The mPINC survey: Impacting US maternity care practices

Abstract The Centers for Disease Control and Prevention administered the original Maternity Practices in Infant Nutrition and Care (mPINC) survey, a census of all US birth facilities, from 2007 to 2015 to monitor infant feeding‐related maternity care practices and policies. The purpose of this paper is to describe the many uses of mPINC data. Hospitals, organizations and governments (federal, state and local) have used the mPINC survey as a tool for improving care among the populations they serve. Nationally, the mPINC survey has been used to document marked improvements in infant feeding‐related maternity care. Researchers have used the mPINC data to examine a variety of questions related to maternity care practices and policies. The newly revised mPINC survey (2018) has been designed to capture changes that have occurred over the past decade in infant feeding‐related US maternity care. Hospitals, organizations, governments and researchers will be able to continue using this important tool in their efforts to ensure US maternity care practices and policies are fully supportive of breastfeeding.


| INTRODUCTION
Breastfeeding is considered the optimal feeding method for most infants (American Academy of Pediatrics: Section on Breastfeeding, 2012) and reduces chronic diseases in women (Feltner et al., 2018). Evidence-based maternity care practices are known to improve breastfeeding exclusivity (Perrine, Scanlon, Li, Odom, & Grummer-Strawn, 2012) and duration (DiGirolamo, Grummer-Strawn, & Fein, 2008). The first hospital in the United States designated as part of the Baby-Friendly Hospital Initiative, a programme designed to recognize hospitals offering evidence-based maternity care as outlined in the Ten Steps to Successful Breastfeeding (Ten Steps), was in 1996, but widespread participation in the initiative was low for many years (Baby-Friendly USA, 2016). Although local (Kovach, 2002) and state (Rosenberg, Stull, Adler, Kasehagen, & Crivelli-Kovach, 2008) efforts had been made to assess hospital practices supportive of breastfeeding, there were no national or regional data on implementation of these practices despite the critical role hospitals were known to play in supporting breastfeeding. Thus, in the fall of 2003, the Centers for Disease Control and Prevention (CDC) hosted a meeting of experts to explore the feasibility of national surveillance of maternity care practices related to breastfeeding, including data needs, potential barriers to data collection and potential methods for surveillance.  . The domains of care were scored from 0 to 100 to generate seven subdomain scores, which were averaged to calculate a total mPINC score. Both the mPINC survey and scoring algorithm are available by emailing: mpinc@cdc.gov.
All hospitals and free-standing birth centres in the United States and Territories (American Samoa, Guam, Northern Mariana Islands, Puerto Rico and the US Virgin Islands) (hereafter, "states" unless otherwise noted) with maternity beds in the year prior to the survey were eligible to participate. The main facility switchboard was contacted with a request to speak with the manager of the Mother-Baby unit.
Once reached, this manager confirmed the hospital's eligibility and identified the most appropriate person to receive the survey. Hospitals were encouraged to get input from key staff, as needed, when completing the survey. Completed surveys were submitted either electronically or via paper with response rates ranging from 82% to 83% for all five cycles (range of participating facilities: 2,582-2,742).
Given the high response rate of this census, data on maternity practices and policies covering 74% to 82% of US births were obtained ( The purpose of this paper is to describe the many uses of mPINC data, including for national and state surveillance of infant feedingrelated maternity care practices, for hospital quality improvement, and for public health research.

Key messages
• The mPINC survey, a biennial census of US maternity care facilities, has been used for surveillance, quality improvement and research.
• Data from the mPINC survey have captured improvements in US maternity care practices and policies.
• The mPINC survey has been a valuable tool for hospitals, organizations, governments and researchers to improve breastfeeding support provided to mothers and infants.
COVID-19 pandemic, routine maternity care practices (e.g., maternalinfant rooming-in) were disrupted over concerns of SARS-CoV-2 transmission from an infected mother to her newborn. A supplemental questionnaire was quickly developed and is currently being administered to participating mPINC hospitals in order to understand how the pandemic has influenced hospital practices.

| mPINC: Data for individual hospital improvement
Each facility that participated in the mPINC survey received six copies of an individualized Benchmark Report, which was sent to the person who originally received the survey as well as to key leadership posi-

| mPINC: Data for public health research
In addition to the uses described above, mPINC data have also been used for public health research. For example, several studies have linked mPINC scores with breastfeeding data from other sources and have found breastfeeding outcomes were positively correlated with mPINC scores. Two studies (Barrera, Beauregard, Nelson, & Perrine, 2019;Patterson, Keuler, & Olson, 2018) linked mPINC data to in-hospital exclusive breastfeeding rates collected by The Joint Commission, finding that hospitals with higher mPINC scores had higher in-hospital exclusive breastfeeding rates. Self-reported maternal data from the Pregnancy Risk Assessment Monitoring System were linked with mPINC to demonstrate that women who delivered in hospitals with higher mPINC scores were more likely to be breastfeeding and to be exclusively breastfeeding at 8 weeks postpartum . The mPINC data from 48 hospitals in Alabama were linked with infant feeding data from the newborn screening database, finding infants delivered in hospitals, which had Structural and Organizational Aspects of Care Delivery, a mPINC subdomain, scores higher than the state mean were more likely to be breastfeeding .

| Limitations
Despite the benefits of the mPINC survey, it does have certain limitations. The mPINC data are self-reported by hospital staff.
Although CDC specifically asks for the contact information of the person with the most knowledge about these practices to send the survey to, and encourages input of other key staff, the validity and reliability of these data have not been formally evaluated nor has the scoring algorithm. As described above, studies have shown that higher mPINC scores are associated with exclusive breastfeeding at hospital discharge (Barrera et al., 2019) and any breastfeeding at 8 weeks

| CONCLUSION
The mPINC survey has been a valuable surveillance, quality improvement and research tool in tracking and helping improve maternity care practices, and policies in the US. Facilities have used their facility-specific Benchmark Reports to identify potential strengths and weakness in the care they provide to mothers and newborns. States and organizations have used state-specific data to make programmatic decisions to support breastfeeding among their constituents. National, aggregate data have captured marked improvements in infant feeding-related maternity care practices and policies in the United States. We anticipate that the new mPINC survey will continue to be an impactful tool for hospitals, organizations, governments and researchers in their efforts to ensure US maternity care practices and policies are fully supportive of breastfeeding.

ACKNOWLEDGMENTS
The authors would like to acknowledge the many collaborators and stakeholders who have provided input to CDC and the hospitals, and all of the participating hospital and birth centre staff who have made the mPINC survey successful. We would also like to acknowledge the CDC staff who have made contributions to the mPINC survey over the years, in particular Lawrence Grummer-Strawn, Katherine Shealy, Paulette Murphy and Kelley Scanlon and to thank Battelle for their partnership in implementing the survey.