For many years the International Liaison Committee on Resuscitation has recommended the use of tactile stimulation for initial management of infants born with inadequate respiratory effort at birth without systematically examining its effectiveness.
Systematic review to compare the effectiveness of tactile stimulation with routine handling in newly born term and preterm infants.
Medline, Embase, Cochrane CENTRAL, along with clinical trial registries.
Randomized and non-randomized studies were included based on predetermined criteria.
Data were extracted independently by authors. Risk of Bias in Nonrandomized Studies of Interventions (ROBINS-I) was used to assess risk of bias in non-randomized studies. Grading of Recommendations, Assessment, Development and Evaluations (GRADE) was used to assess the certainty of evidence.
Among 2455 unique articles identified, 2 observational studies were eligible and qualitatively summarized. Because one of the studies was at critical risk of bias, only the other study including 243 preterm infants on continuous positive airway pressure with clinical indications for tactile stimulation was analyzed. It showed a reduction in tracheal intubation in infants receiving tactile stimulation compared with no tactile stimulation (12 of 164 vs 14 of 79, risk ratio of 0.41 [95% confidence interval 0.20 to 0.85]); however, the certainty of evidence was very low.
The available data were limited and only from observational studies.
A potential benefit of tactile stimulation was identified but was limited by the very low certainty of evidence. More research is suggested to evaluate the effectiveness as well as the optimal type and duration of tactile stimulation.
Globally, 2.4 million neonatal deaths occurred in 2019.1 About 25% of these deaths are due to intrapartum-related events.2 Failure to initiate breathing after birth is the primary reason for death in infants with birth asphyxia.3 One of the most common interventions to stimulate breathing at birth is tactile stimulation.
Few studies have tried to elucidate whether and how tactile stimulation stimulates breathing. Possible mechanisms are related to changing the arousal state or activating mechanoreceptors or proprioceptors that directly stimulate breathing and adduct the vocal cords via reflex arcs.4,5 Animal studies show that licking, rolling, and biting are strategies used by mothers to ensure regular breathing of their pups at birth.6–8
For decades, tactile stimulation has been suggested in the initial steps of stabilization of the newborn infant in the treatment recommendations from the International Liaison Committee on Resuscitation (ILCOR). In 1999, the statement was: “Appropriate stimulation includes drying the infant with a towel, flicking the bottoms of the feet, or gently rubbing the back. Such tactile stimulation may initiate spontaneous respirations in newly born infants who are experiencing primary apnea.”9 In 2006, 2010, 2015, and 2020, ILCOR neonatal resuscitation algorithms all contained initial steps in stabilization, which included “dry and provide warmth, position, assess the airway, stimulate to breathe.”10–13 North American, European, Australian-New Zealand, and Japanese guidelines on neonatal resuscitation include tactile stimulation to facilitate respiratory efforts at birth.14–17 These recommendations are largely based on many years of historical practice, experience and expert opinion. The World Health Organization recommends that newly born babies who do not breathe spontaneously after thorough drying should be stimulated by rubbing the back 2 to 3 times before clamping the cord and initiating positive pressure ventilation (PPV), but they grade this as a weak recommendation based on Guideline Development Group’s consensus in the absence of published evidence.18 A Delphi panel of 18 experts estimated that immediate newborn assessment and stimulation would reduce both intrapartum-related and preterm deaths by 10%. The authors conclude that the quality of evidence for stimulation at birth is low, partly because it is considered the standard of care.19
Since 2015, ILCOR has used the grading of recommendations assessment, development and evaluation (GRADE) process to appraise evidence.20 Because the effectiveness of tactile stimulation to facilitate breathing at birth has never been systematically evaluated by ILCOR, it was decided to assess the evidence for tactile stimulation during postnatal stabilization.
Methods
The systematic review protocol was registered on January 22, 2021 in the International Prospective Register of Systematic Reviews (CRD 42021227768).
Review Question
The question evaluated by the systematic review was: Among newborns at birth with inadequate respiratory effort, does tactile stimulation that occurs in addition to routine handling with measures to maintain temperature, compared with routine handling with measures to maintain temperature, change any of the following primary outcomes: establishment of spontaneous breathing without PPV, time to the first spontaneous breath or crying, and/or time to heart rate ≥100 beats per minute.
Population: the eligible infants were term or preterm newborn infants immediately after birth with absent, intermittent, or shallow respirations.
Intervention: Any tactile stimulation performed within 60 seconds after birth and defined as any of the following: rubbing the chest or sternum, rubbing the back, rubbing the soles of the feet, flicking the soles of the feet or a combination of these methods. This intervention should be done in addition to routine handling with measures to maintain temperature.
Control: Routine handling with measures to maintain temperature, defined as care taken soon after birth, including positioning, drying, and additional thermal care.
Outcomes: The prespecified 3 primary outcomes were the establishment of spontaneous breathing without positive pressure ventilation (yes or no), time to the first spontaneous breath or crying from birth, and time to heart rate ≥100 beats per minute from birth. The following secondary outcomes were studied: survival as reported by authors, neurodevelopmental outcomes, intraventricular hemorrhage (only in preterm infants <34 weeks), oxygen and/or respiratory support at admission to a neonatal special or ICU, and admission to a neonatal special or ICU for those not admitted by protocol based on gestational age and/or birth weight.
Study Design: Randomized controlled trials (RCTs) and nonrandomized studies (nonrandomized controlled trials, interrupted time series, controlled before-and-after studies, and cohort studies) were eligible for inclusion. Unpublished studies (conference abstracts, trial protocols) and animal studies were excluded.
Timeframe: All years and all languages were included provided there was an English abstract.
The exploration of subgroups was planned for gestational age (<34, 34–36, and ≥37 weeks), cord management (early, delayed cord clamping and cord milking), settings (high and low resource health facility), and method of stimulation (type, number and/or duration of stimuli), according to availability of data.
Literature Search
Search strings were developed by an information specialist (JR) for the following databases that were searched from their inception until December 6, 2020: Medline, Embase, and Cochrane Central Register of Controlled Trials (CENTRAL). The following trial registries were also searched: International Clinical Trials Registry Platform, and US Clinical Trials Registry. The search was rerun on June 3, 2021, and on September 17, 2021.
The search strategies used a comprehensive combination of subject headings and keywords for tactile stimulation, physical stimulation, rubbing, flicking, stimulation, touch, stabilization, neonatal resuscitation, neonatal transition, infant newborn, infant premature, neonate, newly born, delivery, and birth. No language limits were applied. The full electronic search strategy is shown in Supplemental Table 4. References from included studies and review articles were also hand searched, and ILCOR Neonatal Life Support Task Force was consulted.
Study Selection and Data Extraction
Titles and abstracts were independently screened by 2 authors (M.F.A. and E.F.). Agreement between both screeners was 98.4% (95% confidence interval [CI]: 97.8% to 98.9%). Conflicts were resolved by a third author (R.G.). These 3 authors then screened full text articles and the discrepancies were resolved by consensus. Four authors (R.G., M.F.A., E.F., and T.I.) independently extracted the data on study designs, study population, interventions, and outcomes.
Assessment of Risk of Bias and Data Analysis
Three authors (R.G., M.F.A., and E.F.), with an adjudication by the fourth author (T.I.), independently assessed the risk of bias of each included study using the risk of bias in nonrandomized studies of interventions (ROBINS-I) tool21 as low, moderate, serious, or critical according to the following domains: confounding, selection of participants into the study, classification of interventions, deviations from intended interventions, missing data, and selection of the reported result.
For studies without critical risk of bias in which authors compared either primary or secondary outcomes between infants receiving tactile stimulation and those not, a meta-analysis with random-effect Mantel-Haenszel methods was planned. If only a single study provided outcome data, the risk ratio (RR) and the 95% CIs were calculated with unconditional maximum likelihood estimation using R version 3.6.1 (R Foundation for Statistical Computing: Vienna, Austria).
The certainty of evidence was judged as high, moderate, low, or very low for each outcome based on the GRADE approach by judging the risk of bias, indirectness, imprecision, inconsistency, and other consideration (eg, publication bias).20 Because we used ROBINS-I, an initial assumption of high certainty of evidence was applied for observational studies as well as for RCTs according to the recommendation from the GRADE group.22
Narrative Review
Because few eligible studies were found, those reporting any outcome of tactile stimulation in newly born infants that did not comply with the inclusion criteria of this systematic review were summarized in a narrative review.
Results
The search found 2455 unique articles after removing 430 duplicate records (Fig 1). The title and abstract screening identified 31 articles for full-text screening. Eight studies reported any outcome of tactile stimulation in newly born infants.23–30 Among them, only 2 studies, both observational, met eligibility criteria.24,25 The characteristics of these and the 6 ineligible studies included for narrative review only are summarized in Table 1.
Characteristics of the 8 Studies That Reported Any Outcome of Tactile Stimulation in Newly Born Infants, According to Patients, Intervention, Control, and Outcomes of interest for This Systematic Review
| Study | Patients | Intervention | Control | Primary Outcome | Secondary Outcome | Eligibility for SR | ||
|---|---|---|---|---|---|---|---|---|
| Newborn at Birth With Inadequate Respiratory Effort | TS in Addition to Routine Handling With Measures to Maintain Temperature | Routine Handling With Measures to Maintain Temperature | Spontaneous Breath Without PPV | Time to First Breath or Crying | Time HR >100 Beats per Minute | O2 and/or Respiratory Support at Admission | ||
| Katheria 201623 | Yes | Yes | No | NA | NA | NA | NA | No |
| Dekker 201724 | Yes | Newborns on CPAP | Newborns on CPAP | NA | NA | NA | Tracheal intubation | Yes |
| Baik-Schneditz 201825 | No | Yes | No | NA | NA | NA | Respiratory support 15 min after birth | Yes |
| Dekker 201826 | Yes | Yes | No | NA | NA | NA | Respiratory function parameters after birth and oxygenation until NICU admission | No |
| Gaertner 201827 | Not clear: resuscitation anticipated | No | No | NA | NA | NA | NA | No |
| Pietravalle 201828 | Yes | Yes | No | Only for TS group | NA | NA | NA | No |
| Van Henten 201929 | No | No | No | NA | NA | NA | NA | No |
| Cavallin 202130 | Yes | Yes | No | Yes | Yes | NA | NA | No |
| Study | Patients | Intervention | Control | Primary Outcome | Secondary Outcome | Eligibility for SR | ||
|---|---|---|---|---|---|---|---|---|
| Newborn at Birth With Inadequate Respiratory Effort | TS in Addition to Routine Handling With Measures to Maintain Temperature | Routine Handling With Measures to Maintain Temperature | Spontaneous Breath Without PPV | Time to First Breath or Crying | Time HR >100 Beats per Minute | O2 and/or Respiratory Support at Admission | ||
| Katheria 201623 | Yes | Yes | No | NA | NA | NA | NA | No |
| Dekker 201724 | Yes | Newborns on CPAP | Newborns on CPAP | NA | NA | NA | Tracheal intubation | Yes |
| Baik-Schneditz 201825 | No | Yes | No | NA | NA | NA | Respiratory support 15 min after birth | Yes |
| Dekker 201826 | Yes | Yes | No | NA | NA | NA | Respiratory function parameters after birth and oxygenation until NICU admission | No |
| Gaertner 201827 | Not clear: resuscitation anticipated | No | No | NA | NA | NA | NA | No |
| Pietravalle 201828 | Yes | Yes | No | Only for TS group | NA | NA | NA | No |
| Van Henten 201929 | No | No | No | NA | NA | NA | NA | No |
| Cavallin 202130 | Yes | Yes | No | Yes | Yes | NA | NA | No |
CPAP, continuous positive airflow pressure; HR, heart rate; NA, not available; PPV, positive pressure ventilation; SR, systematic review; TS, tactile stimulation.
The 2 eligible studies were both single-center observational studies that analyzed video recorded resuscitation of preterm infants. The first was conducted in the Netherlands by Dekker et al.24 It was a retrospective study and initially included 245 infants with gestational age <32 weeks and compared those who did and did not receive tactile stimulation in the first 7 minutes after birth. Tactile stimulation was defined as observable rubbing and/or flicking of the back or feet that occurred after drying, placing the infant in a plastic bag, covering the head with a cap, placing the pulse oximeter probe, and applying the continuous positive airway pressure (CPAP) interface. If tactile stimulation was performed, inferences were made about indication if there was evidence of apnea, irregular breathing, bradycardia, or hypoxia. Among the 164 infants who received tactile stimulation, each received a median of 3 stimuli (interquartile range [IQR]:1–5), with a median duration of each stimulus of 8 seconds (IQR: 4–16). The median time after birth at the first episode of tactile stimulation was 114 seconds (IQR: 73–182). Tracheal intubation in the delivery room after birth was performed in 12 of 164 (7%) of the infants who received tactile stimulation and in 14 of 79 (18%) of those who did not. Data on tracheal intubation could not be retrieved in 2 nonstimulated infants (authors’ information).
The other eligible study was prospective and was performed in an Austrian center.25 It also used video-recordings of resuscitation of both preterm and term infants born by cesarean section and observed whether infants received tactile stimulation during the first 15 minutes after birth. Tactile stimulation was applied at 1 or more of 3 sites: the sternum, feet or back, and was provided after initial drying and application of monitoring devices. The 43 infants who received tactile stimulation had a median of 1 stimulus (range: 1–7), with a median duration of 15 seconds (range: 5–63). Respiratory support in the first 15 minutes after birth was provided in 24 of 43 (56%) stimulated neonates (preterm: 14 of 18 [78%]; term: 10 of 25 [40%]) and in 31 of 57 (54%) nonstimulated infants (preterm: 29 of 33 [88%]; term: 2 of 24 [8%]).
Risk of Bias of Eligible Studies
The evaluation of risk of bias for both studies is shown in Fig 2.31 The study of Baik-Schneditz et al25 was considered at critical risk of bias, mainly due to confounding by indication (indication bias). We made this judgement because, in term infants in the study, receiving stimulation was associated with a much higher rate of subsequently receiving respiratory support, suggesting that depressed or apneic infants were more likely to receive stimulation than those with good respiratory effort. Dekker et al24 attempted to infer the reasons why stimulation was provided and noted that the rate of intubation was higher in the nonstimulated group regardless of presence of an indication. Hence, we considered this study to be less susceptible to indication bias. However, there was no adjustment for other potential confounders (gestational age, delivery mode, respiratory condition at birth, etc), so the study was still considered at high risk of bias.
Assessment of Risk of Bias by ROBINS-I tool for both studies included in the systematic review.
Assessment of Risk of Bias by ROBINS-I tool for both studies included in the systematic review.
Data Analysis and GRADE Assessment
The study by Baik-Schneditz was not eligible for data analysis due to its critical risk of bias according to ROBINS-I.21 Therefore, only the study by Dekker et al24 was analyzed and evaluated with GRADE methods. Infants who received tactile stimulation had lower rates of tracheal intubation than those who did not receive the stimulation: 12 of 164 (7%) vs 14 of 79 (18%), with estimated RR of 0.41 (95% CI 0.20 to 0.85); very low certainty of evidence. The certainty of evidence was downgraded for very serious risk of bias and indirectness, and serious imprecision, while upgraded for the strong association (Table 2).
GRADE Evidence Profile Table for the Study Included in the Systematic Review (Dekker et al 2017)
| Certainty Assessment | No. Patients | Effect | Certainty | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Outcome (Studies) | Risk of Bias | Inconsistency | Indirectness | Imprecision | Others | Intervention | Comparison | Relative (95% CI) | Absolute (95% CI) | |
| Intubation (1 observational study) | Very seriousa | Not serious | Seriousb | Seriousc | Strong association | 12 of 164 (7.3%) | 14 of 79 (17.7%) | RR 0.41 (0.20 to 0.85) | 105 fewer per 1000 (from 142 fewer to 27 fewer) | ⨁○○○ Very low |
| Certainty Assessment | No. Patients | Effect | Certainty | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Outcome (Studies) | Risk of Bias | Inconsistency | Indirectness | Imprecision | Others | Intervention | Comparison | Relative (95% CI) | Absolute (95% CI) | |
| Intubation (1 observational study) | Very seriousa | Not serious | Seriousb | Seriousc | Strong association | 12 of 164 (7.3%) | 14 of 79 (17.7%) | RR 0.41 (0.20 to 0.85) | 105 fewer per 1000 (from 142 fewer to 27 fewer) | ⨁○○○ Very low |
Dekker J, Martherus T, Cramer SJE, van Zanten HA, Hooper SB, Te Pas AB. Tactile stimulation to stimulate spontaneous breathing during stabilization of preterm infants at birth: A retrospective analysis. Front Pediatr. 2017;5:61. CI, confidence interval; RR, risk ratio.
Tactile stimulation (TS) should be visible on videos (selection bias), reporting is not complete for non TS infants and not controlled for possible confounders.
All the infants in the study were on CPAP before receiving tactile stimulation.
The sample size did not reach optimal information size.
Narrative Review
Among the 6 articles that reported data on the impact of tactile stimulation but did not meet the eligibility criteria of this systematic review, 4 studies were from high resource settings23,26,27,29 and the other 2 were from low resource settings.28,30 The summary of the study designs and results is shown in Table 3.
Summary of Design and Results of Studies Reviewed but Not Included in Systematic Review.
| Author | Infants Included | Study Characteristics | Results |
|---|---|---|---|
| Katheria 201623 | Preterm infants with 23–31 wk, n = 150 | Single center RCT in USA comparing delayed cord clamping with (V-DCC) and without ventilation (DCC-only). Video recordings were evaluated. Among the 150 infants, 45 received TS for apnea, defined as rubbing the back after drying. | Among the 125 infants born by cesarean section, the DCC-only infants, compared with the V-DCC group received TS more frequently (50 vs 22%; P = .001) and longer (mean: 41 vs 20 s; P = .002). However, there was no difference in the time to first breath (mean: 25 vs 27 s; P = .627). For the 25 infants born vaginally, there were no significant differences between the groups. |
| Dekker 201826 | Preterm infants with 27–32 wk, n = 44 | Single center RCT in Netherlands comparing infants randomized to receive repetitive TS (n = 21; 10 s of stimulation alternating with 10 s at rest; median stimulation episodes = 8 [IQR 7–10]) vs standard TS (n = 23; median episodes = 3 [IQR 36]). Video recordings for at least 4 min after birth. TS was defined as rubbing the back or the soles of the feet | At the start of transport to the NICU, those in the repetitive stimulation group had higher oxygen saturation levels (87% ± 3% vs 82% ± 9%, P = .007) and lower oxygen requirements (0.28 [IQR 0.23–0.35] vs 0.34 [IQR 0.29–0.44], P = .036). There were no significant differences between groups in median min vol at 1–4 min, need for PPV or surfactant administration, and maximum oxygen requirement during resuscitation. |
| Gaertner 201827 | Preterm and term infants with resuscitation anticipated: <30 wk: n = 60; ≥30 wk: n = 60 | Single center retrospective observational study in Australia. Video recordings for the first 5 min after birth were reviewed to examine the use of TS defined as drying, chest or back rubs, and flicking of feet, after cord clamping | TS was performed in 75 of 120 (63%) infants: <30 wk: 21 of 60 (35%) and ≥30 wk: 54 of 60 (90%). First stimulation was provided at a median of 19 s (IQR 15–24) with no difference between the two groups. The most common response to TS was limb movement followed by infant cry and facial grimace. Truncal stimulation was associated with more crying and movement than foot flicking |
| Pietravalle 201828 | Term infants that needed resuscitation (apnea, hypotonia or both), n = 150 | Single center retrospective observational study in Mozambique. Video recordings until the end of resuscitation were reviewed to examine the use of TS, defined as rubbing of back, chest or abdomen, or flicking of feet, after drying | TS was performed in 102 of 150 (68%) infants. First stimulation was provided at a median of 134 s (IQR 53–251), with only 28% receiving TS in the first minute after birth. Only 9 (9%) infants who received TS responded with spontaneous breathing without need for PPV |
| Van Henten 201929 | Preterm infants <37 wk (median 34 wk), n = 40 | Single center prospective observational study in the Netherlands. Video recordings for the first 10 min after birth were examined to evaluate the TS characteristics, defined as drying, rubbing the sternum or back and/or flicking the feet | TS was performed in 36 of 40 (90%) infants; 19 of 40 (48%) were stimulated more than once. Drying was the initial form of TS in 28 infants. First stimulation was provided at a median of 15 s (IQR 10–40). No association was found between TS and the time to first breath |
| Cavallin 202130 | Preterm and term infants with BW >1500 g that did not cry at birth, n = 186 | Single center RCT in Uganda comparing back rubbing versus foot flicking to provide tactile stimulation. Success of stimulation was defined as achievement of effective crying preventing the need of PPV. Video recordings at birth and during the transition period. | Success of TS occurred in 68 of 93 (82%) in the back rubbing group and in 68 of 93 (73%) in the foot flicking group (adjusted RR 1.19; 95% CI 0.86 to 1.66). Time to first cry was not different between the two groups (adjusted mean difference −7 s, 95% CI −35 to 21). No procedure-associated complications arose during the study. |
| Author | Infants Included | Study Characteristics | Results |
|---|---|---|---|
| Katheria 201623 | Preterm infants with 23–31 wk, n = 150 | Single center RCT in USA comparing delayed cord clamping with (V-DCC) and without ventilation (DCC-only). Video recordings were evaluated. Among the 150 infants, 45 received TS for apnea, defined as rubbing the back after drying. | Among the 125 infants born by cesarean section, the DCC-only infants, compared with the V-DCC group received TS more frequently (50 vs 22%; P = .001) and longer (mean: 41 vs 20 s; P = .002). However, there was no difference in the time to first breath (mean: 25 vs 27 s; P = .627). For the 25 infants born vaginally, there were no significant differences between the groups. |
| Dekker 201826 | Preterm infants with 27–32 wk, n = 44 | Single center RCT in Netherlands comparing infants randomized to receive repetitive TS (n = 21; 10 s of stimulation alternating with 10 s at rest; median stimulation episodes = 8 [IQR 7–10]) vs standard TS (n = 23; median episodes = 3 [IQR 36]). Video recordings for at least 4 min after birth. TS was defined as rubbing the back or the soles of the feet | At the start of transport to the NICU, those in the repetitive stimulation group had higher oxygen saturation levels (87% ± 3% vs 82% ± 9%, P = .007) and lower oxygen requirements (0.28 [IQR 0.23–0.35] vs 0.34 [IQR 0.29–0.44], P = .036). There were no significant differences between groups in median min vol at 1–4 min, need for PPV or surfactant administration, and maximum oxygen requirement during resuscitation. |
| Gaertner 201827 | Preterm and term infants with resuscitation anticipated: <30 wk: n = 60; ≥30 wk: n = 60 | Single center retrospective observational study in Australia. Video recordings for the first 5 min after birth were reviewed to examine the use of TS defined as drying, chest or back rubs, and flicking of feet, after cord clamping | TS was performed in 75 of 120 (63%) infants: <30 wk: 21 of 60 (35%) and ≥30 wk: 54 of 60 (90%). First stimulation was provided at a median of 19 s (IQR 15–24) with no difference between the two groups. The most common response to TS was limb movement followed by infant cry and facial grimace. Truncal stimulation was associated with more crying and movement than foot flicking |
| Pietravalle 201828 | Term infants that needed resuscitation (apnea, hypotonia or both), n = 150 | Single center retrospective observational study in Mozambique. Video recordings until the end of resuscitation were reviewed to examine the use of TS, defined as rubbing of back, chest or abdomen, or flicking of feet, after drying | TS was performed in 102 of 150 (68%) infants. First stimulation was provided at a median of 134 s (IQR 53–251), with only 28% receiving TS in the first minute after birth. Only 9 (9%) infants who received TS responded with spontaneous breathing without need for PPV |
| Van Henten 201929 | Preterm infants <37 wk (median 34 wk), n = 40 | Single center prospective observational study in the Netherlands. Video recordings for the first 10 min after birth were examined to evaluate the TS characteristics, defined as drying, rubbing the sternum or back and/or flicking the feet | TS was performed in 36 of 40 (90%) infants; 19 of 40 (48%) were stimulated more than once. Drying was the initial form of TS in 28 infants. First stimulation was provided at a median of 15 s (IQR 10–40). No association was found between TS and the time to first breath |
| Cavallin 202130 | Preterm and term infants with BW >1500 g that did not cry at birth, n = 186 | Single center RCT in Uganda comparing back rubbing versus foot flicking to provide tactile stimulation. Success of stimulation was defined as achievement of effective crying preventing the need of PPV. Video recordings at birth and during the transition period. | Success of TS occurred in 68 of 93 (82%) in the back rubbing group and in 68 of 93 (73%) in the foot flicking group (adjusted RR 1.19; 95% CI 0.86 to 1.66). Time to first cry was not different between the two groups (adjusted mean difference −7 s, 95% CI −35 to 21). No procedure-associated complications arose during the study. |
BW, birth weight; DCC, delayed cord-clamping; IQR, Interquartile range; PPV, positive pressure ventilation; RCT, randomized controlled trail; TS, tactile stimulation.
Discussion
Two observational studies were eligible for this systematic review, but 1 had a critical risk of bias.25 Therefore, only the study of Dekker et al24 was quantitatively analyzed and assessed by the GRADE method. This study showed a reduction in tracheal intubation in infants receiving tactile stimulation compared with no tactile stimulation (12 of 164 vs 14 of 79, RR of 0.41 [95% CI 0.20 to 0.85]); however, the evidence was of very low certainty.
This systematic review was designed to examine the impact of tactile stimulation in term and preterm newborn infants immediately after birth with absent, intermittent, or shallow respirations. In the 6 studies included in the systematic and narrative reviews, only 4 reported data from the infants with insufficient breathing just after birth,23,24,28,30 but 3 of them could not be included in the systematic review because they lacked a comparison group.23,28,30 The study of Dekker et al24 included in the systematic review analyzed apneic, hypoxic and/or bradycardic preterm infants. The caveat is that all newly born infants in this study were put on CPAP before tactile stimulation, in contrast to the common practice of tactile stimulation before CPAP or PPV.14–18 This may limit the generalizability of the study results. In this context, neither the systematic nor the narrative review found consistent data regarding the population or outcomes of interest.
Clinical evaluation of physical stimulation at birth became feasible recently, with delivery room video recordings that allowed the evaluation of neonatal practices as part of qualitative improvement initiatives and/or clinical research.32 In this context, most studies discussed in this review retrospectively analyzed the use of tactile stimulation in video recordings of neonatal resuscitation, but they were not designed to evaluate its effectiveness and used different techniques of stimulation.23–25,27–29 The types of stimulation reported were drying, rubbing and/or flicking parts of the infant’s body, namely the back, chest, sternum, abdomen, and/or feet. Across the studies, median time of the first tactile stimulation after birth (15–134 seconds), median number of tactile stimulation episodes (1–4 per infant), and median duration of each tactile stimulation (8–20 seconds) were variable. This inconsistency posed difficulties in determining its effects. Only 1 study evaluated standardized tactile stimulation: back rubbing or foot flicking repeated 2 to 3 times for a maximum of 3 to 5 seconds, but the efficacy of procedures was not compared with a control group without stimulation.30 In relation to video recordings, some studies require investigators in the room,25,30 which may cause selection and performance biases, while others use motion activated, fixed-mount cameras.23,24,26–29
Tactile stimulation is recommended as part of a bundle of procedures to stimulate infants undergoing transition at birth to reduce the need for interventions, such as PPV, tracheal intubation, and CPAP. This is particularly pertinent in low resource settings that may lack the necessary trained health care providers and equipment to provide respiratory care. In a cohort study of 5889 livebirths in Tanzania, research assistants monitored birth and the response of birth attendants to a depressed baby. Among the livebirths, resuscitation was attempted in 920, and 50.1% (461) of them commenced spontaneous respirations after tactile stimulation and/or upper airway suctioning, without receiving face mask ventilation.3 Msemo et al, also in Tanzania, reported on the implementation of the Helping Babies Breathe (HBB) program in 8 hospitals in 2009, which was associated with a sustained 47% reduction in early neonatal mortality within 24 hours and a 24% reduction in fresh stillbirths after 2 years. According to the authors, the HBB approach is simple, emphasizing immediate drying and stimulation, an intervention that can be readily implemented at any delivery, whether in health care facilities or not.33 In the same study, it is interesting to note that infants in primary apnea probably do respond to the immediate bundle of interventions after birth which decreases the need for facial mask ventilation, while those in secondary apnea do not respond.33 KC et al observed 22 752 births in Nepal, Bangladesh, and Tanzania. After the initial action of drying, 5330 (23.4%) did not cry. Tactile stimulation was provided to 38.6% of these infants and suction to 35.8%. Despite these interventions, 3860 (72.4%) did not establish respirations within 1 minute and received further stimulation and suction. Most of those who did not respond to stimulation did receive bag and mask ventilation, but only 1% within the recommended 1 minute after birth.34 This study highlights a possible adverse effect of tactile stimulation, as also suggested by Pietravalle et al, in Mozambique.28 Over-emphasis on stimulation could lead to it being used for an extended duration, thereby delaying important decisions to initiate ventilation and impairing outcomes. Furthermore, the ascertainment of the true effect of tactile stimulation in all these studies was difficult because of the number and variable timing of cointerventions.
The major reason for exclusion of many studies in this systematic review was the lack of a control group. This applied to all studies in the narrative review. Designing a clinical trial with a control group for whom tactile stimulation was withheld may be difficult or impossible because tactile stimulation has been widely recommended, and it is regarded as noninvasive, safe, and effective. Dekker et al reported that “colleagues of the neonatal team are very reluctant to not stimulate infants as tactile stimulation is one of the most basic interventions during neonatal resuscitation. It is difficult to achieve clinical equipoise for a study in which tactile stimulation is compared to the omission of stimulation.”26 Given the lack of sufficient evidence for effectiveness of tactile stimulation found in this review, and the potential for harm from over-zealous or prolonged tactile stimulation, studies with a ‘no tactile stimulation’ control group would be needed and could be justified.
Regarding outcomes, this systematic review was designed to evaluate rates of spontaneous breathing without PPV, time to first breath or crying, time to achieve heart rate >100 beats per minute, and several neonatal morbidities and long-term outcomes. Unfortunately, except for spontaneous breathing without PPV reported by Pietravalle et al and Cavallin et al and time to first cry reported by Cavallin et al only in infants that received tactile stimulation,28,30 none of these outcomes were described in the studies included in this review. The closest approximations to these outcomes were tracheal intubation in one study and receipt of respiratory support 15 minutes after birth in the other.24,25 The outcome measures for future studies need careful consideration to ensure that they are patient focused, clinically important, and sufficiently similar between studies to allow meta-analysis. The outcome measures selected for the systematic review may be a good starting point but may require use of video recording of neonatal resuscitation and preplanning to achieve early placement of monitoring equipment. Possible adverse effects should also be consistently reported, as a case report of soft tissue trauma, with bruises and scratches on the infant’s back, has been described during/after tactile stimulation.35 The potential for tactile stimulation to cause critical delays in initiation of other resuscitation interventions may also need formal assessment.
Strengths of our review include rigorous literature searching and systematic review by using a preregistered protocol, assessment of certainty of evidence by using GRADE,20 and input from experts from the ILCOR Neonatal Life Support Task Force. The additional narrative review provided a wider view on the current research evidence on tactile stimulation and contributed to identification of knowledge gaps and research needs.
Limitations include the absence of fully eligible RCTs or observational studies. In the included studies, small sample size, no adjustments for potential confounders, and the variation in reported outcomes limited the potential to synthesize evidence. Also, it should be noted that the single study included in the systematic review was done in a high-income country within a center that has highly trained neonatologists and, therefore, the result may not be applicable in different settings.
Conclusions
This review showed that, at present, there is a wide heterogeneity in the methods of tactile stimulation, such as time of initiation after birth, duration of each stimulus, total duration of the procedure, type and number of applied stimuli, and body region stimulated. The very limited available data suggest a possible benefit to tactile stimulation in decreasing the need for tracheal intubation in preterm infants, but the certainty of evidence was very low. Some authors raised the possibility of harm in the form of soft tissue trauma or that provision of tactile stimulation could delay the initiation of assisted ventilation, which could have adverse consequences for response to resuscitation and other important outcomes. More research is suggested to evaluate the effectiveness, optimal type and duration, and benefits of tactile stimulation. This review did not find evidence to change the current practice of providing tactile stimulation before the initiation of PPV to infants with absent, intermittent, or shallow respirations immediately after birth. In accordance with contemporary guidelines,13 the provision of tactile stimulation should not cause delays in the commencement of assisted ventilation within a minute from birth if spontaneous breathing is not established.
Acknowledgments
We thank Jenny Ring (Information Specialist, Melbourne) for her valuable support. The following additional International Liaison Committee on Resuscitation Neonatal Life Support Task Force members provided input on the review protocol, the interpretation of the results, and the manuscript as experts in neonatal resuscitation: Walid El-Naggar, MD, Dalhousie University (Halifax, Nova Scotia, Canada); Jorge Fabres, MD, Universidad Catolica de Chile (Santiago, Chile); Joe Fawke, MD, Leicester Royal Infirmary (Leicester, UK); Elizabeth E. Foglia, MD, University of Pennsylvania (Philadelphia, Pennsylvania); Vishal S. Kapadia,MD, University of Texas Southwestern Medical Center (Dallas, Texas); Mandira D. Kawakami, MD, Federal University of Sao Paulo (Sao Paulo, SP, Brazil); Han-Suk Kim, MD, Seoul National University College of Medicine (Seoul, Korea); Henry C. Lee, MD, Stanford University (Stanford, California); Chris McKinlay, MD, University of Auckland (Auckland, New Zealand); Yacov Rabi, MD, University of Calgary (Calgary, Alberta, Canada); Charles Roehr, MD, John Radcliffe Hospital, Oxford University Hospitals (United Kingdom); Georg M Schmölzer, MD, University of Alberta (Edmonton, Canada); Takahiro Sugiura, MD, Toyohashi Municipal Hospital (Toyohashi, Aichi, Japan); Daniele Trevisanuto, MD, University of Padova (Padova, Italy); Gary M Weiner, MD, University of Michigan (Ann Arbor, Michigan).
FUNDING: No external funding.
A complete list of group members appears in the acknowledgments.
Drs Guinsburg, de Almeida, Finan and Isayama prepared the protocol, screened studies, abstracted data, completed risk-of-bias and GRADE evaluations, completed the analysis, and prepared the manuscript; Drs Perlman, Wyllie, Liley and Wyckoff reviewed the protocol and the analysis, and edited the manuscript; and all authors critically reviewed and revised the manuscript for important intellectual content, approved the final manuscript as submitted, and agree to be accountable for all aspects of the work.
- CI
confidence interval
- CPAP
continuous positive airway pressure
- GRADE
grading of recommendations, assessment, development and evaluations
- HBB
Helping Babies Breathe
- ILCOR
International Liaison Committee on Resuscitation
- IQR
interquartile range
- PPV
positive pressure ventilation
- RCT
randomized controlled trial
- ROBINS-I
risk of bias in nonrandomized studies of interventions
- RR
risk ratio
Comments