Evaluation and Development of Potentially Better Practices for Improving Family-Centered Care in Neonatal Intensive Care Units
Objective. Technological and scientific advances have progressively decreased neonatal morbidity and mortality. Less attention has been given to meeting the psychosocial needs of the infant and family than on meeting the infant’s physical needs. Parents’ participation in making decisions and caring for their child has often been limited. Environments designed for efficient technological care may not be optimal for nurturing the growth and development of sick neonates or their families. Eleven centers collaborating on quality improvement tried to make the care of families better by focusing on understanding and improving family-centered care.
Methods. Through internal process analysis, review of the evidence, collaborative learning, and benchmarking site visits to centers of excellence in family-centered care, a list of potentially better practices was developed. Choice of which practices to implement and methods of implementation were center specific. Improvement goals were in 3 areas: parent-reported outcomes, staff beliefs and practices, and clinical outcomes in length of stay and feeding practices. Measurement tools for the first 2 areas were developed and pilots were conducted.
Results. Length of stay and feeding outcomes were not different before the collaboration (1998) and at the formal end of the collaboration (2000).
Conclusions. Prospective parent-reported outcomes are being collected, and the staff beliefs and practices questionnaire will be repeated in all centers to determine the impact of the project in those areas.
- family-centered care
- parental involvement
- collaborative quality improvement
- neonatal intensive care
- NIC/Q 2000
KEY POINTS OF ARTICLE
Family-centered care best practices are not limited to specific guidelines for family’s presence but include beliefs about family-centered care by leaders and the organization.
Making improvements in family-centered care may require changing the beliefs and culture of the unit.
Measurement tools for parent-reported outcomes and staff beliefs and practices were developed and piloted in this project.
Clinical outcomes of length of stay and feeding were not different before the project began and in the second year of the project (2000).
APPLYING LESSONS LEARNED TO PRACTICE
Family-centered care is a concept that must be integrated into the culture and functioning of a neonatal intensive care unit (NICU).
An environment that supports the presence and involvement of families may enhance family-centered care.
A high level of collaboration with families is more dependent on the attitudes of the care providers and the relationships that they establish with the families than on the physical facilities.
The absence of useful measurement tools at the outset of the project was a barrier but presented an opportunity to use the strength of the collaboration to develop them.
The field of neonatal intensive care has changed dramatically in the past 40 years. Technological and scientific advances have progressively decreased neonatal morbidity and mortality. However, the NICU environment is one of high stress, crises, and turbulent emotions for the families of premature and ill neonates.1–8 Less attention has been focused on finding the best ways to meet the psychosocial needs of the infant and family than on meeting the infant’s physical needs.9,10 Parents play the central role in providing for most children’s emotional, physical, social, and developmental needs, yet historically they have been limited in participating in their child’s care in the NICU.11–13 There is growing recognition that environments designed to make delivery of technological care efficient for staff may not be optimal for nurturing the growth and development of sick neonates or for their families.14–16
Family-centered care places the needs of the individual infant in the context of the family and redefines the relationship between parents and caregivers. Information sharing and collaboration are cornerstones of family-centered care, and they shape a unit’s culture, policies, programs, and facility design as well as the day-to-day interactions between caregivers and families.17 The potential benefits of family-centered care include improved satisfaction with care, decreased parental stress, increased parental comfort and competence with postdischarge care, improved success with breastfeeding, shortened hospital lengths of stay, decreased readmissions postdischarge, and increased staff satisfaction.18–21
The overall aim of this project was to use Vermont Oxford Network Neonatal Intensive Care Unit Quality Improvement Collaborative Year 2000 (NIC/Q 2000) quality improvement methods to foster early and ongoing family involvement in care, education, and administrative processes in the NICU. There were 3 areas that we believed would be improved: parent-reported outcomes, nursery staff beliefs and practices about family-centered care, and clinical outcomes. At the beginning of this project, there were no available tools for standardized outcome measurements in the first 2 areas. Part of this project involved developing these tools. Comparative data before and after the project are therefore unavailable. In the third area of clinical outcomes, we used the population of 401- to 1500-g birth weight infants to reflect improvements. Two reasons for selecting this population of infants were that they have long and complex hospital stays, and historical and current data would be available in the Vermont Oxford Network database on very low birth weight (VLBW) infants. Setting clinical improvement goals was difficult because of the indirect and complex relationships between improved family-centered care and clinical outcomes. Length of stay and feeding outcomes were chosen because of previously reported effects or hypothesized relations, as in support of feeding choices.18,19 The incidence of chronic lung disease has been reported to be reduced with individualized developmental care.22–24 Because there is some overlap in family-centered care and developmental care, the incidence of chronic lung disease at 36 weeks’ corrected age was analyzed, although no specific improvement goals were set. Clinical improvements were based on comparing data from 1998 to outcomes in 2000, except for the support of feeding plans, which was measured only for 2000.
The process for improving family-centered care was similar to that used by all of the NIC/Q 2000 focus groups. This included a detailed internal process analysis by each center, an ongoing review of the evidence base for proposed improvements, collaborative learning by conference calls, focus group activities at joint meetings, and planning and conducting benchmarking site visits to centers of excellence in family-centered care. One goal of these activities was to develop a list of potentially better practices (PBPs) for improving family-centered care. These evolving concepts were translated into quality improvement cycles that were individualized at each center. This article describes the steps taken to define a list of potentially better family-centered care practices, the development of measurement tools for parent-reported outcomes and staff beliefs and practices, and the comparative clinical outcomes.
Forming the Focus Group
Eleven centers in collaboration with the Institute for Family-Centered Care (IFCC) formed the focus group (Appendix A). A neonatologist led the group, which was also facilitated and supported by Vermont Oxford Network NIC/Q 2000 project faculty. Initial work included the definition of goals for the project, a review of the literature on family-centered care practices, and the development of a process for selecting benchmarks sites of excellence in family-centered care for visits. Biweekly teleconferences and communication via electronic mail listserv facilitated collaboration and project management.
Participating centers completed the “Family-Centered Care in Newborn Intensive Care Units: A Self-Assessment Inventory” developed by the IFCC.25 This survey helps a center assess policies, programs, practices, and the environment of the NICU. We believed that 2 additional areas that were covered less thoroughly by the inventory but also keys to the success of family-centered care were administrative support and continuity of care in the community. The group therefore expanded the survey to include assessment of strengths and opportunities for improvement in these areas as well. Completion of the expanded inventory tool created a common ground for dialogue among centers, as well as helped centers identify individual improvement goals. (After benchmark site visits were completed, the survey was edited to the shorter version in Appendix B.)
Benchmark Site Visits
Identifying benchmark sites of excellence in family-centered care was challenging. There are no standard rating scales for success in delivering family-centered care. Twelve potential centers were identified using published reports of programs, recommendations from the IFCC, the NIC/Q 2000 Collaborative Learning Directory, which identifies areas of clinical and organizational excellence, and suggestions by members of the focus group.18,26 Representatives of each of the 12 potential benchmark centers were interviewed by telephone using a standard set of questions (Table 1). Each of the 8 content areas was scored using a scale of 0 to 5. Four centers with a composite score of 37 or higher out of a maximum possible score of 40 were then contacted to arrange benchmark site visits. The 4 sites selected were the University of Massachusetts Memorial Medical Center (Worcester, MA), Rainbow Babies and Children Hospital (Cleveland, OH), Utah Valley Regional Medical Center (Provo, UT), and Evergreen Hospital Medical Center (Kirkland, WA).
Each center was asked to complete the expanded self-assessment of family-centered care (Appendix B) and return the responses before the site visit. Benchmark centers were also asked to provide key information about family-centered care, including vision and philosophy statements, visitation policies, and job descriptions. Centers were asked whether they would be willing to share specific Vermont Oxford Network VLBW outcomes data for comparison with the family-centered care focus group centers. For each visit, an individual from our focus group collaborated with a contact person at the benchmark site to clarify goals and objectives and to outline the agenda and details for the visit. Each center in the focus group decided which benchmark visits they would make. The composition and size of the visit team was also individual choice, although the most effective visiting teams were multidisciplinary, consisting at a minimum of physicians, nurses, and parents. Many centers also brought respiratory care technicians or other NICU staff. All 11 centers made at least 1 site visit, and most visited all 4. A typical site visit began with an informational presentation including a description of the NIC/Q 2000 project and the specific goals of the family-centered care focus group. Comparative outcome data on clinical outcomes were presented without identification of centers. The subsequent agenda for the day varied among sites but included tours of the unit(s) and support services departments (eg, physical/occupational therapy, family resource center, child care). Breakout sessions with families, administrators, and staff were excellent opportunities for group learning. It was particularly important for site visitors to have an opportunity to spend unstructured time on the unit talking with staff and families. At the end of the day, representatives from the benchmark site were asked to list the top 10 practices in their NICU that they believed supported and promoted family-centered care. The visiting centers then spent an hour discussing what had been learned during the visit. Assignments were made to compile all of the information and write a summary report of each visit.
Outcomes of care as reported by parents were important for assessing the impact of improvements in family-centered care. Although patient satisfaction surveys are now commonly used by hospitals to identify areas for improvement, few have been designed specifically for families of NICU patients.27–29 Available instruments were not particularly oriented toward family-centered care, were prohibitively expensive, or were insufficiently developed to use in the time frame of this project. The lack of acceptable tools for this project led us to design one. This tool for parents of NICU patients was based on a well-established and validated group of health status surveys developed by the Dartmouth COOP project.30,31 The tool is web-based and uses templates and formats of questions used for older age patients (http://www.howsyourbaby.com).
An initial draft of the survey questions was reviewed for content validity with multidisciplinary input from all centers, from representatives from the IFCC, and through focus groups with families of previous NICU patients. A pilot trial of the survey was conducted at 6 of the 11 focus group centers. One site with poor computer access conducted the pilot on paper. A coordinator at each site asked families to complete the computer-based survey. The goals of the pilot were to validate content and determine clarity, ease of use, and general impressions about format. The time to complete each survey was recorded by the site coordinator. After revisions, the survey was extended into a more prolonged trial in which 10 of the 11 centers are participating.
Questions that were designed specifically to address goals of improving family-centered care were as follows:
How well do you know your infant (personality; likes and dislikes; ways my infant uses to calm himself/herself)?
During your infant’s hospital stay, how often did you feel like a parent?
How ready do you feel you are to care for your infant after discharge from the hospital?
How do you feel about feeding your infant?
How have you felt about breastfeeding? (for mothers planning to breastfeed)
NICU Care Provider Questionnaire
One area expected to reflect improvements in family-centered care was NICU staff beliefs and practices related to family-centered care.32,33 A multidisciplinary subgroup from all 11 hospitals and a representative from the IFCC reviewed the available literature and measurement tools.34–36 Desirable characteristics were that the tool be comprehensive, appropriate for all staff who interact directly with infants and families, and able to measure change over time. Because no currently available tool completely met these criteria, a new tool was developed.
The process for developing a new staff beliefs and practices questionnaire began with generating a list of beliefs and practices that the subgroup believed reflected family-centered care. This list served as the basis for a draft questionnaire. The draft underwent several revisions after review by families of NICU graduates, NICU staff, and experts in survey development and family-centered care. A pilot trial of the questionnaire was then administered to 72 care providers in February 2000. Analysis of the data from the pilot and feedback from the 72 respondents was used to develop a final version of the questionnaire.
Quantitative clinical improvement goals for this population were 1) a 10% decrease in hospital length of stay, 2) a 10% decrease in the gestational age at which full oral feedings were achieved, and 3) support of family feeding plans resulting in 85% of infants at discharge being fed according to the feeding plan the family declared as their choice within 48 hours of admission. The incidence of chronic lung disease was also compared, but no specific improvement goal was set. We used data from the Vermont Oxford Network VLBW database for the clinical outcomes of length of stay and chronic lung disease. Length of stay was defined as the number of days to initial discharge home. Infants who were transferred to another hospital rather than discharged from the hospital were excluded. Chronic lung disease was defined as need for oxygen supplementation at 36 weeks’ corrected age. The incidence was imputed for infants who were no longer in the member hospital using a published algorithm in the Vermont Oxford Network Annual NICU Quality Management Report.37
Feeding data were obtained by retrospective chart review for infants who had a birth weight of 401 to 1500 g and were discharged from the hospital in 1998. Feeding data for infants who had the same birth weight and were discharged from the hospital in 2000 were collected prospectively and entered by each center into an Internet-based data entry program designed for this project. Definitions for data elements were the same for both periods. Gestational age is the Vermont Oxford Network database definition in completed weeks and days; date of full oral (PO) feeds is the date of the last gavage feeding + 1; gestation at full PO feeds is calculated by adding the weeks and days from birth to the date of full PO feeds to the gestational age at birth. Admission feeding plan was defined as the type of feeding the parent(s) intended to use after discharge from the hospital. The declaration of parent preference for feeding must be obtained within 48 hours of admission. The admission feeding plan was categorized as breast milk (the parents plan to feed breast milk exclusively, either by nursing or feeding expressed breast milk by bottle), combined (the parents plan to feed breast milk but plan to use formula when unable to breast feed), formula (the parents plan formula feeding only), and unknown (clear declaration of parent preference for postdischarge feeding was unavailable within 48 hours of admission, or parents have not decided). Discharge feeding plan was defined as breast milk (feeding at discharge is exclusively breast milk by nursing taking expressed breast milk from a bottle, nursing aid, or other artificial source or any combination), combined (feeding at discharge includes any combination of breast milk and formula supplement), or formula (feeding at discharge is by formula exclusively). Concordance between admission and discharge feedings plans for breastfeeding was defined as the percentage of mothers who had an admission feeding plan of either breast milk or combined and whose discharge feeding plan was breast milk or combined. Concordance for formula feeding was the percentage of mothers who had an admission plan of formula feeding and whose discharge feeding plan was formula feeding.
PBPs and Implementation
The PBPs developed through this collaborative quality improvement process are listed in Table 2. Ten content areas were identified. Within each content area is a list of more specific practices. Decisions about which practices to implement as well as how to prioritize them and which strategies to use to make improvements was at the discretion of each participating center. Table 2 also contains a summary of the implementation status among the centers for each of the practices at the conclusion of the formal NIC/Q 2000 project (September 2000). For each practice, the following possible statuses are tabulated: no plan (there are no plans to institute this practice), plan only (this practice will be instituted in the future, but no active changes in process), active change (improvement cycles addressing this practice are in progress but not completed), and complete (the practice is fully in place and only being monitored).
For the pilot trial, 76 surveys were completed at 6 sites. Of parents who completed the pilot surveys, 90.8% believed that all (48.7%) or most (42.1%) of the questions addressed important issues. Seventy-five of the 76 parents surveyed believed that the questions were clear (57.9%) or mostly clear (40.8%). Of the 69 parents who used the computer form of the survey, 95.7% believed the survey was very easy (66.7%) or easy (29.0%) to use. When asked whether they would be interested in reading more about the areas identified by the survey as concerns, 53.9% would like more information on the computer, 86.8% would like the information as printed handouts, and 75% would like information as a book. The average time for completion of the survey by computer varied from 10 to 15 minutes, with the variation mostly related to differences in computer hardware and connection speeds. The longest recorded time to complete was 25 minutes. Pilot center contacts reported few problems with administering the on-line survey.
NICU Care Provider Questionnaire
A pilot trial of the NICU Care Provider Questionnaire was conducted at 10 of the 11 centers. A contact at each center selected individuals to complete the 81-item questionnaire and an accompanying feedback form. For determining whether the questionnaire was relevant across disciplines, the pilot included nurses, neonatologists, unit secretaries, respiratory therapists, pharmacists, physical and occupational therapists, social workers, paid parents, and parent volunteers. Pilot participants represented a range of ages as well as years of experience in neonatal intensive care.
Data from 72 pilot questionnaires and feedback forms were analyzed and used to guide a final revision. Respondents evaluated the questionnaire on a scale of 1 to 4, where 1 = strongly agree and 4 = strongly disagree. Results for 3 attributes of the questionnaire were 1) directions were clear and easy to follow (mean: 2.0), 2) questions were easily understood (mean: 2.0), and 3) items were relevant to family-centered care (mean: 1.8). The questionnaire was completed on average in 26 minutes. Ten percent of the pilot participants believed that the questionnaire took too long to complete. The development subgroup also believed that reducing the time for completion would increase acceptance and return rate for the questionnaire. A detailed review led to deleting redundant questions and removing those that were unclear. The final questionnaire uses a 7-point Likert scale to elicit responses across 5 global areas: nursery environment, preparedness for traditional care and family-centered care, beliefs in family-centered care, family-centered practices, and level of organizational support. In addition, 3 open-ended questions ask respondents to identify barriers and challenges in practicing family-centered care, changes that they wish to make in their individual practice, and necessary resources to improve their work with families.
In May 2000, the revised 72-item questionnaire was distributed to all staff at the 11 family-centered care focus group centers, and 1143 surveys were completed. The questionnaire will be repeated at the same centers by the end of 2002.
Clinical outcomes are shown in Table 3. There was no difference in the gestational age or birth weight for VLBW infants who were admitted to focus group hospitals in 1998 compared with 2000. The length of stay to initial discharge home was the same for both years. The incidence of chronic lung disease was 31.4% in 2000, compared with 27.3% in 1998. For both the outcomes of length of stay and chronic lung disease, great variability existed among centers. The mean length of stay decreased in 7 of the 11 centers and increased in 4. The incidence of chronic lung disease increased in 7 centers and decreased in 4. Statistical analysis was not performed. There was no difference in the mean corrected gestational age at which full oral feedings was reached between 1998 and 2000 (36.2 weeks for both periods). The median corrected age for reaching full oral feedings varied among centers from 34.7 to 36.6 weeks. The second feeding outcome was the concordance between admission and discharge feeding plans. This outcome was calculated from prospectively collected data. The concordance for mothers who planned to breastfeed for all centers was 67%, with a range among centers from 51% to 91%. Concordance for formula feeding was 94% with little variation among centers. For breastfeeding and formula feeding combined, the concordance was 74%, with a range among centers from 68% to 87%.
Using quality improvement strategies to improve the practice of family-centered care proved challenging. Family-centered care is a concept that must be integrated into the culture and functioning of an intensive care unit.38 At the beginning of this collaborative project, individual centers were at different stages of belief in the values of family-centered care and in practicing according to them. This diversity was apparent from the internal reviews and as the participating centers learned about each other’s strengths and challenges through collaboration. The time frame for this project was relatively short to document changes objectively in behavior and culture. The benchmarking site visits were conducted in June and November of 1999, with the draft of the PBPs completed in December. Few of the practice concepts had been implemented completely by the time of the formal end of the project (Table 2). Vision and philosophy of care statements that articulate the unit’s commitment to family-centered care were in place in 6 of the 11 centers, with 1 actively in the process of creating them and 4 centers in the planning stages. Seven centers had policies that welcomed families 24 hours a day, including during rounds and change of shift, and the other 4 centers were either actively changing or planning such a policy. However, only 2 centers believed that families were truly viewed as integral members of the care team and as primary caregivers to their infants rather than as visitors. Another practice that was initiated or planned in all centers was the practice of families serving on family advisory boards for the unit. For many centers starting such advisory boards, there was inadequate time to measure actual changes that those boards might have stimulated. No other PBPs were completely implemented in even a majority of participating centers, although active change projects were in progress or planned for most. Many of the centers have identified major deficits in the physical environment of the NICU and have initiated planning processes for renovation or new construction to address those deficits. It became clear particularly through site visits that an environment that supports the presence and involvement of families could enhance family-centered care. It was equally clear that a high level of collaboration with families was more dependent on the attitudes of the care providers and the relationships that they established with the families of infants in their units than on the physical facilities. Several centers discovered that including families as full participants in their infant’s care depends on effective functioning of the multidisciplinary care team. Those centers have initiated projects to improve teamwork, projects that in many ways are as difficult as planning and building new physical facilities.
Literature supports family-centered care as a compelling philosophy for improving communication and relationships with families.1 Firm evidence to support an impact on physical outcomes is sparse. Families tell us that they want to have a voice in how they will participate in their children’s care, but there are few tools to measure how successful we are in giving them that voice.39,40 Preliminary evidence from the web-based parent survey developed for this project suggests that there will be great variation among centers in how successfully we practice family-centered care (data not shown). Comparative data should lead to better insights into areas for improvements. Success at supporting breastfeeding for mothers of VLBW infants is extremely limited. The proportion of women who provided breast milk for their VLBW infant varied from 46% to 91% among centers. Most of these mothers were providing expressed breast milk. Success at breastfeeding VLBW infants by suckling at the breast at discharge was rare.
The outcomes chosen at the beginning of this project were somewhat arbitrary and were based more on hypothesis than on existing evidence. It is not particularly surprising that this project was not associated with demonstrable improvements in length of stay, feeding, or chronic lung disease. Changes in parent-reported outcomes and in staff beliefs are perhaps more likely, but tools to measure those outcomes before and after the project did not exist. We plan to repeat the NICU Care Provider Questionnaire at all of these centers at a time more likely to reflect changes that have been initiated, as well as to measure the parent-reported outcomes over time. We believe that 1 of the successes of this project will be to make these tools available to other centers that are interested in improving their care and in establishing some baseline data from centers participating in this focus group that can be used for comparison.
The PBPs that this focus group discovered are a combination of generic principles of leadership and teamwork as well as specific guidelines for creating structures to foster collaboration with families across all areas of functioning of the NICU. Applying their practices will lead to a more fulfilling work environment for care providers and an environment where families are respected as essential partners in the care of their children.
- Affleck G, Tennen H, Rowe J. Infants in Crisis: How Parents Cope With Newborn Intensive Care and Its Aftermath. New York, NY: Springer-Verlag; 1991
- Pinch WJ, Spielman ML. Parental perceptions of ethical issues post-NICU discharge. West J Nurs Res.1993;15 :422– 437
- Stainton MC, Harvey S, McNeil D. Understanding Uncertain Motherhood: A Phenomenological Study of Women in High-Risk Perinatal Situations. Calgary, Alberta, Canada: University of Calgary; 1995
- ↵Cuttini M, Rebagliato M, Bortoli P, et al. Parental visiting, communication, and participation in ethical decisions: a comparison of neonatal unit policies in Europe. Arch Dis Child Fetal Neonatal Ed.1999;81 :F84– F91
- ↵Committee to Establish Recommended Standards for Newborn ICU Design. Recommended standards for newborn ICU design. J Perinatol.1999;19(pt 2) :S2– S12
- ↵Hanson JL, Jeppson ES, Johnson BH, Thomas J. Newborn Intensive Care: Resources for Family-Centered Practice. Bethesda, MD: Institute for Family-Centered Care; 1997
- ↵Forsythe P. New practices in the transitional care center improve outcomes for babies and their families. J Perinatol.1998;18(6 pt 2 suppl) :S13– S17
- ↵Meyer EC, Coll CT, Lester BM, Boukydis CF, McDonough SM, Oh W. Family-based intervention improves maternal psychological well being and feeding interaction of preterm infants. Pediatrics.1994;93 :241– 246
- Als H, Lawhon G, Brown E, et al. Individualized behavioral and environmental care for the very low birth weight preterm infant at high risk for bronchopulmonary dysplasia: neonatal intensive care unit and developmental outcome. Pediatrics.1986;78 :1123– 1132
- ↵Institute for Family-Centered Care. Family-Centered Care in Newborn Intensive Care Units: A Self-Assessment Inventory. Bethesda, MD: Institute for Family-Centered Care; 1997
- ↵Vermont Oxford Network. NIC/Q 2000 Collaborative Learning Directory. Burlington, VT: Vermont Oxford Network
- ↵Conner JM, Nelson EC. Neonatal intensive care: satisfaction measured from a parent’s perspective. Pediatrics.1999;103(1 suppl) :336– 349
- ↵McWilliam PJ. Brass Tacks: A Self-Rating of Family-Centered Practices in Early Intervention: Neonatal Intensive Care. Chapel Hill, NC: Frank Porter Graham Child Development Center, University of North Carolina; 1992
- King S, Rosenbaum P, King G. The Measure of Processes of Care (MPOC): A Means to Assess Family-Centered Behaviors of Health Care Providers. Hamilton, Ontario, Canada: Neurodevelopmental Clinical Research Unit, McMaster University; 1995
- ↵Horbar JD, Rogowski J, Plsek PE, et al. Collaborative quality improvement for neonatal intensive care. NIC/Q Project Investigators of the Vermont Oxford Network. Pediatrics.2001;107 :14– 22
- ↵Harrison H. The principles of family-centered neonatal care. Pediatrics.1993;92 :643– 650
- Copyright © 2003 by the American Academy of Pediatrics