Tom Sullivan | March 23, 2022
Editor’s note: This article is based on a meeting of the Health Evolution Work Group on Leveraging Data to Improve Health Equity. The Health Evolution Forum is underwritten by Leadership Partners AmeriHealth Caritas and Change Healthcare as well as Work Group Partner MetaCx.
If the first step to solving a problem is typically recognizing and acknowledging its existence, the ongoing civil injustice issues and pandemic have provided enough evidence to eradicate any lingering denial of the persistent inequities in the U.S. health care system. The next step in addressing those problems is to better understand them through data.
More and more organizations, in fact, are prioritizing efforts to improve the data they collect about race, ethnicity, language and sex. Particular to health care, “organizations are increasingly collecting these data to meet regulatory requirements and build a foundation for monitoring racial and ethnic disparities, as well as disparities in quality of care due to language barriers.”1
Despite Affordable Care Act requirements that the federal government collect population data on race, ethnicity and language, however, it has been understood for years that data sets remain incomplete.2
To address that problem, approximately 50 organizations have signed the Health Equity Pledge developed by Health Evolution Fellows to increase data collection, strengthen partnerships to dismantle collection and stratification efforts, and drive more consistent adoption of data standards and definitions for stratification and analysis.
“We know statistically and geographically in the US that the number one driving factor for disparities and outcomes is race,” said Tosan O. Boyo, Senior Vice President of Hospital Operations at John Muir Health and a Fellow in the Health Evolution Forum.
During a recent meeting of Fellows participating in the Roundtable on Community Health and Advancing Health Equity’s Work Group on Leveraging Data to Improve Health Equity, Boyo and executives from organizations on the frontlines of improving data collection shared their experience and insights gleaned thus far.
Deborah Kilday, MSN, Principal, Women and Infant Services, Premier, joined the meeting to discuss an arrangement with the U.S. Department of Health and Human Services (HHS) and a multitude of hospitals to collect and analyze data to more effectively address maternal infant mortality, among the most-recognized disparities to date.
Focusing on maternal mortality data
In 2020, the CDC found that 861 women died of maternal causes in the United States, an increase over the 754 who died in 20193 and the maternal mortality rate for Black women was nearly 3 times the rate for White women.4
At the end of 2020 the U.S. Department of Health and Human Services’ Office on Women’s Health embarked on an action plan for reducing maternal mortality and morbidity that includes a diverse group of stakeholders.
The national strategy involves a partnership with Premier5 that Kilday described as a two-pronged arrangement. One prong is HHS leveraging Premier’s nationally representative standardized research database to examine overall U.S. discharges from the acute care setting, both inpatient and outpatient, to analyze factors impacting maternal infant outcomes including social determinants, racial and ethnic disparities.
The research database is nearly 20 years deep, houses an average of one million births a year dating back to 2008 and more than a decade of deliveries, with more than 1 billion patient encounters and also includes demographic disease state information about billed services, medications, laboratory tests and additional hospital and patient characteristics, Kilday said.
The second prong of the initiative is a collaborative of some 220 hospitals6 representing all 50 states and the district of Columbia that have signed on to share their de-identified aggregated process and outcomes data with Premier which, in turn, shares that de identified data with HHS. The agency can use the information to learn about what is driving outcomes at the facility level and to help identify risk factors and understand how quality improvement efforts in the implementation of standardized practices and processes influence overall outcomes.
“The entire collaborative has an overarching theme of health equity, so we can stratify data looking for care variations and disparities and we can measure social determinants of health,” Kilday said. “We stratify by race, for example, and we can stratify more granularly than what can be seen at the CDC level to augment CDC and national data.”
Premier equips hospitals participating in the collaborative with evidence-based care practices and a perinatal collaborative outcome dashboard that more than 1,300 hospitals are using.7 Every hospital in the collaborative has the same standardized reporting dashboard with more than 150 metrics that the hospitals control all the way down to the patient-provider level to create reports that help with performance improvements, Kilday added.
Stratifying outcomes by looking at key drivers of in-hospital delivery related deaths across approximately 10 million births and accounting for patient attributes like co-morbidities and complications enabled Premier to determine, for instance, that Medicaid patients have higher odds of death at the time of delivery than mothers covered by commercial plans.
“If you really want to see what’s pushing the needle and driving the outcomes,” Kilday said, “you have to become more granular with standardized metrics.”
While the HHS and Premier initiative highlights some of the considerable potential for using data and analytics to address inequities, there are a number of challenges that span many of the existing disparities today, including patients declining to have their data collected, systems opting out of such efforts because of data complexity, and incorporating additional types of data to collect.
Since developing a survey instrument to collect race and ethnicity data, the University of California, San Francisco, has found that many patients go through the process smoothly. That said, it’s not rare for some to object because they’re undocumented, have belief-driven issues with submitting their information or have questions about what will happen to the data after it’s collected, according to Russ Cucina, MD, Chief Health Information Officer, UCSF.
“Obviously, we would only use it for health care purposes but not all patients are reassured by that,” Cucina said. To address concerned patients, UCSF developed a 30-page toolkit with scripts for each scenario that the front desk workers responsible for data collection can use.
UCSF also provides patients with a multi-lingual paper form that enables them to self-identify so that not every data collection effort has to be verbally interactive, which Cucina noted has been key because throughput at registration is important and allowing patients to self-identify rather than adding to the question-and-answer verbal interaction has helped with scalability and completeness.
John Muir’s Boyo said that data complexity is one of the key reasons for health systems to either decide against collecting data on race and ethnicity or opt-out of such initiatives after starting them.
“In my current and previous roles, one of our goals was to eliminate or reduce the chances of systems opting out. We want to make sure that everyone isn’t trying to eat the whole elephant,” Boyo added. “It really comes down to what a system can commit to at this point of the journey that a lot of systems can do in the same way across the country.”
Another challenge, of course, is broadening data collection work to incorporate additional types of information, such as economic related data, education level, socioeconomic status, and sexual orientation and gender identity (SOGI) data.
UCSF, for its part, is working to determine how best to move forward with additional data types. “We are working on how to capture and document the broader set of social determinants of health,” Cucina adds.
Call to action: Pledge to improve data collection
Despite challenges, the cases of HHS, Premier and more than 200 hospitals demonstrate the potential progress that can be made by harnessing data to better understand and ultimately address existing inequities including but by no means limited to maternal mortality.
CEOs and leaders interested in taking steps to enhance data collection practices can learn more about and sign the Health Equity Pledge. Health Evolution is also making available the following resources for advancing health equity:
The Health Evolution Roundtable on Community Health and Advancing Health Equity, along with the Roundtable on Next Generation IT in Health Care and the Roundtable on New Models of Care Delivery, will convene at the upcoming 2022 Summit* in sessions that will include a discussion at the intersection of data and health equity, addressing bias and barriers upstream in the data pipeline.
*The Health Evolution Summit will take place April 6-8, 2022 at The Ritz-Carlton Laguna Niguel and Waldorf Astoria Monarch Beach Resort & Club hotels in Dana Point, CA. View the agenda or Apply to attend.
Sources & Citations:
1 Centers for Medicare and Medicaid Services, Inventory of resources for the standardized demographic and language data collection
2 Health Affairs, Data on race, ethnicity, and language mostly incomplete for manager care plan members
3 Centers for Disease Control and Prevention, Maternal mortality rates in the United States 2020
4 CDC, Maternal mortality rates in the United States 2020
5 Health and Human Service Department, HHS announces 200 hospitals participating in new infant and maternal care collaborative
6 Premier, HHS Perinatal Improvement Collaborative Aims to Improve Outcomes for mothers and babies across the nation
7 Premier, Maternal Health, Bundles of Joy