- Email: [email protected]
Dietitian Involvement in the Neonatal Intensive Care Unit: More Is Better
RESEARCH Current Research
Continuing Education Questionnaire, page 1240 Meets Learning Need Codes 4150, 5000, 5010, and 5060
Dietitian Involvement in the Neonatal Intensive Care Unit: More Is Better IRENE E. OLSEN, PhD, RD; DOUGLAS K. RICHARDSON, MD, MBA; CHRISTOPHER H. SCHMID, PhD; LYNNE M. AUSMAN, DSc, RD; JOHANNA T. DWYER, DSc, RD
I. E. Olsen is an assistant professor, Cincinnati Children’s Hospital Medical Center and the University of Cincinnati College of Medicine, Center for Epidemiology & Biostatistics and Division of Neonatology, Cincinnati, OH. At the time of the study, she was a neonatal dietitian, Department of Nutrition, Beth Israel Deaconess Medical Center, Boston, MA and PhD candidate, Gerald J. and Dorothy R. Friedman School of Nutrition Science and Policy, Tufts University, Boston, MA. D. K. Richardson was a neonatologist and Director of Research, Department of Neonatology, Beth Israel Deaconess Medical Center, Boston, MA, and a professor, Department of Maternal and Child Health, Harvard School of Public Health, Boston, MA. C. H. Schmid is a senior statistician, Institute for Clinical Research and Health Policy Studies, Tufts-New England Medical Center, Boston, MA. L. M. Ausman is a professor and Dean of Students, Gerald J. and Dorothy R. Friedman School of Nutrition Science and Policy, Tufts University, Boston, MA; adjunct professor, School of Medicine, Tufts University, Boston, MA; and scientist, Jean Mayer US Department of Agriculture Human Nutrition Research Center on Aging, Tufts University, Boston, MA. T. Dwyer is a scientist and professor in the ]?⬎Gerald J. and Dorothy R. Friedman School of Nutrition Science and Policy, in the School of Medicine, and in the Jean Mayer US Department of Agriculture Human Nutrition Research Center on Aging, all at Tufts University, Boston, MA; she is also a professor with the Department of Maternal and Child Health, Harvard School of Public Health, Boston, MA; and director of the Frances Stern Nutrition Center, TuftsNew England Medical Center, Boston, MA. Address correspondence to: Irene E. Olsen, PhD, RD, Assistant Professor, Cincinnati Children’s Hospital Medical Center and the University of Cincinnati College of Medicine, Center for Epidemiology & Biostatistics and Division of Neonatology, Cincinnati Children’s Hospital Medical Center (MLC 5041), 3333 Burnet Ave, Cincinnati, OH 45229. E-mail: [email protected] Copyright © 2005 by the American Dietetic Association. 0002-8223/05/10508-0004$30.00/0 doi: 10.1016/j.jada.2005.05.012
1224
Journal of the AMERICAN DIETETIC ASSOCIATION
ABSTRACT Objective Describe the level of registered dietitian (RD) involvement in neonatal intensive care units (NICUs) and associations with NICU nutrition practices. Design Questionnaires were mailed to 820 NICUs in the United States with two follow-up mailings to nonresponders. Abbreviated phone surveys were conducted with a random sample of 10% of nonresponders. A nutrition care score was devised based on a sum of 10 survey questions (range 0 to 10) to summarize the intensity of reported practices. Subjects/setting Directors of NICUs in the United States and RDs associated with them. Statistical analyses 2, analysis of variance, Bonferroni and Duncan multiple range tests, regression. Results Respondents from 417 (54%) of the 772 NICUs eligible for the study provided completed questionnaires. Among NICUs responding, 76% involved RDs in care (41% employed full- or part-time RDs, 35% employed consult RDs), and 24% had no RD. NICUs with full- or part-time RDs provided fewer kilocalories and more protein parenterally, and more kilocalories and protein enterally. NICUs with less RD involvement were more likely to provide full-term infant feedings (eg, unfortified breast milk, full-term formula) to very-low-birth-weight infants. Mean nutrition care score varied with RD involvement from 4.6⫾1.7 (mean⫾standard deviation) for NICUs with a consult RD and 4.7⫾1.4 for NICUs employing no RD to 5.6⫾1.7 for NICUs with a full- or part-time RD (overall P⬍.001). Conclusions More involvement of RDs in NICUs increased the intensity of important aspects of nutrition care that may improve outcomes of very-low-birth-weight infants in NICUs. These findings highlight the importance of RDs as NICU team members. J Am Diet Assoc. 2005;105:1224-1230.
R
egistered dietitians (RDs) have special expertise on the nutrition issues of very-low-birth-weight (VLBW) infants. While other members of the neonatal intensive care unit (NICU) team focus on infant daily care, procedures, and life-saving events, the RD’s
© 2005 by the American Dietetic Association
primary focus on evidence-based, individualized nutrition care helps to ensure that this is provided consistently to each infant. RD involvement in the NICU, such as in the design of NICU nutrition practice protocols and monitoring tools or direct nutrition interventions, is associated with improved nutritional intake and infant growth (1-5), shortened hospital length of stay (1,3,5,6), and reduced related costs (3,4). However, a 1990 survey of American NICUs reported that only 38% of responding NICUs involved an RD in the care of their infants (7), suggesting that the potential benefits of RD involvement were not fully realized in NICUs. This study describes the current level of RD involvement in NICUs in the United States and how reported nutrition practices differ by the level of RD involvement. Our hypothesis was that NICUs with a full- or part-time RD would be more likely to frequently monitor growth and use aggressive nutrition practices for their VLBW infants than those with less RD involvement. In addition, we expected NICU nutrition practices to be more intensive overall when the NICU RD had received more specialty experience, training, and education related to clinical practice. METHODS The Beth Israel Deaconess Medical Center’s Committee on Clinical Investigations approved this study. Data were collected using an 8-page, mailed survey questionnaire focused on nutrition personnel and practices in NICUs caring for VLBW infants (⬍1,500 g at birth). The questionnaire was pilot-tested in six New England NICUs (8). A personalized letter describing the study and requesting return of the questionnaire in the self-addressed, stamped envelope was sent in May 1999 to all of the 820 NICUs listed in the American Academy of Pediatrics NICU Directory (1999). Nonresponders were sent a reminder postcard after 4 weeks and a second mailing of the questionnaire after 8 weeks. Return of the survey questionnaire was also encouraged through two NICU-related electronic mailing lists and the American Dietetic Association Pediatric Nutrition dietetic practice group. The study was closed in November 1999. At that time, a random sample of 10% of nonresponders was selected for comparison to the NICUs that returned surveys and an abbreviated phone survey describing the NICU, its personnel, and practices related to nutrition was administered. The survey questionnaire collected information about the characteristics of the NICUs (eg, number of beds, performs surgery or not), nutrition practices (eg, timing of initiation of parenteral nutrition, use of human milk fortifier in breast milk), growth surveillance (eg, anthropometrics and frequency), and which personnel were involved in making nutrition-related decisions. Assessment of the level of RD NICU involvement was coded as: full- or part-time RD, consult RD (defined as no assigned hours to the NICU), or no RD. If full- or part-time RDs were involved in the NICU, the respondent was asked to give the questionnaire to the RD to complete, including details on training, experience, and education. Directors of NICUs who responded that they did not employ a full- or part-time RD were requested to complete the survey
themselves and to comment on the reasons for the lack of RD involvement in the NICU. In light of the expanding body of literature on the benefits of aggressive nutrition practices for VLBW infants available in 1999 when the survey was conducted (9-16), a nutrition care score was created to help summarize the intensity of the NICUs’ nutrition and growth surveillance practices and to facilitate comparisons between groups. Criteria for the nutrition care score (Table 1) were selected by one of the authors (I.E.O.) based on a review of recommendations for NICU nutrition practices in the relevant literature (9-16) and clinical experience, and then critiqued by a neonatologist (D.K.R.) and another NICU RD. One point was assigned to each of the 10 criteria for a possible score of 0 to 10 (least to most). More frequent growth surveillance and aggressive nutrition practices were assigned a point for intensity of care. The nutrition care score was calculated for all NICUs that provided a response for each of the 10 criteria. Data analyses were conducted with SAS 8.0 (SAS, Cary, NC, 1999). NICU characteristics, nutrition practices, nutrition care score, and its criteria were compared among RD groups: no RD, consult RD, and full- or parttime RD. Within the full- or part-time RD group, the nutrition care score and its criteria were also examined by the number of hours spent in the NICU, years of NICU and total clinical experience, and whether the RD had a graduate degree. Significance was determined at the .05 level. Overall comparisons of means and frequencies were conducted using analysis of variance and 2, respectively. Multiple pairwise comparisons of RD groups used Duncan Multiple Range test for means and Bonferroni procedure for frequencies. Regression was used to test the nutrition care score for potential confounding of NICU size and an interaction between RD group and NICU size (ie, if the differences in nutrition care score by RD group were explained by NICU size and/or varied by NICU size, respectively). RESULTS Of the 820 NICUs listed in the American Academy of Pediatrics directory, eight were declared ineligible because respondents indicated that either they did not care for VLBW infants (n⫽5) or that the NICU had closed (n⫽3). The 395 NICUs that did not respond to the survey included 37 NICUs with uncorrectable addresses. Of the remaining nonresponders, a random phone survey (n⫽37) found that one NICU did not care for VLBW infants, two were closed, and one was a duplicate listing. Applying these percentages of ineligible NICUs to the group of unreturned surveys, we projected an additional 40 NICUs would have been ineligible for our study. We received 417 completed questionnaires, and therefore our final response rate was 54% (417 of 772). Among the 417 responding NICUs, 76% involved RDs: 41% employed full- or part-time RDs, 35% employed consult RDs, and 24% had no RD. Among the 244 NICUs without a full- or part-time RD, the three most common reasons given by NICU directors for the lack of RD involvement were lack of need (49%), lack of funding (35%), and lack of availability of a trained NICU RD (25%). Table 2 compares NICU size and characteristics with respect to the level of RD involvement. NICUs with full-
August 2005 ● Journal of the AMERICAN DIETETIC ASSOCIATION
1225
Table 1. Comparison of the percent of NICUsa that met each criterion of the nutrition care scoreb, used to summarize the intensity of the NICUs’ nutrition and growth surveillance practices, among NICUs with differing levels of registered dietitian (RD) involvement Consultant RD (nⴝ145)
No RD (nⴝ99)
Please tell us if you: 1. Plot and trend weight daily 2. Plot and trend head circumference weekly or more 3. Plot and trend length weekly or more 4. Parenteral nutrition started within first 48 hours of life 5. Enteral nutrition started within first 72 hours of life 6. Parenteral nutrition goal: 90-110 kcal/kg/d 7. Parenteral nutrition goal: ⱖ3.5 g protein/ kg/d 8. Provide ⬎24 kcal/oz enteral feedings 9. Use enteral protein supplement 10. Iron supplement started at: ⱕ14 days of life, or ⱕ“full feeds” Nutrition Care Score (mean⫾standard deviation)a
n
Total responses
%
36
94
38
86 67
92 92
65
Full-/Part-time RD (nⴝ172)
Total responses
%
Total responses
%
P valuec
57
137
42
76
166
46
.48
93 73
124 95
139 138
89 69
145 121
164 164
88 74
.41 .62
95
68
92
143
64
123
169
73
.27
40
99
40
63
145
43
66
172
38
.66
45
87
52x
54
122
44x
113
160
71y
⬍.001
17 33 6
90 98 97
19x 34x 6x
22 54 10
127 144 144
17x 37x 7x
54 117 37
164 171 172
33y 68y 21y
⬍.01 ⬍.001 ⬍.001
31
95
8
56
142
14
73
169
18
.24
n
4.7⫾1.4x
n
4.6⫾1.7x
⬍.001
5.6⫾1.7y
a
NICU⫽neonatal intensive care unit. Nutrition care score was calculated only for sites that provided responses for all 10 criteria (n⫽78, 113, 149 for no RD, consult, and full-/part-time RDs, respectively; one survey missing level of RD involvement). % represent the number of NICUs that met the criteria divided by the number of NICUs that responded to the question. c P value for overall comparison of frequencies by 2. Multiple pairwise comparisons of RD groups used Bonferroni procedure. xy Different symbols indicate significant differences. b
Table 2. Comparison of NICUa size and selected characteristics among 417 NICUsb with differing levels of registered dietitian (RD) involvement Characteristics
No RD (nⴝ99)
Consultant RD (nⴝ145)
Full-/Part-time RD (nⴝ172)
P valuec
NICU bedsd (mean⫾standard deviation)
20.7⫾10.6x
24.7⫾16.2y
35.6⫾17.6z
⬍.001
NICU characteristics: Part of pediatric department Performs surgery on neonates in hospital Performs ECMOe in NICU Accepts infants transferred from other NICUs Transfers infants out to other NICUs Affiliated with academic medical center a
n
Total responses
%
n
Total responses
%
n
Total responses
%
83 59 4 82 94 63
97 97 97 96 97 96
86 61x 4x 85x 97 66
133 99 15 127 132 102
143 143 142 143 143 142
93 69xy 11x 89x 92 72
157 133 57 163 159 126
170 168 163 169 168 164
92 79y 35y 96y 95 77
.18 ⬍.01 ⬍.001 ⬍.01 .31 .23
NICU⫽neonatal intensive care unit. Total sample⫽417 surveys; rows add to a maximum of 416 responses due to a missing RD status for one survey. % represent the number of NICUs that provided an affirmative answer divided by the number of NICUs that responded to the question. c P value for overall comparison of means by analysis of variance and frequencies by 2. Multiple pairwise comparisons of RD groups used Duncan’s Multiple Range test for means and Bonferroni procedure for frequencies. d Data from the American Academy of Pediatrics’ directory were used to fill in missing NICU bed counts for 29 surveys. e ECMO⫽extracorporeal membrane oxygenation. xyz Different symbols indicate significant differences. b
1226
August 2005 Volume 105 Number 8
Table 3. Selected enteral nutrition feeding practices for very-low-birth-weight infants compared among 417 NICUsa with differing levels of registered dietitian (RD) involvement All NICUsb (nⴝ417)
Consultant RD (nⴝ145)
No RD (nⴝ99)
Total responses % n
Total responses %
n
Total responses %
Full-/Part-time RD (nⴝ172) P valuec
n
Use ⬎24 kcal/oz enteral feedings Use enteral protein supplements Use fortified human milk Criteria for use of fortified human milk Birth weight Gestational age Growth performance Criteria for when to start HMFd Total volume of enteral feedings Criteria for when to stop HMF At or close to discharge from NICU Achieved weight Corrected gestational age
205 414 53 414 406 417
49 33 98 13 6 97 97 94 99
34x 54 144 6x 10 144 95 142 145
37x 117 171 7x 37 172 98 169 172
68y ⬍.001 21y ⬍.001 98 .23
282 417 230 417 238 417
68 63 99 55 43 99 57 59 99
64xy 87 145 43x 71 145 60 85 145
60x 132 172 49x 115 172 59 94 172
77y ⬍.01 67y ⬍.001 55 .67
316 417
76 67 99
68x 106 145
73xy 142 172
83y ⬍.05
160 417 191 417 101 417
38 32 99 46 50 99 24 17 99
32x 50 17
33x 46 25
46x ⬍.05 43 .49 27 .18
48 145 67 145 37 145
n
Total responses %
Characteristic
79 172 74 172 46 172
a
NICU⫽neonatal intensive care unit. Total sample⫽417 surveys; rows add to a maximum of 416 responses due to a missing RD status for one survey. % represent the number of NICUs that met the criteria divided by the number of NICUs that responded to the question. c P value for overall comparison of frequencies by chi-square. Multiple pairwise comparisons of RD groups used Bonferroni procedure. d HMF⫽human milk fortifier. xy Different letters indicate significant differences. b
or part-time RDs were larger, more likely to accept transfers, to perform surgery within the hospital, and to provide extracorporeal membrane oxygenation. Nutritional Intake Parenteral Nutrition. Of the respondents, 69% (280/408) started parenteral nutrition within 48 hours of life and 76% (311 of 409) started intravenous lipids within 72 hours. Differences were not evident by level of RD involvement in these practices (P⫽.27 and .95, respectively). However, goals for energy, protein, and lipids differed significantly by the level of RD involvement. For the no RD vs consult RD vs full-/part-time RD groups, reported energy goals (mean⫾standard deviation) were 109⫾14 (n⫽87) vs 110⫾16 (n⫽122) vs 104⫾14 (n⫽160) with a P⫽.0009; reported protein goals were 3.0⫾0.5 (n⫽90) vs 3.0⫾0.4 (n⫽127) vs 3.2⫾0.4 (n⫽164) with a P⫽.0003; and reported lipid goals were 3.0⫾1.0 (n⫽87) vs 2.8⫾0.6 (n⫽124) vs 3.0⫾0.5 (n⫽159) with a P⫽.04. NICUs with full-/part-time RDs had lower goals for parenteral energy (104 kcal/kg/day) and higher goals for parenteral protein intakes (3.2 g/kg/day) than the other groups. Enteral Nutrition. Enteral feedings for VLBW infants were started by 3 days of life in 40% (n⫽169) and by day 4 in an additional 24% (n⫽101) of the 417 NICUs reporting, but timing did not vary by the level of RD involvement. Almost half of the NICUs provided VLBW infants with high energy density (eg ⬎24 kcal/oz) breast milk or preterm formula feedings; the practice was much more frequent in NICUs with full-/part-time RDs. Enteral protein supplement use was low. Among nonusers 19% (n⫽78) of
the NICUs reported that they were aware of the benefits of extra protein but were not yet using protein supplements; NICUs with consulting RDs were significantly more likely to provide such responses (P⫽.002). The most frequently reported feedings for these VLBW infants were fortified breast milk, defined as breast milk with added human milk fortifier (n⫽395, 95% of NICUs) and preterm formula (n⫽406, 97%). Use of fortified breast milk with human milk fortifier was essentially universal, and there were no differences between NICUs based on level of RD involvement (Table 3). More than half of the responding NICUs used the criteria of birth weight, gestational age, and/or growth performance for deciding to use fortified breast milk for feeding VLBW infants. The total volume of enteral feedings fed was the primary criterion for when to start human milk fortifier. On average, NICUs were more likely to wait until feedings were up to 121 to 150 cc/kg/day (85 of 207, 41%) although 39% (43 of 109) with full-/part-time RDs reported starting human milk fortifier when feedings were up to only 80 to 100 cc/kg/day (data not shown). Less frequently used criteria for when to start human milk fortifier were gestational age, day of life, and birth weight. In decisions about stopping human milk fortifier, NICU respondents most often considered achieved weight (usually using a cutoff of 1,801 to 2,000 g), “at or close to NICU discharge” and/or corrected gestational age (most frequently using a cutoff of 35 to 36 weeks). NICUs with no RD involvement were more likely to stop the human milk fortifier sooner and at lower achieved weights. “Non-preterm diets,” including unfortified breast milk, transitional formulas, and full-term formulas, tended to be used
August 2005 ● Journal of the AMERICAN DIETETIC ASSOCIATION
1227
less frequently to feed VLBW infants among the NICUs with a full-/part-time RD. Nutrition Care Score The nutrition care score for the 341 NICUs ranged from 0 to 10 with a mean of 5.1 points (Table 1). Overall, the nutrition care scores were relatively low compared with a “perfect” score for intensive nutrition care practices. When compared by level of RD involvement, differences among the mean nutrition care scores were small but significant, and increased by level of RD involvement. When further adjusted for NICU size there was essentially no change in the nutrition care scores between RD groups (ie, no confounding), and the nutrition care score differences between RD groups did not vary by NICU size (ie, no interaction). Overall, more of the NICUs with full-/part-time RDs reported that they provided fewer parenteral kilocalories, more enteral kilocalories, and more parenteral and enteral protein to the VLBW infants they cared for than did NICUs with less RD involvement. Within the full-/part-time RD group, number of hours spent in the NICU, years of NICU and total clinical experience, and whether the RD had a graduate degree did not further distinguish the differences in nutrition care score or its criteria (data not shown). Registered Dietitians in the NICUs Of the 172 full-/part-time RDs who completed our survey, 81% (n⫽139) described their involvement in the NICU to be either “part of the NICU health care team and involved in making nutrition decisions” or “involved in patient care and provides recommendations to the NICU health care team that are usually followed.” The mean number of RD hours per week assigned to the NICUs was 28.5⫾19.0, and involvement ranged from 2 to 100 hours per week. On average, 1.3⫾0.6 (range 1 to 4) RDs shared these hours within an NICU. An RD-to-infant bed ratio was used for staffing in only 7 of the 172 NICUs; ratios ranging from 1:21 to 1:60. Fifteen percent (n⫽63) of the 417 responding NICUs reported that they were planning to or were in the process of increasing the RD participation in their NICU; such responses were highest among NICUs with full-/part-time RDs (P⫽.001).
The number of consulting RDs in NICUs today may be helpful if such individuals later increase their involvement to full-/part-time status and become full members of the NICU team. We assumed that all full-/part-time dietitians working in the NICUs were RDs, and this was specifically indicated by 87% (n⫽149) of the 172 respondents. Fifty-three percent (n⫽91) of the RDs were licensed or certified by their state, 17% (n⫽29) were certified in pediatric nutrition, and 28% (n⫽49) in nutrition support. Their NICU
1228
August 2005 Volume 105 Number 8
experience averaged 5.7⫾5.1 (range 0 to 29) years and clinical experience was 12.2⫾7.2 years (range 1 month to 40 years). Forty-eight percent (n⫽82) of RDs had a master’s degree, usually in nutrition, and two had a doctorate in nutrition. Fifty-nine percent (n⫽102) had obtained additional training, most frequently through neonatal conferences (26%, n⫽45) or practicums (22%, n⫽38), and less frequently through neonatal nutrition fellowships, mentored training, college courses/independent study, and lectures. However, pursuit of these additional training opportunities was unrelated to the number of years of NICU or clinical experience (P⫽.08, .92, respectively). Telephone Interview of Nonresponders All of the 37 NICUs randomly selected from the nonrespondent group were successfully contacted by telephone. The nonresponding NICUs had fewer beds than the responders (n⫽33; 21⫾15 vs 28⫾17, P⫽.03), but level of RD involvement was similar. The 22 nonresponding NICUs without a full-/part-time RD were less likely to report lack of funding as the reason, and more likely to report “don’t know” or lack of need than the responding NICUs. The nonresponding NICUs were also less likely than responders to plot infants’ growth curves for weight and head circumference (85% vs 96%, P⫽.007; and 85% vs 94%, P⫽.03, respectively), and less likely to use more than 24 kcal/oz enteral feedings (21% vs 49%, P⫽.002). They also scored lower on a modified nutrition care score constructed from a subset of six of the 10 nutrition care score criteria (Table 1, criteria numbers 1, 2, 3, 8, 9, 10). DISCUSSION Seventy-six percent of the NICUs responding to our survey involved an RD to some degree in providing care; substantially more than the 38% in a similar 1990 survey of NICUs in the United States (7). However, if consult RDs had been excluded from our estimate, RD involvement would have only been 41%. We suspected that the consult RDs’ lack of consistent time in the NICU might limit their impact on practices there, and this was supported by our data. Therefore, to more fairly test RD impact on practices, full- and part-time RDs were analyzed separately from consult RDs. Full-time and parttime RDs were combined because we expected their NICU involvement to be more consistent and that the number of hours they spent in the NICU would increase with NICU size and acuity (measured by intensive care beds). The relationship between NICU size and full-/parttime RD involvement was confirmed (r⫽0.53, P⬍.0001). For most comparisons there were substantial differences between the practices reported by NICUs with a full-/ part-time RD and those by NICUs with no RD or a consult RD, suggesting that more NICUs could benefit from the involvement of full-/part-time RDs in the nutritional care of VLBW infants. Full-/part-time RDs were especially likely to be involved in the larger NICUs, which are often higher in patient acuity. Perhaps their specialty skills were recognized as being especially needed in such settings. The number of consulting RDs in NICUs today may be helpful if such individuals later increase their involvement to
full-/part-time status and become full members of the NICU team. Overall, RD presence in NICUs seems to be increasing. The barriers of perceived lack of need, funding, and trained RDs reported in 1990 (7) still seem to be hindering RD involvement in more NICUs. The perceived lack of need for an RD may be related in part to the design of our study, which consisted of all NICUs caring for VLBW infants, including smaller NICUs with lower acuity and less need for intensive nutritional care. However, this perception of the respondents, who were usually NICU directors, underscores the need for more documentation and communication by RDs to emphasize the positive impact RD services can have on improving practices and outcomes for VLBW infants. There were interesting variations in the nutrition-related practices reported by this survey. Consistent with nutrition recommendations in 1999 (9-16), NICUs with more RD involvement provided VLBW infants with less kilocalories and more protein parenterally and more kilocalories and protein enterally compared with NICUs with less RD involvement. The use of enteral protein supplements was low in all settings. However, there was a trend toward increased use overall and prevalence in NICUs as level of RD involvement increased. The reported initiation of parenteral and enteral nutrition was later than recommended for VLBW infants at that time (11,16,17) regardless of the level of RD involvement. Early and intensive parenteral and enteral nutrition were not routinely practiced in these NICUs when the survey was done in 1999. However, NICUs commonly fortified breast milk and used preterm formula for VLBWs. Fortified breast milk was used earlier and longer, and “non-preterm diets” were less likely to be reported as appropriate feeding options for VLBW infants in NICUs with full-/ part-time RDs than those with less RD involvement. The nutrition care score, summarizing the intensity of growth surveillance and nutrition practices, showed the positive impact of a full-/part-time RD in the NICU, yet nutrition care scores differed only by 20% between groups. Several factors may have contributed to making the differences smaller than they really were, including: response bias, design of the nutrition care score, and timing of the survey. First, the response rate (54%) in our survey was better than that of two other national NICU surveys [37% (18) and 42% (7)]. However, differences between our responders and nonresponders suggest that returned surveys were not a representative sample of all US NICUs. We suspect that our findings likely underestimated the true differences in nutrition practices among NICUs. The NICUs that were most interested in nutrition may have been more likely to respond to our survey, attenuating differences in practices that otherwise might exist by level of RD involvement. Second, although our survey questionnaire was pilottested (8), the nutrition care score may not have been optimally designed and adequately tested for its ability to detect the RD’s positive impact on care. The nutrition care score was designed as a summary score of the intensity of NICU growth surveillance and nutrition practices, such that a highly intensive score would represent close monitoring of growth and aggressive
nutrition practices. Although more intensive care would be considered good care for most VLBW infants, this might not be the case for all infants due to the need to individualize care. As a result, the nutrition care score may not have detected the actual prevalence of good vs intensive care, and thus may underrepresent the positive impact of RD involvement. Third, the focus on neonatal nutrition has burgeoned over the years since our survey (19-24) and the benefits of more aggressive nutrition intervention have become better known. This has likely led to more routine use of these practices, and thus clinical care today may reflect recommendations for good care more closely than in the past. Thus, we suspect that differences in nutrition practices due to RD involvement would be even greater today than they were in 1999. A final limitation of this survey study is the implicit assumption that with more nutrition involvement, and explicitly with more RD involvement, outcomes for VLBW infants will improve. In fact, additional research is needed before such a causal chain is demonstrated and the number of full-/part-time RDs in the NICU for optimal care is established. Therefore, although our findings suggest that more is better in terms of the physical presence of the RD on the NICU team, more evidence documenting the value of more specialty experience, training, and education for NICU RDs is still needed. CONCLUSIONS In more than 75% of NICUs studied, RDs were involved to some degree in providing care. However, less than 50% of NICUs employed RDs on a full- or part-time basis, indicating that there is much room for improvement in making RDs NICU team members. RD involvement in the NICU increased the intensity of important nutrition practices that may improve outcomes in VLBW infants, including the provision of fewer kilocalories and more protein parenterally, and avoidance of full-term infant diets. RD involvement and effectiveness in the NICU are still limited by NICU directors’ lack of awareness of the benefits provided by an NICU RD and possibly also by the lack of availability of trained NICU RDs. The credibility and effectiveness of NICU RDs may be enhanced if job descriptions emphasize the need for an advanced degree in nutrition science as well as specialty training that includes mentored clinical experience in neonatal and high-risk infant nutrition. This work is dedicated to the memory of the late Douglas K. Richardson, MD, MBA, who contributed substantially to this project and manuscript. Our thanks to all of the NICUs that responded to our survey and to our research assistants and other support personnel involved in this study. We thank Melody Thompson, MS, RD, of Ross Products Division, Abbott Laboratories, for sharing her experiences in conducting a previous NICU nutrition survey and her feedback on the survey instrument developed for this study. Lastly, we express our appreciation to the ADA Pediatric Nutrition dietetic practice group for providing us with NICU RD contact information and publicizing the survey in their newsletter.
August 2005 ● Journal of the AMERICAN DIETETIC ASSOCIATION
1229
Partial funding for this study was provided by Ross Products Division, Abbott Laboratories, Columbus, OH. Dr Olsen was the Ellen and Ronald Block Scholar at Tufts University, School of Nutrition Science and Policy. This material is based on work partially supported by the US Department of Agriculture, under agreement No. 58-1950-9-001. Any opinions, findings, conclusions or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the US Department of Agriculture. References 1. LaBarre DJ, Maher MM, Raye JR, Rowe JC. Effect of monitored feeding protocol on VVLBW infants [abstract]. J Am Diet Assoc. 1987;87:A-58. 2. Elsaesser KR. Dietitian intervention in neonatal intensive care reduces errors and improves clinical outcomes [abstract]. J Am Diet Assoc. 1998;98:A-21. 3. Kuzma-O’Reilly B, Duenas ML, Greecher C, Kimberlin L, Mujsce D, Miller D, Walker DJ. Evaluation, development, and implementation of potentially better practices in neonatal intensive care nutrition. Pediatrics. 2003;111:E461-E470. 4. Kuzma-O’Reilly BA. Effectiveness of nutrition appraisal in the NICU [abstract]. Pediatr Res. 2000;47: 409A. 5. Hudak LJ, Worobey J. Monitoring nutritional intake of the preterm infant: testing a nutrition management protocol. Top Clin Nutr. 1998;14:67-76. 6. Stave VS, Robbins S, Fletcher AB. A comparison of growth rates of premature infants prior to and after close nutritional monitoring. Clin Proc CHNMC. 1979;35:171-180. 7. Thompson M, Price P, Stahle DA. Nutrition services in neonatal intensive care: A national survey. J Am Diet Assoc. 1994;94:440-441. 8. Olsen I, Awnetwant E, Cardi G, Gallagher L, McKinley L, O’Donnell E, Pryzstac L, Warren L, Dwyer J, Richardson D. Variations in nutrition personnel and practice among 7 neonatal intensive care units [abstract]. J Am Diet Assoc. 1997;97:A-30. 9. Committee on Nutrition, American Academy of Pediatrics. Nutritional needs for preterm infants. In: Kleinman RE, ed. Pediatric Nutrition Handbook. 4th ed. Elk Grove Village, IL: American Academy of Pediatrics; 1998:55-87. 10. Tsang RC, Lucas A, Uauy R, Zlotkin S, eds. Nutritional Needs of the Preterm Infant: Scientific Basis and Practical Guidelines. Baltimore, MD: Williams & Wilkins; 1993.
1230
August 2005 Volume 105 Number 8
11. Groh-Wargo S, Thompson M, Hovasi Cox J, Hartline JV, eds. Nutritional Care for High-Risk Newborns. 2nd ed. Chicago, IL: Precept Press, Inc; 1994. 12. Baugh N, Recupero MA, Kerner JA. Nutritional requirements for pediatric patients. In: Christensen ML, ed. The A.S.P.E.N. Nutrition Support Practice Manual. Silver Spring, MD: American Society for Parenteral and Enteral Nutrition; 1998:24-1–24-13. 13. Carlson SJ, Johnson KJ, Cress GA, Connolly NW, Ziegler EE. Higher protein intake improves growth of VLBW infants fed fortified breast milk. Pediatr Res. 1999;45:278A. 14. Schulze KF, Stefanski M, Masterson J, Spinnazola R, Ramakrishnan R, Dell RB, Heird WC. Energy expenditure, energy balance, and composition of weight gain in low birth weight infants fed diets of different protein and energy content. J Pediatr. 1987;110:753759. 15. Ziegler EE. Protein in premature feeding. Nutrition. 1994;10:69-71. 16. Thureen PJ. Early aggressive nutrition in the neonate. Pediatr Rev. 1999;20:E45-E55. 17. Koo WWK, McLaughlin K, Saba M. Nutrition support for the preterm infant. In: Christensen ML, ed. The A.S.P.E.N. Nutrition Support Practice Manual. Silver Spring, MD: American Society for Parenteral and Enteral Nutrition; 1998:26-1–26-16. 18. Baker SF, Smith BJ, Donohue PK, Gleason CA. Skin care management practices for premature infants. J Perinatol. 1999;19(6 Pt 1):426-431. 19. Ziegler EE, Thureen PJ, Carlson SJ. Aggressive nutrition of the very low birthweight infant. Clin Perinatol. 2002;29:225-244. 20. Klein CJ. Nutrient requirements for preterm infant formulas. J Nutr. 2002;132(suppl 1):S1395-S1577. 21. Denne SC. Protein and energy requirements in preterm infants. Semin Neonatol. 2001;6:377-382. 22. Olsen IE, Richardson DK, Schmid CH, Ausman LM, Dwyer JT. Intersite differences in weight growth velocity of extremely premature infants. Pediatrics. 2002;110:1125-1132. 23. Thureen PJ, Melara D, Fennessey PV, Hay WW Jr. Effect of low versus high intravenous amino acid intake on very low birth weight infants in the early neonatal period. Pediatr Res. 2003;53:24-32. 24. Clark RH, Thomas P, Peabody J. Extrauterine growth restriction remains a serious problem in prematurely born neonates. Pediatrics. 2003;111:986990.
Continuing Education Questionnaire, page 1240 Meets Learning Need Codes 4150, 5000, 5010, and 5060
Dietitian Involvement in the Neonatal Intensive Care Unit: More Is Better IRENE E. OLSEN, PhD, RD; DOUGLAS K. RICHARDSON, MD, MBA; CHRISTOPHER H. SCHMID, PhD; LYNNE M. AUSMAN, DSc, RD; JOHANNA T. DWYER, DSc, RD
I. E. Olsen is an assistant professor, Cincinnati Children’s Hospital Medical Center and the University of Cincinnati College of Medicine, Center for Epidemiology & Biostatistics and Division of Neonatology, Cincinnati, OH. At the time of the study, she was a neonatal dietitian, Department of Nutrition, Beth Israel Deaconess Medical Center, Boston, MA and PhD candidate, Gerald J. and Dorothy R. Friedman School of Nutrition Science and Policy, Tufts University, Boston, MA. D. K. Richardson was a neonatologist and Director of Research, Department of Neonatology, Beth Israel Deaconess Medical Center, Boston, MA, and a professor, Department of Maternal and Child Health, Harvard School of Public Health, Boston, MA. C. H. Schmid is a senior statistician, Institute for Clinical Research and Health Policy Studies, Tufts-New England Medical Center, Boston, MA. L. M. Ausman is a professor and Dean of Students, Gerald J. and Dorothy R. Friedman School of Nutrition Science and Policy, Tufts University, Boston, MA; adjunct professor, School of Medicine, Tufts University, Boston, MA; and scientist, Jean Mayer US Department of Agriculture Human Nutrition Research Center on Aging, Tufts University, Boston, MA. T. Dwyer is a scientist and professor in the ]?⬎Gerald J. and Dorothy R. Friedman School of Nutrition Science and Policy, in the School of Medicine, and in the Jean Mayer US Department of Agriculture Human Nutrition Research Center on Aging, all at Tufts University, Boston, MA; she is also a professor with the Department of Maternal and Child Health, Harvard School of Public Health, Boston, MA; and director of the Frances Stern Nutrition Center, TuftsNew England Medical Center, Boston, MA. Address correspondence to: Irene E. Olsen, PhD, RD, Assistant Professor, Cincinnati Children’s Hospital Medical Center and the University of Cincinnati College of Medicine, Center for Epidemiology & Biostatistics and Division of Neonatology, Cincinnati Children’s Hospital Medical Center (MLC 5041), 3333 Burnet Ave, Cincinnati, OH 45229. E-mail: [email protected] Copyright © 2005 by the American Dietetic Association. 0002-8223/05/10508-0004$30.00/0 doi: 10.1016/j.jada.2005.05.012
1224
Journal of the AMERICAN DIETETIC ASSOCIATION
ABSTRACT Objective Describe the level of registered dietitian (RD) involvement in neonatal intensive care units (NICUs) and associations with NICU nutrition practices. Design Questionnaires were mailed to 820 NICUs in the United States with two follow-up mailings to nonresponders. Abbreviated phone surveys were conducted with a random sample of 10% of nonresponders. A nutrition care score was devised based on a sum of 10 survey questions (range 0 to 10) to summarize the intensity of reported practices. Subjects/setting Directors of NICUs in the United States and RDs associated with them. Statistical analyses 2, analysis of variance, Bonferroni and Duncan multiple range tests, regression. Results Respondents from 417 (54%) of the 772 NICUs eligible for the study provided completed questionnaires. Among NICUs responding, 76% involved RDs in care (41% employed full- or part-time RDs, 35% employed consult RDs), and 24% had no RD. NICUs with full- or part-time RDs provided fewer kilocalories and more protein parenterally, and more kilocalories and protein enterally. NICUs with less RD involvement were more likely to provide full-term infant feedings (eg, unfortified breast milk, full-term formula) to very-low-birth-weight infants. Mean nutrition care score varied with RD involvement from 4.6⫾1.7 (mean⫾standard deviation) for NICUs with a consult RD and 4.7⫾1.4 for NICUs employing no RD to 5.6⫾1.7 for NICUs with a full- or part-time RD (overall P⬍.001). Conclusions More involvement of RDs in NICUs increased the intensity of important aspects of nutrition care that may improve outcomes of very-low-birth-weight infants in NICUs. These findings highlight the importance of RDs as NICU team members. J Am Diet Assoc. 2005;105:1224-1230.
R
egistered dietitians (RDs) have special expertise on the nutrition issues of very-low-birth-weight (VLBW) infants. While other members of the neonatal intensive care unit (NICU) team focus on infant daily care, procedures, and life-saving events, the RD’s
© 2005 by the American Dietetic Association
primary focus on evidence-based, individualized nutrition care helps to ensure that this is provided consistently to each infant. RD involvement in the NICU, such as in the design of NICU nutrition practice protocols and monitoring tools or direct nutrition interventions, is associated with improved nutritional intake and infant growth (1-5), shortened hospital length of stay (1,3,5,6), and reduced related costs (3,4). However, a 1990 survey of American NICUs reported that only 38% of responding NICUs involved an RD in the care of their infants (7), suggesting that the potential benefits of RD involvement were not fully realized in NICUs. This study describes the current level of RD involvement in NICUs in the United States and how reported nutrition practices differ by the level of RD involvement. Our hypothesis was that NICUs with a full- or part-time RD would be more likely to frequently monitor growth and use aggressive nutrition practices for their VLBW infants than those with less RD involvement. In addition, we expected NICU nutrition practices to be more intensive overall when the NICU RD had received more specialty experience, training, and education related to clinical practice. METHODS The Beth Israel Deaconess Medical Center’s Committee on Clinical Investigations approved this study. Data were collected using an 8-page, mailed survey questionnaire focused on nutrition personnel and practices in NICUs caring for VLBW infants (⬍1,500 g at birth). The questionnaire was pilot-tested in six New England NICUs (8). A personalized letter describing the study and requesting return of the questionnaire in the self-addressed, stamped envelope was sent in May 1999 to all of the 820 NICUs listed in the American Academy of Pediatrics NICU Directory (1999). Nonresponders were sent a reminder postcard after 4 weeks and a second mailing of the questionnaire after 8 weeks. Return of the survey questionnaire was also encouraged through two NICU-related electronic mailing lists and the American Dietetic Association Pediatric Nutrition dietetic practice group. The study was closed in November 1999. At that time, a random sample of 10% of nonresponders was selected for comparison to the NICUs that returned surveys and an abbreviated phone survey describing the NICU, its personnel, and practices related to nutrition was administered. The survey questionnaire collected information about the characteristics of the NICUs (eg, number of beds, performs surgery or not), nutrition practices (eg, timing of initiation of parenteral nutrition, use of human milk fortifier in breast milk), growth surveillance (eg, anthropometrics and frequency), and which personnel were involved in making nutrition-related decisions. Assessment of the level of RD NICU involvement was coded as: full- or part-time RD, consult RD (defined as no assigned hours to the NICU), or no RD. If full- or part-time RDs were involved in the NICU, the respondent was asked to give the questionnaire to the RD to complete, including details on training, experience, and education. Directors of NICUs who responded that they did not employ a full- or part-time RD were requested to complete the survey
themselves and to comment on the reasons for the lack of RD involvement in the NICU. In light of the expanding body of literature on the benefits of aggressive nutrition practices for VLBW infants available in 1999 when the survey was conducted (9-16), a nutrition care score was created to help summarize the intensity of the NICUs’ nutrition and growth surveillance practices and to facilitate comparisons between groups. Criteria for the nutrition care score (Table 1) were selected by one of the authors (I.E.O.) based on a review of recommendations for NICU nutrition practices in the relevant literature (9-16) and clinical experience, and then critiqued by a neonatologist (D.K.R.) and another NICU RD. One point was assigned to each of the 10 criteria for a possible score of 0 to 10 (least to most). More frequent growth surveillance and aggressive nutrition practices were assigned a point for intensity of care. The nutrition care score was calculated for all NICUs that provided a response for each of the 10 criteria. Data analyses were conducted with SAS 8.0 (SAS, Cary, NC, 1999). NICU characteristics, nutrition practices, nutrition care score, and its criteria were compared among RD groups: no RD, consult RD, and full- or parttime RD. Within the full- or part-time RD group, the nutrition care score and its criteria were also examined by the number of hours spent in the NICU, years of NICU and total clinical experience, and whether the RD had a graduate degree. Significance was determined at the .05 level. Overall comparisons of means and frequencies were conducted using analysis of variance and 2, respectively. Multiple pairwise comparisons of RD groups used Duncan Multiple Range test for means and Bonferroni procedure for frequencies. Regression was used to test the nutrition care score for potential confounding of NICU size and an interaction between RD group and NICU size (ie, if the differences in nutrition care score by RD group were explained by NICU size and/or varied by NICU size, respectively). RESULTS Of the 820 NICUs listed in the American Academy of Pediatrics directory, eight were declared ineligible because respondents indicated that either they did not care for VLBW infants (n⫽5) or that the NICU had closed (n⫽3). The 395 NICUs that did not respond to the survey included 37 NICUs with uncorrectable addresses. Of the remaining nonresponders, a random phone survey (n⫽37) found that one NICU did not care for VLBW infants, two were closed, and one was a duplicate listing. Applying these percentages of ineligible NICUs to the group of unreturned surveys, we projected an additional 40 NICUs would have been ineligible for our study. We received 417 completed questionnaires, and therefore our final response rate was 54% (417 of 772). Among the 417 responding NICUs, 76% involved RDs: 41% employed full- or part-time RDs, 35% employed consult RDs, and 24% had no RD. Among the 244 NICUs without a full- or part-time RD, the three most common reasons given by NICU directors for the lack of RD involvement were lack of need (49%), lack of funding (35%), and lack of availability of a trained NICU RD (25%). Table 2 compares NICU size and characteristics with respect to the level of RD involvement. NICUs with full-
August 2005 ● Journal of the AMERICAN DIETETIC ASSOCIATION
1225
Table 1. Comparison of the percent of NICUsa that met each criterion of the nutrition care scoreb, used to summarize the intensity of the NICUs’ nutrition and growth surveillance practices, among NICUs with differing levels of registered dietitian (RD) involvement Consultant RD (nⴝ145)
No RD (nⴝ99)
Please tell us if you: 1. Plot and trend weight daily 2. Plot and trend head circumference weekly or more 3. Plot and trend length weekly or more 4. Parenteral nutrition started within first 48 hours of life 5. Enteral nutrition started within first 72 hours of life 6. Parenteral nutrition goal: 90-110 kcal/kg/d 7. Parenteral nutrition goal: ⱖ3.5 g protein/ kg/d 8. Provide ⬎24 kcal/oz enteral feedings 9. Use enteral protein supplement 10. Iron supplement started at: ⱕ14 days of life, or ⱕ“full feeds” Nutrition Care Score (mean⫾standard deviation)a
n
Total responses
%
36
94
38
86 67
92 92
65
Full-/Part-time RD (nⴝ172)
Total responses
%
Total responses
%
P valuec
57
137
42
76
166
46
.48
93 73
124 95
139 138
89 69
145 121
164 164
88 74
.41 .62
95
68
92
143
64
123
169
73
.27
40
99
40
63
145
43
66
172
38
.66
45
87
52x
54
122
44x
113
160
71y
⬍.001
17 33 6
90 98 97
19x 34x 6x
22 54 10
127 144 144
17x 37x 7x
54 117 37
164 171 172
33y 68y 21y
⬍.01 ⬍.001 ⬍.001
31
95
8
56
142
14
73
169
18
.24
n
4.7⫾1.4x
n
4.6⫾1.7x
⬍.001
5.6⫾1.7y
a
NICU⫽neonatal intensive care unit. Nutrition care score was calculated only for sites that provided responses for all 10 criteria (n⫽78, 113, 149 for no RD, consult, and full-/part-time RDs, respectively; one survey missing level of RD involvement). % represent the number of NICUs that met the criteria divided by the number of NICUs that responded to the question. c P value for overall comparison of frequencies by 2. Multiple pairwise comparisons of RD groups used Bonferroni procedure. xy Different symbols indicate significant differences. b
Table 2. Comparison of NICUa size and selected characteristics among 417 NICUsb with differing levels of registered dietitian (RD) involvement Characteristics
No RD (nⴝ99)
Consultant RD (nⴝ145)
Full-/Part-time RD (nⴝ172)
P valuec
NICU bedsd (mean⫾standard deviation)
20.7⫾10.6x
24.7⫾16.2y
35.6⫾17.6z
⬍.001
NICU characteristics: Part of pediatric department Performs surgery on neonates in hospital Performs ECMOe in NICU Accepts infants transferred from other NICUs Transfers infants out to other NICUs Affiliated with academic medical center a
n
Total responses
%
n
Total responses
%
n
Total responses
%
83 59 4 82 94 63
97 97 97 96 97 96
86 61x 4x 85x 97 66
133 99 15 127 132 102
143 143 142 143 143 142
93 69xy 11x 89x 92 72
157 133 57 163 159 126
170 168 163 169 168 164
92 79y 35y 96y 95 77
.18 ⬍.01 ⬍.001 ⬍.01 .31 .23
NICU⫽neonatal intensive care unit. Total sample⫽417 surveys; rows add to a maximum of 416 responses due to a missing RD status for one survey. % represent the number of NICUs that provided an affirmative answer divided by the number of NICUs that responded to the question. c P value for overall comparison of means by analysis of variance and frequencies by 2. Multiple pairwise comparisons of RD groups used Duncan’s Multiple Range test for means and Bonferroni procedure for frequencies. d Data from the American Academy of Pediatrics’ directory were used to fill in missing NICU bed counts for 29 surveys. e ECMO⫽extracorporeal membrane oxygenation. xyz Different symbols indicate significant differences. b
1226
August 2005 Volume 105 Number 8
Table 3. Selected enteral nutrition feeding practices for very-low-birth-weight infants compared among 417 NICUsa with differing levels of registered dietitian (RD) involvement All NICUsb (nⴝ417)
Consultant RD (nⴝ145)
No RD (nⴝ99)
Total responses % n
Total responses %
n
Total responses %
Full-/Part-time RD (nⴝ172) P valuec
n
Use ⬎24 kcal/oz enteral feedings Use enteral protein supplements Use fortified human milk Criteria for use of fortified human milk Birth weight Gestational age Growth performance Criteria for when to start HMFd Total volume of enteral feedings Criteria for when to stop HMF At or close to discharge from NICU Achieved weight Corrected gestational age
205 414 53 414 406 417
49 33 98 13 6 97 97 94 99
34x 54 144 6x 10 144 95 142 145
37x 117 171 7x 37 172 98 169 172
68y ⬍.001 21y ⬍.001 98 .23
282 417 230 417 238 417
68 63 99 55 43 99 57 59 99
64xy 87 145 43x 71 145 60 85 145
60x 132 172 49x 115 172 59 94 172
77y ⬍.01 67y ⬍.001 55 .67
316 417
76 67 99
68x 106 145
73xy 142 172
83y ⬍.05
160 417 191 417 101 417
38 32 99 46 50 99 24 17 99
32x 50 17
33x 46 25
46x ⬍.05 43 .49 27 .18
48 145 67 145 37 145
n
Total responses %
Characteristic
79 172 74 172 46 172
a
NICU⫽neonatal intensive care unit. Total sample⫽417 surveys; rows add to a maximum of 416 responses due to a missing RD status for one survey. % represent the number of NICUs that met the criteria divided by the number of NICUs that responded to the question. c P value for overall comparison of frequencies by chi-square. Multiple pairwise comparisons of RD groups used Bonferroni procedure. d HMF⫽human milk fortifier. xy Different letters indicate significant differences. b
or part-time RDs were larger, more likely to accept transfers, to perform surgery within the hospital, and to provide extracorporeal membrane oxygenation. Nutritional Intake Parenteral Nutrition. Of the respondents, 69% (280/408) started parenteral nutrition within 48 hours of life and 76% (311 of 409) started intravenous lipids within 72 hours. Differences were not evident by level of RD involvement in these practices (P⫽.27 and .95, respectively). However, goals for energy, protein, and lipids differed significantly by the level of RD involvement. For the no RD vs consult RD vs full-/part-time RD groups, reported energy goals (mean⫾standard deviation) were 109⫾14 (n⫽87) vs 110⫾16 (n⫽122) vs 104⫾14 (n⫽160) with a P⫽.0009; reported protein goals were 3.0⫾0.5 (n⫽90) vs 3.0⫾0.4 (n⫽127) vs 3.2⫾0.4 (n⫽164) with a P⫽.0003; and reported lipid goals were 3.0⫾1.0 (n⫽87) vs 2.8⫾0.6 (n⫽124) vs 3.0⫾0.5 (n⫽159) with a P⫽.04. NICUs with full-/part-time RDs had lower goals for parenteral energy (104 kcal/kg/day) and higher goals for parenteral protein intakes (3.2 g/kg/day) than the other groups. Enteral Nutrition. Enteral feedings for VLBW infants were started by 3 days of life in 40% (n⫽169) and by day 4 in an additional 24% (n⫽101) of the 417 NICUs reporting, but timing did not vary by the level of RD involvement. Almost half of the NICUs provided VLBW infants with high energy density (eg ⬎24 kcal/oz) breast milk or preterm formula feedings; the practice was much more frequent in NICUs with full-/part-time RDs. Enteral protein supplement use was low. Among nonusers 19% (n⫽78) of
the NICUs reported that they were aware of the benefits of extra protein but were not yet using protein supplements; NICUs with consulting RDs were significantly more likely to provide such responses (P⫽.002). The most frequently reported feedings for these VLBW infants were fortified breast milk, defined as breast milk with added human milk fortifier (n⫽395, 95% of NICUs) and preterm formula (n⫽406, 97%). Use of fortified breast milk with human milk fortifier was essentially universal, and there were no differences between NICUs based on level of RD involvement (Table 3). More than half of the responding NICUs used the criteria of birth weight, gestational age, and/or growth performance for deciding to use fortified breast milk for feeding VLBW infants. The total volume of enteral feedings fed was the primary criterion for when to start human milk fortifier. On average, NICUs were more likely to wait until feedings were up to 121 to 150 cc/kg/day (85 of 207, 41%) although 39% (43 of 109) with full-/part-time RDs reported starting human milk fortifier when feedings were up to only 80 to 100 cc/kg/day (data not shown). Less frequently used criteria for when to start human milk fortifier were gestational age, day of life, and birth weight. In decisions about stopping human milk fortifier, NICU respondents most often considered achieved weight (usually using a cutoff of 1,801 to 2,000 g), “at or close to NICU discharge” and/or corrected gestational age (most frequently using a cutoff of 35 to 36 weeks). NICUs with no RD involvement were more likely to stop the human milk fortifier sooner and at lower achieved weights. “Non-preterm diets,” including unfortified breast milk, transitional formulas, and full-term formulas, tended to be used
August 2005 ● Journal of the AMERICAN DIETETIC ASSOCIATION
1227
less frequently to feed VLBW infants among the NICUs with a full-/part-time RD. Nutrition Care Score The nutrition care score for the 341 NICUs ranged from 0 to 10 with a mean of 5.1 points (Table 1). Overall, the nutrition care scores were relatively low compared with a “perfect” score for intensive nutrition care practices. When compared by level of RD involvement, differences among the mean nutrition care scores were small but significant, and increased by level of RD involvement. When further adjusted for NICU size there was essentially no change in the nutrition care scores between RD groups (ie, no confounding), and the nutrition care score differences between RD groups did not vary by NICU size (ie, no interaction). Overall, more of the NICUs with full-/part-time RDs reported that they provided fewer parenteral kilocalories, more enteral kilocalories, and more parenteral and enteral protein to the VLBW infants they cared for than did NICUs with less RD involvement. Within the full-/part-time RD group, number of hours spent in the NICU, years of NICU and total clinical experience, and whether the RD had a graduate degree did not further distinguish the differences in nutrition care score or its criteria (data not shown). Registered Dietitians in the NICUs Of the 172 full-/part-time RDs who completed our survey, 81% (n⫽139) described their involvement in the NICU to be either “part of the NICU health care team and involved in making nutrition decisions” or “involved in patient care and provides recommendations to the NICU health care team that are usually followed.” The mean number of RD hours per week assigned to the NICUs was 28.5⫾19.0, and involvement ranged from 2 to 100 hours per week. On average, 1.3⫾0.6 (range 1 to 4) RDs shared these hours within an NICU. An RD-to-infant bed ratio was used for staffing in only 7 of the 172 NICUs; ratios ranging from 1:21 to 1:60. Fifteen percent (n⫽63) of the 417 responding NICUs reported that they were planning to or were in the process of increasing the RD participation in their NICU; such responses were highest among NICUs with full-/part-time RDs (P⫽.001).
The number of consulting RDs in NICUs today may be helpful if such individuals later increase their involvement to full-/part-time status and become full members of the NICU team. We assumed that all full-/part-time dietitians working in the NICUs were RDs, and this was specifically indicated by 87% (n⫽149) of the 172 respondents. Fifty-three percent (n⫽91) of the RDs were licensed or certified by their state, 17% (n⫽29) were certified in pediatric nutrition, and 28% (n⫽49) in nutrition support. Their NICU
1228
August 2005 Volume 105 Number 8
experience averaged 5.7⫾5.1 (range 0 to 29) years and clinical experience was 12.2⫾7.2 years (range 1 month to 40 years). Forty-eight percent (n⫽82) of RDs had a master’s degree, usually in nutrition, and two had a doctorate in nutrition. Fifty-nine percent (n⫽102) had obtained additional training, most frequently through neonatal conferences (26%, n⫽45) or practicums (22%, n⫽38), and less frequently through neonatal nutrition fellowships, mentored training, college courses/independent study, and lectures. However, pursuit of these additional training opportunities was unrelated to the number of years of NICU or clinical experience (P⫽.08, .92, respectively). Telephone Interview of Nonresponders All of the 37 NICUs randomly selected from the nonrespondent group were successfully contacted by telephone. The nonresponding NICUs had fewer beds than the responders (n⫽33; 21⫾15 vs 28⫾17, P⫽.03), but level of RD involvement was similar. The 22 nonresponding NICUs without a full-/part-time RD were less likely to report lack of funding as the reason, and more likely to report “don’t know” or lack of need than the responding NICUs. The nonresponding NICUs were also less likely than responders to plot infants’ growth curves for weight and head circumference (85% vs 96%, P⫽.007; and 85% vs 94%, P⫽.03, respectively), and less likely to use more than 24 kcal/oz enteral feedings (21% vs 49%, P⫽.002). They also scored lower on a modified nutrition care score constructed from a subset of six of the 10 nutrition care score criteria (Table 1, criteria numbers 1, 2, 3, 8, 9, 10). DISCUSSION Seventy-six percent of the NICUs responding to our survey involved an RD to some degree in providing care; substantially more than the 38% in a similar 1990 survey of NICUs in the United States (7). However, if consult RDs had been excluded from our estimate, RD involvement would have only been 41%. We suspected that the consult RDs’ lack of consistent time in the NICU might limit their impact on practices there, and this was supported by our data. Therefore, to more fairly test RD impact on practices, full- and part-time RDs were analyzed separately from consult RDs. Full-time and parttime RDs were combined because we expected their NICU involvement to be more consistent and that the number of hours they spent in the NICU would increase with NICU size and acuity (measured by intensive care beds). The relationship between NICU size and full-/parttime RD involvement was confirmed (r⫽0.53, P⬍.0001). For most comparisons there were substantial differences between the practices reported by NICUs with a full-/ part-time RD and those by NICUs with no RD or a consult RD, suggesting that more NICUs could benefit from the involvement of full-/part-time RDs in the nutritional care of VLBW infants. Full-/part-time RDs were especially likely to be involved in the larger NICUs, which are often higher in patient acuity. Perhaps their specialty skills were recognized as being especially needed in such settings. The number of consulting RDs in NICUs today may be helpful if such individuals later increase their involvement to
full-/part-time status and become full members of the NICU team. Overall, RD presence in NICUs seems to be increasing. The barriers of perceived lack of need, funding, and trained RDs reported in 1990 (7) still seem to be hindering RD involvement in more NICUs. The perceived lack of need for an RD may be related in part to the design of our study, which consisted of all NICUs caring for VLBW infants, including smaller NICUs with lower acuity and less need for intensive nutritional care. However, this perception of the respondents, who were usually NICU directors, underscores the need for more documentation and communication by RDs to emphasize the positive impact RD services can have on improving practices and outcomes for VLBW infants. There were interesting variations in the nutrition-related practices reported by this survey. Consistent with nutrition recommendations in 1999 (9-16), NICUs with more RD involvement provided VLBW infants with less kilocalories and more protein parenterally and more kilocalories and protein enterally compared with NICUs with less RD involvement. The use of enteral protein supplements was low in all settings. However, there was a trend toward increased use overall and prevalence in NICUs as level of RD involvement increased. The reported initiation of parenteral and enteral nutrition was later than recommended for VLBW infants at that time (11,16,17) regardless of the level of RD involvement. Early and intensive parenteral and enteral nutrition were not routinely practiced in these NICUs when the survey was done in 1999. However, NICUs commonly fortified breast milk and used preterm formula for VLBWs. Fortified breast milk was used earlier and longer, and “non-preterm diets” were less likely to be reported as appropriate feeding options for VLBW infants in NICUs with full-/ part-time RDs than those with less RD involvement. The nutrition care score, summarizing the intensity of growth surveillance and nutrition practices, showed the positive impact of a full-/part-time RD in the NICU, yet nutrition care scores differed only by 20% between groups. Several factors may have contributed to making the differences smaller than they really were, including: response bias, design of the nutrition care score, and timing of the survey. First, the response rate (54%) in our survey was better than that of two other national NICU surveys [37% (18) and 42% (7)]. However, differences between our responders and nonresponders suggest that returned surveys were not a representative sample of all US NICUs. We suspect that our findings likely underestimated the true differences in nutrition practices among NICUs. The NICUs that were most interested in nutrition may have been more likely to respond to our survey, attenuating differences in practices that otherwise might exist by level of RD involvement. Second, although our survey questionnaire was pilottested (8), the nutrition care score may not have been optimally designed and adequately tested for its ability to detect the RD’s positive impact on care. The nutrition care score was designed as a summary score of the intensity of NICU growth surveillance and nutrition practices, such that a highly intensive score would represent close monitoring of growth and aggressive
nutrition practices. Although more intensive care would be considered good care for most VLBW infants, this might not be the case for all infants due to the need to individualize care. As a result, the nutrition care score may not have detected the actual prevalence of good vs intensive care, and thus may underrepresent the positive impact of RD involvement. Third, the focus on neonatal nutrition has burgeoned over the years since our survey (19-24) and the benefits of more aggressive nutrition intervention have become better known. This has likely led to more routine use of these practices, and thus clinical care today may reflect recommendations for good care more closely than in the past. Thus, we suspect that differences in nutrition practices due to RD involvement would be even greater today than they were in 1999. A final limitation of this survey study is the implicit assumption that with more nutrition involvement, and explicitly with more RD involvement, outcomes for VLBW infants will improve. In fact, additional research is needed before such a causal chain is demonstrated and the number of full-/part-time RDs in the NICU for optimal care is established. Therefore, although our findings suggest that more is better in terms of the physical presence of the RD on the NICU team, more evidence documenting the value of more specialty experience, training, and education for NICU RDs is still needed. CONCLUSIONS In more than 75% of NICUs studied, RDs were involved to some degree in providing care. However, less than 50% of NICUs employed RDs on a full- or part-time basis, indicating that there is much room for improvement in making RDs NICU team members. RD involvement in the NICU increased the intensity of important nutrition practices that may improve outcomes in VLBW infants, including the provision of fewer kilocalories and more protein parenterally, and avoidance of full-term infant diets. RD involvement and effectiveness in the NICU are still limited by NICU directors’ lack of awareness of the benefits provided by an NICU RD and possibly also by the lack of availability of trained NICU RDs. The credibility and effectiveness of NICU RDs may be enhanced if job descriptions emphasize the need for an advanced degree in nutrition science as well as specialty training that includes mentored clinical experience in neonatal and high-risk infant nutrition. This work is dedicated to the memory of the late Douglas K. Richardson, MD, MBA, who contributed substantially to this project and manuscript. Our thanks to all of the NICUs that responded to our survey and to our research assistants and other support personnel involved in this study. We thank Melody Thompson, MS, RD, of Ross Products Division, Abbott Laboratories, for sharing her experiences in conducting a previous NICU nutrition survey and her feedback on the survey instrument developed for this study. Lastly, we express our appreciation to the ADA Pediatric Nutrition dietetic practice group for providing us with NICU RD contact information and publicizing the survey in their newsletter.
August 2005 ● Journal of the AMERICAN DIETETIC ASSOCIATION
1229
Partial funding for this study was provided by Ross Products Division, Abbott Laboratories, Columbus, OH. Dr Olsen was the Ellen and Ronald Block Scholar at Tufts University, School of Nutrition Science and Policy. This material is based on work partially supported by the US Department of Agriculture, under agreement No. 58-1950-9-001. Any opinions, findings, conclusions or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the US Department of Agriculture. References 1. LaBarre DJ, Maher MM, Raye JR, Rowe JC. Effect of monitored feeding protocol on VVLBW infants [abstract]. J Am Diet Assoc. 1987;87:A-58. 2. Elsaesser KR. Dietitian intervention in neonatal intensive care reduces errors and improves clinical outcomes [abstract]. J Am Diet Assoc. 1998;98:A-21. 3. Kuzma-O’Reilly B, Duenas ML, Greecher C, Kimberlin L, Mujsce D, Miller D, Walker DJ. Evaluation, development, and implementation of potentially better practices in neonatal intensive care nutrition. Pediatrics. 2003;111:E461-E470. 4. Kuzma-O’Reilly BA. Effectiveness of nutrition appraisal in the NICU [abstract]. Pediatr Res. 2000;47: 409A. 5. Hudak LJ, Worobey J. Monitoring nutritional intake of the preterm infant: testing a nutrition management protocol. Top Clin Nutr. 1998;14:67-76. 6. Stave VS, Robbins S, Fletcher AB. A comparison of growth rates of premature infants prior to and after close nutritional monitoring. Clin Proc CHNMC. 1979;35:171-180. 7. Thompson M, Price P, Stahle DA. Nutrition services in neonatal intensive care: A national survey. J Am Diet Assoc. 1994;94:440-441. 8. Olsen I, Awnetwant E, Cardi G, Gallagher L, McKinley L, O’Donnell E, Pryzstac L, Warren L, Dwyer J, Richardson D. Variations in nutrition personnel and practice among 7 neonatal intensive care units [abstract]. J Am Diet Assoc. 1997;97:A-30. 9. Committee on Nutrition, American Academy of Pediatrics. Nutritional needs for preterm infants. In: Kleinman RE, ed. Pediatric Nutrition Handbook. 4th ed. Elk Grove Village, IL: American Academy of Pediatrics; 1998:55-87. 10. Tsang RC, Lucas A, Uauy R, Zlotkin S, eds. Nutritional Needs of the Preterm Infant: Scientific Basis and Practical Guidelines. Baltimore, MD: Williams & Wilkins; 1993.
1230
August 2005 Volume 105 Number 8
11. Groh-Wargo S, Thompson M, Hovasi Cox J, Hartline JV, eds. Nutritional Care for High-Risk Newborns. 2nd ed. Chicago, IL: Precept Press, Inc; 1994. 12. Baugh N, Recupero MA, Kerner JA. Nutritional requirements for pediatric patients. In: Christensen ML, ed. The A.S.P.E.N. Nutrition Support Practice Manual. Silver Spring, MD: American Society for Parenteral and Enteral Nutrition; 1998:24-1–24-13. 13. Carlson SJ, Johnson KJ, Cress GA, Connolly NW, Ziegler EE. Higher protein intake improves growth of VLBW infants fed fortified breast milk. Pediatr Res. 1999;45:278A. 14. Schulze KF, Stefanski M, Masterson J, Spinnazola R, Ramakrishnan R, Dell RB, Heird WC. Energy expenditure, energy balance, and composition of weight gain in low birth weight infants fed diets of different protein and energy content. J Pediatr. 1987;110:753759. 15. Ziegler EE. Protein in premature feeding. Nutrition. 1994;10:69-71. 16. Thureen PJ. Early aggressive nutrition in the neonate. Pediatr Rev. 1999;20:E45-E55. 17. Koo WWK, McLaughlin K, Saba M. Nutrition support for the preterm infant. In: Christensen ML, ed. The A.S.P.E.N. Nutrition Support Practice Manual. Silver Spring, MD: American Society for Parenteral and Enteral Nutrition; 1998:26-1–26-16. 18. Baker SF, Smith BJ, Donohue PK, Gleason CA. Skin care management practices for premature infants. J Perinatol. 1999;19(6 Pt 1):426-431. 19. Ziegler EE, Thureen PJ, Carlson SJ. Aggressive nutrition of the very low birthweight infant. Clin Perinatol. 2002;29:225-244. 20. Klein CJ. Nutrient requirements for preterm infant formulas. J Nutr. 2002;132(suppl 1):S1395-S1577. 21. Denne SC. Protein and energy requirements in preterm infants. Semin Neonatol. 2001;6:377-382. 22. Olsen IE, Richardson DK, Schmid CH, Ausman LM, Dwyer JT. Intersite differences in weight growth velocity of extremely premature infants. Pediatrics. 2002;110:1125-1132. 23. Thureen PJ, Melara D, Fennessey PV, Hay WW Jr. Effect of low versus high intravenous amino acid intake on very low birth weight infants in the early neonatal period. Pediatr Res. 2003;53:24-32. 24. Clark RH, Thomas P, Peabody J. Extrauterine growth restriction remains a serious problem in prematurely born neonates. Pediatrics. 2003;111:986990.