Abstract
Background
Early detection of developmental delays in children can significantly help them realise their full potential. The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-V), is the system in use at the moment to detect neurodevelopmental delays among children in the United States and other Western nations. However, due to cultural differences, there is a pertinent need for a content-validated module in the context of low- and middle-income countries, including India.
Purpose
The aim of this study was to develop and validate Teacher’s Evaluation of Neurodevelopmental Delays (TEDD) tool based on the criterions and definitions of neurodevelopmental disorders as mentioned in DSM-V and then synced with developmental milestones mentioned in International Classification of Diseases, 10th revision (ICD-10) and Early Childhood Care and Education (ECCE)/ New York City Early Education Centre (NYCE) framework, and items were worded in behavioural terms
Methods
We did a thorough review of the literature for the development of TEDD tool and used modified Delphi technique to content validate it. Data from nine experts, such as doctors, clinical psychologists, special educators, speech and language and applied behaviour analysis therapist were used for the analysis.
Results
Analysis resulted into 28 items being retained which can be applied in the Indian context.
Conclusion
This study has shown good content validity of the TEDD tool. Future studies are being planned to rule the feasibility and reliability of this tool.
Keywords: Neurodevelopmental disorders, teachers, preschool children, evaluation tool, India
Introduction
The early childhood period is from prenatal development to eight years of age. 1 Early childhood is crucial as experiences that a child receives during this period which lays the foundation for brain development and functioning till adulthood. Across the developmental lifespan, neurodevelopment is a dynamic interrelationship between cognitive, brain genetic, emotional and behavioural processes. Neurodevelopmental disorders and disabilities can result from significant and ongoing disturbance to this dynamic process due to environmental and genetic risks. 2 Infants and young toddlers who experience delays in their skill development are said to have neurodevelopmental delays (NDD). Early childhood neurodevelopment studies how the central nervous system develops and works (CNS). The CNS continues to develop for years after birth and begins doing so early in embryonic life. Therefore, an NDD is the outcome of aberrant CNS development, which can happen at any time, from in utero through the early developmental stage up to the age of about five. Atypical neurodevelopment impacts can affect a single area of functioning (like language) or several areas (e.g., motor, language and cognitive). 3 NDDs disproportionately impact children living in poor- and low-income communities. The range of childhood NDDs must be better understood, and this calls for reliable screening methods that are feasible and based on widely recognised disease classifications. Childhood disability is a crucial component of the global development agenda by the United Nations General Assembly and the Agenda for Sustainable Development. 4
Diagnostic and Statistical Manual of Mental Disorders, Third Edition (DSM-III) introduced ‘developmental disorders’ for the first time for the category that consisted of Autistic Disorder. 5 In DSM-V, ‘Neurodevelopmental disorders’ were added as an overarching disorder category. Over the past 10 years, the diagnostic framework for neurodevelopmental disorders, particularly attention-deficit/hyperactivity disorder (ADHD) and autism spectrum disorder (ASD), has changed. The DSM-5 is the system in use at the moment in the United States (DSM-V). Neurodevelopmental disorders got even greater importance in International Classification of Diseases, 11th revision (ICD-11), the most recent update of the WHO’s ICD since they were incorporated into the heading of the chapter on psychiatry. This work focuses on NDDs (as opposed to disorders) as it is generally assumed to be a precursor to the more intense disorders.
It is challenging to comprehend the prevalence of NDD globally due to several issues. The capacity to determine accurate incidence rates can be complicated by obstacles to evaluation and healthcare access in low- and lower-middle-income countries (LMICs) and for specific populations in high-income countries (HICs). Between 2014 and 2016, around 4.67% of children in the United States between the ages of 3 and 7 were determined to have an NDD, excluding those with neurodevelopmental disorders like autism and intellectual disability (ID). 6 According to a more extensive definition of developmental delay used in the United Kingdom for a cohort of children born between 2000 and 2002, an estimated 10% of infants under the age of nine months had mild delays, with an additional 2% having more severe delays. 7 According to data from 2009, 7.0% of children in Australia between the ages of 0 and 14 years satisfied the criteria for a disability encompassing developmental delays, health issues and sensory and motor impairments. Once more, these estimates change depending on how NDD is defined, how data is collected and the ages of the children being researched. The greatest birth cohort is in India (about 26 million). About one in eight children aged two to nine years living in India may have at least one neurodevelopmental disorder. Given the extensive prevalence of recognised NDD risk factors, India is believed to have a significant NDD burden. Arora and colleagues examined approximately 4,000 children aged 2–9 in their study at Kangra, Dhenkanal, Palwal, North Goa and Hyderabad, five geographically distinct sites in India, for common NDD, such as hearing impairment (HI), speech and language disorders, autism spectrum disorders and ID. 8 About 0.4–4.3% of younger age category children were found to have two or more NDDs. Besides, site-specific prevalence of any of the seven NDDs in children aged two to six years ranged from 2.9 to 18.7%. Pooled estimates for NDD including all sites were 9.2% (95% confidence interval [CI]: 7.5–11.2). 8
Given the reluctance as well as ignorance (in some cases) of the parents to get timely assessment of NDD-related symptoms, teachers are the second most important stakeholder to judge them. Therefore, we believe that an assessment tool for teachers will aid early detection of NDD among children. In this regard, we proposed the TEDD (Teacher’s Evaluation of Neurodevelopmental Delays) tool and this work attempts to establish the content validity of the TEDD tool. The content validation is the validation of TEDD’s items to improve and support preschool teachers’ knowledge and skills for early identification and screening of children with NDDs. Based on the framework of the tool, each item of TEDD was developed based on the criterions and definitions of neurodevelopmental disorders as mentioned in DSM-V and then synced with developmental milestones mentioned in ICD-10 and Early Childhood Care and Education (ECCE)/ New York City Early Education Centre (NYCE) framework, and items were worded in behavioural terms. Expert consensus is a standard method that is used to develop new measures. The author vetted the constructs and the items with an expert panel.
The research questions of this stage are:
Are the measured dimensions suitable for screening NDD based on expert consensus?
Based on the expert’s consensus, what are the psychometrically valid measurement items for various NDD dimensions?
Methods
Content Validity Basis Feedback from Expert Panel
In order to achieve the objectives of this stage and to answer the research questions, a screening tool was developed. Escobar-Pérez and Cuervo-Martnez 9 (p. 29) define content validation by expert judgement ‘as an educated view from individuals with a track record in the subject who are considered qualified experts and who can provide information, evidence, judgements, and assessments’. Evaluating an instrument through expert judgement involves asking a group to decide or express their opinion on a particular aspect. Experts have a critical role in explaining, adding and/or changing the necessary aspects. 10
Modified Delphi Method
In this research, the modified Delphi method is used to gain consensus about the face validity/content validity, gain insight into the tool’s usability among a panel of experts and analyse the collected data. The modified Delphi technique is an adaptation of the Delphi technique for the purpose of estimating validity of the questionnaire. This research uses the modified Delphi technique as the author carefully selected the items as per the framework. The original Delphi is based on iterative, one-on-one interviews done consecutively with the experts. Modified Delphi is built on assembling these same people and bringing up the concerns for a deliberate discussion and simultaneous consensus among all participants.
Participants
In this study, a purposive sampling technique was used for sample selection. Although there are no hard and fast criteria for sample size in qualitative research and data collecting utilising the Delphi technique, a sample size of eight people is considered an acceptable minimum. 11 A sample of nine subject matter experts was chosen to constitute the panel because of their academic background, knowledge, training expertise and experience in child development, neurodevelopmental disorders, diagnosing and tool development. Participants were required to meet three inclusion criteria:
Identify as MD/RCI (Rehabilitation Council of India) Licenced, Registered certification of an International Board of the specific field of study.
Knowledge and experience in the subject field: The experts should have relevant academic knowledge and professional experience in early child development.
A minimum of 15+ years of experience.
Having effective communication skills. Able to contribute opinion to the need of the study.
The experts were selected based on their experience and content knowledge in medical, behaviourism and special educational needs. Professionals in the field of diagnosis and/or expertise in child development and related disorders, namely, psychiatry/child psychiatry/clinical psychology/early interventions were approached by the author on the telephone or virtual call and explained the details of this study. All the experts have been actively involved in research, written articles and authored materials on mental health and have had their private practice for more than two decades. The experts were independent of those who developed the item pool. Informed consent was implied by responding to a detailed email written to each one of them providing a brief overview of the study and their roles and responsibilities.
The participants were MD doctors, RCI-licenced clinical psychologists and special educators, subject matter experts consisting of child and general psychiatrists (MD – psychiatry/child and adolescent psychiatry), developmental paediatrician (MD – developmental Paediatrics), licensed clinical psychologists (RCI licenced), speech and language and applied behaviour analysis therapist. The mean work experience of the panel was 24.83 years (standard deviation [SD]: 10.37 years). Table 1 lists the details of the panel.
Table 1. Judges, Field of Experience/Academic Training and Work Experience.
| Judge | Field of Expertise/Academic Training | Work Experience (Years) |
| 1. | MD, General Psychiatry and Research | 25 |
| 2. | MD, Child Psychiatry | 24 |
| 3. | Licensed Clinical Psychologist and Professor and Head (PhD, MPhil) | 30 |
| 4. | RCI Licensed Clinical Psychologist and Special Educator (MPhil, MS Clinical Psychology) | 22 |
| 5. | Speech Language Pathologist and US Board Certified Behaviour Analyst (ABA) (MASLP) | 30 |
| 6. | Developmental and Behavioural Paediatrician, Asst. Professor (MBBS, MD, Specialisation in NDD, Conduct Disorders, ADHD, Autism, Developmental Delays, SLDs) | 18 |
| 7. | MD, Psychiatry (AIIMS) | 12 |
| 8. | Developmental Paediatrician (MD Paediatrics, Specialisation in NDD, Behavioural Medicine) | 10 |
| 9. | MD, Psychiatry, Sr. Professor, PGIMS Department of Psychiatry, Rohtak (MD, DPM, DNB Psychiatry) | 13 |
Questionnaire Design
A Delphi questionnaire was used for data collection. A total of six questionnaires for each NDD were prepared. Each questionnaire consists of six sections. Section A is the expert’s demography. Section B describes the instructions that will be provided on TEDD to the teachers during administration and the expert’s views and comments. Section C is the DSM-V criterion of the specific NDD. Section D is the expert’s view on items on Form A (36–48 months), Section E is the expert’s view on items on Form B (49–60 months), Section F is the expert’s view on items on Form C (61–72 months). The questionnaires consisted of a total of 273 items across all six questionnaires combined. Each item was also assigned one of the five domains of TEDD, namely, cognitive, social-emotional, daily living/adaptive, and sensory-motor. The questionnaire involved five-point scale options: 5 = essential, 4 = important, 3 = don’t know/depends, 2 = unimportant and 1 = should not be included. More scale response options may result in less skewed data.
Data Collection
Once panel members were recruited, they were sent an electronic link to the online version of all the questionnaires, containing each of the original items for the five domains/dimensions of the tool, hosted by Google Forms, along with a Word document containing categories and indicators to be used to validate each item and a PDF copy of DSM-V. The protocol is mentioned in Table 2.
Table 2. The 5-Point Protocol of Rating the Items in the Questionnaire.
| Rating | Indicators. | Points |
| Essential | The item is essential, clear, concise and highly representative of the domain/disorder with which it is associated, that is, it must be included. | 5 |
| Important | The item is essential, clear, concise and representative of the domain/disorder with which it is associated to a significant extent. | 4 |
| Depends/Rewrite | The item needs to be rewritten and/or moved to a different domain to which it is better matched. It might depend on any additional comments. | 3 |
| Unimportant | The item is either unimportant or unclear and not wholly representative of the domain/disorder with which it is associated. | 2 |
| Should not be included | The item is least important, unclear, not concise and not representative of the domain/disorder with which it is associated and, hence, should not be included. | 1 |
Development of Item Pool
After preparing the initial item pool through the deductive method (i.e., using theory to guide item generation given by DeVellis and Thorpe, 12 the items on the pool were discussed over multiple Zoom calls with inputs and guidance from a licenced clinical psychologist and child psychiatrist. The author used the theories and the books mentioned above to guide the development of the initial items. Each NDD in the DSM-V criterion was thoroughly studied, and the corresponding developmental domain/s that fit best were chosen for each criterion. Further, each of the items/milestones for the domain/s was selected from the ICD-10, DSM-V and ECCE/NYCE framework. During these discussions, some of the items were deleted, while some others were added and/or modified. For example, duplicate items were removed, ambiguous language and complex-to-understand words were changed to easily understandable words, and technical words were removed and substituted with simple words. Some items were shortened, while others were made more comprehensive. Some of the items were modified to be more culturally relevant. Rural and urban settings (children and teachers) were considered, especially while formulating and deciding items related to ID and other relevant domains. In the end, 273 items were finalised after the first round of discussion that reflected the proposed six-dimensional construct of NDD. At this point, the items’ redundancy was considered so that a comprehensive and large pool of items was developed.
Statistical Analysis and Conducting the Modified Delphi Method
The SPSS Statistics 24.0 software was used for data analysis. Fleiss’ ‘κ’ was used to analyse the degree of agreement among the experts, as this is an analytical statistic that allows three or more raters to independently judge a sequence of items using an instrument with a fixed number of ordinal categories. 13 This coefficient has a minimum value of 0 and a maximum value of 1. The Fleiss values were interpreted using Landis and Koch’s scale, 14 which numerically represents the strength of agreement among observers (Table 3).
Table 3. Fleiss’ κ Values and Strength of Agreement.
| Fleiss’ κ | Strength of Agreement |
| 0.00 | Poor |
| 0.1–0.20 | Slight |
| 0.21–0.40 | Fair |
| 0.41–0.60 | Moderate |
| 0.61–0.80 | Substantial |
| 0.81–1.00 | Almost perfect |
Content Validation by Expert Judgement
For the evaluation of the tool, the proportion of possible agreements occurring in each dimension was taken into account for Fleiss’ κ.
According to Fleiss, the Fleiss κ measures the degree of agreement between more than two ‘categorical’ judges. Lower numbers are evaluated on a scale and a Fleiss κ of 1 shows complete agreement among all judges. Fleiss’ κ is employed when all the conditions are fulfilled:
The response variable being assessed by two or more raters is categorical.
The two or more categories of the response variable being assessed by the raters must be mutually exclusive, which has two components, and the response variable being assessed must have the same number of categories.
The same number of categories must be present for each rater for the response variable that is being assessed. In other words, the rating scale must be the same for all raters.
Content Validation by Expert’s Judgement: Disorder-Wise
The calculation of Fleiss considered each disorder’s proportion of probable agreements while evaluating the original instrument. The judges determined that the strength of agreement was moderate for communication disorder and fair for each other disorder, as indicated in Table 3. 14 P value less than .05 was used as the statistical benchmark for the findings, and a 95% confidence interval for each disorder indicates that the agreement is positively accurate. Table 4 shows the Fleiss, degree of agreement and statistical significance of the disorders.
Table 4. Fleiss’ κ and Statistical Significance of the Neurodevelopmental Disorders of the Instrument.
| Disorder | Fleiss’ κ | Level of Agreement | Ρ Value |
| ADHD | 0.259 | Fair | .01 |
| ASD | 0.231 | Fair | .02 |
| Communication disorders | 0.424 | Moderate | .00 |
| Intellectual disability | 0.205 | Fair | .01 |
| Motor disorders | 0.238 | Fair | .03 |
| SLD | 0.242 | Fair | .00 |
Note: ADHD, attention deficit hyperactivity disorder; ASD, autism spectrum disorders; SLD, specific learning disability.
Content Validation by Expert’s Judgement: Domain-Wise
The calculation of Fleiss considered each domain proportion of probable agreements while evaluating the original instrument. The judges determined that the strength of agreement was moderate for daily living and language, speech and communication domains and fair for each other domains like cognitive, sensory motors and social-emotional, as indicated in Table 3. P value less than .05 was used as the statistical benchmark for the findings, and a 95% confidence interval for each domain indicates that the agreement is positively accurate. Table 5 shows the Fleiss, degree of agreement and statistical significance of the domains.
Table 5. Fleiss’ κ and Statistical Significance of the Developmental Domains of the Instrument.
| Domain | Fleiss’ κ | Level of Agreement | Ρ Value |
| Cognitive (CG) | 0.343 | Fair | .02 |
| Daily living/adaptive (DL) | 0.421 | Moderate | .01 |
| Sensory motor (SM) | 0.232 | Fair | .04 |
| Social emotional (SE) | 0.256 | Fair | .00 |
| Language speech and communication (LSC) | 0.484 | Moderate | .01 |
Results
A total of 28 items were retained for Form A, 31 items for Form B and 32 items for Form C were retained post content validity procedure. The finalised scale items are shown in Table 6. The items of the content valid TEDD tool are applicable and adapted according to the Indian context.
Table 6. Pool of Items Retained After Content Validation.
| TEDD’s domains | Items |
| Form A – cognitive (CG) |
|
| Form A – sensory motor (SM) |
|
| Form A – social emotional (SE) |
|
| Form A – language speech and communication (LSC) |
|
| Form A – daily living/adaptive (DL) |
|
| Form B – cognitive |
|
| Form B – daily living/adaptive |
|
| Form B – sensory motor |
|
| Form B – social emotional |
|
| Form B – language, speech and communication |
|
| Form C – cognitive |
|
| Form C – sensory motor |
|
| Form C – social emotional |
|
| Form C – language, Speech and communication |
|
| Form C – daily living/adaptive |
|
Discussion and Conclusion
This article is an attempt to develop content validity of the TEDD tool by obtaining the opinion of the experts who have been diagnosing NDD among children in the Indian context. The study lacks in terms of other validity and reliability checks to establish its applicability and accuracy in diagnosing NDD among children. The TEDD tool developed by the authors has good acceptable range of content validity. Future studies must determine the validity and reliability for the TEDD tool with larger samples and also determine a gold standard for early detection of NDD among preschoolers in LMICs. There is a dire need to develop such tools for other countries having higher cohort of children suffering from NDD and the incumbent tools like DSM-V are not applicable due to cultural differences.
Lastly, this work uses Fliess’ κ to detect the consensus among different raters. Fleiss’ κ is particularly advantageous when there are more than two raters, making it versatile for settings where multiple experts provide evaluations as an objective measure of inter-rater reliability. While Fliess’ κ provides a numerical value for agreement, interpreting this value requires understanding the context and the specific categories used. This can sometimes be complex and future researchers should be mindful about the application of Fleiss’ κ in different context. A good way to determine the suitability of Fleiss’ κ is to rely on the past literature and prepare a taxonomy of the contexts where it has been applied successfully.
The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding: The authors received no financial support for the research, authorship and/or publication of this article.
ORCID iD: Sunanda Kolhe
https://orcid.org/0000-0003-1994-5752
Authors’ Contributions
All authors contributed to the study’s conception and design. Research conceptualization, data collection, analysis, and plagiarism checks were performed by Sunanda Kolhe, and Anand Prakash. The first manuscript draft was prepared by Sunanda Kolhe, Anand Prakash, and Maxim Pereira. All authors reviewed the results and approved the final version of the manuscript.
Statement of Ethics and Informed Consent
Due permission was obtained from the concerned authorities of the schools where the study was conducted. Confidentiality was maintained by assigning a code to each participating student’s document. The data were stored with a security code. Participants were informed of the confidentiality of their responses, and all doubts of the teachers were clarified before the author interviewed them to fill out the TEDD tool.
References
- 1.Block RW, Dreyer BP, Cohen AR, et al. An agenda for children for the 113th Congress: R ecommendations from the Pediatric Academic Societies. Pediatrics 2013; 131(1): 109–119. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Boivin MJ, Kakooza AM, Warf BC, et al. Reducing neurodevelopmental disorders and disability through research and interventions. Nature 2015; 527(7578): S155–160. [DOI] [PubMed] [Google Scholar]
- 3.Villagomez AN, Muñoz FM, Peterson RL, et al. Neurodevelopmental delay: C ase definition & guidelines for data collection, analysis, and presen tation of immunization safety data. Vaccine 2019; 37(52): 7623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Assembly G. Resolution adopted by the General Assembly on 11 September 2015 . A/RES/69/315 15 September 2015. New York: United Nations. [Google Scholar]
- 5.(American Psychiatric Association). APA Diagnostic and statistical manual of mental disorders. 4th ed. Washington, DC: APA, 2000. text revision. [Google Scholar]
- 6.Zablotsky B, Black LI and Blumberg SJ.. Estimated prevalence of children with diagnosed developmental disabilities in the United States, 2014– 2016, 2017. [PubMed] [Google Scholar]
- 7.Parsons S and Platt L.. Disability among young children: Prevalence, heterogeneity and socio-economic disadvantage. Centre for Longitudinal Studies, Institute of Education. London: University of London, 2013. [Google Scholar]
- 8.Arora NK, Nair MK, Gulati S, et al. Neurodevelopmental disorders in children aged 2–9 years: P opulation-based burden estimates across five regions in India. PLoS Medicine 2018;15(7): e1002615. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Escobar-Pérez J and Cuervo-Martínez Á.. Validez de contenido y juicio de expertos : U na aproximación a su utilización. Avances en medición 2008; 6(1): 27–36. [Google Scholar]
- 10.Garrote PR and del Carmen Rojas M.. La validación por juicio de expertos : D os investigaciones cualitativas en Lingüística aplicada. Revista Nebrija de lingüística aplicada a la enseñanza de lenguas. 2015; 18: 124–139. [Google Scholar]
- 11.Ogbeifun E, Agwa-Ejon J, Mbohwa C, et al. The Delphi technique: A credible research methodology. 2016 [Google Scholar]
- 12.DeVellis RF and Thorpe CT.. Scale development: T heory and applications . New Delhi: Sage Publications, 2021. [Google Scholar]
- 13.McHugh ML. Interrater reliability: T he kappa statistic. Biochemia Medica 2012; 22(3): 276–282. [PMC free article] [PubMed] [Google Scholar]
- 14.Landis JR and Koch GG.. The measurement of observer agreement for categorical data. Biometrics 1977; 33(1): 159–174. [PubMed] [Google Scholar]
