(1) Quantify and compare the family's and the nurse's perception regarding the family's discharge preparedness. (2) Determine which elements contribute to a family's discharge preparedness.
We studied the families of all the infants discharged from a neonatal intensive care unit after a minimum of a 2-week admission. The families rated their overall discharge preparedness with a 9-point Likert scale on the day of discharge. Independently, the discharging nurse evaluated the family's discharge preparedness. Families were considered discharge ‘prepared’ if they rated themselves and the nurse rated their technical and emotional preparedness as 7 on the Likert scale.
We had 867 (58%) family–nurse pairs who completed the survey. Most families (87%) were prepared for discharge as assessed by the concordant questionnaire (Likert scores of 7 by the parent and the nurse). In multivariate analysis, confidence in their child's health and maturity (odds ratios, OR=2.5 95% confidence interval, CI (1.2, 5.3)), their readiness for their infants to come home (OR=2.9 95% CI (1.0, 8.3)), and selecting a pediatrician (OR=4.2 95% CI (1.6, 11.0)) were statistically significant.
Assistance with pediatrician selection and home preparation may improve the percentage of families prepared for discharge.
Discharge preparation in the neonatal intensive care unit (NICU) is critically important.1 Even for families with healthy term infants, poor discharge preparation has been linked to problems at home and increased unscheduled health care use.2, 3 Poor discharge planning has the potential for even worse outcomes for preterm infants because of their poor cues to needs and ongoing medical problems after hospital discharge, in particular, those with lung disease and others at the highest risk for rehospitalization.4, 5, 6, 7, 8
Rehospitalization and other post-discharge health care use constitute a significant proportion of the preterm infants care costs.9 A recent report from the Institute of Medicine estimates the annual preterm birth costs at US $26.2 billion with 16.9 billion associated with infant medical care.9 Adequate parental education can reduce the risk of readmission by ensuring that the parents seek medical attention appropriately, administer medications and other therapies correctly, and show confidence in the home management of non-acute medical problems.1
Interventions to teach parents about the skills needed should result in parents being discharge prepared.1, 3, 7, 10, 11 However, discharge readiness is decided by both the parents and the clinical staff.2 Full-term infant studies indicate that, despite discharge teaching, some parents do not feel adequately prepared.2, 3 Among preterm infants, this is a crucial aspect of neonatal care that has had only limited research. We are only aware of one group exploring this issue for the NICU patients for whom the skill level needed by parents is generally higher.7
We sought to augment the published literature by first quantifying how prepared the discharge families felt on leaving the hospital. Then, we compared the family's self-perception with that of the nurse's regarding discharge preparedness. Finally, we wanted to determine which elements contribute the most to a family's discharge preparedness.
The discharge process
In our NICU, discharge planning begins shortly after admission. There is standard discharge teaching for all families of newborns, which includes basic infant care skills such as feeding, bathing, and temperature assessing. Families are also instructed in signs of illness and criteria for which to call their pediatrician. Families are asked the name of their pediatrician and encouraged to make a follow-up appointment before NICU discharge. As part of the discharge process, NICU families are offered instruction in cardiopulmonary resuscitation. Families of medically complex infants receive the standard teaching as well as special teaching, which targets their infant's complex needs. The nurse who provided the teaching documents all of the discharge teaching on the discharge teaching checklist. The attending physician determines the timing of discharge based on their knowledge of the infant's physiologic competency and family readiness through incorporating the perceptions of the nursing and social work staff.
Conceptualization of this study
We examined decision making by families and the nursing staff and characterized the decision-making process between this pair of informants by obtaining their respective discharge preparedness perception on the discharge day. The hypothesized correlates of discharge preparedness incorporated infant, mother, and nursing descriptive characteristics, as well as family technical skills and emotional elements. We assumed that by discharging the infant, the attending physician had deemed the infant/family ready for discharge. Thus, we did not elicit the physician's evaluation.
Beth Israel Deaconess Medical Center is a major teaching hospital of the Harvard Medical School with approximately 5000 deliveries annually. The NICU, which averages 800 admissions per year, has 40 intermediate and intensive care beds (levels II and III) with an average daily census of 38 admissions. More than 80 clinical nurses, 8 neonatal nurse practitioners, 11 neonatal respiratory therapists, 14 neonatologists, and 18 neonatal–perinatal fellows staff the NICU.
Instrument design and implementation
The study instrument was derived from a discharge preparedness assessment instrument used for healthy term babies in the Life Around Newborn Discharge Study described elsewhere.2, 12 We adapted their instrument for preterm infants based on recommendations from the American Academy of Pediatrics13 about essential parenting capabilities and topics raised by Sheikh et al.7
We studied families of all infants discharged from our NICU after a minimum of a 2-week admission. On discharge day, families rated their overall discharge preparedness on a 9-point Likert scale with the following anchors: ‘Not at all prepared’ and ‘Very prepared’. Questions focused on the standard teaching received by all families. Families then appraised their discharge preparedness on 14 individual questions. Eleven questions addressed technical expertise13 (infant care skills as well as preterm infant needs) and were divided into four response categories: ‘This does not apply’, ‘Not at all prepared’, ‘Somewhat prepared’, and ‘Prepared’. Three questions addressed emotional discharge preparedness determinates (e.g. confidence in ability) in three response categories: ‘Not at all prepared’, ‘Somewhat prepared’, and ‘Very prepared’.
In addition, on discharge day, the discharging nurse independently evaluated the family's overall emotional and technical discharge preparedness using a 9-point Likert scale. Each discharging nurse characterized herself as being a primary nurse, a member of the infant's nursing team, familiar with the family, and/or familiar with the infant. Families and nurses were blinded to each other's responses. Before use in this study, we pre-tested the instrument for readability, clarity, and face validity on a convenience sample of five nurses and five families.
We then conducted an 8-month prospective observational pilot study of 120 family–nurse dyads. We collected descriptive characteristics of the participants and non-participants in this pilot study. There were no differences between the participants and non-participants in the pilot study. The pilot data were used as the basis for the Beth Israel Deaconess Medical Center NICU discharge preparedness quality improvement initiative.
Quality improvement data were then prospectively collected between November 2003 and April 2007. All families and discharging nurses were given the discharge preparedness instrument as part of the discharge day process. On a monthly basis, the percentage of families that reported feeling unprepared for discharge was reported to the NICU Leadership Committee. This quality improvement monthly report only contained aggregate summary data.
To review individual family and nurse responses, we obtained approval from the Beth Israel Deaconess Medical Center Institutional Review Board to conduct a retrospective data analysis of discharge preparedness quality improvement data. We were allowed to conduct this study after excluding the pilot data and removing the major identifiers from the quality improvement data. In an effort to protect the nursing staff, the Institutional Review Board explicitly forbade the collection of more specific demographic information about the discharging nurse beyond that presented.
Staff reviewed medical records for maternal age, parity, and gravidity as well as infant gender, gestational age, birth and discharge weights, 1 and 5 min Apgar scores, and length of stay. ‘Medically complex’ infants were considered to be any with conditions that required more special teaching such as discharge with a gastrostomy tube, central venous line, ileostomy/colostomy, tracheostomy, ventriculoperitoneal/subgaleal shunt, home medications (not including vitamins or iron), home oxygen, or a cardiorespiratory monitor. These conditions were infrequent.
Discharge preparedness definition
Families were considered discharge ‘prepared’ if they rated themselves and the nurse rated their technical and emotional preparedness 7 on the Likert scale. Families were considered discharge ‘unprepared’ if they rated their overall level of preparedness or their nurse rated their technical or emotional level of preparedness <7. ‘Prepared’ was a dichotomous variable.
Discharge preparedness global assessment
We calculated an average technical score weighted according to applicability of each family's technical skills and an average family emotional score based on their responses to the emotional questions. We multiplied these scores (proportions) by 10 so that they would be of the same order of magnitude as the measured family's overall preparedness score for ease of comparison. The family's overall preparedness score and the nurse's technical and emotional scores were continuous variables for the correlation analyses.
We performed non-parametric Spearman correlations between the family's overall self-assessment, the calculated family technical and emotional scores, and the nurse's technical and emotional family assessments.
Discharge preparedness correlates
Using logistic regression, we investigated the bivariate and multivariate relationships between discharge preparedness and its correlates. Variables significant at the 0.05 level in the bivariate analysis were entered in the multivariate analysis. Subsequently, multivariate models were examined using stepwise, forward selection, and backward elimination model-building strategies—all yielding the same final model.14 Only those variables that remained significant at the 0.05 level were retained in the final model. Unadjusted and adjusted odds ratios (OR) with corresponding 95% confidence intervals (CI) and P-values are provided for these analyses. For model building, only those with complete data were included. Finally, we tested the effect of the correlates of discharge preparedness on being discharge prepared by adding them to the logistic regression model individually. We calculated the relative contribution of each significant correlate by examining the marginal increase in the model χ2-test accounted for by each correlate as it was added and removed from a model containing all the significant correlates.15, 16 All statistical analyses were done using SAS version 9.1 (SAS Institute Cary, NC 2002).
There were 1492 eligible discharges during the study period. Families returned 1059 (71%) and nurses returned 992 (66%) discharge questionnaires. We had complete data for analyses for 867 family–nurse pairs (response rate of 58%). Demographic information on the non-participants was not routinely collected as part of the quality improvement protocol and could not be retrospectively acquired.
Mother and infant descriptive information is presented in Table 1. Thirty-eight percent of mothers were from minority racial groups (non-White and/or Hispanic). The median (range) maternal age was 34 years (16–47 years) and 57% had not given birth earlier. Mothers were respondents 81% of the time—alone (54%) with other parent (27%). We know from other quality improvement measures during their NICU hospitalization that 90% of our mothers are part of a couple.
At discharge time, 13% of families were ‘unprepared’ as defined by a Likert response of <7 by either the family member or nursing assessment. The assessment of unprepared was concordant between the family and nurse in <1% of all cases. However, most families were deemed unprepared because of the nurse's emotional or technical assessments. Among the families, 3% reported feeling less than prepared, but the nurses reported 9% of the families were emotionally and 9% of the families were technically less than prepared (P-values of <0.001 and 0.003, respectively).
Discharge preparedness global assessment
The majority of families reported feeling prepared for discharge (97%). The families’ mean (standard deviation) overall preparedness self-assessment score was 8.4±0.9. They had a technically preparedness score of 9.6±0.8 and an emotionally preparedness score of 9.4±1.3 calculated from the technical and emotional questions, respectively. The families’ overall preparedness score was highly correlated with calculated technical and emotional scores (both had P-values <0.0001).
Comparing prepared families to those who were not, showed a highly significant difference in the overall preparedness self-assessment scores 8.5±0.7 and 7.7±1.4, respectively (P<0.0001). The technical (9.6±0.8 and 9.4±1.2, respectively) and emotional (9.5±1.2 and 8.7±2.0, respectively) preparedness scores also had statistical significance P-values=0.004 and <0.0001, respectively.
The family self-assessed their overall degree of discharge preparedness (8.4±0.9) higher than the nurse's technical (mean 8.3±0.9) and emotional (mean 8.3±0.9) assessments (both P<0.0001). The family's calculated technical and emotional scores correlated, respectively, with the nurse's technical and emotional assessment (both P<0.0001). Despite the highly statistically significant difference between these two measures, the clinical interpretation is limited to the nursing staff not feeling as confident with the families as the families did with themselves.
Discharge preparedness correlates
Mother, infant and discharging nurse descriptive characteristics
Table 1 provides discharge prepared and unprepared mother, infant, and discharging nurse descriptive characteristics. Families with heavier and/or more medically complex infants were more likely to feel prepared for discharge. We were unable to distinguish prepared and unprepared families based on any maternal or discharging nurse characteristic. We noted that the discharging nurse often did not have a long-term relationship with the family: She was the primary nurse (16%) or a member of the nursing team (21%) infrequently; she was familiar with the infant (58%) and the family (56%) only a slight majority of the time. Nurses who reported families as unprepared were less likely to report being familiar with the infant or family (Table 1). Although not statistically significant, at the time of discharge among the ‘unprepared’, a lower percentage of nurses reported familiarity with infants and families.
Family self-reported technical correlates of discharge preparedness
Table 2 provides the family's self-reported technical discharge preparedness correlates. For this table when the skill was applicable, we collapsed the responses ‘somewhat prepared’ and ‘not at all prepared’ into ‘not completely prepared’. Prepared families were more confident with several things: bottle feeding, baby care skills, expectations of infant bowel/bladder function, preparation of a bed for the infant, recognition signs of illness in the infant, selection of a pediatrician, arrangement for help at home, and understanding special programs for preterm infants.
Family self-reported emotional correlates of discharge preparedness
Table 3 provides the family's self-reported emotional discharge preparedness correlates. For this table, we dichotomized the responses into very confident/very ready or not very confident/not very ready (i.e. consolidating somewhat confident/ready and not at all confident/ready). Families were more likely to be prepared at discharge when very confident with their infant's breathing and heart rate, their infant's health and maturity, and their readiness for their infants to come home. Physiological ‘maturity’ implies that infants are able to coordinate breathing and oral feeds, ingest adequate volumes and gain weight, maintain normal body temperature in an open environment, and maintain a stable cardiorespiratory function.1
Discharge preparedness multivariate correlates
The bivariate and multivariate results are presented in Table 4. After controlling for infant characteristics, technical skills, and emotional responses, the factors remaining statistically significant included preparedness with selecting a pediatrician (OR=4.2, 95% CI (1.6, 11.0)), their readiness for their infants to come home (OR=2.9, 95% CI (1.0, 8.3)), and feeling confident with their infant's health and maturity (OR=2.5, 95% CI (1.2, 5.3)). The significant correlates of discharged preparedness assessed by the uniquely attributable χ2-test were preparedness with selecting a pediatrician (88%), their readiness for their infants to come home (11%), and feeling confident with their infant's health and maturity (1%).
We found that most families are prepared for NICU discharge. Usually, the discharging nurse rated the family's discharge preparedness lower than the family's own rating. Discharge preparedness was associated with infant, but not maternal or discharging nurse, characteristics. Controlling for other significant factors, the family was prepared for discharge when confident with their pediatrician selection, the infant's health and maturity, and their home environment.
Little published literature exists on NICU discharge preparedness perceptions. We examined discharge preparedness from the family's and the nurse's perspective. Fewer than 1% of families were unprepared by all assessments confirming the discrepancy in the discharge preparedness between the nurse's evaluation and the family's self-assessment noted in some literature.7, 17 Generally, families rated their discharge preparedness higher than the nurse's rating. It is probable that the nurse and family used different evaluation criteria. We speculate that the family felt comfortable in providing infant care in the protected NICU hospital environment. The nurse's evaluation likely represents a subjective opinion of how the family will fare without support. That nurse's subjective discharging opinion may have been biased negatively relying solely on the family's performance on the discharge day because she was often unfamiliar with the family. Although the difference between the nursing and family scores was statistically significant, its clinical significance is uncertain. This might be better detected in future studies that correlate discharge preparedness with long-term outcomes.
The American Academy of Pediatrics recommends that criteria for hospital discharge timing be based on physiological maturity and stability instead of weight attainment.1 Specifically, infants are discharged when they are able to coordinate breathing and oral feeds, ingest adequate volumes and gain weight, maintain normal body temperature in an open environment, and maintain stable cardiorespiratory function.1
Although awaiting attainment of these milestones, most NICUs expect parents to master necessary baby care skills by discharge and routinely assess the parent's level of competence.13, 18 Most families in our study felt adept with routine technical skills taught as part of the discharge process. In a recent qualitative study, Broedsgaard identified that parents felt their discharge teaching contributed to the majority of their confidence in caring for their preterm infant(s) and the infant's well-being after discharge.19 Discharge teaching is a key step in infant introduction into an existing family unit, thus helping families make informed decisions, have basic technical baby care skills, and recognize signs and develop appropriate responses to illness.20, 21, 22 Families actively seek this knowledge.23 Although discharge teaching is the responsibility of the entire medical team, the bulk of the teaching is conducted by the nursing staff.6, 7, 13, 20 The high rate of family preparedness in this study suggests adequate discharge teaching.
We noted a disparity between the overall discharge preparedness scores of the ‘prepared’ and ‘unprepared’ families. We also noted a disparity between the technically and emotionally preparedness scores of the ‘prepared’ and ‘unprepared’ families. The most pronounced disparity was with the emotional scores, which for prepared families was 9.5±1.2 and for unprepared families was 8.7±2.0. This incongruity would imply that some unmeasured, likely emotional factor, may exist. The limited number of items queried by our instrument did not allow us to explore this hypothesis further. We speculate that earlier preterm infant exposure and local/family/community support may also contribute.22 Qualitative research methods such as open-ended questions might better delineate hard-to-quantify items including family adaptability, optimism, resiliency, and cohesiveness.24
Another limitation is that we did not assess educational attainment or couple status, which have been shown in other studies to be related to perceived discharge readiness.3 These factors likely have a limited impact in our population because the majority is part of a couple and has some college education or more. Unlike the pilot study, we did not collect data on the families and nurses not included in this study. In the pilot study, the families and nurses included were almost identical to the families not in the study, but selection bias cannot be excluded in any study with <100% participation.
The discharging nurse may not have been the best source of evaluation of family preparedness. We were limited to the discharging nurse's assessment of the family because we wanted the family to complete all discharge teaching before evaluation. We accepted this limitation, given the unpredictability of nursing schedules/assignments and fluctuations of the daily NICU census, making a primary nurse or a nursing team member inconsistently available at discharge, especially for those babies posted for home discharge.
Despite the limitations, we found that most families were prepared for discharge. We found a distinctive set of characteristics (family was confident with their pediatrician, home environment, and the infant's health and maturity), associated with being prepared for discharge, which distinguishes our data from some of the published reports. Other studies have prospectively examined discharge preparedness on discharge day in term infants.2, 3 However, for preterm infants, this issue has only been addressed retrospectively 4–6 weeks after discharge.7 This study adds to the literature because we combine the concepts of discharge day preparedness for families with preterm infants. Furthermore, our results would imply that efforts to assist families with pediatrician selection and home preparation may improve the percentage of families actually prepared for discharge. Having the families make the first pediatrician appointment before NICU departure, may be a way to strongly encourage pediatrician selection.
There is still a gap in our understanding of how discharge preparedness affects long-term outcomes of preterm infants and their families. Some suggest that discharge teaching prevents some difficulties at home after discharge.2, 3 We do not know for a fact whether the ‘prepared’ families had fewer difficulties than the ‘unprepared’ families. Researchers need to further elucidate the impact of discharge preparedness on the discharge day by correlating it with future outcomes including infant development and growth, readmission to the hospital, immunization attainment, and family adaptation and continued confidence. This gap could be addressed by conducting a longitudinal follow-up study in conjunction with a discharge readiness assessment.
Funding for the project was from NRSA institutional training grant AHRQ T32 HS 000063 to the Harvard Pediatric Health Services Research Fellowship Program and the Harvard School of Public Health Department of Maternal Child Health Educational training grant MCHP 2T 76MC 00001-47. An early version of the pilot work for this manuscript was presented in abstract and poster form at the Pediatric Academic Societies’ Annual Meeting in San Francisco, California in May 2004. We have no conflict of interest.