National Fatality Review Case Reporting System: Twenty Years of Data Collection

In 2002, the Maternal and Child Health Bureau at the Health Resources and Services Administration (HRSA), an agency of the US Department of Health and Human Services, funded the National Fatality Review Case Reporting System (NFR-CRS). The development of the NFR-CRS was part of the goals and objectives for the cooperative agreement creating the Fetal, Infant, Child Death Review program.¹ The Michigan Public Health Institute was awarded that cooperative agreement and created the National Center for Fatality Review and Prevention, which houses the NFR-CRS. Through expanded funding from HRSA, the National Center for Fatality Review and Prevention started supporting data collection and technical assistance to FIMR teams in 2015.

The purpose of NFR-CRS is to provide CDR and FIMR teams with a user-friendly, web-based data system for collecting and analyzing data. CDR and FIMR teams enter data into the NFR-CRS, which allows for the “ongoing systematic collection, storage, analysis, and interpretation of data essential to the planning, implementation, and evaluation of public health practice closely integrated with the timely dissemination of these data to those who need to know.”² Data entered in the NFR-CRS can help identify risk factors that contribute to fetal demise and child death and are the basis for prevention.^3–5 In addition to data collection and storage, the NFR-CRS is designed to provide CDR and FIMR teams with easy-to-use standardized reports that highlight key risk and protective factors for their jurisdiction. The NFR-CRS also offers the ability to document prevention recommendations and the prevention initiatives that are being developed and implemented. One of the NFR-CRS’s roles is to address gaps in the understanding of causes and contributing factors to fetal, infant, and child deaths by providing access to data to researchers for analysis and publication. Researchers can apply to the National Center for Fatality Review and Prevention’s data dissemination committee. Once the application is approved, de-identified data can be shared, analyzed, and published.

Since its inception, there have been 9 revisions of the NFR-CRS. Most notably, a new section titled Life Stressors was added in 2020 and all cause-of-death-specific sections (eg, firearm, motor vehicle and other transport, drowning) were revised in 2022. In addition to the revisions to CDR questions, in 2017, a workgroup of >25 state and local FIMR coordinators and abstractors was convened to advise the National Center for Fatality Review and Prevention on incorporating additional FIMR-specific questions into NFR-CRS with official adoption in 2018. As of April 1, 2024, the NFR-CRS contains >286 000 deaths reviewed by CDR teams and >13 800 deaths reviewed by FIMR teams.

The HRSA funding provides the foundation of the NFR-CRS; however, other governmental agencies, such as the Centers for Disease Control and Prevention (CDC) and the National Institutes of Health, have provided additional financial support to expand the use of CDR processes and the NFR-CRS to better understand infant and child deaths. For instance, in 2009, the CDC and the National Institutes of Health started funding population-based surveillance in certain jurisdictions using the NFR-CRS for the compilation of data for the Sudden Unexpected Infant Death and Sudden Death in the Young Case Registry, the largest multistate source of sudden unexpected infant death and sudden death in the young incidence data.^6–7 

The 4 elements of data quality applicable to CDR and FIMR data are as follows:

• Completeness: the extent to which the expected data elements are answered/data are entered and not missing;

• Consistency: the uniformity and agreement of data elements across cases;

• Accuracy: the extent to which the entered data reflect the known facts about the child, family, and circumstances surrounding the death;

• Timeliness: the extent to which the data are up to date.

In 2015, the National Center for Fatality Review and Prevention implemented a data quality initiative (DQI) for CDR programs.

CDR program coordinators from 7 states were convened to identify priority variables for monitoring data quality. Variables were included on the basis of their importance for developing and monitoring prevention strategies and making system changes. Because of the large number (>100) of priority variables identified during this initial process, a smaller subset of 48 variables was classified as core variables, and benchmarks for proportion missing and unknown were established for variables in this subset. These core variables were suggested as a minimum set of variables for monitoring data quality by CDR programs with limited financial or staff resources. A similar DQI for FIMR was initiated in 2020, with FIMR coordinators from 5 states identifying a list of 37 priority variables for monitoring FIMR data quality. The lists of CDR and FIMR priority variables are available at https://ncfrp.org/data/data-dissemination/.

A 5-year (2015–2020) retrospective analysis of the CDR DQI revealed overall improvement (11% decline) in the average proportion of missing or unknown responses for all priority variables; this improvement was even greater (19% decline) for the 48 core variables.⁸ In addition, more states met the benchmarks set for proportion missing and unknown for the core variables. To build on these improvements, data quality continues to be a focus of the National Center for Fatality Review and Prevention, with individualized data quality technical assistance offered to states, and an NFR-CRS update focused on data quality released in June 2024.⁸ 

In 2021, Data Explorer, an easy-to-use data query feature in which data can be summarized in multiple formats, was added to the NFR-CRS. Data Explorer was loosely modeled after CDC Wide-Ranging Online Data for Epidemiologic Research.⁹ There are 2 options for displaying data in Data Explorer. The first option displays data in a bar chart (saved as a PDF). The second option provides individual case rows of data in an Excel format. Both options allow for on-demand availability and the immediate use of data. NFR-CRS users only have access to data from their jurisdiction or state. Data Explorer allows each NFR-CRS user to access and manipulate their data without the assistance of epidemiologists, which increases the availability of relevant and easily interpretable data products for local CDR and FIMR teams.

The NFR-CRS was expanded in 2020 to include questions about life stressors (circumstances that may have impacted the child or family). The initial 6 domains were as follows:

• Social/economic: Racism, food insecurity, lack of childcare, and poverty are examples;

• Relationships: Breakup, bullying as a victim or perpetrator, family discord, and argument with parents or caregivers are examples;

• School: School failure, pressure to succeed, and extracurricular activities are examples;

• Transitions: End of the school year, transition to or from the child welfare system, and release from the hospital are examples;

• Trauma: Previous abuse (emotional or physical), rape or sexual assault, and family or domestic violence are examples;

• Other: An open text box for additional information to be collected.

In 2022 a medical life stressor domain (including questions about lack of money for care, multiple providers not coordinated, limitations of health insurance, and lack of provider–family compatibility) was added, and the other 6 domains were expanded. Additionally, NFR-CRS contains proxy variables for adverse childhood experiences. Examples include history of abuse, history of substance use, or history of domestic violence.⁸ Adverse childhood experiences are helpful for teams to identify common themes, including disparities and other specific challenges, in their communities.

Although data collected in the NFR-CRS are unique and not available in other data systems, there are several limitations to the data.^9–11 The most significant limitation is that most states do not review and enter all fetal, infant, or child deaths. As a result, the NFR-CRS data are not population-based and cannot be used to calculate rates. Additionally, the NFR-CRS data cannot be compared between states or local teams or over time because of the variation in deaths reviewed, the timing of reviews, and fluctuating participation by states.

Questions within the NFR-CRS have evolved, which creates challenges for data entry, collection, analysis, and interpretation. For example, when the NFR-CRS was piloted in 2005, there was not an opioid overdose epidemic in the United States, and electric vehicles, including scooters and bicycles, did not exist. Questions have been added to the NFR-CRS to capture the potential impact of these societal changes. Although newer questions reflect changing challenges to injury prevention, data analysis must take into account newer records that include these additional response options and earlier records that do not. To maintain historic data while accommodating changes, analysts must be aware that a lower proportion of a response option may represent the expansion of data collection options over time.

Another limitation of the NFR-CRS is the high proportion of missing data, particularly for some data elements that are challenging to collect. State and local teams have varying access to data sources and funding to support data collection, entry, and analysis. Consequently, the amount of missing data differs between and within fatality review teams. In addition, some of the elements in the NFR-CRS are subjective, for example, household income (recorded as low, medium, or high) and life stressors such as racism or housing instability. Because these fields either require the NFR-CRS user to access the data dictionary or use their “own” definitions, there are differences in how some data are documented in the NFR-CRS. Although missing and inconsistent data remain limitations, significant progress has been and will continue to be made to improve the quality, consistency, and completeness of the NFR-CRS data. Turnover in staff who enter data in the NFR-CRS is another limitation of the data. Staff turnover often delays data entry into the NFR-CRS and creates variability between staff members entering data.

Despite the limitations described in the previous section, the NFR-CRS data are a valuable source of unique information about how and why infants and children die. This breadth and quantity of data are not available from any other single source. As such, the NFR-CRS data can be used to help understand complex challenges such as sleep-related infant deaths, youth suicide, and fetal demise. For instance, the collection of information on the caregiver, supervisor, and person responsible allows for the identification of risk and protective factors that span the social-ecological model, from individual factors to family and community factors. Additionally, the NFR-CRS captures a blend of quantitative and qualitative data. For example, an interview with a primary caregiver in many FIMR reviews incorporates the view of the family, the community systems, and opportunities for prevention.

An additional strength of the NFR-CRS is the nimble nature of the system to adapt to emerging changes. For example, during the coronavirus disease 2019 pandemic, the NFR-CRS was modified 3 times in 1 year to add and update variables about coronavirus disease 2019. This dexterity aids in collecting information about emerging public health crises.

Another significant strength of the NFR-CRS data is that they can be used to influence quality improvement. For example, a CDR or FIMR team may struggle to collect information about the position of an infant in a sleep-related death. However, the identification of this missing information during the review process and emphasizing its importance for prevention initiatives will often result in agency policy change so that law enforcement investigative practices include the collection of that information.

Lastly, CDR and FIMR teams typically review deaths within 6 to 12 months, which allows fatality review teams to access their data as soon as it is entered into the NFR-CRS. As a result, emerging issues in a community can be identified quickly.

The National Center for Fatality Review and Prevention will continue, through ongoing outreach and education, to encourage CDR and FIMR teams not using the NFR-CRS to participate. The goal is for all US fatality review teams to use the NFR-CRS. In addition, the National Center for Fatality Review and Prevention will continue to support current NFR-CRS users in reviewing and entering a higher proportion of deaths in their jurisdiction.

The National Center for Fatality Review and Prevention and its partners will continue to expand the capacity of the NFR-CRS to conduct surveillance on public health issues for which it is uniquely equipped, such as child deaths in the context of natural disasters. The National Center for Fatality Review and Prevention will also continue to support activities that improve the quality and visibility of the data contained in the NFR-CRS. The National Center for Fatality Review and Prevention will explore strategies for reducing data entry burden, such as exploring possible data linkages through application programming interfaces between the NFR-CRS and other data systems that collect fatality review information, such as the National Association for Public Health Statistics and Information Systems’ State and Territorial Exchange of Vital Events or the CDC’s National Violent Death Reporting System.

The NFR-CRS was, if not the first, one of the first web-based data collection tools for fatality review teams. The NFR-CRS has a long history of evolving to support the ever-changing needs of CDR and FIMR programs. The NFR-CRS will continue to grow and leverage technological advances to improve all aspects of data collection, analysis, and reporting. Improved funding for CDR and FIMR programs at the state and local levels would have a significant positive impact on the NFR-CRS data quality. Although the NFR-CRS has documented limitations, the data collected through the fatality review process provides an irreplaceable view of fetal, infant, and child mortality and the complicated factors surrounding these deaths.

CDC

Centers for Disease Control and Prevention

CDR

child death review

DQI

data quality initiative

FIMR

fetal and infant mortality review

HRSA

Health Resources and Services Administration

NFR-CRS

National Fatality Review Case Reporting System