Background. In 1978, kangaroo mother care (KMC) was proposed as a caring alternative for low birth weight (LBW) infants. We are reporting here early outcomes of a randomized, controlled trial comparing KMC to traditional care.
Method. An open randomized, controlled trial was conducted in a large tertiary care hospital. All newborn infants ≤2000 g, surviving the neonatal period and being eligible for a minimal care unit, were included. A total of 1084 newborns ≤2000 g were followed, and 746 were randomized—382 to KMC and 364 to traditional care. KMC infants were discharged after randomization, regardless of weight or gestational age. Infants spent 24 hours per day in an upright position, in skin-to-skin contact, and attached to the mother's chest. After randomization, control infants remained at the minimal care unit until meeting usual discharge criteria. Both groups are being followed up to 12 months of corrected age; 679 (90%) were available for evaluation when they reached term (40 to 41 weeks of postconceptional age). The present paper reports early outcomes (when reaching term) including mortality, infectious episodes, hospital stay after eligibility, and growth and feeding patterns.
Results. Both study groups were similar regarding all baseline variables but weight at eligibility. The risk of dying was similar in both groups (relative risk = 0.59, 95% confidence interval 0.22–1.6). There were no differences in growth indices. Nosocomial infections were more frequent in control infants. Hospital stay after eligibility was shorter in KMC, primarily for infants ≤1800 g.
Conclusions. These results show that KMC is a safe approach to the care of clinically stable LBW infants. Our findings provide the necessary scientific support to a method that is already incorporated in the care of LBW infants at many hospitals around the world and at different levels of care.
- kangaroo mother care
- skin-to-skin contact
- early discharge of low birth weight infant
- controlled trial
In 1978, the Kangaroo Mother Program was proposed by Dr E. Rey and developed further by others (H. Martı́nez, L. Navarrete, et al) at one of the largest obstetric facilities (Instituto Materno Infantil) in Santa Fe de Bogotá, Colombia. Kangaroo mother care (KMC) is an alternative to the expensive and seldom available traditional method of care for low birth weight (LBW) infants. KMC, as originally defined, comprises three major components: 1) kangaroo position, 2) kangaroo feeding policy, and 3) kangaroo discharge policy (early discharge in kangaroo position). Individual components of the KMC, excluding early discharge, have been evaluated.1-23 Our research group conducted an observational controlled assessment of the original Rey-Martı́nez KMC, including early discharge while in kangaroo position, between 1990 and 1993.24 The results of that study raised questions about early somatic growth, exclusive breastfeeding, and early discharge from hospital.
In response to those unsolved issues, we undertook two actions. First, guidelines and procedures of the KMC were modified, placing emphasis on early weight-gain monitoring, feeding supplementation, and evaluation of readiness of the mother and child couple to cope with early discharge from the hospital. Second, in 1993, we started a randomized, controlled trial (RCT) of the three components of the kangaroo intervention to assess not only survival, but other important variables such as growth and development, morbidity (both episodic and chronic conditions) during the first year of life, need for readmission after early neonatal discharge from the hospital, and quality of mother-to-infant bonding. We present here the primary results when all infants reached 40 to 41 weeks of conceptional age (WCA; gestational age + postbirth age = 40 to 41 weeks). The RCT was performed in a different hospital from the one where the method was developed, and the KMC method tested here included the modifications introduced by our team.
Study Population (Inclusion and Exclusion Criteria)
The study population comprised live newborn infants with birth weights ≤2000 g, born at Clı́nica San Pedro Claver (CSPC), which is one of the largest tertiary care obstetrics facility in Colombia, delivering 13 500 to 15 000 live infants per year. Almost a fourth of them have birth weights ≤2500 g and 8% of the total are <2000 g. An infant could be included in the study if he/she had a mother or a relative (living in the same household) able to understand and willing to follow the general program instructions (care for a premature baby and compliance with a 1-year follow-up schedule) according to a semistructured interview, and also eligible for KMC (see eligibility criteria for KMC in Study Interventions). At this time, evaluation of mothers' willingness to perform the specifics of the kangaroo intervention was not made. Exclusion criteria included being referred to another institution, plans to leave Bogotá in the near future, life-threatening or major malformations, early-detected major conditions arising from perinatal problems (severe hypoxic–ischemic encephalopathy, pulmonary hypertension, etc), and parental or family refusal to comply with the follow-up program, or, for those assigned to the kangaroo group, refusal to comply with the specifics of the intervention. This limitation in the design had the potential for introducing an imbalance between study groups, but none of the mothers assigned to kangaroo refused to participate.
According to our previous study,24 death rate among eligible patients under traditional care at the CSPC was ∼7%. To detect a twofold increase in the risk of dying, a sample of 215 subjects per group would be needed (α 0.05, two-tailed test, power 80%). Assuming a death rate among controls as high as 10%, at least 328 subjects per group were needed to detect a twofold increment in the risk of dying. We expected that only 70% of the newborn infants <2000 g would be eligible. Therefore, we planned to evaluate at least 940 infants ≤2000 g to identify a sufficient number of eligible patients.
Primary outcomes were mortality and growth (weight, height, and head circumference). Mortality was measured as the proportion of randomized children dying by 40 to 41 WCA among children with known life status at 40 to 41 WCA. Secondary outcomes were 1) length of hospital stay in survivors, measured as the sum of all days spent in hospital from eligibility to 40 to 41 WCA, in living children with complete follow-up; 2) overall incidence of infection measured as all infectious episodes requiring antibiotic treatment, regardless of whether treated in the inpatient or outpatient setting; 3) incidence of severe infection measured as nosocomial infections requiring systemic antibiotics or infections detected after discharge and requiring hospital admission for treatment; and 4) proportion of infants totally or partially breastfed, at 40 to 41 WCA. Partial breastfeeding was defined as the complement of breast milk with formula ordered by study personnel and covering up to 30% of the estimated nutritional requirements of the infant, as well as when the mother reported administration of nonprescribed formula in addition to breast milk. WCA 40 to 41 was chosen because it represents the time when LBW infants reach the conceptional age of term delivery. The study interventions compared here are expected to support LBW infants while they mature enough to face extrauterine life in similar conditions and with similar care requirements as those born at term.
Recruitment and Allocation
An observer especially assigned by the study reviewed the records of the delivery room and admission book daily and identified all newborn infants with birth weight or weight at admission ≤2000 g born at or referred to CSPC. Infants admitted to the neonatal care unit (NCU) as well as infants not admitted to the NCU (ie, those placed in a crib in the mother's hospital room) were followed and assigned a consecutive identification number.
Infants admitted to the NCU were followed by an attending physician, who assessed eligibility for the KMC intervention. Those children not admitted to the NCU were assessed for eligibility before mother and baby were discharge.
Once an infant was identified as eligible, he or she was randomly assigned to KMC or to the control intervention. Stratified block randomization lists were prepared in advance using a random numbers list (permutations of the number 16). There were four randomization lists according to birth weight: ≤1200 g, 1201 to 1500 g, 1501 to 1800 g, and 1801 to 2000 g. Blocks of four infants (two assigned to KMC and two to control) were prepared using a random number table.
Once the attending physician decided that a child was eligible, the person managing the allocation list (N.C.) was informed of the identification number of the eligible infant. The person managing randomization was aware of only the identification number, weight at birth (to consult the appropriate list according to weight strata), and whether the infant was a twin or triplet. The unit of randomization was the product of each delivery rather than individual infants. Therefore, twins or triplets were allocated to the same treatment to which the first of them who was eligible had been allocated. Each product of a delivery was allocated to the treatment corresponding to that registered in his/her weight stratum list according to the order in which each one had been declared eligible. In cases of twins or triplets, to consult the appropriate allocation list, the information on weight of the first eligible member of the set was used.
For infants assigned to KMC, parents were interviewed by a study nurse and invited to participate. If they accepted, the mother and infant were transferred to the Clı́nica del Niño, another Social Security Hospital where the ambulatory adaptation to KMC took place. Once in the Kangaroo Mother Program at the Clı́nica del Niño, parents were trained in KMC, and the adaptation of the mother and infant pair to kangaroo positioning and feeding was monitored.The mother and infant pair could stay at the program as long as necessary to demonstrate that the adaptation was appropriate.
Informed consent was not asked of parents of infants assigned to the control intervention. This procedure, proposed by Meinert and Tonascia28 (randomization before seeking consent to participate), was chosen for the following two reasons. First, early discharge is very appealing to parents, and it is very likely that a sizable proportion of families would have asked to be assigned to KMC. This is not a theoretic possibility; such a condition could introduce a significant contamination bias in the study. In fact, a kangaroo vs traditional trial in Guatemala had to be terminated because of contamination of the control intervention. Second, according to the Ethics Committee (Javeriana University, Bogotá), there were no ethical problems because those assigned to the control intervention received the usual care currently provided at the participant institution. At that time, the intervention was regarded as the best available care for LBW infants.
The interventions used were a modified version of the Rey-Martı́nez KMC. Infants <2000 g at birth are discharged as soon as they overcome major adaptation problems to extrauterine life, receive proper treatment for infection or concomitant condition, suck and swallow properly, and achieve a positive weight gain regardless of actual weight or gestational age. Mothers are used as “incubators” and as the main source of food and stimulation; babies are kept 24 hours a day in a strict upright position, in skin-to-skin contact, firmly attached to the mother's chest. In that way, the infant's temperature remains within the normal range by the mother's body heat. Babies must be breastfed regularly, and premature' formula supplements are administered to guarantee adequate weight gain if necessary. Infants are examined daily until weight gain reaches at least 20 g per day. Infants remain in the kangaroo position until they no longer accept it by demonstrating discomfort, pushing out limbs, and crying and fussing when mothers try to return them to the upright position.
Control (traditional care) intervention included keeping the infant in an incubator at the San Pedro Claver Hospital until he or she is able to regulate temperature and is thriving (ie, appropriate weight gain). Infants are discharged according to current hospital practice at CSPC (usually not before his or her weight is ∼1700 g). At the time of the study, the current practice at the San Pedro Claver NCU was to restrict parents' access to their babies severely.
Data Collection and Follow-up
After discharge, both groups had access to the same ambulatory care and follow-up. To minimize contamination bias, kangaroo infants' visits took place in the morning and control infants' visits in the afternoon. Ambulatory care included administration of metoclopramide29 and supplemental vitamins (A, D, E, C) to all infants and prophylactic aminophylline to infants <34 WCA at entry to the study.
Mothers' interviews and clinical chart reviews provided baseline data. Baseline information and follow-up data were collected prospectively at the neonatal unit and at scheduled visits. Data were entered into a computerized data base, after which both manual and computerized data cleaning activities were performed. Given that baseline data concerned variables and events occurring before randomization, observers collecting baseline information were unaware of the study group to which the infant belonged. Clinicians assessing infants at follow-up could not be blinded, given that kangaroo mothers were holding their infants in the kangaroo position at time of the interviews and clinical exam.
Comparisons between study groups for discrete variables were performed using χ2 or Fisher's exact tests. Continuous variables were compared by t test or nonparametric tests when appropriate. Adjustments for potential confounders and exploration for interactions were accomplished using ANOVA and/or multiple linear regression for continuous variables; robust regression (RR) ANOVA/ANCOVA and multiple RR26 for continuous variables with nonnormal distributions; and logistic regression for dichotomous outcome variables. The analysis was performed from the intention-to-treat perspective (effectiveness as opposed to efficacy), which means that all subjects were analyzed according to the group to which they were allocated, regardless of compliance with treatment or contamination of the intervention.
Between September 20, 1993 and September 20, 1994, a total of 13 560 live births were delivered at the CSPC. Of them, 3350 (24.7%) corresponded to LBW infants (birth weight ≤2500 g). A total of 1084 (8%) live newborn infants weighing ≤2000 g were assessed and followed to determine eligibility. A total of 307 (28%) infants were declared ineligible before randomization for reasons such dying before eligibility (160 infants, 15%); major malformation and dimorphic syndromes (7 infants, 1.5%); early detection of severe sequel of neonatal conditions including cerebral palsy, severe encephalopathy, bronchopulmonary dysplasia, etc (29 infants, 6.1%); referral to other institutions because of insufficient number of beds (131 infants, 12%); and other reasons (10 infants, 2.1%). The remaining 777 (72%) were randomized to one of the two interventions. Thirty-one infants were subsequently withdrawn, leaving only 746 in the study. These withdrawals were necessary because some of the conditions that precluded admission in the study (primarily neurologic impairment; toxoplasmosis, other rubella, CMV, herpes simplex infections, etc) were not detectable at time of randomization. When severe neurologic problems were detected during the first 2 weeks after initiation of the competing interventions or when reports from serologic exams demonstrating intrauterine infection developed after randomization, one of the authors (J.G.R.) was informed. This author, who was not involved in clinical care or direct data collection and who was completely blind to the treatment status of the subject, excluded all patient with a diagnosis of neurologic impairment or proof of intrauterine infection (congenital rubella, toxoplasmosis, etc) within 2 weeks after randomization. All others detected after that period were not excluded. These procedures for withdrawal after randomization were conducted in accordance with recommended procedures for RCT with prolonged follow-up.30 Fourteen (45%) of the withdrawn infants were in the KMC group, and 17 (55%) were in the control group.
Survival rates, weight at birth, gestational age, and Apgar score at 1 minute were higher in eligible compared with excluded infants. There were no differences in the distribution of other important variables (such as sociodemographic, pregnancy, labor, delivery, and newborn infant characteristics). Regarding the 131 referred infants, the distribution of all their baseline variables was similar to that for infants included in the study. We have no data on early survival of this group of referred infants.
Of the 746 included infants, 382 were randomized to the KMC and 364 to the traditional care group. None of the families allocated to the kangaroo group refused to participate. Nevertheless, and according to self-report, 11% of the kangaroo mothers did not comply with all the components of the intervention. Regardless of compliance, all subjects received the same follow-up. Complete follow-up data were obtained for 679 (91%) infants at the 40 to 41 WCA visit, 16 of whom died before this visit. Of the 67 infants for whom follow-up was not obtained, 20 were evaluated at subsequent scheduled visits and another 10 were documented to be alive by a telephone interview. No additional information was available for the remaining 36 subjects (4.8%).
The distribution of sociodemographic variables reflects the composition and characteristics of the population served by the Social Security in Bogotá. There were no differences regarding these characteristics between the two study groups: mother's age (kangaroo, 27.3 years; control, 27.4 years), proportion of stable couples (kangaroo, 83%; control, 86%), mother's education level (elementary school or less: kangaroo, 22%; control, 22%; secondary school: kangaroo, 56%; control, 56%; postsecondary school: kangaroo, 22%; control, 22%), mother's employment (office work: kangaroo, 11%; control, 13%; physical labor: kangaroo, 33%; control, 30%; housewife: kangaroo, 35%; control, 38%; other: kangaroo, 21%; control, 19%), and median per capita monthly income (in Colombian pesos, 1994) (kangaroo, Col$70 000; control, Col$ 70 000).
The high prevalence of pathologic conditions such as preeclampsia (38%) during pregnancy and the high rate of cesarean sections (Table1), reflect the fact that our study population was the high-risk segment of deliveries served by CSPC. There were no differences in the distribution of these variables, with the exception of number of multiple pregnancies (22 more in the kangaroo group; Table 1).
Regarding the characteristics of eligible newborn infants, there was a slight and statistically nonsignificant excess of females in the traditional care group. Nevertheless, the exploratory analysis showed that it did not confound the results. All other relevant characteristics at birth were distributed evenly (Table2).
Weights at eligibility were slightly lower in the kangaroo group. This difference was taken into account in the multivariate analysis (Table3).
Most frequent diagnoses at primary discharge in the study population were wet lung, 24%; hyaline membrane disease, 15%; neonatal sepsis, 11%; necrotizing enterocolitis, 5%; and nosocomial infection (suspected and/or confirmed) before eligibility, 10%. These diagnoses were distributed evenly between study groups.
Number and proportion of deaths were lower in the kangaroo infants, although this difference was not statistically significant (kangaroo, 6/364 [1.6%]; control, 10/345 [2.9%];, relative risk = 0.59 (95% confidence interval = 0.22–1.6).
Although the number and proportion of infectious episodes from eligibility to 41 weeks were similar (kangaroo, 49 [14%]; control, 44 [14%], P = .25), the distribution according to severity was different. The proportions of mild to moderate infectious episodes that could be treated as on an outpatient basis were 6.7% and 2.8% in the kangaroo and control groups, respectively (P = .019). Proportion of nosocomial infections after eligibility and before primary discharge was higher in the control group (kangaroo, 3.8%; control, 7.8%, P = .026). In addition, the number of total infectious episodes that had to be treated in the hospital was lower in the kangaroo group, although the difference was not statistically significant (kangaroo, 7.6%; control, 11%, P = .17). A few kangaroo subjects14 had nosocomial infections after eligibility, because several patients remained in the hospital after being declared eligible because of administrative and socioeconomic reasons. In some of them, nosocomial infections appeared after eligibility but before discharge. In others, infectious symptoms were evident within 3 days after discharge, meeting the criteria for nosocomial infection.
Somatic growth indices were almost identical in the two groups. Body weight at 40 WCA was similar to that usually observed in term newborn infants in Bogotá (2600 m above sea level) (Table4).
Exclusive and partial breastfeeding were highly prevalent in the two groups (46% exclusive and 48% partial) at 40 to 41 WCA. The proportion of subjects receiving only formula was small but statistically significantly higher in the traditional care group (Table4).
After adjusting for weight at eligibility (RR ANOVA/ANCOVA), there was an average saving of 1.1 days in total hospital stay from eligibility to 40 to 41 WCA in the KMC group (Table5). In addition, two-way ANOVA stratifying by birth weight showed that the saving in hospital stay was clearly related to weight at birth; the maximum saving was observed in infants <1200 g (8.6 vs 14.86 days), whereas for infants ≥1800 g, the hospital stay was virtually identical (2.86 vs 2.77 days) (Table 5).
Readmissions for primary apneas (kangaroo, 4; control, 4), hypoglycemia (kangaroo, 5; control, 9), and aspiration (kangaroo, 5; control 6) were not different between the 2 groups.
Most health facilities for high-risk deliveries and infants in large cities of developing countries are not very different from the San Pedro Claver Hospital and the Clı́nica del Niño in Bogotá. Despite the existence of properly trained personnel and good quality equipment (including advanced technology equipment) centralized in a few referral institutions, the huge demand for tertiary care far exceeds capacity, leading to a suboptimal quality of care for many. The complete KMC, including early discharge, skin-to-skin contact, and good quality nutrition based primarily on breastfeeding, has the largest potential for benefit in this environment. We would like to stress that the version of the KMC that we used is an alternative to practices in a minimal care unit (MCU). There are no published RCTs evaluating the entire KMC as described here, including the early discharge component.
Sloan et al25 reported an open RCT conducted in an obstetrics hospital in Quito, Ecuador. The study involved 300 infants <2000 g, who survived the adaptation to extrauterine life. Skin-to-skin contact and nearly exclusive breastfeeding were offered to 140 hospitalized infants, whereas 160 control infants were cared for at an MCU (incubators and/or heated cribs). The study was interrupted before collecting the proposed sample size because of an excess of serious illness among control infants. The authors reported that the inpatient kangaroo method saved health care costs, did not affect growth indices, and during the first 6 months of life, diminished the risk for serious illnesses, perhaps by changing health care patterns (including more frequent visits to health care providers) among “kangaroo families.” On the other hand, we are reporting early outcomes of infants in KMC (at 40 to 41 WCA), making it impossible to evaluate whether our data support the morbidity reduction found in Ecuador. Economic consequences of the intervention cannot be compared either. In the study by Sloan et al, kangaroo care was provided to hospitalized patients, whereas in our study, KMC was delivered outside the hospital, making it impossible to extrapolate the economic findings of the study in Ecuador to our study population.
Previously, we conducted an observational study comparing two facilities,24 one that provided the entire Rey-Martı́nez KMC, whereas the other kept infants in an MCU until standard discharge criteria were met. Although there were no differences in mortality rates, that study raised questions regarding insufficient early growth, increased morbidity from infection, and the relevance and nature of eventual saving in hospital stay in kangaroo infants.
The early results of the present RCT provide the following answers:
We can confirm that KMC is not associated with additional risk of dying, compared with traditional care infants, by 40 to 41 WCA. In fact, we found a reduction in mortality in KMC that is not statistically significant. This finding implies that survival under KMC is as good as or even better than that observed under traditional care, making KMC an excellent option for caring for LBW infants primarily in the developing world.
We also demonstrated that there is no reduction in early physical growth with the KMC. Although this finding differs from that of both of our previous studies of KMC and that of Dı́az-Rosello,27 we supplemented breast milk with formula for infants with daily weight gains of <20 g/day, independent of the assigned intervention.
Early discharge under KMC did not increase readmission attributable to metabolic and noninfectious respiratory complications of LBW infants, such as primary apnea, hypoglycemia, and bronchoaspiration. In fact, the episodes of hypoglycemia (defined as blood glucose ≤30 mg/dL) in both groups occurred in infants who were discharged without appropriate parent training in preterm infant feeding and/or kangaroo position and nutrition adaptation (infants allocated to KMC not being transferred to the Clı́nica del Niño but discharged directly to home, control infants discharged too early, and/or without appropriate information given to their parents). These incidents occurred occasionally and despite protocol violations in those cases, subjects were not excluded from analysis (intention to treat).
Bronchoaspiration after discharge was a major cause of morbidity and readmission in our previous study.24 Obstructive apnea and bronchoaspiration were attributed to variations of the kangaroo position that were used in our first study. For the present RCT, we had mothers keep babies strictly in an up-right position with the abdomen against the mother's chest. Before the initiation of this study and associated with this change in the guidelines, we documented a dramatic drop in the frequency of clinically evident episodes of obstructive apnea and/or bronchoaspiration.
Although the number of infectious episodes was similar in the two study groups, kangaroo infants' infections were less severe, most of them requiring only ambulatory care. Traditional care infants had a higher number of nosocomial infections and greater need for inpatient care of their infections.
KMC reduced the number of total days in the hospital and diminished the overcrowding of the neonatal unit, which is a particular problem in hospitals in developing countries. The reduction in hospital days was more marked for newborns with lower birth weights and nonexistent for those >1800 g. This is attributable to the fact that in the study institution, traditional care implies discharge when the infant's weight is ∼1700 to 1800 g. In other places where infants are discharged at higher weights, it is likely that the saving in hospital stay will extend to infants >1800 g.
Ambulatory care of infants discharged early under KMC includes close monitoring by qualified health personnel (neonatologists, registered nurses, etc) and an efficient follow-up system, including home visits and capability to readmit sick infants to a tertiary care hospital. Therefore, for understanding the impact of KMC on the use of health resources, a formal economic analysis is needed.
In our setting, KMC is initiated once a LBW infant is stable and ready to for transfer to a minimal care neonatal unit. There are reports in the literature about the earlier use of skin-to-skin contact, one of the components of KMC. This earlier use of skin-to-skin contact for limited amounts of time seems to have additional benefits not only in the mother's healing process2 but also in the stabilization of vital parameters,1,4,8,16,17 promotion of breastfeeding, and faster weight gain,12,13 and can be started very early, almost immediately, after birth6,14 even in intubated infants.10 There is room for potential improvement in the outcomes of our LBW infants and in the use of resources by the earlier use of skin-to-skin contact. We plan to evaluate the effects of the introduction of skin-to-skin contact in our neonatal intensive care unit (NICU) as one of the next steps in the development of our KMC research program.
Finally, our report has several limitations to assess fully the potential role of KMC in promotion and maintenance of breastfeeding. First, our data are limited to breastfeeding practices when babies reach 40 to 41 WGA. It is important to know whether the difference we found in feeding patterns survives during the first year of life. Second, data were collected using the categories described in “Methods,” which makes it difficult to compare our findings with those obtained through other measures. In the future, we plan to use the set of breastfeeding definitions developed for consistent use worldwide under the auspices of the Interagency Group for Action on Breastfeeding Definitions.30 Third, for some babies (the smaller and the sicker), the kangaroo intervention started too late in their extrauterine life to have had a large influence in the initiation and maintenance of breastfeeding. As mentioned before, the initiation of skin-to-skin contact while the infant is at the NICU, together with the establishment of hospital policies favoring breastfeeding at all levels (the “friendly hospital” initiative31,32), might have a more significant impact on the proportion and duration of breastfeeding than would have the KMC alone.
This study provides evidences showing that when using KMC (including early discharge) as an alternative to an MCU in hospitals with good technology but limited resources, kangaroo infants do better when reaching term (40 to 41 WGA) than babies under usual care. The evidence shows that 1) mortality and growth indices are similar, 2) total episodes of infection are similar but the spectrum of severity differs, favoring kangaroo infants, 3) there is an average 1.1 day savings in hospital stay for kangaroo infants and saving are larger in lower birth weight infants weights, and 4) there is a small but significant difference in early breastfeeding patterns. Previous studies (including our own evaluation of the Rey-Martı́nez KMC) raised doubts about early growth of infants receiving KMC. As mentioned above, our data show conclusively that KMC (including the use of formula to supplement breast milk, when necessary) does not jeopardize early growth of LBW infants. These early results are important for those already using KMC as well as for others considering it as an alternative to the care of LBW infants. Nevertheless, data regarding long-term outcomes (at least at 1 year of age corrected for gestational age) including the quality of mother-to-child bonding and emotional development are necessary to complete the evaluation of benefits and risks of this technique. Continued follow-up of these infants should provide answers to these important questions.
This study was funded jointly by the Instituto de Seguros Sociales de Colombia and the World Laboratory (ONG, Lausanne, Switzerland, Project Number MCD13).
We thank all members of the Kangaroo Research team: A. Mondragón, MD; R. Gómez, MD; M. Cristo, Psychologist; E. Vélez, Psychologist; M. Girón, Social Worker; R. Martı́nez, RN; F. A. Gómez, RN; and Mrs. M. V. Jiménez. Without their commitment and devotion, this study could not have been developed.
- Received November 19, 1996.
- Accepted February 27, 1997.
Reprint requests to (N.C.) Carrera 7a #46-20, Apto 2001, Santa Fe de Bogotá DC, Colombia.
- KMC =
- kangaroo mother care •
- LBW =
- low birth weight •
- RCT =
- randomized control trial •
- WCA =
- weeks of conceptional age •
- CSPC =
- Clı́nica San Pedro Claver •
- NCU =
- neonatal care unit •
- RR =
- robust regression •
- MCU =
- minimal care unit •
- NICU =
- neonatal intensive care unit •
- SD =
- standard deviation
- Bergman NJ,
- Jurisoo LA
- Drosten-Brooks F
- Ludington Hoe SM, Anderson GC, Simpson S, Hollingsead A
- Mondlane RP,
- da Graca AM,
- Ebrahim GJ
- ↵Schmidt E, Wittreich G. Care of the abnormal newborn: a random controlled trial study of the “kangaroo method” of care for low birth weight newborns. In: World Health Organization, ed. WHO Interregional Conference on Appropriate Technology Following Birth. Trieste, Italy: WHO; 1986
- Wahlberg V
- Whitelaw A
- Whitelaw A,
- Heisterkamp G,
- Sleath K,
- Acolet D,
- Richards M
- Charpak N,
- Ruiz Peláez JG, Charpak Y
- ↵Hamilton LC. Regression with Graphics: A Second Course in Applied Statistics. Belmont: Duxbury Press; 1992;183–212
- ↵Diaz-Rosello JL, Bellman M. Report of Pan American Health Organization/World Health Organization: early discharge–ambulatory care program for low birth weight infants (kangaroo method). Montevideo, Uruguay: Centro Latinoamericano de Perinatologı́a; 1985
- ↵Meinert CL, Tonascia S. Clinical Trials. Design, Conduct and Analysis. New York, NY: Oxford University Press; 1986:208–216
- ↵Voyer ML. Prematurite I, II, III. In: Encyclopedie Medico-Chirurgical de Pediatrie. Paris, France: Elsevier; 1996
- Kyenkya-Isabirye M
- Copyright © 1997 American Academy of Pediatrics