No additional studies or data were found on the updated search to December 2004.
Recurrent apnea is common in preterm infants, particularly at very early gestational ages. These episodes of loss of effective breathing can lead to hypoxemia and bradycardia, which may be severe enough to require resuscitation including use of positive pressure ventilation or other treatments. Physical stimulation is often used to restart breathing and it is possible that repeated stimulation, such as with an oscillating mattress or other kinesthetic stimulation, might be used to treat infants with apnea and prevent its consequences.
In preterm infants with apnea, to determine whether the use of kinesthetic stimulation leads to clinically important reductions in apnea and bradycardia, use of mechanical ventilation or continuous positive airways pressure, or neurodevelopmental disability, without clinically important side effects.
The standard search strategy of the Cochrane Neonatal Review Group was used. This included searches of the Oxford Database of Perinatal Trials, Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 4, 2004), MEDLINE, PREMEDLINE, CINAHL, and EMBASE up to December 2004. Searches were performed of previous reviews including cross references, abstracts, conferences, symposia proceedings, expert informants, and journal handsearching mainly in the English language.
All trials using random or quasi-random patient allocation that compared kinesthetic stimulation to placebo or no treatment in preterm infants with apnea of prematurity.
Standard methods of the Cochrane Collaboration and its Neonatal Review Group were used with separate evaluation of trial quality, data extraction by both authors and synthesis of data using relative risk and weighted mean difference. As all three included trials were crossover trials, the data were extracted from all exposure periods and combined where appropriate. Measures of severity of apnea as well as the response to treatment were consistent with an evaluation of 'clinical apnea', as defined by the American Academy of Pediatrics.
Three crossover studies (Korner 1978; Tuck 1982; Jirapaet 1993) were identified that compared a form of kinesthetic stimulation to control for the treatment of apnea of prematurity. No study reported a clinically important reduction (>50%) in apnea. Using a lower threshold (>25%), Korner 1978 reported less apnea and bradycardia in infants whilst on an oscillating water bed. Tuck 1982 demonstrated a reduction in frequency of apneas (> 12 seconds) associated with bradycardia (< 100 bpm), apneas associated with hypoxia (TcP02 < 50 mmHg), and apneas requiring stimulation in infants on a rocking bed. Individual patient data were not available from the author to determine if there was an important reduction in clinical apnea. No outcome could be extracted from the study using a 'vertical pulsating stimulus' by Jirapaet 1993 that was consistent with the definition of clinically important apnea. Jirapaet 1993 reported no infants required resuscitation or ventilation. Adverse events such as death, intraventricular hemorrhage and neurodevelopmental disability were not reported.
There is insufficient evidence to recommend kinesthetic stimulation as treatment for clinically significant apnea of prematurity. Previous reviews have suggested that kinesthetic stimulation is not effective at preventing apnea of prematurity (Henderson-Smart 2005) and is not as effective as theophylline at treating clinically significant apnea of prematurity (Osborn 2005).
Apnea in infants has been defined as a pause in breathing of greater than 20 seconds or an apneic event less than 20 seconds associated with bradycardia and/or cyanosis (Nelson 1978). Recurrent episodes of apnea are common in preterm infants and the incidence and severity increases at lower gestational ages. Although apnea can occur spontaneously and be attributed to prematurity alone, it can also be provoked or made more severe if there is some additional insult such as infection, hypoxemia or intracranial pathology (Henderson-Smart 1995).
If prolonged, apnea can lead to hypoxemia and reflex bradycardia which may require active resuscitative efforts to reverse. There are clinical concerns that these episodes might be harmful to the developing brain or cause dysfunction of the gut or other organs. Frequent episodes may be accompanied by respiratory failure of sufficient severity as to lead to intubation and the use of intermittent positive pressure ventilation (IPPV).
Physical stimulation by nursing staff is commonly used to arouse the apneic infant and so stimulate breathing. This raises the question of whether frequent physical stimuli might reduce the number of apneic events. Furthermore, some believe that the preterm infant is deprived of the frequent stimuli that would be felt in utero and that substituting these with an oscillating mattress to provide kinesthetic stimulation might improve growth and development.
In preterm infants with apnea, whether the use of kinesthetic stimulation leads to clinically important reductions in apnea and bradycardia, use of mechanical ventilation or continuous positive airways pressure (CPAP), or neurodevelopmental disability, without clinically important side effects.
All trials using random or quasi-random patient allocation were included.
Preterm infants with recurrent clinical apnea with or without associated bradycardia, cyanosis or hypoxia.
Kinesthetic stimulation (various forms of oscillating mattresses or other repetitive stimulation involving moving the baby) used as treatment for apnea of prematurity compared to placebo or no treatment.
Measures of severity of apnea as well as the response to treatment must be consistent with an evaluation of 'clinical apnea' as defined by the American Academy of Pediatrics (see above). A reduction in the frequency of 'clinical apnea' that might be expected to affect clinical practice is taken as a greater than 50% reduction in the frequency of 'clinical apnea' compared to the control.
Outcomes:
1) Persisting apneas +\- bradycardia;
2) Mean rates of apnea/bradycardia;
3) Hypoxemic episodes associated with apnea;
4) Failure of treatment as indicated by use of additional measures such as use of pharmacological stimulation, IPPV or CPAP;
5) Death before hospital discharge;
6) Rate of intraventricular haemorrhage;
7) Neurodevelopmental status at follow up.
The standard search strategy of the Cochrane Neonatal Review Group was used. This included searches of the Oxford Database of Perinatal Trials, Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 4, 2004), previous reviews including cross references, abstracts, conferences and symposia proceedings, expert informants, journal hand searching mainly in the English language. MEDLINE (1966-December 2004), EMBASE (1980-2004) and CINAHL (1982-December 2004) were searched. Abstracts of the Society for Pediatric Research were hand searched for the years 1996 to 2004 inclusive.
The Cochrane Controlled Trials Register was searched using search terms 'kinesthetic', 'water bed', 'air bed', 'oscillating', '(apnoea or apnea) and (infant or preterm or neonate or newborn)', 'rocking', and 'vertical pulsating'.
MEDLINE was searched using MeSH headings 'apnea and infant-premature'; and text words 'kinesthetic', 'water bed', 'air bed', 'oscillating', '(apnea or apnoea) and (infant or preterm or neonate or newborn)', 'rocking', and 'vertical pulsating'.
The Oxford Database of Perinatal Trials was searched using search term 'apnea'.
Standard methods of the Cochrane Collaboration and its Neonatal Review Group were used. The methodological quality of each trial was reviewed by the second author blinded to trial authors and institution(s).
Each author extracted the data separately, then compared and resolved differences.
The standard method of the Neonatal Review Group to synthesise data was followed, using the fixed effects model to calculate relative risk (RR), risk difference (RD) and mean difference (MD) or weighted mean difference (WMD) where appropriate. As all three trials were crossover trials, the data were extracted from all exposure periods and combined where appropriate.
Ten trials were identified, seven of which failed to meet inclusion criteria (see table 'Characteristics of Excluded Studies'). Three trials were included (Korner 1978; Tuck 1982; Jirapaet 1993) that compared a form of kinesthetic stimulation to control for the treatment of apnea of prematurity. Details of each study have been entered into the table of included studies. All three studies enrolled preterm infants with established apnea with exclusion of secondary causes, and not currently on any treatment with a respiratory stimulant. In the study by Korner 1978, it was unclear as to how many apneas were required to be eligible for enrolment. In the study by Tuck 1982, infants were eligible if they had > 2 apneas in 24 hours. In the study by Jirapaet 1993, infants were eligible if they had > 3 apneas within six hours.
All three studies were crossover trials with alternating periods of treatment and control in the infants studied. Tuck 1982 is a two-period crossover trial, whereas Korner 1978 and Jirapaet 1993 were multiple crossover trials, each studying four consecutive alternating periods.
The type of kinesthetic stimulation differed between the three trials. Korner 1978 examined the effect of an irregularly oscillating water bed (12-14/min); Tuck 1982 examined the effect of a regularly rocking bed tray (10-22/min); and Jirapaet 1993 examined the effect of a 'vertical pulsating stimulus' (blood pressure cuff placed under the thorax inflated 12-20/min).
Apnea and bradycardia were recorded by nursing staff (monitors set for apnea > 20 seconds duration and bradycardia < 80 bpm) and by polygraphic recording of apnea (> 10 seconds) and bradycardia (< 80 bpm) in the study by Korner 1978. Apnea (> 12 seconds), bradycardia (< 100 bpm) and transcutaneous oxygen were recorded by polygraphic recording (with the researcher at the bedside) in the study by Tuck 1982. Apnea and bradycardia were recorded by nursing staff (monitors set for apnea > 15 seconds and bradycardia < 100 bpm), and by polygraphic recording (apnea > 15 seconds and bradycardia < 100 bpm) in the study by Jirapaet 1993.
Data could be extracted only from the study by Korner 1978 to calculate the number of apneas (> 20 seconds) in 24 hours.
Details of the methodological quality of each trial are given in the table of included studies.
All three trials were crossover trials, with Tuck 1982 using a two-period crossover, and Korner 1978 and Jirapaet 1993 using multiple crossover periods. In two trials there was random order of treatment (Tuck 1982; Jirapaet 1993). In the third trial (Korner 1978) the method of allocation was unclear. Tuck 1982 used a coin toss to determine order (letter from author), but the actual methods of randomization or alternation were unclear in the other two studies.
Due to the nature of the treatments, blinding of intervention was not possible in any of the studies.
All studies had complete follow-up of enrolled infants.
Polygraphic recordings, but not nursing documented apnea and bradycardia, were assessed blindly in the study by Jirapaet 1993. Blinding of the assessment of polygraphic recordings was unclear in the study by Tuck 1982. Nursing documented apnea and bradycardia was not blinded in the study by Korner 1978.
See 'summary of analyses'.
Clinically significant reductions in apnea: No study reported a clinically
significant reduction in apnea consistent with the prespecified criteria
in this review. For apnea > 20s, Korner 1978
using the oscillating water bed reported no significant difference in number
of infants with > 4 / day or > 9 / day. For apnea associated with bradycardia,
Korner 1978 reported no significant difference
in number of infants with > 4 / day or > 9 apneas / day. For apnea
associated with bradycardia, Korner 1978 reported no significant difference in the number of infants having a < 50% reduction in frequency of apnea. Tuck 1982
using the rocking bed reported frequency of apneas (> 12 seconds) associated
with bradycardia (< 100 bpm), number of apneas associated with hypoxia
(TcP02 < 50 mmHg), and number of apneas requiring stimulation in infants
on the rocking bed. However, the events for individual patients were combined
and individual patient data were not available from the author to determine
if there was an important reduction in clinical apnea.
Other reported apnea: Using a lower threshold (>25% reduction), Korner 1978 reported less apnea and bradycardia in infants on the oscillating water bed (RR 0.14, 95% CI 0.02, 0.91). For apnea associated with bradycardia, both Korner 1978 and Tuck 1982 found significant reductions in the number of infants with more frequent episodes. Tuck 1982 reported reductions in the frequency of apneas (> 12 seconds) associated with bradycardia (< 100 bpm) whereas Korner 1978 reported apnea > 15s associated with bradycardia <80bpm / day. Meta-analysis of these two studies found a significant reduction in the number of infants with more frequent episodes (RR 0.11, 95% CI 0.03, 0.42). No outcome could be extracted from the study using the 'vertical pulsating stimulus' by Jirapaet 1993 that was consistent with the definition of clinically important apnea. Jirapaet 1993 reported the number of episodes of apnea > 15s / day and found a significant reduction (MD -15.78, 95% CI -21.20, -10.36) in infants whilst receiving the 'vertical pulsating stimulus'. Korner 1978 reported the number of episodes of apnea > 15s associated with bradycardia <80bpm / day and found no significant difference. Tuck 1982 reported a significant reduction in apnea >12s associated with hypoxia (TcO2 < 50mmHg) (RR 0.12, 95% CI 0.02, 0.85), but no significant difference in the frequency of apnea requiring stimulation (RR 0.20, 0.03, 1.47).
Other events: Jirapaet 1993 reported that no infants required resuscitation or ventilation in either group. Adverse events such as death, intraventricular haemorrhage and neurodevelopmental follow-up were not reported.
There is no evidence from these studies that kinesthetic stimulation is effective in producing a clinically important reduction in clinically important apnea of prematurity (> 20 seconds duration or a shorter apnea associated with bradycardia or cyanosis), or in reducing clinically important outcomes such as need for IPPV, CPAP or respiratory stimulants.
However, using a lower threshold that has uncertain clinical significance, the trials indicate that kinesthetic stimulation reduces the rates of apnea (with or without associated bradycardia and hypoxia) in infants with apnea of prematurity. The results were consistent despite the use of three different methods of providing kinesthetic stimulation.
The limitations of the studies included in this review are their small sample size (the three eligible trials included only 49 babies in total) and the use of a crossover design which removes the ability to measure effects on late outcomes. Because of the small number of infants enrolled in the three trials included in this review, measures of effect are imprecise and a small benefit of kinesthetic stimulation in reducing clinically important apnea is not precluded.
This review is hindered by the clinical heterogeneity of the eligible trials, which limits the ability to synthesize their results. The three trials used different methods of kinesthetic stimulation, and different definitions for apnea and bradycardia. It has not been possible for the reviewers to obtain individual patient data or dichotomous data for some clinically important outcomes. These would be needed for synthesis of data on these outcomes.
This review suggests that kinesthetic stimulation is effective at reducing apnea and bradycardia in infants with apnea of prematurity. Whether it produces a clinically important reduction in clinical apnea of prematurity remains unproven. Previous reviews have suggested that kinesthetic stimulation is not effective at preventing apnea of prematurity (Henderson-Smart 2005) and is not as effective as theophylline in treating clinically significant apnea of prematurity (Osborn 2005).
There is insufficient evidence to recommend kinesthetic stimulation as treatment for clinically important apnea of prematurity.
A further trial of kinesthetic stimulation compared to control using clinically important outcomes and with adequate power is required before kinesthetic stimulation can be recommended to treat clinically important apnea with or without bradycardia.
Dr Tuck provided information about the methods of his study.
None.
| Study | Methods | Participants | Interventions | Outcomes | Notes | Allocation concealment |
| Jirapaet 1993 | Multiple crossover trial with randomized order of treatment. Single center. Blinding of randomization: not stated. Blinding of intervention: not possible. Complete follow up: yes. Blinding of outcome: yes. | Preterm infants with > 2 apneas in 24 hours (n = 29). 29-34 weeks gestation (mean 32.3 weeks). Birth weight 890-2300g (mean 1474g). Age day 2-13 (mean 4.4 days). Secondary causes of apnea excluded. No respiratory stimulant or ventilator support. | Multiple crossover trial with four equal, alternate 6 hour periods of "Vertical pulsating stimulus" ( inflating BP cuff placed under thorax) at 12-20 cycles per minute, and alternate period of no 'VPS'. | Polygraphic recorded apnea > 15 seconds and bradycardia < 100 bpm. Type of apnea (mixed, central and obstructive). Use of IPPV. Need for resuscitation. | B | |
| Korner 1978 | Multiple crossover trial of kinesthetic treatment. Single center trial. Method of allocation unclear: ? quasi-random with alternation of first exposure. Blinding of randomisation: not stated. Blinding of intervention: not possible. Complete follow up: yes. Blinding of outcome measurement: No for nursing observations of apnea/bradycardia. Not clear for polygraphic recording. | Preterm infants with apnea (n = 8). Gestation 27-32 weeks (mean 30 weeks). Birth weight 1072-1650g (mean 1270g). Age 7-28 days (mean day 15). No oxygen therapy, no medications and secondary causes of apnea excluded. | Multiple
crossover design with four equal, alternate 6 hour periods of irregularly
oscillating water bed 12-14 cycles per minute, and non-oscillating water bed. | Nursing recorded apnea/bradycadia (monitor alarms for apnea > 20 seconds and bradycardia < 80 bpm). Polygraphic recorded apnea > 10 seconds with bradycardia (moderate 80-120 bpm, severe < 80bpm). Duration and distribution of sleep and wake cycles. | Criteria for frequency and severity of apnea for enrolment not stated. | B |
| Tuck 1982 | Two-period crossover trial with randomized order of treatment. Single center study. Blinding of randomization: yes: 'tossing a coin'. Blinding of intervention: not possible. Complete follow up: yes. Blinding of outcome measurement: no, observer at bedside. | Preterm infants with > apneas in 24 hours (n = 12). Gestation 26-32 weeks (mean 29.5). Birth weight 800-1700g (mean 1210g). Age 2-45 days (mean 14 days). Four in oxygen, no theophylline, no secondary cause of apnea. | Single crossover with two consecutive, equal alternate 4-8 hour periods of regularly rocking bed (using bellows under the tray) at 10-22 cycles per minute, and non-rocking bed. | Polygraphic recorded apnea (> 12 seconds), bradycardia (< 100 bpm) and transcutaneous oxygen concentration. Need for stimulation. | B |
| Study | Reason for exclusion |
| Frank 1973 | Reported 11 methods of stimulating infants with apnea. Stimulus only applied during apnea. |
| Garcia 1993 | Randomised crossover trial. Ineligible comparison. Compared tactile (moderate shaking of the leg) to taste stimulation of infants with apnea. |
| Groswasser 1995 | Crossover study using a rocking mattress in infants at 39 weeks postmentrual age with obstructive sleep apnea. |
| Kattwinkel 1975 | Ineligible comparison of kinesthetic stimulation and CPAP in preterm infants with apnea. Kinesthetic stimulation given by rubbing the extremeties for 5 out every 15 minute period for total of 3 hours. Method of allocation to treatment not stated. |
| Korner 1981 | Study abandoned. Data incomplete. |
| Lovell 1999 | Randomised study of vibrotactile stimulation of preterm infants with apnea. Excess losses as only one of 7 infants randomised had apnea. |
| Svenningsen 1995 | Historical controls used. |
Jirapaet V, Subramanian KNS. The effect of vertical pulsating stimulus on apnea of prematurity. Journal of the Medical Association of Thailand 1993;76:319-26.
Korner 1978 {published data only}
Korner AF, Guilleminault C, Van den Hoed J, Baldwin RB. Reduction of sleep apnea and bradycardia in preterm infants on oscillating water beds: a controlled polygraphic study. Pediatrics 1978;61:528-33.
Tuck 1982 {published and unpublished data}
Tuck SJ, Monin P, Duvivier C, May T, Vert P. Effect of a rocking bed on apnoea of prematurity. Archives of Disease in Childhood 1982;57:475-7.
Frank UA, Bordiuk JM, Borromeo-McGrail V, Saltzman MB, Keitel HG. Treatment of apnea in neonates with an automated monitor-actuated apnea arrestor. Pediatrics 1973;51:878-83.
Garcia 1993 {published data only}
Garcia AP, White-Traut R. Preterm infants' responses to taste/smell and tactile stimulation during an apneic episode. Journal of Pediatric Nursing 1993;8:245-52.
Groswasser 1995 {published data only}
Groswasser J, Sottiaux M, Rebuffat E, Simon T, Vandeweyer M, Kelmanson I, Blum D, Kahn A. Reduction in obstructive breathing events during body rocking: a controlled polygraphic study in preterm and full-term infants. Pediatrics 1995;96:64-8.
Kattwinkel 1975 {published data only}
Kattwinkel J, Nearman HS, Fanaroff AA, Katona PG, Klaus MH. Apnea of prematurity. Comparative therapeutic effects of cutaneous stimulation and nasal continuous positive airway pressure. Journal of Pediatrics 1975;86:588-92.
Korner 1981 {published data only}
Korner AF. What we don't know about water beds and apneic preterm infants [letter]. Pediatrics 1981;68:306-7.
Lovell 1999 {published data only}
Lovell JR, Eisenfeld L, Rosow E, Adam J, Lapin C, Bronzino JD. Vibrotactile stimulation for treatment of neonatal apnea: a preliminary study. Connecticut Medicine 1999;63:323-5.
Lovell JR, Eisenfeld L, Rosow E, Adam J, Lapin C, Bronzino JD. Vibrotactile stimulation for treatment of neonatal apnea: a preliminary study. Neonatal Intensive Care 1999;12:36-7, 39-41.
Svenningsen 1995 {published data only}
Svenningsen NW, Wittstrom C, Hellstrom-Westas L. OSCILLO-oscillating air mattress in neonatal care of very preterm babies. Technology and Health Care 1995;3:43-6.
* indicates the primary reference for the study
Henderson-Smart DJ. Recurrent apnoea. In: Ed Yu VYH, editor(s). Bailliere's Clinical Paediatrics. Vol. 3 No. 1 Pulmonary Problems in the Perinatal Period and their Sequelae. London: Bailliere Tindall, 1995:203-22.
Henderson-Smart DJ, Osborn DA. Kinesthetic stimulation for preventing apnea in preterm infants. In: The Cochrane Database of Systematic Reviews, Issue 2, 2005.
Nelson NM. Members of the Task Force on rolonged Apnea of the American Academy of Pediatrics. Prolonged apnea. Pediatrics 1978;61:651-2.
Osborn DA, Henderson-Smart DJ. Kinesthetic stimulation vs theophylline for apnea in preterm infants. In: The Cochrane Database of Systematic Reviews, Issue 2, 2005.
Osborn DA, Henderson-Smart DJ. Kinesthetic stimulation for treating apnea in preterm infants. In: The Cochrane Database of Systematic Reviews, Issue 1, 1999.
Osborn DA, Henderson-Smart DJ. Kinesthetic stimulation for treating apnea in preterm infants. In: The Cochrane Database of Systematic Reviews, Issue 2, 2002.
| Comparison or outcome | Studies | Participants | Statistical method | Effect size |
|---|---|---|---|---|
| 01 Kinesthetic stimulation vs control | ||||
| 01 Apnea (>20s) > 4/day | 1 | 16 | RR (fixed), 95% CI | 1.00 [0.69, 1.45] |
| 02 Apnea (>20s) > 9/day | 1 | 16 | RR (fixed), 95% CI | 0.86 [0.53, 1.38] |
| 03 Apnea and bradycardia > 4/day | 1 | 16 | RR (fixed), 95% CI | 0.88 [0.67, 1.14] |
| 05 Apnea and bradycardia > 9/day | 1 | 16 | RR (fixed), 95% CI | 0.75 [0.50, 1.12] |
| 06 < 50 % reduction in frequency of apnea/bradycardia | 1 | 16 | RR (fixed), 95% CI | 0.75 [0.50, 1.12] |
| 07 < 25% reduction in frequency of apnea/bradycardia | 1 | 16 | RR (fixed), 95% CI | 0.14 [0.02, 0.91] |
| 08 Apnea / bradycardia more frequent | 2 | 40 | RR (fixed), 95% CI | 0.11 [0.03, 0.42] |
| 09 Episodes of central apnea > 15 seconds (number/day) | 1 | 58 | WMD (fixed), 95% CI | -15.78 [-21.20, -10.36] |
| 10 Episodes of apnea > 15s & bradycardia < 80 (number/day) | 1 | 16 | WMD (fixed), 95% CI | -9.42 [-20.77, 1.93] |
| 11 More frequent apnea with hypoxia (TcO2<50mmHg) | 1 | 24 | RR (fixed), 95% CI | 0.13 [0.02, 0.85] |
| 12 More frequent apnea requiring stimulation | 1 | 24 | RR (fixed), 95% CI | 0.20 [0.03, 1.47] |
| 13 Use of IPPV or CPAP | 0 | 0 | RR (fixed), 95% CI | No numeric data |
| 14 Need for resuscitation for apnea | 0 | 0 | RR (fixed), 95% CI | No numeric data |
| The review is published as a Cochrane review in The
Cochrane Library, Issue 2, 2005 (see http://www.thecochranelibrary.com for
information). Cochrane reviews are regularly updated as new evidence emerges
and in response to comments and criticisms, and The Cochrane Library should
be consulted for the most recent version of the Review. |