Pregnancy-Related Mortality Due to Infection

Abstract

OBJECTIVE:

To describe demographic and clinical characteristics, preventability, Maternal Mortality Review Committee (MMRC)–determined contributing factors, and example recommendations for pregnancy-related deaths due to infection using data from MMRCs.

METHODS:

We used pregnancy-related death data from MMRCs in 29 states occurring during 2017–2019 with MMRC-determined underlying cause of death of infection. We describe the distribution of demographic and clinical characteristics, present the most frequent contributing factor classes, and provide example MMRC prevention recommendations.

RESULTS:

Ninety-one pregnancy-related infection deaths were identified, and MMRCs determined that 86.4% of deaths were preventable among 88 deaths for which MMRCs determined preventability. Most deaths occurred within 42 days of delivery (69.3%). Additional clinical information was available for many deaths. Group A streptococci were the most frequently identified pathogen (34.0%, 18/53) and genital tract was the most frequently identified source (47.9%, 35/73) of the infection. The most frequent health care encounter before death was hospitalization (50.7%, 36/71). More than half of decedents (69.1%, 47/68) had a health care encounter less than 7 days before death. The five most frequent contributing factor classes were clinical skill/quality of care (18.6%), delay (10.1%), knowledge (10.1%), lack of continuity of care (9.6%), and lack of access/financial resources (7.8%). The MMRC prevention recommendations occur at multiple levels, addressing frequent contributing factor classes.

CONCLUSION:

Most pregnancy-related deaths due to infection are preventable. Example MMRC recommendations highlight prevention opportunities, including improving patient and clinician knowledge regarding clinical signs and symptoms of serious infections, implementing obstetric sepsis protocols, and enhancing care coordination within and across systems.

A report from Maternal Mortality Review Committees (MMRCs) in 36 states identified infection as a leading underlying cause of pregnancy-related death, accounting for 9.2% of pregnancy-related deaths in 2017–2019.¹ This category includes both pelvic and nonpelvic infections. Sepsis is a life-threatening organ dysfunction caused by a dysregulated host response to infection.² A previous analysis using hospital discharge and readmission data focused on maternal sepsis, examining nationwide estimates from 2013 to 2016.³ Sepsis occurred in 0.04% of all delivery hospitalizations or readmissions within 42 days, and 23% of all maternal deaths were sepsis related.³ Nationally representative data from delivery hospitalizations provide evidence that Black, Native American, and Asian/Pacific Islander women have the highest rates of sepsis.⁴ In addition, increases in maternal age are associated with increased risk for sepsis: Women aged 40 years or older have sepsis rates 4.0 times those of women aged 25–29 years.⁴ To date, the guidelines for management of serious infections in pregnancy and postpartum have focused on acute hospital-based management.^5,6 A broader understanding of the contributors to pregnancy-related infection deaths may be helpful for identifying additional opportunities for prevention.

The MMRCs provide an in-depth understanding of pregnancy-related mortality through detailed reviews of deaths during and within 1 year of pregnancy by multidisciplinary groups of clinical and nonclinical subject matter experts.⁷ At the state or jurisdiction level, MMRCs access a wide variety of sources, including prenatal care, birth, hospitalization and other medical records, autopsy reports, social service records, and birth and death records, to accurately identify and comprehensively review deaths. The MMRCs create specific and actionable recommendations to inform local, state, and national prevention strategies based on their comprehensive reviews.⁸

Data from MMRC review of pregnancy-related infection deaths, including underlying causes, timing in relation to pregnancy, and contributing factors, provide rich information to inform future preventive approaches. The purposes of this analysis are to describe demographic and clinical information among pregnancy-related deaths due to infection, to describe MMRC-identified contributing factors for these deaths, and to share example recommendations made by MMRCs for reducing preventable infection deaths among the most frequently identified contributors.

METHODS

For this descriptive study using Maternal Mortality Review Information Application data shared with the Centers for Disease Control and Prevention (CDC) by MMRCs, we analyzed data from the 29 states with pregnancy-related deaths with infection as the MMRC-determined underlying cause of death occurring among their residents from 2017 to 2019. In some states, only partial years of data were available. The MMRCs determine a pregnancy-associated death (death during or within 1 year of the end of pregnancy) to be pregnancy related if the death was from a pregnancy complication, a chain of events initiated by pregnancy, or the aggravation of an unrelated condition by the effects of pregnancy. The underlying cause of death is defined as the disease or injury that initiated the chain of events leading to the death. The MMRC-determined underlying cause of death is coded with a standardized list of 21 major categories and 69 sub-categories and may vary from the underlying cause of death documented on the official death record because of the multidisciplinary review and additional information that may be available to the MMRC.

Race and ethnicity data were derived from the birth or fetal death record if available and from the death record when a birth or fetal death record was unavailable or had missing data and classified as Hispanic, non-Hispanic American Indian or Alaska Native, non-Hispanic Asian, non-Hispanic Black, non-Hispanic Native Hawaiian and Other Pacific Islander, non-Hispanic White, or non-Hispanic multiple races. Age at death was based on information from the death record.

Methods for determining the timing of death in relation to pregnancy have been described previously.¹ We classified the timing of death in relation to the end of pregnancy into five mutually exclusive categories: 1) pregnant at the time of death; 2) death occurred on the day of delivery, if the death occurred within 24 hours of delivery; 3) death occurred between 1 and 6 days after the end of pregnancy; 4) death occurred between 7 and 42 days after the end of pregnancy; or 5) death occurred 43–365 days after the end of pregnancy.

The MMRCs document preventability by determining whether the death was preventable (“yes” or “no”) and whether there was “no chance,” “some chance,” or “good chance” to alter the outcome. A death is considered preventable if the MMRC determines that there was at least some chance of the death being averted by one or more reasonable changes to patient or family, health care professional, facility, system, or community factors.

Source of infection, primary pathogen, type of last health care encounter, time from last health care encounter, antibiotic receipt, and time from symptomatic presentation to antibiotic initiation were determined from clinical review of the case narratives and autopsy reports if available by two co-authors (L.M.H., N.T.J.). Place of death was ascertained from the death record and from the clinical review of the case narratives if place of death was marked as unknown on the death record.

If a pregnancy-related death is determined to be preventable, MMRCs use free text to describe contributing factors and recommendations. For each contributing factor described, MMRCs select a corresponding contributing factor class from a standardized list of 27 specific classes on the Maternal Mortality Review Information Application Committee Decisions Form; each death may have multiple contributing factor classes identified. The MMRCs also make recommendations for preventing future pregnancy-related deaths that address each contributing factor they identify. For this analysis, we analyzed contributing factor classes and reviewed all MMRC recommendations among the five most frequent contributing factor classes. Contributing factor classes can have more than one recommendation, and the same recommendation can be associated with multiple contributing factor classes. Example MMRC recommendations were selected that address each of the five most frequent contributing factor classes.

Descriptive statistics were calculated as counts and percentages. Missing, unknown, and nonapplicable values are reported but not included in the calculation of distribution percentages. All analyses were performed with SAS 9.4. This activity was reviewed by the CDC, was deemed not human research, and was conducted consistent with applicable federal law and CDC policy.

RESULTS

Demographic characteristics of the 91 individuals with pregnancy-related deaths with an underlying cause of infection are shown in Table 1. Most deaths occurred among non-Hispanic White women (54.4%, n=49), women aged 25–39 years (72.5%, n=66), and women residing in urban areas (77.0%, n=57) (Table 1). Pregnancy-related deaths due to infection occurred most frequently in the postpartum period; 27.3% (n=24) occurred 1–6 days postpartum, and 36.4% (n=32) occurred 7–42 days postpartum (Table 1).

Table 1.

Demographic Characteristics of Individuals With Pregnancy-Related Deaths With an Underlying Cause of Infection, Residents of 29 States, 2017–2019

Characteristic	Pregnancy-Related Infection Deaths (N=91)*
Timing of death in relation to pregnancy
During pregnancy	14 (15.9)
Day of delivery	5 (5.7)
1–6 d postpartum	24 (27.3)
7–42 d postpartum	32 (36.4)
43–365 d postpartum	13 (14.8)
Unknown	3 (—)
Race and ethnicity
Hispanic	15 (16.7)
Non-Hispanic American	1 (1.1)
Indian or Alaska Native
Non-Hispanic Asian	0 (0.0)
Non-Hispanic Black	23 (25.6)
Non-Hispanic Native	0 (0.0)
Hawaiian and Other Pacific Islander
Non-Hispanic White	49 (54.4)
Non-Hispanic multiple races	2 (2.2)
Missing	1 (—)
Age (y)
15–19	4 (4.4)
20–24	17 (18.7)
25–29	24 (26.4)
30–34	21 (23.1)
35–39	21 (23.1)
40–44	4 (4.4)
45 or older	0 (0.0)
Education
12th grade or less, no diploma	14 (16.3)
High school diploma or high school equivalency certificate	27 (31.4)
Some college credit, no degree	21 (24.4)
Associate or bachelor’s degree	21 (24.4)
Advanced degree	3 (3.5)
Missing	5 (—)
Urbanicity of place of last residence
Urban	57 (77.0)
Rural	17 (23.0)
Missing	17 (—)

Data are n (%).

^*Percentages may not add to 100% because of rounding. Unknown and missing values were not included in the calculation of distribution percentages.

Among pregnancy-related infection deaths, 18 (19.8%) had no documented source of infection. The genital tract was the most common source of infection among the 73 deaths with a documented source (47.9%, 35/73) (Table 2). Respiratory system was the second most frequent source, observed in 17.8% of decedents (13/73). Streptococcus pyogenes (group A streptococci) was the primary pathogen in 34.0% (18/53) among 53 deaths with an identified pathogen. Group A streptococci were identified in blood cultures in 11 deaths, urine culture in one, culture of ascites in one, clinical presentation in three, and unknown in two (data not shown). Escherichia coli was the second most frequently identified pathogen, occurring in 15.1% of deaths (8/53) (Table 2). A primary pathogen was not reported for 41.8% of deaths (38/91). Receipt of antibiotics was documented in 51 deaths. Among the 17 with documented intervals, the interval from symptomatic presentation to antibiotic initiation was beyond 1 hour for 15 of 17 women.

Table 2.

Clinical Characteristics for Individuals With Pregnancy-Related Deaths With an Underlying Cause of Infection, Residents of 29 U.S. States, 2017–2019

Characteristic	Pregnancy-Related Infection Deaths (N=91)*
Method of delivery for live births (n=59)
Vaginal	29 (50.0)
Cesarean	29 (50.0)
Missing	1 (—)
Source of infection
Genital	35 (47.9)
Respiratory	13 (17.8)
Gastrointestinal	7 (9.6)
Urinary	7 (9.6)
Skin and soft tissue	5 (6.8)
Other†	6 (8.2)
Missing	18 (—)
Primary pathogen
Streptococcus pyogenes (group A streptococci)	18 (34.0)
Escherichia coli	8 (15.1)
Anaerobic bacteria	4 (7.5)
Influenza	4 (7.5)
Staphylococcus aureus‡	3 (5.7)
Clostridioides difficile	2 (3.8)
Multiple pathogens	6 (11.3)
No growth or negative culture	4 (7.5)
Other	4 (7.5)
Missing	38 (—)
Place of death
Home	9 (9.9)
Hospice	2 (2.2)
Hospital inpatient	63 (69.2)
Hospital outpatient or ED	16 (17.6)
Other	1 (1.1)
Did delivery occur during terminal hospitalization?
Yes	35 (42.7)
No§	47 (57.3)
NA‖	8 (—)
Missing	1 (—)
Type of last health care encounter before death or terminal hospitalization
Hospitalization	36 (50.7)
ED visit	13 (18.3)
Labor and delivery triage visit	4 (5.6)
Office visit	18 (25.4)
Missing	20 (—)
Time from last health care encounter to death (d)
Less than 1	8 (11.8)
1–2	17 (25.0)
3–4	15 (22.1)
5–6	7 (10.3)
7 or more	21 (30.9)
Missing	23 (—)
Were antibiotics received?
Yes	51 (89.5)
No	6 (10.5)
NA, death occurred before hospitalization	7 (—)
Missing	27 (—)
Time to antibiotic treatment (h; n=51)¶
Less than 1	2 (11.8)
1–2	5 (29.4)
3–4	3 (17.6)
5–6	1 (5.9)
7–8	1 (5.9)
9–10	1 (5.9)
11–12	1 (5.9)
More than 12	3 (17.6)
Missing	34 (—)

ED, emergency department; NA, not applicable.

Data are n (%).

^*Percentages may not add to 100% because of rounding. Missing and NA values were not included in the calculation of distribution percentages.

^†Includes endocarditis, meningitis, and bloodstream.

^‡Includes methicillin-resistant staphylococcus.

^§Delivery occurred before and separate from the terminal hospitalization.

^‖Died while pregnant or delivered and died outside the hospital.

^¶Defined as time from symptomatic presentation to antibiotic initiation among women who received antibiotics.

More than half of decedents (69.1%, 47/68) with information on last health care encounter had a health care encounter within 7 days of death or terminal hospitalization. Of those 47 with an encounter before 7 days, the encounter was a hospitalization (48.9%, n=23), an office visit (27.7%, n=13), an emergency department (ED) visit (17.0%, n=8), or labor and delivery triage visit (6.4%, n=3) (Appendix 1, available online at http://links.lww.com/AOG/E488). Twenty-five women died either at home (9.9%, n59) or in an ED or hospital outpatient setting (Table 2).

The MMRCs determined preventability in 96.7% of the 91 deaths; of these, 86.4% were determined to be preventable. There were 397 MMRC-identified contributing factor classes among the 72 preventable deaths that had contributing factors and recommendations (median 4, range 1–31) contributing factor classes per preventable death. The five most frequently identified contributing factor classes were clinical skill/quality of care (18.6%, n574), delay (10.1%, n540), knowledge (10.1%, n540), lack of continuity of care (9.6%, n538), and lack of access/financial resources (7.8%, n531) (Table 3). These five contributing factor classes accounted for 56.2% of the contributing factor classes identified by MMRCs. Example MMRC recommendations for prevention of future deaths for each of the five most frequent contributing factor classes are listed in Table 3.

Table 3.

Example Maternal Mortality Review Committee Recommendations to Address the Five Most Frequent Contributing Factor Classes Among Preventable Pregnancy-Related Infection Deaths

Contributing Factor Class	Contributing Factor Class Description	Example MMRC Recommendations*
Clinical skill/quality of care	Personnel were not appropriately skilled for the situation or did not exercise clinical judgment consistent with standards of care (health care professional or facility perspective).	Recommend that every hospital have an evidence-based sepsis protocol in place for pregnant and postpartum women Health care professionals should receive continuing education on recognizing obstetric emergencies in the outpatient setting and educate patients and their families on recognizing signs and symptoms of those emergencies. Recommend that EDs should educate health care professionals of modified obstetric early warning signs in ED triage and other settings of care Facilities should conduct root-cause analysis or other investigations to identify gaps in care that led to sentinel events and incorporate adjustments to fill in those gaps. Health care professionals should be educated as to the signs and symptoms of sepsis and should follow appropriate protocols for treating sepsis. Support maternal sepsis education, sepsis bundle implementation. and education throughout maternal care facilities
Delay	The health care professional or patient was delayed in referring or accessing care, treatment, or follow-up care or actions.	Training should be provided to all health care health care professionals and staff so that predispositions do not result in a delay in diagnosis and differential diagnoses are considered. Hospitals should have clear policies for EDs to seek consultation from an obstetric health care professional for all pregnant or postpartum women with specific triggers indicative of pregnancy or postpartum complications. Lack of local primary care and a medical home requiring patients to travel long distances can lead to delays in seeking medical care. [States] should enhance novel ways of expanding access to local medical care, including telehealth, phone follow-up for patients seen in ED, and mobile clinics. Health care professionals should follow appropriate protocols for treating sepsis Facilities should ensure that all women, partners, family members are educated about postpartum warning signs and are empowered to speak up and seek emergent care.
Knowledge	Lack of knowledge regarding importance of event, treatment, or follow-up (patient or health care professional perspective)	Health care professionals should familiarize themselves with reports of group A streptococci and their clinical presentation. Hospitals should disseminate educational materials on maternal early warning signs and sepsis in pregnancy to ED clinical team members and support coordination between emergency and maternal health services to implement evidence-based protocols to identify and manage obstetric and postpartum emergencies. Public health agencies and medical professional societies should increase public awareness of symptoms of sepsis by coordinating with national sepsis awareness campaigns as soon as possible. Outpatient health care professionals should ensure that pregnant women are linked to available case management and home visiting programs, such as family case management, Healthy Start, Family Connects.
Lack of continuity of care	Lack of continuity of care. Health care professionals did not have access to a woman’s complete records or did not communicate their status sufficiently.	Health facilities and health care professionals need to provide enhanced care coordination and follow-up for patients facing barriers to treatment, including mental health conditions. All facilities should complete levels-of-care assessment and establish transfer protocols. Health systems should ensure early and adequate postpartum follow-up for women regardless of how a pregnancy ends (live birth, stillbirth, miscarriage). Payers should employ care coordinators to ensure that patients are receiving ongoing care and that patients are enabled to overcome barriers to care. Community organizations should educate the public on Medicaid eligibility for undocumented women during pregnancy and postpartum period and assist with processes of applying for coverage.
Lack of access/ financial resources	Lack of access or financial resources, including systemic barriers (eg, lack or loss of health care insurance, health care professional shortages, and lack of public transportation)	Link postpartum women with health care professional to have 12 wk of follow-up. Focus on education for health care professionals in continuity of care, family planning, and prevention. Facilities should consider transportation and distance to care for patients living in underserved parts of the state. Hospitals should provide financial case management services to ensure that patients are able to maintain insurance coverage. Medicaid could be extended to 1 y for all postpartum women, and well-woman care should be easily accessible during the preconception periods.

MMRC, Maternal Mortality Review Committee; ED, emergency department.

^*The MMRC recommendations were edited slightly for clarity.

DISCUSSION

From our analysis of the 91 pregnancy-related deaths with an underlying cause of infection among residents of 29 states between 2017 and 2019, we found that genital tract infections accounted for 47.9% of deaths with source data available and that group A streptococci were the primary pathogen in 34.0% of deaths with pathogen data available. Many decedents had a recent encounter with the health care system. Although records of the timing of initial antibiotic administration were limited, treatment within the recommended 1 hour⁵ was rare. The MMRCs determined that more than four of five pregnancy-related infection deaths were preventable. The five most common MMRC-identified contributing factor classes included clinical skill/quality of care, delay, knowledge, lack of continuity of care, and lack of access/financial resources.

The MMRCs are uniquely positioned to develop specific and actionable recommendations tailored to jurisdiction contexts. The multidisciplinary composition of MMRCs allows them to understand the societal and environmental factors that contribute to pregnancy-related deaths and to make recommendations for solutions at the clinician, facility, health system, and community levels, which can overlap and intersect. Common among the example recommendations was the need for education of patients and clinicians to improve recognition of signs and symptoms of severe infection and sepsis that can prevent delays in initiation of appropriate treatment.

S pyogenes was the most frequently identified pathogen. In a meta-analysis on invasive group A streptococci outcomes, the annual incidence of invasive group A streptococci in pregnancy and postpartum was 0.12 per 1,000 live births (95% CI, 0.11–0.14, I²5100%) in high-income countries.⁹ The incidence of invasive group A streptococci is reported as 20-fold higher in postpartum compared with nonpregnant women, and invasive disease has been associated with rapid deterioration and death.^10,11 Invasive group A streptococci or streptococcal toxic shock syndrome is associated with 30–45% case fatality.¹² Because of its relative rarity, few clinicians may have experience making this diagnosis. Given the severity of the infection and the need for rapid identification and treatment, increasing familiarity with invasive group A streptococci as an entity among medical personnel providing obstetric services may help prevent delays contributing to pregnancy-related deaths.¹³

Encounters with the health care system, including within 24 hours of death or the terminal hospitalization, along with delays in initiation of antibiotics, were frequent and underscore the benefits of education among all types of clinicians across various settings to enhance early screening, diagnosis, and treatment of sepsis.⁶ Failure to recognize or delay in recognition of sepsis may be more likely in pregnancy because normal physiologic changes result in increased heart rate and lower systemic blood pressure.¹⁴ In addition, fever is not universally present with sepsis, and individuals in septic shock may be hypothermic.¹⁴ A variety of tools have been developed to rapidly identify patients at high risk of developing severe complications who require more aggressive therapy.⁵ For example, implementation of a sepsis alert in electronic medical records was associated with a 44% reduction in the odds of sepsis-related mortality.¹⁵ Pregnancy-adjusted screening tools for sepsis have been developed but need to be optimized before their use can be universally recommended.^16,17

About one in four women in this report died at home or in the hospital ED. This is consistent with findings from a Michigan MMRC report that many patients arrived late for care or died at home without receiving care¹⁸ and may reflect inadequate patient education about when to seek care or barriers experienced in accessing care. A systematic review of health education activities to improve maternal and child health evaluated numerous multimodal strategies, including community-based, mHealth/digital, individualized, and group-based pedagogical methodologies.¹⁹ The study found that the factor most predictive of improving maternal and child health outcomes was whether education began early during pregnancy and continued into the postpartum period.¹⁹ Digital campaigns may facilitate longitudinal education and specifically may improve outreach to disproportionately affected populations.²⁰ The Alliance for Innovation in Maternal Health has developed patient- and clinician-facing materials, including urgent maternal warning signs, which are also used as part of the CDC’s Hear Her campaign to promote an understanding of when to seek care.^21,22

Lack of continuity of care was identified as a frequent contributing factor class, for which example recommendations addressed the need for care escalation and maternal transfer protocols. The concept of risk-appropriate care was introduced to provide an understanding of the maternal care capabilities of maternity hospitals and to encourage the equitable distribution of full-service maternity care systems integrated across community health networks to facilitate transfer according to maternal condition.²³ Although level-of-care assessment is not a requirement, individual maternity hospitals are able to evaluate their own capacity in terms of resources, staffing, and processes to identify constraints in addressing high-risk maternal conditions and to develop systems in place for appropriate transfer. Facilitating continuity of care through pregnancy and postpartum transitions, including support for patient appointment scheduling and structured health care professional handoffs, could improve outcomes.²⁴

Clinical skill/quality of care was the most documented contributing factor class. Example MMRCs recommendations included implementation of protocols and bundles for life-threatening infections. The Alliance for Innovation on Maternal Health created the Consensus Bundle on Sepsis in Obstetric Care.^6,25 The bundle is a condition-specific collection of evidence-informed frameworks intended to be uniformly applied in every episode of relevant care. Bundle elements include the development of interdepartmental and intradepartmental protocols and policies for the assessment, treatment, and escalation of care of patients experiencing obstetric sepsis; implementation of clinician and patient education regarding sepsis warning signs; and engagement with patients and their support network in an open, transparent, and empathetic manner to review diagnosis and treatment plans that are aligned with their health literacy, culture, language, and accessibility. Bundle implementation, often supported by perinatal quality collaboratives,²⁶ may provide a comprehensive initial approach for many health systems to address several of the leading contributing factors, including clinical skill/quality of care, to pregnancy-related infection deaths. Combined approaches using methods from both quality improvement science and implementation science frameworks may promote synergies that enhance improvements in care delivery.²⁷

This study has multiple limitations. These data were from the years preceding the coronavirus disease 2019 (COVID-19) pandemic and therefore may not reflect changes in clinical practice and health care delivery that have occurred subsequently. The data are aggregated from individual jurisdiction-based review committees. Although MMRCs use a standardized review process, variations in review may have existed. Maternal Mortality Review Information Application data are based on the availability and completeness of abstracted data. Not every state contributed data, and partial years of data were included; thus, findings may not be representative of all pregnancy-related infection deaths, nor does this represent a population-based census of pregnancy-related deaths due to infection. Cause-specific pregnancy-related mortality ratios cannot be calculated. Infection source, causative pathogen, antibiotic administration and timing, last health care encounter type and timing, and delivery during terminal hospitalization are not standard variables within Maternal Mortality Review Information Application and were collected centrally from information in case narratives. Variation in the availability of records for abstraction and completeness of abstracted information led to a high percentage of missing information in some variables. Deaths with missing information may differ from those included in the analysis; however, the finding of delayed antibiotic initiation among those who died of sepsis is not unexpected given the association between timely antibiotic initiation and survival.²⁸ We present example MMRC recommendations among only the five most frequently occurring contributing factor classes, which do not represent the full spectrum of the MMRC-identified prevention opportunities. Despite these limitations, this study deepens our understanding of deaths due to infection as an important contributor to pregnancy-related mortality.

Infections are a leading cause of pregnancy-related mortality, and most of these deaths are preventable. Recommendations from MMRCs to address contributing factors highlight the scope and multifaceted nature of prevention strategies. For example, recommendations illustrate prevention opportunities, including educating and training for patients and clinicians to promote early recognition and treatment of infection to prevent progression to sepsis; implementing obstetric sepsis protocols to ensure timely, appropriate treatment and escalation of care; and enhancing care coordination within and across systems. The MMRC recommendations may assist with prioritizing and adapting interventions to achieve the elimination of preventable maternal death resulting from infection in the United States.