Integrating National Violent Death Reporting System Data into Maternal Mortality Review Committees

Abstract

Background:

With the Maternal Mortality Review Information Application (MMRIA) data system, the Centers for Disease Control and Prevention (CDC), alongside Maternal Mortality Review Committees (MMRCs), are developing comprehensive and uniform data collection to eradicate preventable maternal deaths. However, MMRIA is primarily focused on pregnancy-related deaths, and not pregnancy-associated deaths. Currently, the National Violent Death Reporting System Restricted Access Data (NVDRS-RAD) on pregnancy-associated homicides and suicides are not included in MMRIA and by extension the work of most MMRCs. This study examined the NVDRS-RAD data from 2014 to 2017 and argues that the data for pregnancy-associated maternal deaths should be integrated into the work of MMRCs.

Methods:

A secondary data analysis of the NVDRS-RAD from 2014 to 2017 was conducted. Pregnancy-associated mortality was identified using data available within the NVDRS-RAD and categorized according to period of pregnancy and manner of death. Descriptive statistics and risk ratios were calculated. Chi-square tests were also calculated.

Results:

The results indicate that pregnancy and the postpartum period show increased risk for homicide and suicide. Pregnant women were found to be five times more likely to die by homicide than their nonpregnant peers who died by violent means. The relationships between periods of pregnancy and manner of death were all found to be significantly associated although the association was weak.

Conclusions:

Integrating National Violent Death Reporting System data on pregnancy-associated deaths into MMRIA would improve the efficacy of MMRCs and address the intertwined risk factors driving the racial disparities of the United States' maternal mortality rate.

Introduction

Maternal Mortality Review Committees (MMRCs) examine maternal deaths through a multidisciplinary approach and issue important findings on how to reduce maternal mortality. For this reason, the Centers for Disease Control and Prevention (CDC) has made MMRCs a central part of their coordinated efforts to address the maternal mortality crisis in the United States. As of May 2021, there were 49 MMRCs nationwide (46 states and 3 cities) that investigate pregnancy-associated and pregnancy-related deaths. The term “pregnancy-associated death” came out of the creation of the Pregnancy Mortality Surveillance System in 1986. It was intended as an umbrella term to capture all deaths during or within 1 year of pregnancy, regardless of cause, as opposed to the term pregnancy-related death. The CDC defines pregnancy-related death as “the death of a woman during pregnancy or within 1 year of the end of pregnancy from a pregnancy complication, a chain of events initiated by pregnancy, or the aggravation of an unrelated condition by the physiologic effects of pregnancy.”³⁶

With the advent of the Maternal Mortality Review Information Application (MMRIA) in 2017, MMRCs now have a centralized database to ensure uniform data collection and review processes. Review to Action, a resource developed to highlight and promote the work of state-level MMRCs, insists that “The scope of cases for committee review is all pregnancy-associated deaths or any deaths of women with indication of pregnancy up to 365 days, regardless of cause (i.e., motor vehicle accidents during pregnancy, motor vehicle accidents postpartum, suicide, homicide).¹” However, it acknowledges that larger state-level MMRCs often “choose to limit review of deaths to those where the cause of death is classified by one of the obstetric cause of death codes on the death certificate.”¹

Other scholars have noted the challenges to accurate, systematic, and comprehensive maternal mortality data gathering.² We add to this literature by highlighting how integrating the National Violent Death Reporting System (NVDRS) data into MMRIA would give MMRCs a more accurate picture of the scope of maternal mortality. Currently, MMRIA does include information on violent-related deaths, and NVDRS data are sometimes pulled in, but not through direct communication between the two systems. Specifically, the Review to Action guidance advises pulling information on cases from as many of sources as possible and feasible and NVDRS is included as one potential option. However, as indicated in the abstractor manual on the Review to Action site, depending on the authorities under which the MMRC operates, some sources of information could be restricted.³⁸ However, accessing NVDRS data in some jurisdictions requires one individual with access to NVDRS pulling the information to share with the MMRC abstractor. This can increase the time it takes to pull information and review the case, slowing down the impactful changes that could be made. We advocate that direct communication between the two systems would eliminate potential roadblocks to access and be more efficient, especially as both systems have variables from pregnancy up to 1 year postpartum.

The NVDRS is a state-based surveillance system, which compiles information on all types of violent deaths in all settings for all ages. Over 600 unique data elements are collected, from relationship problems to toxicology results to life stressors, such as problems with jobs, finances, or health problems.³ Collecting data for the NVDRS began in 2002 from 6 states and by 2017 had been expanded to include 40 states, Puerto Rico, and the District of Colombia. The states included in the analysis are displayed in Table 1 in the order they were added to NVDRS. The NVDRS also includes Restricted Access Data (NVDRS-RAD), which contains more confidential variables than the NVDRS and short narratives to describe the circumstances related to the death.³

Table 1.

States Included in National Violent Death Reporting System by Year

Year	States added
2002	MA, MD, NJ, OR, SC, VA
2003	AK, CO, GA, NC, OK, RI, WI
2004	KY, NM, UT
2009	MI, OH
2014	AZ, CT, HI, IA, IL, IN, KS, ME, MN, NH, NY, PA, VT, WA
2016	AL, CA, DE, DC, LA, MO, NE, NV, Puerto Rico, WV

In an effort to get a more accurate picture of the scale of maternal mortality in the United States, the U.S. Standard Certificate of Death was revised to include a pregnancy checkbox in 2003, but states were slow to adopt and ramp up use. A 2016 quality assurance pilot in four states (Ohio, Michigan, Louisiana, and Georgia) found a 21% false-positive rate.⁴ A similar study of the pregnancy checkbox in the state of Maryland found that while 98.1% of the maternal deaths directly associated with pregnancy were identified, only 46.7% of deaths caused by pregnancy-associated homicide, suicide, and other causes were identified.⁵

The lack of recognition of the prevalence of pregnancy-associated homicide and suicide through the pregnancy checkbox raises concerns. A systematic review of studies on pregnancy-associated homicide and suicide found that pregnancy-associated homicide continues to be a leading cause of maternal death, and suicide is a significant contributor to pregnancy and postpartum death.⁶ Other studies have found similar results, highlighting the role homicide plays in contributing to maternal mortality.^7
–9 However, these findings are not without controversy. There is evidence that suggests pregnancy and the postpartum period do not pose any greater risk than for other women of reproductive age.²⁶ Much of these discrepancies can be accounted for by the heterogeneity of the members of different state-level MMRCs. For instance, the presence or absence of social workers, mental and behavioral health experts, and community members significantly shape the investigative focus of the reviews.

A similar pattern of the unrecognized impact of pregnancy-associated suicide is seen when it comes to reviewing violence as a contributor to maternal death. A review of self-harm and overdose death studies during the perinatal period found research citing suicide as a leading contributor to maternal death.¹⁵ Additionally, research has found that during the perinatal period, the rate of diagnosed mental health disorders begins to rise.²⁶ This highlights another critical time when intervention could occur if these deaths are reviewed and underscores the importance of having mental health experts and substance use experts as members of MMRCs. Recently, the CDC allocated funding to support six MMRCs (Massachusetts, North Carolina, Ohio, Tennessee, Utah, and Wisconsin) to review all pregnancy-associated overdose deaths in their states.³⁷

Violence, whether it results in homicide or suicide, is a clear contributor to maternal death. To effectively understand and address maternal mortality, violence must be reviewed to the same extent that medical causes of maternal death are. Our research adds to a growing literature that illustrates how the pregnancy-related versus pregnancy-associated dichotomy is a false binary. Pregnancy-associated homicide, suicide, and overdose are all intimately linked to broader social and structural determinants of health. In addition to an individual focus on morbidity, our findings support Kramer et al.'s Call to Action for MMRCs, to employ a health equity framework³¹ to understand how violence prevention is also a maternal mortality issue.⁷

Methods

A secondary data analysis of the National Violent Death Reporting System Restricted Access Data (NVDRS-RAD) from 2014 to 2017 was conducted. IRB approval was not required for this study. Descriptive statistics were calculated for pregnancy-associated homicides and suicides. A pregnancy-associated homicide or suicide was defined as the death of a woman due to homicide or suicide who was pregnant at the time of death or had been pregnant within the previous 365 days. Risk ratios (RRs) for pregnancy-associated homicide and suicide by demographic characteristics were also conducted. RRs were calculated by taking the risk of the suicide, homicide, intimate personal violence, or interpersonal violence among different pregnancy statuses over the risk of these occurrences among nonpregnant/nonpostpartum women. Furthermore, chi-square analysis was conducted to determine the significance of these relationships. An alpha value of 0.05 was used for all statistical tests. The intent of this analysis was to see if certain demographic characteristics and certain pregnancy statuses put women at greater risk for pregnancy-associated violence when compared with their nonpregnant/nonpostpartum counterparts who experienced violent death. SPSS, a statistical software package, version 27 was used to conduct all elements of the analysis.

Pregnancy status was divided into three categories: pregnant, early postpartum (1–42 days postdelivery), and late postpartum (43 days–1 year postdelivery). All women who were indicated in the dataset as pregnant, not otherwise specified, were included in the pregnant category. There were 11,889 cases where pregnancy status was not known. To address this, the narrative descriptions were examined for words relating to pregnancy such as pregnant, pregnancy, breastfeeding, infant, etc. After review, no additional cases were identified; however, these cases were removed from the analysis. Other demographic characteristics included race (White, Black, other), ethnicity (Hispanic, Non-Hispanic), age, education, marital status, and any known history of or current exposure to interpersonal violence and intimate partner violence, respectively. The other category of race included Asian/Pacific Islander, American Indian/Alaskan Native, and persons identifying as one or more races or other. Age was also divided into age categories in the same manner used by previous research,¹⁰ which was less than 20, 20–24, 25–29, 30–34, and greater than or equal to 35. Marital status was categorized as either married or not married. The not married category included those who were divorced or widowed. Education was divided into three categories: less than high school, high school diploma or graduate equivalency degree, and more than high school. Each of these categorizations was done to align with previous research^7,10 to ensure accurate comparisons.

Additional variables utilized in this study were interpersonal violence and intimate partner violence. The interpersonal violence variable assessed if the victim had experience violence in the past month that was distinct and occurred before the violence that killed the victim. The intimate partner violence variable identified cases where the victim's death was directly related to immediate or ongoing conflict or violence between intimate partners. When reviewed for missing data, it was found that there were no missing data for these variables. Additionally, the overall missing data was low throughout the entire sample with 1.9% being the highest amount of missing data for any given variable.

Results

During the 4-year time period of 2014–2017, 19,075 women died by suicide or homicide. Approximately 2.7% (n = 520) of these deaths occurred during pregnancy or the 1 year following pregnancy. The majority (49.4%) of these women were pregnant at the time of their death (n = 257). Around 12.1% (n = 63) of the women were early postpartum and 38.4% (n = 200) were late postpartum. Demographic characteristics of the sample are displayed in Table 2.

Table 2.

Demographic Characteristics of Female Decedents by Pregnancy Status, 2014–2017

	Total (n = 19,075)		Pregnant (n = 257)		Early postpartum (n = 63)		Late postpartum (n = 200)
	N	%	N	%	n	%	n	%
Race
White	14,585	76.5	142	55.3	43	68.3	142	71.0
Black or African American	2,865	15.0	94	36.6	13	20.6	40	20.0
Other	1,578	8.3	20	7.8	7	11.1	18	9.0
Ethnicity
Non-Hispanic	17,277	90.6	213	82.9	55	87.3	186	7.0
Hispanic	1,642	8.6	40	15.6	6	9.5	14	93.0
Age
≥35	10,311	54.7	57	22.2	19	30.2	40	20.0
30–34	2,353	12.5	38	14.8	12	19.0	33	16.5
25–29	2,201	11.7	62	24.1	14	22.2	42	21.0
20–24	2,077	11.0	74	28.8	11	17.4	68	34.0
<20	1,925	10.2	26	10.1	7	11.1	17	8.5
Education level
Less than high school	3,360	18.6	54	21.0	13	20.6	38	19.0
High school graduate or GED	6,244	34.5	110	42.8	18	28.6	70	35.0
Greater than high school	8,472	46.9	79	30.7	26	41.3	84	42.0
Marital status
Married	5,714	30.2	50	19.5	19	30.2	75	37.5
Not married	13,206	69.8	205	80.9	42	66.7	125	62.5
Manner of death
Homicide	5,080	26.6	150	58.4	13	20.6	81	40.5
Suicide	13,995	73.4	107	41.6	50	79.4	119	59.5

GED, graduate equivalency degree.

A chi-square test of independence was performed to examine the association between pregnancy status and manner of death. The sample included 248 pregnant, 63 early postpartum, and 200 late postpartum women, these frequencies were significantly different X² (3, n = 6,070) = 251.8, p ≤ 0.001 (Table 3). These results indicate a significant association between pregnancy status and manner of death. Moving forward, the results will focus first on pregnancy-associated homicide followed by pregnancy-associated suicide.

Table 3.

Chi-Square Results for Pregnant Versus Nonpregnant, by Manner of Death, Intimate Partner Violence, and Interpersonal Violence

Variable	n	X² statistic	p
Manner of death (suicide or homicide)	6,970	235.7	<0.001
Intimate partner violence-related homicide	6,970	165.3	<0.001
Interpersonal violence-related suicide	6,970	33.5	<0.001

Pregnancy-associated homicide

Of the over 19,000 women who died by homicide or suicide, 5,080 died by homicide. Pregnant or postpartum women accounted for 4.8% (n = 244) of these deaths. The majority occurred during pregnancy followed by the late postpartum period and then early postpartum period. Pregnant women, who were victims of homicide, were predominately Black or African American, non-Hispanic, 20–29 years of age, high school graduates, and not married. In ∼53% of pregnancy-associated homicides where the victim was pregnant at the time of death, immediate or ongoing intimate partner violence contributed to the death. In a small percentage of cases, the victim experienced interpersonal violence in the preceding month that was distinct and occurred before the violence that killed the victim. Pregnancy-associated homicides during the early postpartum period accounted for a small percentage (<1%, n = 13). Female descendants in this category tended to be non-Hispanic Whites, age 35+, with greater than high school education, and not married. Homicides occurring in the late postpartum period made up ∼1.6% (n = 81) of all pregnancy-associated homicides. Victims tended to be White, non-Hispanic, 20–24 years of age, with a high school education, and not married. In 61.7% (n = 50) of cases, intimate partner violence was involved.

The RRs for the probability that pregnant, early postpartum, and late postpartum women were more likely to be victims of homicide compared with their nonpregnant/postpartum peers are displayed in Table 3. Findings suggest that women in the pregnant category at the time of death are 5.07 times more likely to die by homicide compared with those who are not in the pregnant category in this sample. Smaller but similar trends were seen for early postpartum (RR = 1.04) and late postpartum (RR = 2.61) women.

Since intimate partner violence was involved in many of the cases of pregnancy-associated homicide, this relationship was further investigated. RRs found that pregnant women were 3.5 times as likely to die from an intimate partner violence (IPV)-related homicide as their nonpregnant counterparts. The late postpartum period was identified also as a time of greater risk with late postpartum women having a RR of 2.77. With both time periods showing heightened risk, these may be points for potential intervention to reduce the risk for pregnant and late postpartum women. Additionally, a chi-square test of independence was run to examine the association between intimate partner violence-related homicide and pregnancy status. The results indicate there is a significant association between intimate partner violence-related homicide and pregnancy status X² (3, n = 6,970) = 165.3, p ≤ 0.001 (Table 3). This suggests there are more intimate partner violence-related homicides in the pregnancy and late postpartum periods than would be expected if there were no association between the two variables. However, the association between the two variables is weak (ϕ_c = 0.172).

Pregnancy-associated suicide

Suicide accounted for 13,995 female deaths reported to the NVDRS-RAD during 2014–2017. Of these suicides, 276 were pregnancy associated. Most of the deaths occurred during the late postpartum period followed by during pregnancy and then the early postpartum period. In each pregnancy status category, women tended to be White, non-Hispanic, and not married. Pregnancy-associated suicide most frequently occurred among those 35 years of age or older, with greater than high school education. Pregnancy-associated suicide was almost never associated with immediate or ongoing intimate partner violence. However, in 6.5%, 4%, and 6.7% of pregnant, early postpartum, and late postpartum suicides, respectively, the NVDRS data reported that women had experienced interpersonal violence in the preceding month.

Since interpersonal violence was a common experience among pregnant, early, and late postpartum women who died by suicide, a chi-square test of independence was conducted. The results indicate there is a significant association between interpersonal violence and pregnancy-associated suicide. The association between these variables was significant, X² (3, n = 16,970) = 33.5, p ≤ 0.001 (Table 3). While this suggests that pregnancy-associated suicide at any time during pregnancy or up to 1 year postpartum is more likely to occur than among nonpregnant or postpartum women, the relationship is weak (ϕ_c = 0.97).

RRs for whether the probability that pregnant, early postpartum, and late postpartum women were more likely to die by suicide compared with their nonpregnant/postpartum peers are displayed in Table 4. The results indicate a reduction in risk between pregnant and late postpartum women dying by suicide compared with their nonpregnant/nonpostpartum counterparts. For early postpartum women the results show a slight reduction in risk (0.96) for dying by suicide compared with their nonpregnant/nonpostpartum counterparts.

Table 4.

Risk Ratios for Manner of Death and Intimate and Interpersonal Violence by Pregnancy Status

Pregnancy status	Homicide		Suicide		Intimate partner violence		Interpersonal violence
Pregnancy status	RR	CI	RR	CI	RR	CI	RR	CI
Pregnant	5.07	4.35–7.26	0.20	0.16–0.25	3.50	2.87–4.27	0.20	0.97–0.41
Early postpartum	1.04	0.57–1.91	0.96	0.52–1.76	0.70	0.27–1.83	0.31	0.78–1.26
Late postpartum	2.61	1.98–3.44	0.38	0.29–0.50	2.77	2.15–3.58	0.20	0.10–0.39

CI, confidence interval; RR, risk ratio.

Discussion

Using the NVDRS-RAD, this analysis examines the relationship between violence and maternal deaths and articulates the need to review pregnancy-associated deaths in a more robust way. The results indicate that pregnancy and the postpartum period show increased risk for pregnancy-associated homicide and suicide, which coincides with other studies.^7,8,11,12,29 Pregnant women are five times more likely to die by homicide than their nonpregnant peers who died by violent means. With both the late postpartum and pregnancy periods showing heightened risk of IPV-related homicide, these may be points for potential intervention to reduce the risk for pregnant and late postpartum women. Other studies have found a higher risk of suicide among pregnant and postpartum women.^13,14 This difference could be due to sample size. However, research has identified several barriers to collecting accurate information on maternal deaths due to suicide suggesting this could be more of an issue than what the data show.³⁰ Therefore, to effectively address maternal mortality in the United States, pregnancy-associated maternal deaths caused by violent or self-inflicted harm must be thoroughly reviewed. To do this, MMRCs should seek to identify all contributing factors at the individual, provider, facility, system, and community levels. Directly linking NVDRS data with MMRIA data would allow for review that is more efficient and could reduce barriers to seeing the full picture of what occurred during a violent maternal death.

Our results also show IPV was involved in 61% of pregnancy-associated homicides, which is similar to the results of other studies.^7,8,16 Furthermore, IPV was found to have a weak but significant association with pregnancy-associated homicide. Interpersonal violence in the preceding month was also found to have a weak but significant association with pregnancy-associated suicides in all pregnancy statuses. Other research has also found a similar relationship¹⁷ as well as connected poor maternal and fetal outcomes with interpersonal violence.^11,18 Interestingly, our results showed a reduction in risk between the pregnancy/postpartum and nonpregnancy/nonpostpartum periods for women who previously experienced interpersonal violence. This relationship should be investigated in future studies. Taken together, there is a clear link between pregnancy-associated violence and interpersonal or intimate partner violence. Violence, whether directly or indirectly contributing to a maternal death, is a critical component that must be assessed and addressed. Neglecting to review the contributing factors to cases of pregnancy-associated homicide or suicide limits the impact activities like maternal mortality reviews can have on effectively addressing the maternal mortality crisis in the United States. While some MMRCs may currently collect data from NVDRS to add to cases being reviewed, there is no record that this is standard practice, and the process may require more than one data analyst to be involved. However, as mentioned previously, the abstractor manual on the Review to Action site states that depending on the authorities under which the MMRC operates, some sources of information could be restricted and recommends that information on cases be pulled from as many sources as possible and feasible, which includes NVDRS as an option. However, this process would most likely still require communication between two or more parties (depending on the MMRC) to access that information causing a potential delay in case review. Streamlining the process of direct access from NVDRS to MMRIA would allow for more timely reviews.

Reviewing pregnancy-associated homicide and suicide deaths and the individual, provider, facility, system, and community factors that may have contributed to the death will lead to a better understanding of where prevention or intervention steps could be taken. For example, screening for intimate partner violence throughout prenatal care and during postpartum care may be one method to provide intervention to women.^6,11,12,19 While the American College of Obstetricians and Gynecologists called for physicians to “screen all women for IPV at periodic intervals, including during obstetric care (at the first prenatal visit, at least once per trimester, and at the postpartum checkup),”²⁰ a recent analysis of the CDC's 2012 Pregnancy Risk Assessment Monitoring System dataset found that only 49.2% of women had been screened for IPV while pregnant.²¹ Furthermore, the United States Preventive Services Taskforce recommends that all women of reproductive age be screened for IPV.²² However, these prevention and intervention steps must go beyond the individual level. There is a need to further study the structural determinants of health that are contributing to pregnancy-associated deaths.^7,34,35 By studying these factors, a clearer picture of the underlying causes of violence will appear and can then be addressed. MMRCs are uniquely positioned to do this type of investigation and make recommendations. Maternal mortality is more than medically associated causes of death and should be examined as such. To do this effectively, MMRCs need to be multidisciplinary. A multidisciplinary committee will allow for diverse perspectives on the incident, a greater understanding of the factors that may have contributed to the death, and where intervention could have occurred to alter the outcome. By expanding their scope of review, MMRCs could begin to identify the numerous causes of maternal mortality and effectively address them.

This study is not without limitations. First, the sample is not nationally representative. Ten states were not included in this analysis as NVDRS-RAD had not begun collecting information from them. Second, there is the chance the results underrepresent the problem. There were several 1,000 cases where pregnancy status was marked as unknown. Despite attempts to identify additional cases through the narrative, no additional cases were found. However, there is still the potential that some of the unknown cases could have been pregnant, especially when it comes to the late postpartum period. Third, there are potential issues surrounding the reliability of intimate partner violence and interpersonal violence. Due to the nature of IPV, many women do not report it. In cases of suicide, it may not be known that IPV was occurring in the woman's life or even looked for as a potential cause of the suicide despite research showing a relationship between suicidal ideation and IPV.^24,25 Therefore, the results could again be underestimating the impact IPV has on pregnancy-associated homicide and suicide. Another potential limitation of the results is the difference in pregnancy status in relation to the periods of time being compared. For example, pregnancy is 9 months, whereas the early postpartum period is one and a half months, and the late postpartum period is ten and a half months. There is the potential that some of the differences seen in suicide and homicide rates during these periods are larger simply because more time is represented in that pregnancy status. Finally, there is the concern around the underidentification of suicide. Research has suggested that medical examiners and coroners may underidentify suicides due to the requirement that positive proof of intent is apparent.³³ Therefore, the results could be different if better identification of suicide occurred.

Conclusions

Our findings echo those of other studies^{7,8,11,12,23,24,27,32} that the relationship between maternal mortality and violence needs to be studied. Our findings highlight the elevated risk pregnant and postpartum women face in relation to violent death, particularly homicide. MMRCs are well suited to review these types of deaths and identify direct and indirect causes of violence; they are also uniquely positioned to make recommendations for changes such as universal screening for intimate partner violence (during prenatal, postpartum, and well-child visits), rapid referrals, safety planning, and wraparound services to support women who screen positive. Presently, many MMRCs focus on medical causes of maternal mortality and not violence and the structural inequalities influencing violent pregnancy-associated deaths. Our results show the need for these types of deaths to be thoroughly reviewed if the United States wants to reduce its high maternal mortality rates. Similar to other recommendations, the authors caution against focusing only on medical causes of maternal mortality,²⁸ but instead broadening the scope of review to include in-depth reviews of violent maternal death. Furthermore, the results of our analysis suggest a need for the NVDRS and MMRIA to talk to one another. This would allow for reviews that are more robust and eliminate the need for MMRCs to have to request this information from the State or other entity. A study found that linking the traditional maternal mortality review data and NVDRS data notably improved detection of violent pregnancy-associated deaths.²⁵ If MMRIA and NVDRS talked to one another, a more complete picture of what happened, including the trimester prenatal care was entered, the type of clinic, type of insurance, etc. could be deduced. Beyond making for a more thorough review by the MMRC, having the two systems talk to one another would allow researchers to better identify trends and risk factors facing pregnant and postpartum women. By incorporating the two suggested approaches, maternal mortality may be reduced in the U.S. as we come to better understand the role violence plays in pregnancy-associated deaths and implement violence prevention efforts.

Footnotes

Disclaimer

The contents of this article are solely the responsibility of the authors. These views do not necessarily represent the official position of the CDC or the American Public Health Association. None of the funders was involved in the study design, analysis, or interpretation of data, writing the report, or the decision to submit the article for publication. The NVDRS is administered by the CDC by participating NVDRS states. The findings and conclusions of this study are those of the authors alone and do not necessarily represent the official position of the CDC or of participating NVDRS states.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported by the Center for Disease Control and Prevention/American Public Health Association NVDRS New Investigator Award.

References

Review to Action. Case selection. 2020. Available at: https://reviewtoaction.org/content/case-selection Accessed July 1, 2021 .

St. Pierre A, Zaharatos J, Goodman D, Callaghan W. Challenges and opportunities in identify, reviewing, and preventing maternal deaths. Obstet Gynecol, 2018; 131:138–142.

Centers for Disease Control and Prevention. National violent death report system, 2019. Available at: https://www.cdc.gov/violenceprevention/datasources/nvdrs/index.html Accessed October 19, 2020 .

Catalano

, Davis

, Petersen

, et al. Pregnant? Validity of the pregnancy checkbox on death certificates in four states, and characteristics associated with pregnancy check-box errors. Am J Obstet Gynecol, 2020; 222:269.e1–269.e8.

Horon

, Cheng

. Effectiveness of pregnancy check boxes on death certificates in identifying pregnancy-associated mortality. Public Health Rep, 2011; 126:195–200.

Shadigian

, Bauer

. Pregnancy-associated death: A qualitative systematic review of suicide and homicide. Obstet Gynecol Survey, 2006; 60:183–190.

Wallace

, Friar

, Herwehe

, Theall

. Violence as a direct cause of and indirect contributor to maternal death. J Women's Health (Larchmt), 2020; 29:1032–1038.

Chang

, Berg

, Saltzman

, Herndon

. Homicide: A leading cause of injury deaths among pregnant and postpartum women in the United States, 1991–1999. Am J Public Health, 2005; 95:471–477.

Cheng

, Horon

. Intimate-partner homicide among pregnant and postpartum women. Obstet Gynecol, 2010; 115:1181–1186.

10.

Wallace

, Hoyert

, Williams

, Mendola

. Pregnancy-associated homicide and suicide in 37 US states with enhanced pregnancy surveillance. Am J Obstet Gynecol, 2016; 215:364.e1–364.e10.

11.

Alhusen

, Ray

, Sharps

, Bullock

. Intimate partner violence during pregnancy: Maternal and neonatal out-comes. J Womens Health (Larchmt), 2015; 24:100–106.

12.

Bronson

, Reviere

. Pregnancy-associated death in Virginia due to homicides, suicides, and accidental overdoses compared with natural causes. Violence Against Women, 2017; 23:1620–1637.

13.

Lindahl

, Pearson

, Colpe

. Prevalence of suicidality during pregnancy and the postpartum. Arch Womens Ment Health, 2005; 8:77–87.

14.

Metz

, Rovner

, Hoffman

, Allshouse

, Beckwith

, Binswanger

. Maternal deaths from suicide and overdose in Colorado, 2004–2012. Obstet Gynecol, 2016; 6:1233–1240.

15.

Mangla

, Hoffman

, Trumpff

, O'Grady

, Monk

. Maternal self-harm deaths: An unrecognized and preventable outcome. Am J Obstet Gynecol, 2019; 221:295–303.

16.

Mehta

, Bachhuber

, Hoffman

, Srinivas

. Deaths from unintentional injury, homicide, and suicide during or within 1 year in pregnancy in Philadelphia. Am J Public Health, 2016; 106:2208–2210.

17.

Palladino

, Singh

, Campbell

, Flynn

, Gold

. Homicide and suicide during the perinatal period: Findings from the National Violent Death Reporting System. Obstet Gynecol, 2011; 118:1056–1063.

18.

Silverman

, Decker

, Reed

, Raj

. Intimate partner violence victimization prior to and during pregnancy among women residing in 26U.S. states: Associations with maternal and neonatal health. Am J Obstet Gynecol, 2006; 195:140–148.

19.

Deshpande

, Lewis-O'Connor

. Screening for intimate partner violence during pregnancy. Rev Obstet Gynecol, 2013; 6:141–148.

20.

Committee on Health Care for Underserved Women. Committee opinion: Intimate partner violence. The American College of Obstetricians and Gynecologists. Obstet Gynecol, 2012; 119:412–417.

21.

Halpern-Meekin

, Costanzo

, Ehrenthal

, et al. Intimate partner violence screening in the prenatal period: Variation by state, insurance, and patient characteristics. Matern Child Health J, 2019; 23:756–767.

22.

United State Preventive Services Task Force. Screening for intimate partner violence, elder abuse, and abuse of vulnerable adults US preventive services task force final recommendation statement. JAMA, 2018; 320:1678–1687.

23.

Lindahl

, Pearson

, Colpe

. Prevalence of suicidality during pregnancy and the postpartum. Arch Womens Ment Health, 2005; 8:77–87.

24.

Metz

, Rovner

, Hoffman

, Allshouse

, Beckwith

, Binswanger

. Maternal deaths from suicide and overdose in Colorado, 2004–2012. Obstet Gynecol, 2016; 6:1233–1240.

25.

Austin

, Vladutiu

, Jones-Vessey

, Norwood

, Proescholdbell

, Menard

. Improved ascertainment of pregnancy-associated suicides and homicides in North Carolina. Am J Prev Med, 2016; 51(5,Suppl. 3):S234–S240.

26.

Mota

, Chartier

, Ekuma

, et al. Mental disorders and suicide attempts in the pregnancy and postpartum periods compared with non-pregnancy: A population study. Canadian J Psychiatry, 2019; 64:482–491.

27.

Samandari

, Martin

, Schiro

. Homicide among pregnant and postpartum women in the United States. Trauma Violence Abuse, 2010; 11:42–54.

28.

Collier

, Molina

. Maternal mortality in the United States: Updates on trends, causes, and solutions. Neoreviews, 2019; 20:e561–e574.

29.

Jocums

, Berg

, Entman

, Mitchell

Jr . Postdelivery mortality in Tennessee, 1989–1991. Obstet Gynecol, 1998; 91(5 Pt 1):766–770.

30.

Orsolini

, Valchera

, Vecchiotti

, et al. Suicide during perinatal period: Epidemiology, risk factors, and clinical correlates. Front Psychiatry, 2016; 7:1–6.

31.

Kramer

, Strahan

, Preslar

, et al. Changing the conversation: Applying a health equity framework to maternal mortality reviews. Am J Obstet Gynecol, 2019; 221:609.e1–609.e9.

32.

Koch

, Geller

. Addressing maternal deaths due to violence: The Illinois experience. Am J Obstet Gynecol, 2017; 217:556.e1–556.e6.

33.

Timmermans

. Suicide determination and the professional authority of medical examiners. Am Sociol Rev, 2005; 70:311–333.

34.

Wallace

, Crear-Perry

, Mehta

, Theall

. Homicide during pregnancy and the postpartum period in Louisiana, 2016–2017. JAMA Pediatr, 2020; 174:387–388.

35.

Crear-Perry

, Correa-de-Araujo

, Lewis Johnson

, McLemore

, Neilson

, Wallace

. Social and structural determinants of health inequities in maternal health. J Womens Health (Larchmt), 2021; 30:230–235.

36.

Centers for Disease Control and Prevention. Pregnancy-related deaths, 2019. Available at: https://www.cdc.gov/reproductivehealth/maternalinfanthealth/pregnancy-relatedmortality.htm Accessed May 12, 2021 .

37.

Centers for Disease Control and

Prevention

, National Center for Chronic Disease Prevention and Health

Promotion

. Rapid assessment of maternal opioid use and overdose, 2015. Available at: https://www.cdc.gov/reproductivehealth/maternalinfanthealth/substance-abuse/opioid-use-disorder-pregnancy/pdf/Rapid-Assessment-of-Maternal-Opioid-Use-and-Overdose_h.pdf. Assessment-of-Maternal-Opioid-Useand-Overdose_h.pdf Accessed May 20, 2021 .

38.

Review to Action. Case abstraction. 2020. Available at: https://reviewtoaction.org/index.php/implement/process-review/case-abstraction Accessed July 8, 2021 .