NIDCAP: New Controversial Evidence for Its Effectiveness

In the early 1980s, Als conceptualized the “synactive” theory of child development and established the Newborn Individualized Developmental Care and Assessment Program (NIDCAP).1,2 This program requires trained caregivers to observe neonatal behaviors before, during, and after caregiving interventions and provide recommendations for care by staff and parents. These recommendations are often generalized to include reductions in noise and light levels, aids to promote flexion and self-regulation, clustering of care, and parental involvement.Authors who have performed meta-analyses regarding NIDCAP have concluded that the evidence to support NIDCAP as a framework within which to provide developmental care to improve medical and developmental outcomes is inconclusive.3–6 Incorporating in meta-analyses the results from the trials reported in this issue of Pediatrics by Maguire et al7 and Peters et al,8 which add 288 infants to the 285 infants previously enrolled in NIDCAP trials,9–15 will provide more precise effect-size estimates, as would meta-regression to adjust for differences in baseline characteristics between study groups.6Maguire et al conclude that NIDCAP is ineffective for outcomes to term, whereas Peters et al state that NIDCAP significantly improves short-term and long-term outcomes. Are there any obvious differences in methodology, study population, intervention, or reporting to explain the differences?In both studies, the randomization sequence was computer generated and allocation to study groups was concealed. Without explanation, Peters et al label the trial as cluster randomized with statistical analyses adjusted accordingly. After allocation, blinding cannot be achieved for outcome assessments in NIDCAP trials that include follow-up examinations, because parents might volunteer group assignment.Were the stated primary outcomes in the publications set a priori or based on results? The Peters et al study was entered in a trials registry in 2007, and the Maguire et al study was entered in 2005; both studies completed recruitment in 2004. In the many previously published abstracts by the Peters et al group, the objectives have included to assess the impact of NIDCAP-based care in very low birth weight infants on length of stay, ventilation days, and chronic lung disease; sedative drug use; sepsis; and neurodevelopmental outcomes.16–19 Maguire et al previously published results regarding parental experiences, parental stress, child behavior, and quality of life in the first year after NIDCAP, with no important differences between the 2 groups noted.20–22Maguire et al based the sample size, and presumably the primary outcome, on Bayley scores at 1 and 2 years of age, whereas Peters et al based sample size on length of stay. Infants in the Maguire et al study would now be older than 2 years, but results of Bayley tests are not presented. In the Peters et al study, the Mental and Psychomotor Developmental Indices at 18 months were not statistically significantly different between the groups (mean difference: 5.60 [95% confidence interval (CI): −1.49 to 12.70] and 3.95 [95% CI: −2.78 to 10.69], respectively). Length of stay depends on postmenstrual age (PMA) and weight at birth, which often favored the NIDCAP group in previous randomized, controlled trials.9–12 A more appropriate outcome is PMA at discharge. Meta-analysis of data from 4 trials7,10,12,13 results in a nonsignificant weighted mean difference in PMA at discharge of −0.50 weeks (95% CI: −1.14 to 0.14).Most intracranial pathology occurs in the first 48 to 72 hours of life23 before the intervention was introduced in either study. Infants were randomly assigned within 48 hours of birth in the Maguire et al study but at a mean age of 4.1 days in the Peters et al study.In the Maguire et al study the intervention consisted of weekly behavioral observations of the infants by a trained, certified NIDCAP developmental specialist, with the first observations and care plans within 48 hours of birth. In the Peters et al study the observations were performed approximately every 2 weeks. In both studies the control group received incubator covers and positioning. Both studies allowed infants to be transferred to step-down units. The intervention does not seem to be more intense in the study by Peters et al.Important baseline characteristics such as PMA and weight at birth were similar in the 2 study groups in the Maguire et al study (difference in means: 0.1 week and 11 g) but favored the NIDCAP group in the Peters et al study (difference in medians: 1 week and 45 g). Adjustments for multiple comparisons would reduce the number of significant outcomes.In summary, because both studies were small, differences in results could be a result of chance or baseline characteristics that favored the NIDCAP group in the Peters et al study. Other sources of bias might be present, because trials that are not double-blinded yield larger effect estimates.24 Before additional research of NIDCAP is initiated, consideration should be given to reports that many NIDCAP behaviors are rarely or never seen among preterm infants, that only a few are associated with stressful/painful interventions, and that clustering of care can result in important behavioral and autonomic reactions.25–27 Any innovative, developmentally sensitive intervention should start at birth in the resuscitation room, where neonates are exposed to excessive noise and light levels and painful/stressful stimuli to body parts (mouth, nose, throat, larynx, and hands) represented by large areas in the sensory cortex.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.