Feasibility and validity of early screening for identifying infants with poor social-communication development in a well-baby clinic system

A Prediction Model of Autism Spectrum Diagnosis from Well-Baby Electronic Data Using Machine Learning

Early detection of autism spectrum disorder (ASD) is crucial for timely intervention, yet diagnosis typically occurs after age three. This study aimed to develop a machine learning model to predict ASD diagnosis using infants' electronic health records obtained through a national screening program and evaluate its accuracy. A retrospective cohort study analyzed health records of 780,610 children, including 1163 with ASD diagnoses. Data encompassed birth parameters, growth metrics, developmental milestones, and familial and post-natal variables from routine wellness visits within the first two years. Using a gradient boosting model with 3-fold cross-validation, 100 parameters predicted ASD diagnosis with an average area under the ROC curve of 0.86 (SD < 0.002). Feature importance was quantified using the Shapley Additive explanation tool. The model identified a high-risk group with a 4.3-fold higher ASD incidence (0.006) compared to the cohort (0.001). Key predictors included failing six milestones in language, social, and fine motor domains during the second year, male gender, parental developmental concerns, non-nursing, older maternal age, lower gestational age, and atypical growth percentiles. Machine learning algorithms capitalizing on preventative care electronic health records can facilitate ASD screening considering complex relations between familial and birth factors, post-natal growth, developmental parameters, and parent concern.

Exploring parent-engaged developmental monitoring of young children before and during the COVID-19 pandemic, Porter Novelli Styles 2019 and 2021

Background

Early identification of developmental delays may have been negatively impacted by the COVID-19 pandemic. Parental engagement in developmental monitoring is a key component to successfully identifying developmental concerns.

Objective

/Hypothesis.The purpose of this project was to understand whether parental engagement in developmental monitoring changed over the course of the COVID-19 pandemic, from Spring 2019 to Fall 2021.

Methods

Survey data was obtained from 2019 SpringStyles and 2021 FallStyles Porter Novelli Public Services ConsumerStyles cross-sectional surveys. Only respondents with at least one child under the age of 8 at the time of the survey were included in the analytic sample (2019 N = 403; 2021 N = 344). Participants were asked several questions about how they monitor their children’s development. Changes in frequency of developmental monitoring from 2019 to 2021 were estimated using chi-squared tests.

Results

In both 2019 and 2021, 89% of parents reported engaging in any type of developmental monitoring. Within the group of parents who engaged in any monitoring, there were no differences across years in the percentage of parents reporting using the methods surveyed, except that a smaller percentage reported comparing their children to others in 2021 (25%) compared to 2019 (36%, p <.002).

Conclusions

Despite major disruptions to families’ lives, there were no significant changes to parents’ overall engagement in developmental monitoring prior to and during the COVID-19 pandemic.

Better together: Developmental screening and monitoring best identify children who need early intervention

BACKGROUND

Widely recommended developmental surveillance methods include developmental monitoring (DM) and development screening (DS). Much research has been done on DS, but very little research has compared the effectiveness of DM and DS together.

OBJECTIVES

To investigate the relationship between DM and DS in Part C early intervention (EI) service receipt.

METHODS

Authors used data from the 2007/2008 and 2011/2012 National Survey of Children's Health (NSCH). Authors report the prevalence of children aged 10 months to 3 years who received (a) DM only, (b) DS only, (c) both DM and DS, and (c) no DM or DS across survey years. Authors compare the odds of EI receipt across these groups.

RESULTS

During both periods, estimated EI receipt prevalence was higher for children receiving both DM and DS (8.38% in 2007/2008; 6.47% in 2011/2012) compared to children receiving no DM or DS (1.31% in 2007/2008; 1.92% in 2011/2012), DM alone (2.74% in 2007/2008; 2.70% in 2011/2012), or DS alone (3.59% in 2007/2008; 3.09% in 2011/2012) (for both time frames, p < .05). From 2007/2008 to 2011/2012, the proportion of children receiving DS only and both DM and DS increased, while children receiving DM only and no DM or DS decreased.

CONCLUSIONS

Children receiving DM and DS together were more likely to receive EI compared to children receiving DM alone, DS alone, or neither DM nor DS. These findings support the AAP recommendations indicating that DM and DS are complementary strategies for improving early identification and linkage to EI for young children.

Infant and dyadic assessment in early community-based screening for autism spectrum disorder with the PREAUT grid

Background

The need for early treatment of autism spectrum disorders (ASD) necessitates early screening. Very few tools have been prospectively tested with infants of less than 12 months of age. The PREAUT grid is based on dyadic assessment through interaction and shared emotion and showed good metrics for predicting ASD in very-high-risk infants with West syndrome.

Methods

We assessed the ability of the PREAUT grid to predict ASD in low-risk individuals by prospectively following and screening 12,179 infants with the PREAUT grid at four (PREAUT-4) and nine (PREAUT-9) months of age. A sample of 4,835 toddlers completed the Checklist for Autism in Toddlers (CHAT) at 24 months (CHAT-24) of age. Children who were positive at one screening (N = 100) were proposed a clinical assessment (including the Children Autism Rating Scale, a Developmental Quotient, and an ICD-10-based clinical diagnosis if appropriate) in the third year of life. A randomly selected sample of 1,100 individuals who were negative at all screenings was followed by the PMI team from three to five years of age to identify prospective false negative cases. The clinical outcome was available for 45% (N = 45) of positive children and 52.6% (N = 579) of negative children.

Results

Of the 100 children who screened positive, 45 received a diagnosis at follow-up. Among those receiving a diagnosis, 22 were healthy, 10 were diagnosed with ASD, seven with intellectual disability (ID), and six had another developmental disorder. Thus, 50% of infants positive at one screening subsequently received a neurodevelopmental diagnosis. The PREAUT grid scores were significantly associated with medium and high ASD risk status on the CHAT at 24 months (odds ratio of 12.1 (95%CI: 3.0–36.8), p < 0.001, at four months and 38.1 (95%CI: 3.65–220.3), p < 0.001, at nine months). Sensitivity (Se), specificity, negative predictive values, and positive predictive values (PPVs) for PREAUT at four or nine months, and CHAT at 24 months, were similar [PREAUT-4: Se = 16.0 to 20.6%, PPV = 25.4 to 26.3%; PREAUT-9: Se = 30.5 to 41.2%, PPV = 20.2 to 36.4%; and CHAT-24: Se = 33.9 to 41.5%, PPV = 27.3 to 25.9%]. The repeated use of the screening instruments increased the Se but not PPV estimates [PREAUT and CHAT combined: Se = 67.9 to 77.7%, PPV = 19.0 to 28.0%].

Conclusions

The PREAUT grid can contribute to very early detection of ASD and its combination with the CHAT may improve the early diagnosis of ASD and other neurodevelopmental disorders.

Autism Spectrum Disorder Screening Instruments for Very Young Children: A Systematic Review

Research on ASD in infancy has provided a rationale for developing screening instruments for children from the first year of life to age of 18 months. A comprehensive literature search identified candidate screening tools. Using methodological probe questions adapted from the Quality Assessment of Diagnostic Accuracy Studies (QUADAS), two Level 1 and three Level 2 screening instruments were reviewed in detail. Research evidence conclusions were that instrument development was in beginning phases, is not yet strong, and requires further development. Clinical recommendations were to continue vigilant developmental and autism surveillance from the first year on but to use the screening instruments per se only for high-risk children rather than for population screening, with considerations regarding feasibility for individual settings, informing caregivers about strengths and weaknesses of the tool, and monitoring new research.

Paternal sperm DNA methylation associated with early signs of autism risk in an autism-enriched cohort

BACKGROUND

Epigenetic mechanisms such as altered DNA methylation have been suggested to play a role in autism, beginning with the classical association of Prader-Willi syndrome, an imprinting disorder, with autistic features.

OBJECTIVES

Here we tested for the relationship of paternal sperm DNA methylation with autism risk in offspring, examining an enriched-risk cohort of fathers of autistic children.

METHODS

We examined genome-wide DNA methylation (DNAm) in paternal semen biosamples obtained from an autism spectrum disorder (ASD) enriched-risk pregnancy cohort, the Early Autism Risk Longitudinal Investigation (EARLI) cohort, to estimate associations between sperm DNAm and prospective ASD development, using a 12-month ASD symptoms assessment, the Autism Observation Scale for Infants (AOSI). We analysed methylation data from 44 sperm samples run on the CHARM 3.0 array, which contains over 4 million probes (over 7 million CpG sites), including 30 samples also run on the Illumina Infinium HumanMethylation450 (450K) BeadChip platform (∼485 000 CpG sites). We also examined associated regions in an independent sample of post-mortem human brain ASD and control samples for which Illumina 450K DNA methylation data were available.

RESULTS

Using region-based statistical approaches, we identified 193 differentially methylated regions (DMRs) in paternal sperm with a family-wise empirical P-value [family-wise error rate (FWER)] <0.05 associated with performance on the Autism Observational Scale for Infants (AOSI) at 12 months of age in offspring. The DMRs clustered near genes involved in developmental processes, including many genes in the SNORD family, within the Prader-Willi syndrome gene cluster. These results were consistent among the 75 probes on the Illumina 450K array that cover AOSI-associated DMRs from CHARM. Further, 18 of 75 (24%) 450K array probes showed consistent differences in the cerebellums of autistic individuals compared with controls.

CONCLUSIONS

These data suggest that epigenetic differences in paternal sperm may contribute to autism risk in offspring, and provide evidence that directionally consistent, potentially related epigenetic mechanisms may be operating in the cerebellum of individuals with autism.