in pregnancy-related violence victimiza- families, there was no direct effect of pregnancy tion. What little research that does exist is incon- on risk for violent victimization after control- clusive, with some researchers finding ling for socioeconomic status, stressful life significant differences by race (e.g., Cokkinides events occurring during the pregnancy year, & Coker, 1998; Dietz et al., 1997; Glander, and age (Jasinski & Kaufman Kantor, 2001). Al- Moore, Michielutte, & Parsons, 1998; Goodwin though the latter study improved on Gelles’s et al., 2000) and others finding no racial/ethnic (1990) study, it was not able to consider the role differences (Campbell, Oliver, & Bullock, 1998; of other pregnancy-related factors in the risk for Renker, 1999; Seguin, 1998; Wiemann et al., violence. More recent research (Jasinski, 2001) 2000). Perhaps even more interesting is the in- has addressed this gap in the literature by using consistency of the direction of the results in two waves of the National Survey of Families studies finding racial differences. For example, and Households (Sweet & Bumpass, 1996; some studies have found White women to be at
Jasinski / PREGNANCY AND DOMESTIC VIOLENCE 53
a greater risk for violence victimization 1995; Martin, English, Clark, Cilenti, & Kupper,
(Glander et al., 1998; McFarlane & Parker, 1996) 1996) and those that find the opposite
or more severe abuse (McFarlane et al., 1992), (Campbell et al., 1992; Campbell, Pugh et al.,
whereas others have found greater levels of 1995; Evins & Chescheir, 1996; Helton, 1986;
vict im iz a t ion a m o n g m in o rit y w o m e n Stewart & Cecutti, 1993; Taggart & Mattson,
(Cokkinides & Coker, 1998; Gazmararian et al., 1996).
1995). One study of 501 patients attending a Among the studies concluding that preg-
low-risk obstetric clinic reported a prevalence nancy provides an interruption in victimization
rate of abuse that was 3 times higher among An- experience, Martin et al.’s (1996) study found
glo American women compared to Hispanic that although 26% of their sample reported life-
American women and 1.6 times higher among time victimization, 23% experienced violence
Anglo American women compared to African only before their current pregnancy, whereas
American women (Berenson et al., 1991). In con- less than 1% experienced violence only during
trast, another study using a sample of 12,612 the current pregnancy and 2% experienced vio-
new mothers selected from the PRAMS found lence both prior to and during the pregnancy.
that non-White mothers had higher rates of vio- Evidence from the PRAMS 1996 Surveillance
lence than mothers who were White. A more re- Report (1999) indicates that a greater propor-
cent study using a national sample found that tion of women reported less physical violence
pregnant Hispanic women were significantly during their pregnancy than before the onset of
more likely than Hispanic women who were not pregnancy. In another, more recent study, none
pregnant to be victims of any assaults, particu- of the women who reported abuse reported that
larly minor assaults by their male partners. the abuse began at the time of the pregnancy
Among Anglo women, pregnancy was associ- (Hedin, 2000).
ated with severe wife assault only (Jasinski & In contrast to this research, others suggest
Kaufman Kantor, 2001). Similar to issues re- that violence may actually escalate with preg-
garding prevalence of pregnancy-related vio- nancy (Campbell et al., 1992; Campbell, Pugh
lence, possible explanations for some of these et al., 1995; Helton, 1986; Stewart & Cecutti,
inconsistencies in the research literature include 1993; Taggart & Mattson,
different types of samples, different assess- 1996). In one study, al- Regardless of the
ments of victimization, and confounding race most one third of the exact dynamics of
and socioeconomic status. Whether or not race/ women who were abused pregnancy-related
ethnicity is a significant risk factor for in their current preg- violence, most of the
pregnancy-related violence remains to be seen. nancy said the abuse in- research finds that
creased in pregnancy women who were
What Are the Motives/Risk Factors for (Berenson et al., 1991). In abused while they
Pregnancy-Related Violence? addition, Campbell and were pregnant had a
Alford (1989) found that history of victimization
Although the body of research examining many victims of marital (Glander et al., 1998;
pregnancy-related violence has grown in recent rape have reported sexual Horrigan, Schroeder,
years and the issue has moved to the forefront of assaults during preg- & Schaffer, 2000;
such organizations as the National Center for nancy or soon after their Smikle et al., 1996).
Injury Prevention, there has been relatively little delivery of the baby. This would suggest
examination of the epidemiology of risk for What can be concluded that women who
abuse among pregnant women. Consequently, from this contradictory have a history of
definitive policy solutions remain absent. Ex- evidence? Regardless of victimization should
amining research in this area shows two types of the exact dynamics of be identified as an
empirical findings: those studies that have pregnancy-related vio- at-risk group, with
found pregnancy to be a time of respite for some lence, most of the re- specific intervention
previously abused women (Campbell et al., search finds that women efforts targeted to
1998; Campbell, Pugh, Campbell, & Visscher, who were abused while them.54 TRAUMA, VIOLENCE, & ABUSE / January 2004
TABLE 2: Correlates of Pregnancy-Related Violence nancy. One factor that has emerged as a
consistent risk factor for violence is low socio-
Correlate of Violence Author, Year
economic status (measured with educational
Low socioeconomic status Martin et al., 2001 le ve ls , in com e , a n d /or e m p loy m e n t )
Cokkinides & Coker, 1998
(Cokkinides & Coker, 1998; Evins & Chescheir,
Goodwin et al., 2000
Curry & Harvey, 1998 1996; Gazmararian et al., 1995; Goodwin et al.,
Low levels of social support Glander et al., 1998 2000). It also appears as if women who are
Wiemann et al., 2000 abused do not have the same levels of social
Curry & Harvey, 1998
Curry, 1998 support as do women who are not abused
First time parenting Jasinski, 2001 (Glander et al., 1998; Wiemann et al., 2000). Each
Unexpected or unwanted Cokkinides et al., 1999 of these two factors, low socioeconomic status
pregnancy Goodwin et al., 2000
Jasinski, 20001
and low levels of social support, may also be re-
Gazmararian et al., 1999 lated to elevated levels of stress and in combina-
Race/ethnicity Goodwin et al., 2000 tion may increase the risk for violence. It may also
Glander et al., 1998
be appropriate to consider other pregnancy-
Cokkinides & Coker, 1998
Dietz et al., 1997 related factors that may increase the level of
Jasinski & Kaufman Kantor, stress experienced by a couple and conse-
2001 quently increase the risk for IPV. For example,
Older age Hedin, 2000
Horrigan et al., 1999 first-time parents may feel more stress related to
Youth Martin et al., 2001 the pregnancy than couples that have already
Goodwin et al., 2000 had children. At least one study has examined
Hedin, et al., 1999
Muhajarine & D’Arcy, 1999
the relationship between IPV and first-time
Alcohol use Grimstad & Backe, 1997 parenting (Jasinski, 2001). This study found that
Martin et al., 1996 having a first child was associated with violence
Curry, 1998
cessation. The implications of this finding, how-
ever, are somewhat double sided. On one hand,
this is consistent with prior work suggesting
they were pregnant had a history of victimiza-
that the birth of a child may provide a time of re-
tion (Glander et al., 1998; Horrigan, Schroeder, spite for previously abused women (Campbell,
& Schaffer, 2000; Smikle et al., 1996). This would Harris, & Lee, 1995; Campbell, Oliver, &
suggest that women who have a history of vic- Bullock, 1993). At the same time, it would be
timization should be identified as an at-risk inappropriate to suggest that women victims
group, with specific intervention efforts tar- actively try to get pregnant as a strategy to stop
geted to them. At the same time, it appears that violence. In addition, couples expecting an un-
although some women suffer abuse inordi- planned or unwanted child may be facing a
nately, the patterns and risk markers for abuse greater level of stress compared to those cou-
among these women have not been conclu- ples who have children that were planned,
sively identified (Petersen, Saltzman, et al., consequently increasing the risk for violence
1997). This gap in the research literature makes (Cokkinides, Coker, Sanderson, Addy, &
the development of comprehensive preven- Bethea, 1999). Goodwin et al.’s (2000) study, for
tion and intervention programs extremely example, found that the prevalence of abuse
difficult. was greater when the male partner did not want
What is known about patterns of risk for vio- the baby. In this study, women with unintended
lence is a result of research focusing on charac- pregnancies were 2.5 times more likely to ex-
teristics of the mother or the pregnancy as po- perience abuse compared to women with in-
tential risks (See Table 2 for a summary of these tended pregnancies. This is consistent with
risk factors). This body of work has produced Jasinski’s (2001) study finding that persistent
several consistent patterns of risk that could be violence was more likely to occur among cou-
used to develop prevention programs aimed at ples in which the male partner perceived that
reducing violence experienced during preg- the pregnancy of his female partner occurredJasinski / PREGNANCY AND DOMESTIC VIOLENCE 55 sooner than intended. Other researchers have tax family functioning, and the cumulative ef- also found similar results. Specifically, fect of multiple stressors can affect levels of Gazmararian and associates found that vio- marital discord. Stress may also affect the ability lence rates were highest for unwanted pregnan- to process information effectively and the selec- cies and lowest for intended pregnancies tion of particular conflict resolution behaviors (Gazmararian, Arrington, Bailey, Schwarz, & in given circumstances and may add to frustra- Koplan, 1999). Possible explanations for this tion and ultimately to violence. Violence tends pattern of behavior include jealousy of the un- to be higher when certain conditions are pres- born child and the perception that the preg- ent, such as a high level of conflict and stress in nancy would interfere with the woman’s role as the family, and intervening variables such as caretaker for her partner (Campbell et al., 1993; belief in the legitimacy of violence to deal with Campbell, Harris, et al., 1995). It is also possible family members who do wrong (Straus, 1980, that a pregnancy not planned by the male part- 1990b). In one study, for example, although ner might represent something that he could not there was no direct effect of pregnancy on risk control and therefore increases the risk for for violent victimization after controlling for violence. socioeconomic status, stressful life events, and Normative transitions associated with the age, life stressors were significantly associated entrance or exit into a social role (e.g., parent- with increased odds of IPV for Hispanic hood) may also increase the risk for victimiza- individuals only (Jasinski & Kaufman Kantor, tion as being associated with stress (Lavee, 2001). This suggests the possibility that other McCubbin, & Olson, 1987; Pearlin, Lieberman, stressful events co-occurring with pregnancy Menaghan, & Mullan, 1981). As such a transi- may contribute to the increased risk for wife tion, pregnancy, or the anticipation of parent- assault among Hispanic women. hood for both new and experienced parents, may increase the level of stress in the family and What Are the Consequences the risk for violence (Curry & Harvey, 1998). In of Pregnancy-Related Violence? addition to creating new strains, pregnancy or the birth of a child may intensify preexisting Aside from the disturbing fact that women strains such as low socioeconomic status. Sev- who are pregnant are physically and sexually eral studies have also found that young preg- abused, there are a number of consequences of nant women are more likely to have been pregnancy-related violence both for the unborn abused than older pregnant women (Hedin, child and for the pregnant Grimstad, Moller, Schei, & Janson, 1999; mother. These conse- One of the relatively Muhajarine & D’Arcy, 1999; Parker et al., 1994; quences include late entry consistent empirical Stewart & Cecutti, 1993), suggesting that the into prenatal care, low findings in the combination of pregnancy and youth may be birth weight babies, pre- research literature on particularly stressful. This is also consistent mature labor, unhealthy pregnancy-related with the domestic violence literature finding maternal behaviors, fetal violence is the delay that youth is a consistent risk factor for victim- trauma, and health issues of prenatal care ization. In addition, the cumulative effect of for the mother (See Table among victims of multiple stressors can affect parental percep- 3 for a summary of these violence (Dietz et al., tions of newborns, family environment, and at- consequences). Although 1997; Gazmararian titudes toward parenting (Fisher, Fagot, & Leve, a significant body of re- et al., 1995; Goodwin 1998) and levels of marital discord. A variety of search finds that violence et al., 2000; studies have found more life stress among during pregnancy is asso- McFarlane et al., physical child abusers (Chan, 1994; Conger, ciated with negative ma- 1992; Parker, 1993; Burgess, & Barrett, 1979), an association pre- ternal outcomes, it has Parker et al., 1994; dicted by most models of family violence. Stress also been suggested that Parker, McFarlane, associated with financial hardships and chronic many of the factors asso- Soeken, Torres, & poverty and unemployment has the potential to ciated with increased risk Campbell, 1993).
56 TRAUMA, VIOLENCE, & ABUSE / January 2004
TABLE 3: Consequences of Pregnancy-Related Violence be a risk factor for pregnancy complications
such as low birth weight babies and premature
Consequence Author, Year
labor.
Delayed prenatal care Dietz et al., 1997
Goodwin et al., 2000
Low birth weight infants. Although there is
Low birth weight Campbell et al., 1999
(bivariate only) general agreement that abuse is associated with
Currey & Harvey, 1998 delays in prenatal care, the same level of agree-
Fernandez & Krueger, 1999 ment does not exist in other areas of research on
Bullock & McFarlane, 1989
Parker et al., 1994 the outcomes of pregnancy-related violence. It
Premature labor Rachana et al., 2002 has been argued, for example, that battered
Cokkinides et al., 1999 women are more likely than nonbattered
Fernandez & Krueger, 1999
Shumway et al., 1999
women to give birth to preterm and low birth
Fetal trauma Rachana et al., 2002 weight infants (Bullock & McFarlane, 1989;
Jacoby, Gorenflo, Black, Campbell et al., 1999; Curry & Harvey, 1998;
Wunderlich & Eyler, 1999
Parker et al., 1994). In one study of the 100 pa-
Connolly et al., 1997
Renker, 1999 tients who were victims of domestic violence,
Berrios & Grady, 1991 16% had low birth weight babies compared to
Health issues for mother Cokkinides et al., 1999 6% of the 389 patients who were not domestic
Horrigan et al., 2000
Parker et al., 1994 violence victims (Fernandez & Krueger, 1999).
Other researchers have found that the percent-
age of victims with low birth weight babies was
twice as high as that of nonvictims (Bullock &
for victimization (e.g., youth, alcohol use, pov- McFarlane, 1989). In addition to the violence
erty) are the same factors associated with nega- experienced by pregnant women, low birth
tive maternal outcomes (Cokkinides et al., weight may also be associated with late entry
1999). into prenatal care as well as other unhealthy be-
haviors by the mother (e.g., smoking, poor nu-
Later entry into prenatal care. One of the rela- trition) (Bohn & Holz, 1996).
tively consistent empirical findings in the re- In contrast, there are also a number of studies
search literature on pregnancy-related violence that have not found any relationship between
is the delay of prenatal care among victims of violence and low birth weight infants. For ex-
violence (Dietz et al., 1997; Gazmararian et al., ample, Cokkinides et al. (1999) found that vio-
1995; Goodwin et al., 2000; McFarlane et al., lence was not significantly associated with low
1992; Parker, 1993; Parker et al., 1994; Parker, birth weight. Their study used the South
McFarlane, Soeken, Torres, & Campbell, 1993). Carolina PRAMS data from 6,143 women who
This is particularly relevant as one of the goals delivered live infants between 1993 and 1995.
of Healthy People 2010 is that 90% of pregnant Similarly, Shumway and associates’ (1999)
women will begin prenatal care in the first tri- study indicated that birth weight and gesta-
mester. Dietz and associates (1997), for exam- tional age at delivery did not vary significantly
ple, found that abused women were 1.8 times with a history of, or the degree of, violence expe-
more likely to delay prenatal care compared to rienced during pregnancy. Some researchers,
women who were not abused. Other research- however, have suggested that the findings of no
ers have found that abused women are twice as relationship between low birth weight and vio-
likely to begin prenatal care in their third tri- lence may be a function of confounding vari-
mester (McFarlane et al., 1992). In addition, one ables such as low socioeconomic status and
study found that 38% of women in abusive rela- poor nutrition (Bullock & McFarlane, 1989).
tionships registered for prenatal care later than Moreover, studies do not always control for
20 weeks gestation compared to 23% of the gestation length when looking at conse-
women who were not abused (Norton et al., quences such as low birth weight. Differences in
1995). Late entry into prenatal care per se may sample size and type as well as a lack of stan-Jasinski / PREGNANCY AND DOMESTIC VIOLENCE 57
dard cutoff points for what constitutes low birth have been experiencing violence and preg-
weight could also account for differences across nancy loss over a long period of time. Other
studies. researchers have also found an increased risk
for miscarriages among abused women (Berrios
Premature labor. In addition to low birth & Grady, 1991; Renker, 1999). In an investiga-
weight, there is also contradictory evidence re- tion of pregnancy complicated by trauma, Con-
garding the relationship between violence and nolly and associates (1997) found that a greater
premature labor. Berenson and associates percentage of placental abruptions (separation
(1994), for example, found that assaulted of placenta from uterine wall) were related to
women were almost twice as likely to experi- domestic violence compared to motor vehicle
ence preterm labor compared to those who were accidents. Based on these results, the authors
not assaulted. Similarly, Shumway et al. (1999) suggested that violence might be directed
found that women who were abused were 2.3 against the pregnancy. Violence has also been
times more likely to experience preterm labor. associated with fetal injury and death (Bohn,
In addition, an increased risk for preterm labor 1990; Webster et al., 1996).
was associated with more serious violence.
Fernandez and Krueger’s (1999) study found Unhealthy maternal behaviors. In addition to
that of the 100 patients who were victims of do- the direct effects of violence on the health and
mestic violence, 22% had preterm deliveries well-being of the unborn child, violence may
compared to only 9% of the 389 patients who also indirectly contribute to negative conse-
were not victims of domestic violence. Other re- quences by increasing the risk for unhealthy
searchers have found the risk of preterm labor maternal behaviors. For example, several stud-
to be as much as 5 times greater among victims ies have found that abused women are more
of severe abuse compared to women who were likely to smoke than women who are not
not abused (Shumway et al., 1999). There are abused (Cokkinides & Coker, 1998; Cokkinides
also several studies that have not found a rela- et al., 1999; Grimstad et al., 1997; Martin et al.,
tionship between violence and premature labor 1996; McFarlane & Parker, 1996; Wiemann et al.,
(Cokkinides et al., 1999; Grimstad, Schei, Backe, 2000). In addition, much of the same research
& Jacobsen, 1997). Differences in empirical find- has also found an association between violence
ings may be due to a variety of factors, including victimization and alcohol and drug use. Martin
the failure to control for other variables that et al.’s (1996) study of 2,092 prenatal patients in
may contribute to preterm labor as well as dif- North Carolina found that during pregnancy,
ferences in research design. victims were more likely to smoke, drink, and
use drugs. Moreover, after controlling for de-
Fetal trauma. Perhaps one of the most serious mographic factors, victims were more likely to
negative consequences of pregnancy-related vi- be in the more severe substance abuse catego-
olence is fetal trauma (e.g., miscarriage, sponta- ries during pregnancy than women who were
neous abortion, etc.). Research focusing on this not victims of violence. In one of the few studies
type of negative outcome has been relatively with a diverse sample, Berenson and associates
consistent in its findings; experiencing abuse (1991) found that drug use was related to batter-
puts the unborn baby at great risk. Jacoby and ing for White and Black women in their sample
associates’ (1999) study of 100 women receiving but not for Hispanic women. These unhealthy
prenatal care found that women who experi- behaviors may be associated with negative con-
enced any form of abuse were significantly sequences for the unborn child as well as for the
more likely to miscarry (42.3% vs. 16.2%) mother.
(Jacoby, Gorenflo, Black, Wunderlich, & Eyler,
1999). In addition, they found an association be- Health issues for mothers. In addition to the
tween current abuse and at least one spontane- negative health consequences experienced by
ous abortion in the woman’s obstetric history. the unborn child, several studies have found
The authors suggest that these women may that violence is associated with negative health58 TRAUMA, VIOLENCE, & ABUSE / January 2004
consequences for the mother as well. Moreover, Issues for Future Research:
many of these health issues are also relevant for What Do We Need in Future Studies
the health of the unborn child. Bohn and Holz’s on Pregnancy-Related Violence?
(1996) review of the literature identifies health
issues such as an unhealthy diet, severe Although there is a substantial body of re-
postpartum depression, and breastfeeding dif- search focused on pregnancy-related violence,
ficulties that are associated with victimization. there are still many areas in which there is little
Other researchers have empirical evidence. Certainly one consideration
In addition, although f o un d t h a t a bus e d for future research is the inclusion of more eth-
there are a number women suffer from more nically diverse samples to make comparisons
of studies using stress and receive less among ethnic groups possible. Until recently,
primarily hospital support from their part- these types of samples have been virtually non-
samples, ner and others (Curry & existent in the literature. Sample diversity is not
comprehensive Harvey, 1998). In addi- limited, however, to racial or ethnic groups.
research using tion, maternal health is- Many hospital- or clinic-based studies have
probability samples sues such as severe de- used primarily economically disadvantaged
and longitudinal pression (Horrigan et al., samples, making generalizations about all so-
designs are 2000), lower self-esteem cial classes impossible. Future research should
needed to make (Curry & Harvey, 1998), make attempts to include individuals from all
more definitive kidney infections (Cok- social classes. In addition, although there are a
conclusions about kinides et al., 1999), poor number of studies using primarily hospital
the relationship weight gain (Parker et al., samples, comprehensive research using proba-
between pregnancy 1994), anemia (Parker bility samples and longitudinal designs are
and wife assault et al., 1994), and first or needed to make more definitive conclusions
and to understand second trimester bleed- about the relationship between pregnancy and
the array of ing (Parker et al., 1994)
wife assault and to understand the array of fac-
factors involved have all been associated
tors involved and the patterns of relationship
and the patterns with violence victimiza-
violence. Clinic- and hospital-based samples
of relationship tion. Other researchers
are useful for understanding the contextual in-
violence. have focused on the inter-
formation about pregnancy-related violence as
val between pregnan-
well as the consequences of such violence, but
cies, finding that victims
of abuse tend to have a without a comparison group of women who are
very short interval between pregnancies (termed not pregnant, these types of studies cannot
rapid repeat pregnancies) (Jacoby et al., 1999; answer questions regarding risk for victimiza-
Parker et al., 1994). Unlike the body of research tion among women as a group. Finally, there are
that is fairly consistent in its findings regarding a diverse number of disciplines involved in
many maternal health issues, there are mixed pregnancy-related violence research including
findings with regard to cesarean deliveries. For such fields as sociology, psychology, criminal
example, Cokkinides and associates (1999), in justice, nursing, education, and public health,
their study of 6,143 women from the South to name a few. Each of these disciplines offers
Carolina PRAMS, found that abused women a unique perspective on this research topic;
were 1.5 times more likely to deliver by C- however, it is relatively uncommon for re-
section. In contrast, Berenson et al. (1994) found searchers in these fields to work together to de-
no relationship between victimization and ce- velop a multidisciplinary research project.
sarean delivery among their sample of 384 poor Such collaboration could prove to be ex-
pregnant women examined at a low-risk prena- tremely beneficial in increasing the range of
tal clinic in Galveston, Texas. It is possible, how- knowledge on the subject of pregnancy-related
ever, that these observed differences could be a violence and ultimately working toward its
result of different sample types. reduction.Jasinski / PREGNANCY AND DOMESTIC VIOLENCE 59
Issues for Practice: sential among women presenting with trauma
What Can Health Care Providers Do? and non-trauma-related symptoms in hospital
emergency departments (Dienemann et al.,
For many women, pregnancy is often the first 1999). This may be particularly relevant for
point of entry into the health care system and pregnant women. Connolly et al. (1997) sug-
perhaps the first contact with a helping profes- gested that injury prevention programs should
sion. Consequently, professionals who deal be incorporated into all prenatal care programs
with pregnant women and new mothers are in a because not all traumas may be correctly classi-
unique position to screen for marital violence fied as domestic violence. In fact, because a pat-
and initiate intervention (Sampselle, Petersen, tern of greater health care use has been identi-
Murtland, & Oakley, 1992). Evidence also sug- fied among victims of physical or sexual
gests that women want their health care provid- violence (Koss, Koss, & Woodruff, 1991), some
ers to inquire about violence victimization view universal screening of women seeking any
(Webster, Stratigos, & Grimes, 2001). Several health care as essential to comprehensive care.
studies have suggested that screening questions The role of health care professionals does not
should be direct (Naumann, Langford, Torres, end at the screening stage, however. Pregnant
Campbell, & Glass, 1999; Norton et al., 1995). women who are screened for previous violence
Naumann and associates (1999), for example, in their relationship should be provided with
found that although women often find it diffi- information about available services if they
cult to start a conversation about abuse, they should need them either during or after the
will answer direct questions. Others have sug- child is born. Health care providers should also
gested that the very process of assessment can provide f ollow - up s e rvice s t o w om e n
be just as important as a particular form of postpartum in order to prevent any reoccur-
assessment, as it acknowledges that violence rence of violent behavior.
against pregnant women is a very serious issue Although much of the research on pregnancy-
(Parker, McFarlane, Soeken, Silva, & Reel, 1999). related violence suggests the importance of the
Guidelines for assessment have been devel- health care provider in prevention and inter-
oped, however, based on empirical research vention, there is also some evidence to suggest
outside the clinical field. This research suggests that health care professionals do not receive
that the screening process should include thor- enough information and training (Naumann
ough assessments of pregnancy-related stress- et al., 1999). Horan and associates (1998), for ex-
ors, including areas such as mistimed or un- ample, found that only two thirds of the obste-
planned pregnancies. Information from both tricians and gynecologists they studied re-
members of the couple should also be included, ported that they screened patients for domestic
as research has demonstrated the importance of violence (Horan, Chapin, Klein, Schmidt, &
both individuals wanting the pregnancy. In ad- Schulkin, 1998). Among those doctors who
dition, screening should include an assessment screened, one quarter reported that they were
of stressors in addition to pregnancy that may not very confident about their ability. They also
contribute to risk for experiencing wife assault. found that doctors who have been trained in the
Research examining the association between last 15 years conducted domestic violence
stress and violence suggest that interventions screenings at higher rates compared to those
that reduce couple stress and aid in role transi- trained earlier. In addition, female doctors were
tion may have the potential to decrease marital more likely than male doctors to screen for vio-
discord and violence. Interventions may also in- lence at the first prenatal visit. Similarly,
clude a careful assessment of family risk mark- Molliconi and Runyan (1996) found that family
ers (e.g., family of origin exposure, substance practitioners asked fewer than 7% of their fe-
abuse), family stressors, and current conflict male patients about abuse. Domestic violence
management strategies, along with providing training also may not occur while doctors are in
individuals with linkages to appropriate ser- medical school but later during the residency
vices. Screening for domestic violence is also es- period (Horan et al., 1998). These findings are60 TRAUMA, VIOLENCE, & ABUSE / January 2004
TABLE 4: Clinical Recommendations for Assessment of there is another important reason for more com-
Pregnancy-Related Violence
plete training for health care professionals. In a
Recommendation Author, Year recent National Institute of Justice research
brief, Isaac and Enos (2001) discussed the im-
Assessment of stressors Jasinski, 2001
portance of documenting abuse for legal pro-
(pregnancy and non- Jasinski & Kaufman Kantor,
pregnancy-related) 2001 ceedings that may take place. Medical docu-
Curry, 1998 mentation can be used to substantiate assertions
Screen all women seeking Dienemann et al., 1999 of abuse, to obtain protective relief in the form of
health care for domestic Connolly et al., 1997
violence Koss, Koss, & Woodruff, a restraining order, and/or to be eligible for cer-
1991 tain exemptions or statuses related to housing,
Ask multiple direct questions Norton et al., 1995 insurance, and financial assistance. Isaac and
to women about domestic
violence
Enos suggested that health care providers can
Training/education for health Naumann et al., 1999 be of most assistance legally by improving their
care providers Horan, Chapin, Klein, record keeping. Specifically, they suggested that
Schmidt, & Schulkin, 1998
only factual information, rather than summary
Molliconi & Runyan, 1996
Gremillion & Evins, 1994 statements or conclusions, should be docu-
Parsons, Goodwin, & mented; photographs of all injuries should be
Peterson, 2000 taken; any patient statements should be clearly
indicated as such; statements about the reliabil-
ity of the patient should be avoided; and all doc-
consistent with other research that has found
umentation should be legible. Although this is
that most medical textbooks do not contain
helpful information, physicians may still fear le-
much information on domestic violence. For ex-
gal reprisal and the possibility of insurance loss,
ample, Parsons and Moore (1997) found that
and they may be uncertain about the policies of
only 37.5% of obstetric and gynecology texts
their office (Gremillion & Evins, 1994).
and primary care texts contained information
on domestic violence. More promising, almost
two thirds of nursing texts contained domestic SUMMARY AND CONCLUSIONS
violence information. Evidence also exists sug-
Agrowing body of research suggests that vio-
gesting that training can be effective in increas- lence during pregnancy has detrimental conse-
ing the screening rate for domestic violence quences for both the mother and the unborn
(Janssen, Holt, & Sugg, 2002). Gremillion and child. Concerns about these effects have led
Evins’s (1994) review of the literature suggests both researchers and practitioners to take a
that, in addition to these training barriers, phy- closer look at pregnancy-related violence in
sicians must also deal with contemporary social terms of risk factors and consequences as well as
issues including societal tolerance for abuse, de- the physician’s role in prevention and interven-
sensitization, and power inequities in relation- tion efforts. Although differences in research
ships. They must also come to terms with their designs and assessments have made it difficult
own personal factors that might include victim- to definitively conclude that pregnant women
ization, gender bias, and ideal notions of the are at a greater risk for IPV compared to women
family unit. Additional barriers cited by who are not pregnant, the consequences for
Molliconi and Runyan (1996) include the physi- pregnant victims remain serious. What is per-
cian’s personal knowledge of either the patient haps most disconcerting, however, is that many
or her partner. Table 4 summarizes research- practitioners who come into contact with these
based clinical recommendations for assessment victims have not been exposed to any or enough
of pregnancy-related violence. training so that they can ask the right questions
In addition to the obvious improvements in and offer assistance to victims. Moreover, even
health care delivery for pregnant victims of do- those who have received some training express
mestic violence and the importance of preven- concern about their own ability to properly
tion for both the unborn child and the mother, assess the existence of IPV among their patients.Jasinski / PREGNANCY AND DOMESTIC VIOLENCE 61
Certainly the first steps have been taken. Re- greatly from joint projects that unite researchers
searchers are continuing to investigate the dy- with practitioners with the ultimate goal of
namics of pregnancy-related violence and, as healthy women, healthy babies, and violence-
suggested by some of the studies of practi- free relationships.
tioners, are taking a close look at how they inter-
act with patients. Future work would benefit
IMPLICATIONS FOR PRACTICE, POLICY, AND RESEARCH
Practice Research
• All women should be screened for domestic violence • Comparisons between studies are difficult due to in-
in health care settings. consistent use of terminology, definitions, and time
• Health care professionals should be trained in the ar- periods of exposure.
eas of domestic violence dynamics and screening • Research samples need to be ethnically and econom-
techniques. ically diverse.
• Health care providers should work to improve re- • Probability samples and longitudinal research de-
cord keeping so that medical documentation can be signs are needed to better understand the relation-
more useful in legal proceedings. ship between pregnancy and violence.
• Interdisciplinary research teams are needed to pro-
vide a more holistic understanding of the preg-
nancy-violence relationship.
REFERENCES
Berenson, A. B., Stiglich, N. J., Wilkinson, G. S., & Anderson, Campbell, J. C., Oliver, C. E., & Bullock, L. F. (1993). Why
G. D. (1991). Drug abuse and other risk factors for physi- battering during pregnancy? AWHONNS Clinical Issues
cal abuse in pregnancy among White non-Hispanic, in Perinatal and Women’s Health Nursing, 4(3), 343-349.
Black, and Hispanic women. American Journal of Obstet- Campbell, J. C., Oliver, C. E., & Bullock, L. F. (1998). The
rics and Gynecology, 164, 1491-1499. dynamics of battering during pregnancy: Women’s
Berenson, A. B., Wiemann, C. M., Wilkinson, G. S., Jones, explanations of why. In J. C. Campbell (Ed.), Empower-
W. A., & Anderson, G. D. (1994). Perinatal morbidity ing survivors of abuse (pp. 81-89). Thousand Oaks, CA:
associated with violence experienced by pregnant Sage.
women. American Journal of Obstetrics and Gynecology, Campbell, J. C., Poland, M. L., Waller, J. B., & Ager, J. (1992).
170, 1760-1769. Correlates of battering during pregnancy. Research in
Berrios, D. C., & Grady, D. (1991). Domestic violence: Risk Nursing & Health, 15, 219-226.
factors and outcomes. Western Journal of Medicine, 155, Campbell, J. C., Pugh, L. C., Campbell, D., & Visscher, M.
133-135. (1995). The influence of abuse on pregnancy intention.
Bohn, D. K. (1990). Domestic violence and pregnancy: Women’s Health Issues, 5(4), 214-223.
Implications for practice. Journal of Nurse-Midwifery, Campbell, J. C., Torres, S., & Ryan, J. (1999). Physical and
35(2), 86-98. nonphysical partner abuse and other risk factors for
Bohn, D. K., & Holz, K. A. (1996). Health effects of child- low birth weight among full term and preterm babies: A
hood sexual abuse, domestic battering, and rape. Jour- multiethnic case-control study. American Journal of Epi-
nal of Nurse-Midwifery, 41(6), 442-456. demiology, 150(7), 714-726.
Bullock, L., & McFarlane, J. (1989). The birth-weight/ Centers for Disease Control and Prevention. (1999).
battering connection. American Journal of Nursing, 89(9), PRAMA 1996 Surveillance Report. Atlanta, GA: Division
1153-1155. of Reproductive Health, National Center for Chronic
Campbell, J. C. (1986). Nursing assessment for risk of Disease Prevention and Health Promotion, Centers for
homicide with battered women. Advances in Nursing Disease Control and Prevention.
Science, 8(4), 36-51. Chan, Y. C. (1994). Parenting stress and social support of
Campbell, J. C., & Alford, P. (1989). The dark consequences mothers who physically abuse their children in Hong
of marital rape. American Journal of Nursing, 89, 946-949. Kong. Child Abuse & Neglect, 18, 261-269.
Campbell, J. C., Harris, M. J., & Lee, R. K. (1995). Violence Cokkinides, V. E., & Coker, A. L. (1998). Experiencing
research: An overview. Scholarly Inquiry for Nursing physical violence during pregnancy: Prevalence and
Practice, 9(2), 105-126. correlates. Family and Community Health, 20(4), 19-38.You can also read
