1. Objectives

To evaluate whether there is evidence to support the practice of kangaroo mother care (KMC) in low birthweight (LBW) infants as an alternative to conventional neonatal care, prior to or after initial stabilization with conventional care, and to assess the beneficial and adverse effects of KMC

2. How studies were identified

The following databases were searched to June 2016:

• CENTRAL (The Cochrane Library 2016, Issue 6)
• MEDLINE
• EMBASE
• LILACS
• CINAHL
• POPLINE
• ISRCTN Registry
• ClinicalTrials.gov
• WHO International Clinical Trials Registry Platform
• Google Scholar

Additionally, websites of the Kangaroo Foundation and International Kangaroo Care were searched; as were reference lists of identified studies, review articles, and textbooks; and relevant conference and symposium proceedings. Researchers in the field were also contacted and handsearching of relevant journals was performed

3. Criteria for including studies in the review

3.1 Study type

Randomized controlled trials, including cluster randomized trials

3.2 Study participants

LBW infants (defined as birth weight <2500 g), irrespective of their gestational age

3.3 Interventions

Two types of intervention were included: i) KMC compared to conventional neonatal care in LBW infants, irrespective of factors such as duration of the intervention and infant stabilization status; and ii) early onset KMC (within 24 hours of birth) compared with late onset KMC (commencing 24 hours after birth)

(The intervention, KMC, included skin-to-skin contact and encouraged breastfeeding, while standard neonatal care generally involved the infant staying in an incubator or radiant warmer, or both. Intermittent KMC was a combination of skin-to-skin contact and radiant warmer/incubator)

3.4 Primary outcomes

Mortality

• At discharge or at 40 to 41 weeks’ postmenstrual age
• At 6 months of age or at six months follow-up
• At 12 months’ corrected age
• At latest follow-up

Severe infection/sepsis
Severe illness
Infant growth

• Weight at discharge or at 40 to 41 weeks’ postmenstrual age, and six and 12 months’ corrected age
• Weight gain at latest follow-up
• Length at discharge or at 40 to 41 weeks’ postmenstrual age, and at six and 12 months’ corrected age
• Length gain at latest follow-up
• Head circumference at discharge or at 40 to 41 weeks’ postmenstrual age, and at six and 12 months’ corrected age
• Head circumference gain at latest follow-up

Neurodevelopmental and neurosensory impairment

• Psychomotor development
• Cerebral palsy
• Deafness
• Visual impairment

Secondary outcomes included nosocomial infection/sepsis, mild/moderate infection or illness, lower respiratory tract disease, diarrhoea, hypothermia, hyperthermia, length of hospital stay, readmission to hospital, breastfeeding, mother-infant attachment, mother-infant interaction, parental and familial satisfaction, home environment and father involvement, and costs of care

4. Main results

4.1 Included studies

Twenty-one randomized controlled trials, enrolling 3042 infants, were included in this review

• Nineteen studies assessed KMC in LBW infants after stabilization; one assessed KMC in LBW infants before stabilization; and one assessed early onset KMC with late onset KMC in comparatively stable LBW infants
• Five studies included infants from multiple pregnancies; six studies included infants with a birthweight of 1500 g or less
• Eleven studies enrolled infants with a median or mean age of ≤10 days; six studies enrolled infants with a median or mean age of 11 to 20 days; and three studies enrolled infants with a median or mean age of 20 to 32 days
• Among studies assessing KMC in stabilized LBW infants, 16 used intermittent KMC whilst three used continuous KMC. Continuous KMC was used in studies assessing KMC in LBW infants prior to stabilization and in those studies comparing early onset KMC with late onset KMC
• In six studies, the mean or median duration of KMC per day was under two hours; four to seven hours in three studies, eight to 17 hours in seven studies, and ≥20 hours in three studies

4.2 Study settings

• Australia, Colombia, Ecuador, Ethiopia, India (8 trials), Indonesia, Madagascar, Malaysia, Nepal, the United Kingdom of Great Britain and Northern Ireland, the United States of America (3 trials), and one multicentre trial (Ethiopia, Indonesia and Mexico)
• Eleven studies were conducted in neonatal intensive care units (NICU) of tertiary care hospitals, public hospitals, maternity hospitals, or university hospitals; four in neonatal units of university hospitals; two in ”kangaroo wards“ (KMC infants) and neonatal intensive/intermediate care units of tertiary care hospitals (controls); two in newborn nurseries; one in a maternity ward; and one in both a hospital and home setting
• In nineteen studies infants were cared for by doctors and nurses; in one study they were cared for by their mothers whilst being supervised by a trained nurse; and in one study infants were cared for by an experienced nurse

4.3 Study settings

How the data were analysed
Two comparisons were made: i) KMC versus conventional neonatal care in LBW infants; and ii) early versus late onset KMC in relatively stable LBW infants. Data were combined in fixed effect meta-analyses to produce risk ratios (RR) for dichotomous data and mean differences (MD) for continuous data, with corresponding 95% confidence intervals (CI). If substantial heterogeneity was detected (I²≥50%), random effects models were employed. Potential sources of heterogeneity were explored in the following subgroup analyses:

• By type of KMC (intermittent, continuous)
• By daily duration of KMC (<2 hours, 8 to 15 hours, ≥20 hours)
• By infant age at initiation of KMC (≤10 days, >10 days)
• By setting in which the trial was conducted (low- and middle-income countries, high-income countries)
• By infant stabilization status at trial entry (before stabilization, after stabilization)

Sensitivity analyses were also performed in which trials at high risk of bias were excluded

Results
Kangaroo mother care versus conventional neonatal care (20 trials/2969 infants)
Mortality
In pooled analysis, KMC statistically significantly reduced the risk of mortality at discharge or at 40 to 41 weeks’ postmenstrual age by 40% compared with conventional care (RR 0.60, 95% CI [0.39 to 0.92], p=0.021; I²=0%; 8 studies/1736 infants). A statistically significant 33% reduction in the risk of mortality was also observed among KMC infants at latest follow-up (RR 0.67, 95% CI [0.48 to 0.95], p=0.026; I²=0%; 12 studies/2293 infants). Both of these results became non-significant for the following subgroups: intermittent KMC; duration of KMC less than two hours per day and duration of KMC four to 15 hours per day; high-income country setting; infants aged ≥10 days at initiation of the intervention; and infants entered into the trial after stabilization. Mortality at six months of age or at six months follow-up, and at 12 months’ corrected age was not statistically significantly different between treatment and control groups (RR 0.99, 95% CI [0.48 to 2.02], I²=0%; 2 trials/354 infants; and RR 0.57, 95% CI [0.27 to 1.17], 1 trial/693 infants; respectively).

Infection/illness
Among stabilized LBW infants, the KMC intervention significantly reduced the risk of: severe infection/sepsis at latest follow-up (RR 0.50, 95% CI [0.36 to 0.69], p=0.000029; I²=24%; 8 studies/1463 infants), severe illness at six months follow-up (RR 0.30, 95% CI [0.14 to 0.67], p=0.0032; 1 study/283 infants), nosocomial infection/sepsis at discharge or at 40 to 41 weeks’ postmenstrual age (RR 0.35, 95% CI [0.22 to 0.54], p<0.00001; I²=0%; 4 studies/1237 infants), lower respiratory tract disease at six months follow-up (RR 0.37, 95% CI [0.15 to 0.89], p=0.027; 1 study/283 infants), and hypothermia at discharge or 40 to 41 weeks’ postmenstrual age (RR 0.28, 95% CI [0.16 to 0.49], p<0.00001; I²=52%; 9 studies/989 infants). For the outcomes diarrhoea and mild/moderate infection or illness, no statistically significant differences between the treatment groups were observed. The reduction in risk of severe infection/sepsis at latest follow-up was not statistically significant for the continuous KMC subgroup. Length of hospital stay was not significantly reduced with KMC (MD -1.6 days, 95% CI [-3.4 to 0.2], 11 studies/1057 infants), and nor was the risk of readmission to hospital (RR 0.60, 95% CI [0.34 to 1.06], 2 trials/946 infants).

Infant growth
Infants in the KMC group were found to gain 4.08 g/day more in weight (95% CI [2.3 to 5.9 g/day], p<0.00001; I²=86%; 11 studies/1198 infants), to gain 0.21 cm/week more in length (95% CI [0.03 to 0.38 cm/week], p=0.022; I²=89%; 3 studies/378 infants), and to gain 0.14 cm/week more in head circumference (95% CI [0.06 to 0.22 cm/week], p=0.00033; I²=73%; 4 studies/495 infants) than those in the conventional neonatal care group at latest follow-up. Furthermore, one study of 592 infants found that those treated with KMC had a greater head circumference at six months’ corrected age compared to the conventional neonatal care group (MD 0.34 cm, 95% CI [0.11 to 0.57], p=0.0031). No significant differences were observed between groups for other growth outcomes.

Neurodevelopmental and neurosensory impairment
In one trial including 588 infants, no significant differences at 12 months’ corrected age were observed between those receiving KMC and those receiving conventional neonatal care for the outcomes psychomotor development (assessed using Griffith quotients), cerebral palsy, deafness or visual impairment.

Breastfeeding
Mothers of infants provided with KMC were more likely to exclusively breastfeed at discharge or at 40 to 41 weeks’ postmenstrual age (RR 1.16, 95% CI [1.07 to 1.25], p=0.0002; I²=39%; 6 studies/1453 mothers), and at one to three months follow-up (RR 1.20, 95% CI [1.01 to 1.43], p=0.044; I²=76%; 5 studies/600 mothers) compared to mothers of infants receiving conventional care. Moreover, mothers of infants provided with KMC were more likely to breastfeed (exclusive and/or partial) at discharge or at 40 to 41 weeks’ postmenstrual age (RR 1.20, 95% CI [1.07 to 1.34], p=0.0018; I²=80%; 10 studies/1696 mothers), at one to three months follow-up (RR 1.17, 95% CI [1.05 to 1.31], p=0.0061; I²=62%; 9 trials/1394 mothers), and at three months follow-up (RR 1.14, 95% CI [1.06 to 1.23], p=0.00028; I²=41%; 5 studies/924 mothers). These positive effects remained statistically significant in subgroup analyses of studies assessing intermittent KMC. In subgroup analyses, intermittent KMC also resulted in significantly improved rates of any breastfeeding at one to two months follow-up (RR 1.89, 95% CI [1.30 to 2.75], 4 trials/159 mothers) and at six months follow-up (RR 1.50. 95% CI [1.08 to 2.08], 3 trials/143 mothers). No differences between treatment and control groups were observed for other breastfeeding outcomes.

Additional outcomes
Mothers in the KMC intervention arm of one study including 269 mother-infant pairs were found to be more satisfied with the method of care compared to those receiving conventional neonatal care (RR 1.17, 95% CI [1.05 to 1.30]). In addition, in certain analyses of mother-infant attachment, including mother’s sense of competence, mother’s feelings of worry and stress, and mother’s sensitivity and responsiveness to the infant, outcomes were significantly higher in the KMC group in a study of 177 mother-infant dyads. Conversely, the mother’s perception of social support was found to be significantly lower in the KMC group at 14 days (MD -0.47, 95% CI [-0.84 to -0.10]), and among those with infants not admitted to NICU (MD -0.20, 95% CI [-0.39 to -0.01]). In a further study, mother-infant attachment was significantly higher at three months follow-up in the KMC group (MD 6.24, 95% CI [5.57 to 6.91]). Furthermore, in a small study of 30 mother-infant pairs, one indicator–mother’s relationship with the infant–was significantly higher in the KMC group (MD 1.00, 95% CI [0.35 to 1.65]). Mother-infant interaction, as measured by co-regulation, was also improved with KMC in one study of 45 mother-infant pairs (MD 16.38 symmetrical co-regulation, 95% CI [13.61 to 19.15]). In another study, the “HOME environment score” at 12 months’ corrected age was significantly higher in KMC families compared to those who received conventional care (MD 0.79 HOME score, 95% CI [0.74 to 0.84], 338 infants). For remaining outcomes, results were either not statistically different between groups, or were not analysed due to lack of data.

Early versus late onset kangaroo mother care in relatively stable low birthweight infants (1 trial/73 infants)
Mortality
In one study of 73 infants, no significant difference in the risk of mortality was identified between early onset KMC and late onset KMC at four weeks of age (RR 1.95, 95% CI [0.18 to 20.53], p=0.58), or at six months of age (RR 1.00, 95% CI [0.15 to 6.72], p=1.0).

Severe infection/illness
Risk of severe infection was not different between treatment groups at four weeks’ age (RR 0.42, 95% CI [0.12 to 1.49], p=0.18), and nor was the risk of any morbidity (RR 0.49, 95% CI [0.18 to 1.28], p=0.15). However, early onset KMC demonstrated a significant reduction in the length of hospital stay in comparison to late onset KMC (MD -0.9 days, 95% CI [-1.2 to -0.6], p<0.00001). The risk of hypothermia, hyperthermia, and readmission to hospital were not different between treatment groups.

Infant growth
Early onset KMC demonstrated a significant reduction in body weight loss from birth to 24 hours and birth to 48 hours in one trial of 73 infants (MD 39.2 g, 95% CI [11.1 to 67.2], p=0.0062; and MD 43.3 g, 95% CI [5.5 to 81.1], p=0.025; respectively). No significant differences were identified between early onset KMC and late onset KMC for weight gain at 4 weeks’ age (MD 58.9 g, 95% CI [-116.9 to 234.6]), or for risk of stunting (RR 0.83, 95% CI [0.46 to 1.48]), risk of wasting (RR 0.10, 95% CI [0.01 to 1.77), and risk of underweight (RR 0.49, 95% CI [0.21 to 1.14]) at six to 12 months’ of age. For all other growth outcomes, no data were reported.

Breastfeeding
Rates of exclusive breastfeeding did not significantly differ between early and late onset KMC groups at 24 hours of age (RR 1.02, 95% CI [0.67 to 1.57], 1 trial/73 infants), two weeks of age (RR 1.00, 95% CI [0.89 to 1.12], 1 trial/71 infants), four weeks of age (RR 0.94, 95% CI [0.85 to 1.04], 1 trial/67 infants), or six months of age (RR 2.69, 95% CI [0.99 to 7.31], 1 trial/55 infants).

5. Additional author observations*

In general, the methodological quality of the included trials was varied; all studies were at risk of performance bias due to lack of blinding and ten of the 21 trials were at unclear risk of allocation concealment bias. However, sensitivity analyses excluding trials at high risk of bias suggested that the direction of findings and size of the treatment effect were not affected by this. High levels of heterogeneity were detected in analyses of hypothermia, weight gain, breastfeeding, and length of hospital stay, and thus these results should be interpreted with caution. Using GRADE criteria, the quality of the evidence for the comparison KMC versus conventional care for the outcomes mortality at latest follow-up, severe infection/sepsis at latest follow-up, hypothermia at discharge/40 to 41 weeks’ postmenstrual age, weight gain at latest follow-up, and any breastfeeding at discharge/40 to 41 weeks’ postmenstrual age was judged to be of moderate quality. Evidence for the outcomes any breastfeeding at one to three months’ follow-up and psychomotor development at 12 months’ corrected age was rated as low quality. The majority of included trials were carried out in hospitals in low- and middle-income countries among stabilized LBW infants, limiting the generalizability of the findings. However, the advantageous effect of KMC on any breastfeeding (exclusive or partial) at one to two months follow-up was also observed among stabilized LBW infants in high-income countries.

The evidence to date favours the use of KMC in stabilized LBW infants as a substitute to conventional neonatal care in resource-limited settings. Further methodologically robust, randomized controlled trials investigating the effects of early-onset continuous KMC on stabilized and unstabilized infants in low- and middle-income countries are warranted, and further KMC trials are required in high-income settings. In addition, further data on long-term neurodevelopmental and neurosensory outcomes and the costs of care are needed.