A roadmap for applying machine learning when working with privacy-sensitive data: predicting non-response to treatment for eating disorders.

OBJECTIVES: Applying machine-learning methodology to clinical data could present a promising avenue for predicting outcomes in patients receiving treatment for psychiatric disorders. However, preserving privacy when working with patient data remains a critical concern. METHODS: In showcasing how machine-learning can be used to build a clinically relevant prediction model on clinical data, we apply two commonly used machine-learning algorithms (Random Forest and least absolute shrinkage and selection operator) to routine outcome monitoring data collected from 593 patients with eating disorders to predict absence of reliable improvement 12 months after entering outpatient treatment. RESULTS: An RF model trained on data collected at baseline and after three months made 31.3% fewer errors in predicting lack of reliable improvement at 12 months, in comparison with chance. Adding data from a six-month follow-up resulted in only marginal improvements to accuracy. CONCLUSION: We were able to build and validate a model that could aid clinicians and researchers in more accurately predicting treatment response in patients with EDs. We also demonstrated how this could be done without compromising privacy. ML presents a promising approach to developing accurate prediction models for psychiatric disorders such as ED.

Primary Care Triple P for parents of NICU graduates with behavioral problems: a randomized, clinical trial using observations of parent-child interaction.

BACKGROUND: Preterm-born or asphyxiated term-born children show more emotional and behavioral problems at preschool age than term-born children without a medical condition. It is uncertain whether parenting intervention programs aimed at the general population, are effective in this specific group. In earlier findings from the present trial, Primary Care Triple P was not effective in reducing parent-reported child behavioral problems. However, parenting programs claim to positively change child behavior through enhancement of the parent-child interaction. Therefore, we investigated whether Primary Care Triple P is effective in improving the quality of parent-child interaction and increasing the application of trained parenting skills in parents of preterm-born or asphyxiated term-born preschoolers with behavioral problems. METHODS: For this pragmatic, open randomized clinical trial, participants were recruited from a cohort of infants admitted to the neonatal intensive care units of two Dutch hospitals. Children aged 2-5 years, with a gestational age <32 weeks and/or birth weight <1500 g and children with a gestational age 37-42 weeks and perinatal asphyxia were included. After screening for a t-score ≥60 on the Child Behavior Checklist, children were randomly assigned to Primary Care Triple P (n = 34) or a wait-list control group (n = 33). Trial outcomes were the quality of parent-child interaction and the application of trained parenting skills, both scored from structured observation tasks. RESULTS: There was no effect of the intervention on either of the observational outcome measures at the 6-month trial endpoint. CONCLUSIONS: Primary Care Triple P, is not effective in improving the quality of parent-child interaction nor does it increase the application of trained parenting skills in parents of preterm-born or asphyxiated term-born children with behavioral problems. Further research should focus on personalized care for these parents, with an emphasis on psychological support to reduce stress and promote self-regulation. TRIAL REGISTRATION: Netherlands National Trial Register NTR2179 . Registered 26 January 2010.

Brief parenting intervention for parents of NICU graduates: a randomized, clinical trial of Primary Care Triple P.

BACKGROUND: Preterm-born or asphyxiated term-born children who received neonatal intensive care show more emotional and behavioral problems than term-born children without a medical condition. It is uncertain whether regular parenting intervention programs to which the parents of these children are usually referred, are effective in reducing child problem behavior in this specific population. Our objective was to investigate whether a regular, brief parenting intervention, Primary Care Triple P, is effective in decreasing emotional and behavioral problems in preterm-born or asphyxiated term-born preschoolers. METHODS: For this pragmatic, open randomized clinical trial, participants were recruited from a cohort of infants admitted to the neonatal intensive care units (NICU) of two Dutch hospitals. Children born with a gestational age <32 weeks or birth weight <1500 g and children born at a gestational age 37-42 weeks with perinatal asphyxia were included. After screening for a t-score ≥60 on the Child Behavior Checklist (CBCL), children were randomly assigned to Primary Care Triple P (n = 34) or a wait-list control group (n = 33). The primary outcome was child emotional and behavioral problems reported by parents on the CBCL, 6 months after the start of the trial. RESULTS: There was no effect of the intervention on the CBCL at the trial endpoint (t64 = 0.54, P = .30). On secondary measurements of child problem behavior, parenting style, parenting stress, and parent perceived child vulnerability, groups either did not differ significantly or the intervention group showed more problems. In both the intervention and control group there was a significant decrease in emotional and behavioral problems during the trial. CONCLUSIONS: Primary Care Triple P, a brief parenting intervention, is not effective in reducing child emotional and behavioral problems in preterm-born children or term-born children with perinatal asphyxia. TRIAL REGISTRATION: Netherlands National Trial Register (NTR): NTR2179.

Long-term effects of routine morphine infusion in mechanically ventilated neonates on children's functioning: five-year follow-up of a randomized controlled trial.

Newborns on ventilatory support often receive morphine to induce analgesia. Animal experiments suggest that this may impair subsequent cognitive and behavioral development. There are sparse human data on long-term effects of neonatal morphine. We aimed to investigate the effects of continuous morphine administered in the neonatal period on the child's functioning. We conducted a follow-up study among 5-year-olds who, as mechanically ventilated neonates, had participated in a placebo-controlled trial on effects of morphine administration on pain and neurologic outcome. They were now tested on intelligence, visual motor integration, behavior, chronic pain, and health-related quality of life. Univariate analyses showed significantly lower overall intelligence quotient (IQ) scores for children who earlier had received morphine, that is, mean 94 (SD 14.5) versus 100 (SD 12.9) for those who received placebo (P = 0.049). Other between-group differences in outcomes were not found. The statistical difference disappeared after correction for treatment condition, open-label morphine consumption over the first 28 days, and a propensity score for clinically relevant co-variables in multiple regression analyses. However, scores on one IQ subtest, "visual analysis," were significantly negatively related to having received morphine and to open-label morphine consumption the first 28 days. The finding of a significant effect of morphine on the "visual analysis" IQ subtest calls for follow-up at a later age focusing on the higher-order neurocognitive functions. Morphine received in the neonatal period has negative effects on the child's cognitive functioning at the age of 5 years which warrants follow-up at a later age.