The Effect of Swing on Stress and Comfort in Premature Newborns: A Randomised Controlled Trial

ABSTRACT

Objectives

The study was conducted to investigate the effect of swing on stress and comfort in premature newborns who could not be reunited with their mothers.

Methods

The study was conducted in a randomised controlled experimental design. The intervention group included newborns who received swing intervention (n = 64) and the control group included newborns who did not receive swing intervention (n = 64). Research data were collected using Newborn Information Form, Premature Infant Comfort Scale and ALPS neo newborn pain and stress assessment scale. The research data were collected two times, before and after the intervention was completed.

Result

Experimental and control groups were similar regarding descriptive characteristics. There was no difference between the groups after the swing intervention (p > 0.05). It was determined that there was no significant difference between the groups in terms of comfort levels before and after the swing intervention (p > 0.05). When the intra‐group mean scores were compared, it was determined that the comfort level increased in the intervention group before and after the swing intervention, and there was a significant difference between the mean scores (p = 0.028).

Conclusion

It was determined that swing intervention did not affect the stress and comfort level of the newborn.

1. Introduction

Premature newborns are those born alive before 37 weeks of pregnancy [1]. Premature newborns or those born before their development is complete, are usually admitted to the neonatal intensive care unit because they are not ready to live in the outside world [2]. Neonatal intensive care units, whatever the cause (prematurity, sepsis or congenital defects), often cause separation that leads to emotional scars for mothers and babies [3]. It has been documented that premature neonates, especially those exposed to stress or pain, are metabolically, behaviourally and clinically more susceptible [4].

Postnatal parent‐neonatal bonding is very important for the development of the newborn in terms of both immediate and future outcomes [5]. Ensuring mother and baby togetherness provides positive and important benefits in terms of hospital stay, thermal regulation, reduction of pain during the procedure, bonding and sleep [6]. Separation makes the transition to parenthood and family‐centred care more difficult and weakens the sense of unity [7]. Accordingly, neonatal intensive care is a particularly distressing experience for both mother and newborn due to prolonged hospitalisation [8]. This situation may lead to fear, uncertainty and insecurity in the mother, as well as the abstraction of reality, and may lead to a break in the emotional bond with the inability to meet the need for vulnerability, dependence and protection in the newborn [9]. Parents report that physical separation, the intensive care environment, high‐tech equipment and the environment negatively affect their ability to bond with their babies. They also state that it makes them feel that the baby does not belong to them [10, 11].

Although hospital practices vary in Turkey, preventing mother–infant separation and integrating the mother into the care of the baby has a high priority in the neonatal intensive care unit where the study was conducted. In short‐term hospitalisations, the mother is hosted, whereas in long‐term hospitalisations, she can see her baby either in a guesthouse for a fee or by travelling back and forth from her home. However, it may be difficult to see her baby by travelling back and forth due to some obstacles such as transportation and financial difficulties. This situation may cause negative effects on the newborn. It has been determined by clinical observations of researchers that newborns who are used to the smell and voice of the mother but who are no longer visited by their mother are restless and have crying spells. Therefore, to promote better neurological development in premature newborns, various interventions are made with an emphasis on more humane care [4]. One of these interventions, rocking, seems to be an advantageous alternative in supporting the baby's sensory development.

Rhythmic movements in the form of rocking/shaking are one of the effective methods to relax newborns and reduce pain [12]. Pillai et al. found that swinging is one of the most effective methods to reduce pain in their systematic review of non‐pharmacological management of procedural pain [13]. Putting babies to sleep by swinging them to sleep standing up or in a cradle is a very common practice in Turkish society [14]. When the studies are examined, it is stated that it is one of the traditional methods applied to comfort the newborn [14, 15, 16].

It is thought that swing intervention for newborns under intense stress who cannot be brought together with their mothers in line with safe sleep recommendations will positively affect their stress and comfort levels. However, no intervention studies were found in the literature. In this direction, the study was realised to investigate the effect of swing intervention on stress and comfort in premature newborns who could not be brought together with their mothers.

1.1. Hypotheses

1. Swing intervention decreases the stress level in premature newborns.

2. Swing intervention increases the comfort level in premature newborns.

2. Methods

2.1. Design

The study was conducted in a randomised controlled design.

2.2. Setting

The study was conducted in the Level III NICU of a university hospital between January 2024 and June 2024.

2.3. Participants

Newborns admitted to a university hospital's third‐level intensive care unit comprised the study population. A medium effect size was chosen because there is no research on swing intervention in the literature. Based on a medium effect size (f: 0.50), 80% power, and 95% (α error) confidence level, the sample size was calculated to be 128 with 64 subjects per group using the G*Power V.3.1.9.4 tool. The intervention group included newborns who received swing intervention (n = 64) and the control group included newborns who did not receive swing intervention (n = 64). Newborns that met the following requirements were eligible: (a) between 30 + 1 and 36 + 7 weeks; (b) whose mothers had not visited in the previous week; (c) with a 5th‐min Appearance, Pulse, Grimace, Activity, and Respiration (APGAR) score greater than seven; (d) weighing 1500 g or more at birth; (e) without any health issues that would have prevented them from swinging and (f) for whom the parents had signed an informed consent form. Newborns with (a) congenital defects, (b) invasive and non‐invasive respiratory support, (c) severe illnesses that could impair physiological reactions to pain and necessitate the use of sedatives, muscle relaxants or antiepileptic drugs or (d) other conditions that hinder swinging (such as fever or oxygen therapy) were excluded.

2.4. Data Collection Tools

The research data were collected using Newborn Information Form, Premature Infant Comfort Scale and ALPS neo newborn pain and stress assessment scale.

2.4.1. Newborn Information Form

It consists of 11 questions examining the sociodemographic (gender, date of birth, date of hospitalisation, birth week, birth weight, head circumference at birth, diagnosis of hospitalisation, mode of delivery, maternal age, number of pregnancies, follow‐up status during pregnancy, etc.) characteristics of the participants.

2.4.2. ALPS Neo Newborn Pain and Stress Assessment Scale

A vital tool for evaluating pain and stress in both premature and term babies is the Lundqvist et al. [17] scale. It includes observational assessments of five criteria: activity level, tone of voice, breathing, hand/ft movements and facial expression. Different levels of pain and stress are indicated by scores ranging from 0 to > 5, where 0–2 denotes no pain or stress, 3–5 denotes mild pain and stress, and scores > 5 indicate severe pain and stress. With an initial α coefficient of 0.95, the instrument exhibits remarkable internal consistency. Cronbach's α in the study by Ceylan and Bolışık was 0.87 in the pre‐test, 0.89 in the application and 0.72 in the post‐test application [18]. These reliability coefficients were marginally lower but still acceptable. Significant interobserver agreement is further supported by the Kappa analysis results (between 0.83 and 1.00) and an intraclass correlation coefficient (ICC) (r = 0.83–1.00), indicating that the scale is dependable for use in clinical evaluations of newborn discomfort and stress. Because of these qualities, it is a useful tool for efficiently monitoring and controlling stress and discomfort in neonatal care settings.

2.4.3. Premature Infant Comfort Scale

The behavioural and psychological comfort as well as discomfort are assessed using the assessment method created by Ambuel et al. [19] and modified for premature newborns by Monique et al. [20]. Alemdar and Tüfekçi's [21] validity and reliability study in Turkey confirms the scale's efficacy in this population. Seven primary criteria are assessed using the 5‐point Likert scale: mean heart rate, muscular tone, facial motions, respiratory status or weeping, alertness, calmness/anxiety and physical movement. With a total score ranging from 7 (very comfortable) to 35 (least comfortable), a higher score denotes poorer degrees of comfort in this rating system. An intervention is required to lessen pain and increase comfort if the total score is 17 or higher. With a Cronbach's α coefficient of 0.88, which indicates good internal consistency, the scale has strong dependability. Furthermore, the intraclass correlation (ranging from 0.83 to 1.00) and kappa analysis (values between 0.83 and 1.00) show strong interobserver agreement, guaranteeing that various observers would yield consistent scoring results. Because of this, the scale is a trustworthy instrument for determining how comfortable premature babies are.

2.5. Randomisation and Blinding

After the pre‐test was applied to the participants who met the inclusion criteria, the group was determined by randomisation. Stratified randomisation method was used for randomisation. Four blocks were formed according to the number of postnatal days (30^0/7–33^6/7 and 34^0/7–36^6/7) and gender (male and female) of the newborns and 32 newborns were included in each block. Neonates were assigned to groups by an unbiased researcher who was not involved in the study using a simple randomisation random number generation program (Research Randomizer) to the blocks created by stratified randomisation.

The study guarantees the statistician's blindness. Since the researcher was a practitioner, she was not blinded. An independent academician other than the researcher coded the groups as A and B, and an independent statistician analysed the data to avoid bias in the evaluation.

2.6. Interventions

In the intensive care unit, drug treatment is administered every 4 h, vital signs are evaluated every 3 h, and care (change of nappies and bed linen, oral care, nasal cleansing, bathing, catheter change, blood sampling and nutrition) is routinely given according to the need. Incubator temperature is kept between 33°C and 34°C. The incubators are covered with a sheet to prevent newborns from being affected by light and the lights are switched off outside the care hours. To prevent newborns from being affected by noise, care is taken not to talk on the phone and loudly, and the alarm sounds of the devices are not too loud. During the data collection phase of the study, no intervention was made in the treatment or care of newborns. The newborns included in the study had a treatment plan according to the physician's order and general care was provided.

2.6.1. Intervention Group

An automatic infant swing was used for swing intervention and all newborns were kept in the swing for 30 min at Level 2 slow level. The swing can be used from birth, has a belt with 3‐point lock, adjustable control, 5‐stage automatic swing level and three different length timers. It has breathable fabric. The intervention was initiated at least 30 min after feeding after evening care and administered once.

2.6.2. Control Group

No intervention other than routine care was performed and research data were collected at the same time intervals with the experimental group.

2.7. Data Collection

The research data were collected two times: before and after the implementation. The intervention was initiated at least 30 min after feeding after evening care and administered once. Since the scales were evaluated through observation, they were evaluated by the neonatal nurse as well as the researcher.

2.8. Analysis

SPSS version 23 was used to analyse the data (SPSS Inc.). The information was displayed as mean and standard deviation, lowest and maximum values, percentages and figures. The Kolmogorov–Smirnov test was used to assess the data's conformance to a normal distribution. The data in the dependent and independent groups were compared using the Student's t‐test. Effect size values identified by Cohen (1992) as small (d = 0.2), medium (d = 0.5) and large (d = 0.8) were taken into account when assessing the effect size. Two viewers participated in the study, and the dependability of ratings was assessed using an ICC. A p value < 0.05 was deemed statistically significant for all analyses.

2.9. Ethics

This study was confirmed by the KTO Karatay University Ethics Committee for Pharmaceutical and Non‐Medical Device Research (dated 22 December 2023, number 16). All study participants provided verbal informed consent before questionnaire administration, cervical sample collection, and visual inspection procedures. The study was conducted by the Declaration of Helsinki.

3. Results

3.1. Characteristics of Participants

For the swing intervention, 212 newborns were included in the sample. Newborns who were connected to a mechanical ventilator (38 newborns), receiving sedatives, muscle relaxants or antiepileptic drugs (20 newborns), receiving oxygen therapy (26 newborns) and whose mothers were visiting (n = 54) were excluded from the study. All of the remaining 128 neonates met the inclusion criteria. The final sample size consisted of 128 participants randomised into two groups: intervention (n = 64) and control (n = 64) (Figure 1). Table 1 shows the descriptive characteristics. The groups were similar in terms of descriptive characteristics (p > 0.05, Table 1).

FIGURE 1.

CONSORT‐2022 flow diagram.

TABLE 1.

Characteristics and clinical data of neonates.

	Experimental group (n = 64)	Control group (n = 64)	t/X 2	p
Gender
Male	29 (45.3)	29 (45.3)	0.000 a	1.000
Female	35 (54.7)	35 (54.7)
Delivery method
Spontaneous vaginal delivery	39 (60.9)	47 (73.4)	1.736 b	0.188
Caesarean section	25 (39.1)	17 (26.6)
Gestational age at birth, weeks	31.95 ± 2.84	31.70 ± 2.931
Postnatal age at enrollment, days	34.22 ± 1.56	34.22 ± 1.54	0.000 c	1.000
Birth weight, g	1841.02 ± 678.78	1778.66 ± 656.34	0.528 c	0.598
Pregnancy monitoring
Made	60 (93.8)	57 (89.1)	0.812 b	0.418
Not made	4 (6.2)	7 (10.9)

Note: Data are n (%), median (P25–P75) or mean ± SD.

^a Pearson's chi‐squared test.

^b Continuity correction.

^c Independent sample t‐test.

3.2. Comparison of Mean ALPS‐Neo Scores

The ALPS‐Neo mean scores of the intervention and control groups differed before the swing intervention. Following the swing intervention, there was no difference between the groups (p > 0.05, Table 2). The mean scores before and after the swing intervention in both groups did not significantly differ from one another, according to a comparison of within‐group mean scores (Table 2).

TABLE 2.

Intragroup and intergroup comparison of intervention and control groups.

	Time	Intervention group	Control group	t‐test a	p
ALPS‐NEO	Pre‐intervention	2.16 ± 2.39	1.11 ± 1.85	2.768	0.006
	Post‐intervention	1.67 ± 1.47	1.50 ± 1.18	0.729	0.467
	t b /p	1.319/0.192	−1.339/0.185

^a Independent samples test.

^b Paired samples statistics.

3.3. Comparison of the Mean Scores of the Premature Infant Comfort Scale

The groups' comfort levels before and after the swing intervention did not differ significantly (p > 0.05, Table 3). Comparing the intra‐group mean scores revealed a significant difference between the mean scores (p = 0.028, Table 3) and an increase in comfort in the intervention group both before and after the swing intervention.

TABLE 3.

Comparison of the mean scores of the Premature Infant Comfort Scale according to the groups.

	Time	Intervention group	Control group	Test a	p
Comfort	Pre‐intervention	22.70 ± 4.35	23.02 ± 3.56	−0.445	0.657
	Post‐intervention	24.16 ± 2.55	23.69 ± 3.21	0.914	0.362
	Within group Analysis b	t = −2.255 p = 0.028	t = −1.018 p = 0.313

^a Independent samples test.

^b Paired samples statistics.

4. Discussion

Although the union of mother and baby is of great importance in the postnatal period, many factors may prevent this union. Especially neonatal intensive care is one of the most important factors affecting this process. Although there are no restrictions on maternal visits in neonatal intensive care units, mothers living in different provinces or districts have difficulty in making visits, especially because the mother has to travel back and forth from home in long‐term hospitalisations. It has been determined by the clinical observations of the researchers that newborns are restless and have crying spells as a result of the mother's absence. There are no studies on the subject in the literature; the study was conducted to examine the effect of swing intervention on stress and comfort in premature newborns who could not be brought together with their mothers. To determine whether the swing is a feasible method to reduce neonatal stress and pain levels and to increase comfort, the findings of the study are discussed in this section together with the results of the literature.

Swing intervention was found not to affect neonatal pain and stress levels. It has been previously reported that hammock intervention decreased heart and respiratory rate and pain scores and improved sleep‐wakefulness status compared to neonates with traditional positioning methods [22]. Studies have found that alternative positions, such as hammocks, reduce the pain and stress levels of premature newborns [23, 24, 25]. However, the fact that it did not make a difference in this study shows how effective and important the stress that occurs in newborns who initially meet with the mother but cannot come together is. Furthermore, this intervention is an innovative approach to improve models of care in neonatal intensive care and further research is needed to evaluate the long‐term effects of the intervention.

It was found that the swing intervention increased the comfort level in the intervention group. In the literature, it has been reported that methods such as facilitated lying position [26], white noise [27, 28], recorded maternal voice and MiniMuffs [28], therapeutic toys [29], Yakson touch and maternal voice [30], kangaroo mother care [31] have a positive effect on increasing the comfort of preterm newborns. However, it was found that the methods were applied to increase comfort related to interventions such as heel prick blood collection, nasal continuous positive airway pressure application and intravenous cannula insertion. It is not possible for mothers to be present in intensive care units for 24 h. Therefore, it is of great importance to perform interventions that will reduce the stress and increase the comfort of the newborn. This helps premature newborns to reduce their stress, increase their comfort and direct their energy towards coping with stress and their development. However, it is important to conduct studies on the subject in order to demonstrate the level of evidence and to generalise the results.

5. Conclusion and Recommendations

In the study conducted to examine the effect of swing intervention on the stress and comfort level of the newborn, it was determined that the in‐group comfort level increased in the intervention group, although there was no significant difference between the groups. In this regard, it is recommended to conduct experimental studies examining the long‐term results of the swing intervention and comparing it with other non‐pharmacological methods to reduce the pain and stress levels of newborns who cannot constantly be together with their mothers due to the neonatal intensive care environment.

5.1. Limitations

This randomised clinical trial is the first study to examine the effect of rocking on stress and comfort in premature neonates who cannot be reunited with their mothers. The strengths of the study are that it was conducted with a randomised controlled experimental design, stratified randomisation was used, and an independent statistician performed the analysis for data security. Limitations of the study include the lack of existing guidelines on shaking intervention and the lack of evaluation of physiological parameters.

Author Contributions

Conceptualisation of the project: H.K., A.C. and B.A. Visualisation of study design: H.K., A.C. and B.A. Data curation: H.K., A.C. and B.A. Data access and verification, data analysis and interpretation, writing of manuscript: H.K., A.C. and B.A. Review and editing of manuscript: H.K., A.C. and B.A.

Ethics Statement

This study was confirmed by the KTO Karatay University Ethics Committee for Pharmaceutical and Non‐Medical Device Research (dated 22 December 2023, number 16). All study participants provided verbal informed consent before questionnaire administration, cervical sample collection and visual inspection procedures. The study was conducted by the Declaration of Helsinki.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to restrictions, for example, they contain information that could compromise the privacy of research participants.