Article Text
PDF
Abstract
Objectives To determine the effects on weight gain and temperature control of transferring preterm infants from incubators to open cots at a weight of 1600 g versus a weight of 1800 g.
Design Randomised controlled trial.
Setting One tertiary and two regional neonatal units in public hospitals in Queensland, Australia.
Participants 182 preterm infants born with a birth weight less than 1600 g, who were at least 48 h old; had not required ventilation or continuous positive airways pressure within the last 48 h; were medically stable with no oxygen requirement, or significant apnoea or bradycardia; did not require phototherapy; and were enterally fed with an intake (breast milk/formula) of at least 60 ml/kg/day.
Interventions Transfer into an open cot at 1600 or 1800 g.
Main outcome measures The primary outcomes were temperature stability and average daily weight gain over the first 14 days following transfer to an open cot.
Results 90 infants in the 1600 g group and 92 infants in the 1800 g group were included in the analysis. Over the first 72 h, more infants in the 1800 g group had temperatures <36.4°C than the 1600 g group (p=0.03). From post-transfer to discharge, the 1600 g group had more temperatures >37.1°C (p=0.02). Average daily weight gain in the 1600 g group was 17.07 (SD±4.5) g/kg/day and in the 1800 g group, 13.97 (SD±4.7) g/kg/day (p=<0.001).
Conclusions Medically stable, preterm infants can be transferred to open cots at a birth weight of 1600 g without any significant adverse effects on temperature stability or weight gain.
Trial registration: ACTRN12606000518561 (http://www.anzctr.org.au).
Statistics from Altmetric.com
Introduction
A key criterion for discharging preterm infants home from nurseries is their ability to maintain temperature once transferred from an incubator to an open cot. There is a plethora of information on providing neutral thermal environments for preterm infants but little to guide practice on when to transfer infants from incubators to open cots. Practice variation exists between continents, regions and nurseries in the same countries.1 2 A target weight, largely based on tradition or the professional experience of clinicians,3 of 1700–1800 g has been used as a measure for transferring infants from incubators to open cots.4 Delaying transfer until an arbitrary weight has been reached may result in longer hospitalisation than necessary, thus increasing the cost of care.5 This delay may also have a negative impact on the infant and family (delaying parenting and nurturing roles) and on the ability of neonatal nurses to provide support for the social development of the infant.4
What is already known on this topic
There is little evidence from randomised controlled trials on the timing or weight that preterm infants can be transferred from incubators to open cots.
What this study adds
▶ Medically stable preterm infants can be transferred to open cots at weights of 1600 g without apparent adverse effects on temperature or weight gain.
▶ Earlier transfer may not decrease length of hospital stay for infants born very preterm as opposed to those born greater than 32 weeks gestational age.
Prior to the trial, we undertook a systematic review which concluded that there was insufficient evidence to guide clinical practice on the weight to transfer preterm infants from incubators to open cots.4 There was evidence from two quasi-randomised studies suggesting that transfer to an open cot may safely occur at weights ≥16006 or ≥1700 g.7 More recent studies suggest potential transfer at 1500 g8 and at 1600 g.9
Given the conclusion of the systematic review,4 this clinical trial selected specific entry points related to body weight (1600 vs 1800 g) and outcome measures based on the literature,5 clinical relevance and ease of measurement by the nurse at the cot-side (eg, temperature, weight).
Methods
Participants
This prospective, randomised, controlled trial was undertaken between 30 June 2003 and 13 February 2009 in one tertiary and two regional neonatal units in Australia. The study was approved by the ethics committees of each hospital and the University of Queensland.
Preterm infants born less than 1600 g were assessed for eligibility. Inclusion criteria included postnatal age of at least 48 h, medical stability (no oxygen requirement, no significant apnoea or bradycardia), no phototherapy requirement and enteral feed intake of at least 60 ml/kg/day. Infants were excluded if they required ventilation or continuous positive airways pressure within the last 48 h or had a major congenital abnormality.
Randomisation and masking
An independent researcher managed computer-generated randomisation (GraphPad StatMate software; GraphPad Software, San Diego, California, USA) and masking with group allocation cards sealed in opaque envelopes. Infants were stratified by birth weight (< or ≥1000 g), gestational age (GA) at birth (< or ≥34 weeks) and hospital, with random block sizes of 6, 8, 10 or 12.
Written informed consent was obtained from one or both parents. On reaching a weight ≥1600 g, the nurse caring for the infant selected the next envelope in sequence to reveal the group allocation. Infants were randomised into the intervention group (open cot at 1600 g) or control group (open cot at 1800 g). Treatment groups could not be masked.
Procedures
The neonatal units were chosen due to similarity in weaning and discharge practices. Each unit used central temperature control systems to maintain temperatures at 24–26°C with relative humidity ≤55%. On transfer, infants were dressed in a singlet, a cotton full-length jumpsuit, a woollen hat and wrapped in a flannelette sheet and a cotton blanket. A quilt was placed over the infant's bedclothes. Post-transfer axillary temperatures were measured at 1 h, 3 h, then every 3 h until 72 h, and thereafter a minimum of three times a day until discharge. Temperature stability was defined as the axillary temperature range of 36.4–37.1°C.10
If an infant's temperature fell below 36.4°C, a predetermined sequence was instituted: additional clothing, additional bedding, an overhead heater and finally return to an incubator. Failure of transfer was defined as return to an incubator for inability to maintain an axillary temperature above 36.4°C within 3 h of being under an overhead heater or if an overhead heater was required on more than two occasions, for periods of greater than 3 h each time, in a 24-h period.
Each infant was weighed naked every second day, prior to a feed, using electronic scales. An exponential model for calculating growth velocity was used.11
Outcomes and sample size calculation
Primary outcomes were temperature stability after transfer to discharge and average daily weight gain post-transfer over 2 weeks. Secondary outcomes were the proportion of infants requiring an overhead heater, the proportion of infants returned to an incubator, length of stay (LOS) (from randomisation to discharge), postmenstrual age (PMA) and weight at discharge.
An a priori retrospective chart audit conducted by the investigators of infants transferring to open cots at 1800 g revealed the following: (1) the proportion of infants needing an overhead heater was 32%, (2) the proportion of infants returned to an incubator was 6.5%, and (3) a mean ± SD daily weight gain was 31.7 ±11.1 g. Calculations indicated a sample size of 90 infants in each group was required to provide 80% power (with α = 0.05) to detect an increase in overhead heater use from 32% to 52%, an increase in return to incubator from 6.5% to 20% and a difference in weight gain of 4.7 g/day.
Statistical analysis
SPSS software (version 17) was used for statistical analysis. Analyses were by intention to treat. Comparisons between groups were tested for statistical significance using Student's t test, Mann-Whitney, χ2 or Fisher's exact statistics. Subgroup analysis included birth weight and GA.
Results
Of 814 infants screened for eligibility, 102 did not meet the inclusion criteria at the time of reaching 1600 g, 79 declined to participate and 450 infants did not participate for other reasons (figure 1). Participant versus non-participant birth weight (1267 (SD ±255.9) g vs 1285 (SD±221.8) g (p=0.44)) and GA (30.2 (SD± 1.9) weeks vs 30.6 (SD±2.1) weeks (p=0.016)) were similar. While the difference in GA was statistically significant, a difference of 4 days is probably not clinically significant.
Number of infants assessed for trial eligibility, randomised and included in the intention-to-treat analysis.
The remaining 183 were enrolled and randomly assigned to the 1600 g group (n=90) or the 1800 g group (n=93). The parents of one infant randomised to the 1800 g group withdrew from the trial. Baseline characteristics are presented in table 1 and subgroup characteristics in table 2.
Infant baseline characteristics at birth
Subgroup characteristics of studied infants
Characteristics at randomisation (table 3) were similar between groups. Differences were seen at the time of transfer, as expected due to the study design and unit practice (table 3).
Infant characteristics at randomisation and transfer
Feed volumes varied at the time of transfer as the current procedure at the participating hospitals is to grade feeds from second to third hourly at 1600 g. This results in an initial decrease in ml/kg/day before feeds are re-established at 180 ml/kg/day over a median of 18.0 h (IQR 6–24). There was a small but statistically significant difference of 0.7°C (p=0·006) between incubator air temperatures at the time infants were transferred.
Temperature stability over the first 72 h and from post-transfer to discharge differed between the groups (table 4). Over the first 72 h, more infants in the 1800 g group had temperatures <36.4°C than the 1600 g group (p=0.03), as did the subgroup <1000 g and <34 weeks gestation (p=0.03). However, these results should be interpreted with caution as not all infants had axillary temperatures measured at each three hourly time point. Infants transferred at 1800 g had a greater number of temperature measurements omitted, except at the 24 and 72 h time points.
Outcome variables by intervention groups
No difference was seen between the two groups for frequency of temperatures >37.1°C. However, a significant difference was seen from post-transfer to discharge, with the 1600 g group having more episodes per infant of temperatures >37.1°C (p=0·02). There were no significant differences in any of the subgroups.
The average daily weight gain was 17.07 (SD ±4.5) g/kg/day in the 1600 g group and 13.97 (SD ±4.7) g/kg/day in the 1800 g group (p=<0.001, table 4). For the subgroup ≥1000 g and <34 weeks gestation, the 1600 g group had a greater average daily weight gain (p=0.01).
The proportion of infants requiring an overhead heater or those who returned to an incubator was not significantly different between the two groups (table 4). While some infants were not placed under the overhead heater (n=17, 11 in 1600 g group and 6 in 1800 g group) or returned to an incubator (n=3, 2 in 1600 g group and 1 in 1800 g group) as prescribed in the protocol, all data were analysed using the intention to treat. Three infants were returned to an incubator for failure to maintain their temperature >36.4°C. The remaining 10 infants were returned to an incubator for other reasons (table 5).
Reasons for return to an incubator
LOS, PMA at discharge home and weight at discharge home were similar between the groups (table 4).
Post hoc analysis was undertaken for a possible explanation for no difference in LOS. Seventy-one percent of enrolled infants were born at less than 32 weeks and 67% were exclusively breastfeeding on discharge. Bottle feeding infants achieved full suck feeds on average 6 days sooner than the breastfeeding infants.
Discussion
This study demonstrated that medically stable, preterm infants can be transferred to open cots at a weight of 1600 g without adverse effects on temperature stability or weight gain. Birth weight (<1000 g) does not appear to influence success or failure to wean to an open cot. Earlier transfer to an open cot does not necessarily result in shortened LOS.
We identified one randomised9 and two quasi-randomised6 7 controlled trials and five observational studies3 8 12,–,14 assessing transferring preterm infants to standard open cots. Sutter and colleagues7 showed that preterm infants could be safely transferred to open cots at 1700 g but only for infants with birth weights >1000 g. However, they expressed caution in interpreting the results, due to the small number of infants with birth weights <1000 g (n=12). Our study findings indicate that infants with birth weights <1000 g transferred at 1600 g (n=15) were not more likely to fail to maintain their temperature, require an overhead heater or return to an incubator than their 1800 g counterparts (n=17). West and colleagues8 observed similar results. However, given these findings are based on relatively small numbers, the results should be interpreted with caution.
Previous studies3 14 15 suggest thermally challenging infants by weaning the incubator temperature to 28°C facilitates successful transfer. Some nurseries target incubator air temperatures of 29°C.1 We noted a decrease in incubator air temperature with increasing postnatal age, which is consistent with the influence of postnatal skin maturation.17 The majority of infants in both groups had incubator temperatures over 29°C at randomisation (table 3) and were transferred to an open cot irrespective of incubator temperature. Of the infants who failed to maintain their temperature, incubator temperature does not appear to have influenced the success or failure to wean, suggesting that thermal challenge is not necessary.
West and colleagues8 reported more high temperatures per infant than low and suggested that perhaps nurses perceived infants being transferred at lower weights as more vulnerable and added additional wraps. Our findings were similar with the majority (80%) of infants experiencing more high temperatures than low. Although our protocol specified identical clothing, bedding and wrapping requirements for infants on transfer, it did not require documentation of the numbers of layers, so like West and colleagues,8 we are unable to assess if this occurred. Extra clothing and bedding for infants transferred at 1600 g remains a possible explanation for the significant difference between the two groups for the number of temperatures >37.1°C.
An observational study of 2900 infants transferred at mean weight of 1850 g14 found a decreased weight gain of 1 g/day for each additional 100 g an infant weighed prior to transfer. Additionally, they reported that earlier transfer was associated with higher growth velocity. In our study, infants who transferred at 1600 g had a greater average daily weight gain of 3–4 g. West and colleagues8 reported no difference in weight gain rate between the four groups in their study (1500, 1600, 1700 and 1800 g). However, the 1600 g group had a weight gain after transfer of 20.1 g/kg/day as opposed to those transferred at other weights, whose weight gain was approximately 16 g/kg/day. The lack of statistically significant difference reported may be a type 2 error, given there were only 15 infants in each group. We speculate that infants transferred at lower weights are monitored more closely (given our finding of more episodes of higher temperatures), get cold less and have more energy available for weight gain.
In our study, infants weighing 1600 g were transferred to an open cot 7 days earlier than those at 1800 g; however, this did not result in shortened LOS. This is in keeping with earlier studies in which infants have been transferred to standard open cots8 or heated mattress systems in open cots.2 18 A possible explanation is achieving full oral feeds has more influence on the timing of discharge. Bakewell-Sachs and colleagues19 report the feeding milestone is one of the last milestones to be achieved by preterm infants born less than 32 weeks. Zecca and colleagues9 recently reported a significantly shortened LOS for preterm infants transferred at 1600 g, with a mean GA of 32 weeks and mean individual amount of breastfeeding at discharge of 43–46%. The infants in our study were considerably younger than those in the study of Zecca et al9 and a greater percentage were discharged exclusively breastfeeding, which may account for our result of no difference in LOS.
Limitations
Our study could not be blinded so staff and parents may have been more vigilant in ensuring that infants in the 1600 g group were kept warm for the first 72 h post-transfer.
There were protocol violations with regard to temperature measurements and the use of the overhead heater. This resulted in 11 infants in the 1600 g group and 6 in the 1800 g group not being placed under an overhead heater. Some of these infants may have required the next step in the predetermined sequence of return to incubator, however, all infants had multiple serial temperature measurements and ongoing temperatures less than 36.4°C would have been noted and acted upon. All temperatures returned to normal limits, and given the average daily weight gain of those in the 1600 g group, this does not appear to have had adverse effects on weight gain. However, this must be viewed with caution as growth patterns were not measured and we are unable to know what effects these cool episodes may have had on these infants.
Conclusions
Medically stable, preterm infants can be transferred to open cots at a weight of 1600 g without significant adverse effects on temperature stability or weight gain.
Acknowledgments
Thanks to the parents of infants, and the nurses and doctors who participated in this study and to research supervisors for their guidance. Thanks to Mr Kimble Dunster for managing computer-generated randomisation and masking group allocation. This trial was funded by research grants from the Royal Brisbane and Women's Hospital Foundation, Queensland Health and the Queensland Nursing Council.
References
Footnotes
-
Funding Queensland Health, Queensland Nursing Council, Royal Brisbane and Women's Hospital Foundation. The funding sources did not participate in the study design, data collection, data analysis, data interpretation or writing of the report.
-
Competing interests None
-
Ethics approval This study was conducted with the approval of The University of Queensland Royal Brisbane and Women's Hospital, Toowoomba Hospital, Redcliffe and Caboolture Hospitals ethics committees.
-
Provenance and peer review Not commissioned; externally peer reviewed.
-
Correction notice This article has been corrected since it was published Online First. The author affiliations have been corrected.