Introduction

Growth and development begin during pregnancy and continue throughout the years¹. Neurodevelopment is an interactive process between children and their environment², influenced by genetic, environmental, biochemical, and physical factors³. Its ultimate goal is the maturation of the nervous system, achieving the development of brain functions and personality formation⁴; there are critical periods with conditions for acquiring skills⁵.

The World Health Organization (WHO) estimates that 10% of a country's population shows some form of developmental delay². According to the Pan American Health Organization, approximately 250 million children (43%) under the age of five in developing countries are at greater risk of not reaching their full development due to poverty, constituting a public health problem². In Mexico, a prevalence of 6% of children with disabilities has been reported, and approximately 25% of children under the age of five have a developmental delay².

The American Academy of Pediatrics recommends neurodevelopmental evaluations at 9, 18, 30 months, and 4-5 years of age in the absence of risk factors⁶. This assessment evaluates developmental milestones⁷ and identifies warning signs such as regression or persistence of patterns that should have disappeared⁸. As the pediatrician's clinical judgment is not sufficient, it is necessary to use screening tools that assess different areas of development⁹. Several international screening tests are described below (Table 1)⁸,¹⁰-¹⁴.

The available screening tests are very complex and lengthy, making them impractical for health workers at the community level. Therefore, neurodevelopmental assessment is often omitted from the well-child visit¹⁵; a standardized clinical test is needed to make a timely diagnosis⁴. A screening test identifies individuals with suspected disease in an apparently healthy population¹⁶; it should have a sensitivity and specificity >70%¹⁶. In the review of the literature, we found tests developed and validated in Mexico, highlighting the Evaluación del Desarrollo Infantil (EDI), with the lowest bias, and the Valoración Neuroconductual del Desarrollo del Lactante (VANEDELA), with the most documented validation process⁹; the EDI is the most widely used¹⁶ (sensitivity 76.1% and specificity 59.1%)¹⁷ (Table 2)¹⁵,¹⁷-¹⁹.

This study aimed to provide a screening tool for detecting suspected neurodevelopmental delay in children under 1 year of age that is practical and can be used by health professionals and caregivers. It is not intended to replace the existing ones but to be a filter that easily identifies those patients at risk who require more extensive and detailed evaluation.

The test is the KARVI scale, created by Dr. Miguel Angel Karlis Rangel, which evaluates children from 0 to 12 months in five areas: sensory (proprioceptive and fine motor), auditory, visual, emotional (socio-affective), and motor (gross motor). It has two achievements per area per month, which are scored as "Yes" (achieved) or "No" (not achieved), leaving a total of 10 items for each month. The test is composed of observational and verbal items. Each domain is scored individually, resulting in a total score that is classified into four categories: optimal development (two achievements reached), standard development (one achievement reached), lack of developmental stimulation (no achievement reached in 1 month of evaluation in a single domain), and developmental delay (none of the achievements reached in at least 2 consecutive months of evaluation). If no achievement is reached in a certain area, the evaluation of the previous and current month of that individual activity is repeated the following month. The results are color coded (traffic lights) to highlight their importance (blue = optimal, green = standard, yellow = lack of stimulation, and red = developmental delay) (Table 3). Among the advantages of this test are its duration of 5-10 min, its straightforward language, and the fact that it does not require any special material but uses objects that are familiar to the child. A limitation is that it is only used in children under 12 months of age.

During our study, we faced several problems, including that in March 2020, the WHO declared a severe acute respiratory syndrome coronavirus 2 pandemic, which triggered an epidemiological emergency that forced governments to take measures such as social isolation²⁰. As a result, face-to-face consultations were reduced, pathology checks were postponed, treatments were interrupted, and social activities were restricted, increasing problems in early childhood development²¹. Studies that analyze the impact of the pandemic on neurodevelopment have emerged. In Spain, a decrease of up to 15% in neurodevelopmental consultations has been reported²²; in Italy, it is mentioned that paying more attention to children with risk factors for developing neurodevelopmental delays is important²³. This information represents a major challenge for physicians, who must focus on early detection²⁴.

Studies of remote neurological assessment using telemedicine, defined by the WHO as "The delivery of health services using information and communication technologies for the exchange of information for the diagnosis, treatment, and prevention of disease," were identified in the literature²⁵. Telemedicine removes the barriers of time and distance by reaching remote locations and reducing waiting times²⁶. The Internet and electronic devices are helpful in monitoring and diagnosing clinical conditions²⁷; thus, the evaluations performed by telemedicine are not inferior to those performed in person in terms of patient and health professional satisfaction²⁸. There have been publications in which telemedicine has been used to evaluate neurological disorders, and this field of telemedicine has been called "teleneurology"²⁹. In Australia in 2016, a study was conducted to determine whether a mobile phone application could identify the risk of neurodevelopmental delay using the General Movements Assessment scale and concluded that the application facilitated identification³⁰. In Iowa in 2014, parents of children with neurological conditions concluded that telemedicine consultations were as effective as in-person consultations³¹. The world has seen the current situation as an opportunity to develop an alternative to continue neurodevelopmental assessment³². In Spain, a 63% increase in pediatric consultations through telemedicine was demonstrated from March to June 2020, maintaining the follow-up of patients with neurodevelopmental disorders and minimizing the risk of contagion²². After reviewing these studies, we developed the idea of continuing our project through teleneurology.

Methods

We conducted an observational, longitudinal, comparative, inferential, and prospective study to determine the sensitivity of the screening test (KARVI scale) in the neurodevelopment assessment in children under 1 year of age. We decided to compare our KARVI scale (screening test) with the EDI test (gold standard), a test developed and validated in Mexico for detecting neurodevelopmental problems³³. The study hypothesis (alternative) was "The screening tool (KARVI scale) is as sensitive as the EDI test for the timely detection of suspected neurodevelopmental delay in children under 1 year of age," while the null hypothesis was "The screening tool (KARVI scale) is not as sensitive as the EDI test for the timely detection of suspected neurodevelopmental delay in children under 1 year of age."

The study was approved by the Research Ethics Committee of the ITESM School of Medicine (No. P000253-EKARVI2019-CEIC-CR003). Pediatricians and neonatologists in the metropolitan area of Monterrey, Nuevo León, were informed of the project; individuals who met the inclusion criteria were identified and, with prior authorization from the physician, were invited to participate. Informed consent was given to each caregiver, and a signed consent form was obtained before the assessments. Participants were recruited between October 2020 and October 2021.

Inclusion criteria were individuals aged 0-12 months, born at term, previously healthy, without apparent risk factors for neurodevelopmental delay (metabolic or genetic diseases, tumors, cranioencephalic trauma, or neurological infections with sequelae), attending well-child visits and referred by their physicians, whose caregivers and physicians agreed to participate in the evaluation, and who had the means to conduct sessions remotely. Exclusion criteria were individuals older than 12 months, pre-mature births, caregivers who did not wish to participate in the study, children with a previously diagnosed disease associated with any neurodevelopmental delay (metabolic and genetic diseases), healthy children whose neurodevelopment could be affected by a previous disease (tumors, cranioencephalic trauma, and neurological infections with sequelae), incomplete evaluations, and children who did not come for follow-up or who did not have the means to conduct sessions through remote access.

We worked with qualitative variables grouped as positive test (abnormal) or negative test (normal) to obtain dichotomous variables. Optimal development (blue) in the KARVI test and normal development (green) in the EDI test were identified as equivalent variables and were considered negative. Standard development (green), delayed stimulation (yellow), and developmental delay (red) in the KARVI test and developmental delay (yellow) and risk of developmental delay (red) in the EDI test were considered positive.

The following formula was used to calculate the sample size:

Legend: z = 1.98, d = 0.05, P = prevalence of neurodevelopmental problems, Q = 1-p

The following parameters were used:

95% confidence level (C), prevalence of neurodevelopmental problems of 20% (P), and absolute precision of 5% (d).

This gave us a total sample of 250 individuals. However, this sample was modified by the pandemic, making it impossible to conduct the sessions in person. Therefore, it was decided to conduct them remotely using the Zoom^® application, avoiding any physical exposure. A total of 52 children were recruited, two of whom dropped out of the protocol, leaving us with 50 individuals who were evaluated remotely using the Zoom^® application by physicians (pediatric residents and interns) who had been previously trained in using both tests. Four monthly Zoom sessions were conducted to obtain age and anthropometric data; the WHO charts were used to obtain percentiles of weight, height, head circumference (HC), and body mass index. The KARVI scale and the EDI test were then applied, and the results were registered into the database.

Results

Fifty-two children were recruited, of which 3.8% (2/52) dropped out of the study due to lack of time to attend the sessions. In the end, the remaining 96.15% (50/52) constituted our final sample for analysis. Participants were 52% male (26/50) and 48% female (24/50) (Fig. 1).

Figure 1 Total number of participants during the study. 

Results were analyzed using IBM SPSS Statistics^® version 28 software. Descriptive statistics were used to analyze demographic variables and obtain measures of central tendency (Table 4), followed by 2 × 2 and Pearson's χ² tables to compare the degree to which a test can discriminate between individuals with and without neurodevelopment problems. Inferential or comparative statistics were used for parameter estimation and hypothesis testing.

Four consecutive monthly evaluations of the EDI standard test and the KARVI test were performed using digital media: Zoom^®, Skype^®, and Whatsapp^®, by one of the doctors in the study, with a duration of approximately 15-20 min, in which the EDI test was applied through the electronic platform and the KARVI test through Google^® forms. Sensitivity, specificity, false negative/type II error, false positive/type I error, positive predictive value, and negative predictive value of both tests were determined in each evaluation. In evaluation 1, sensitivity was 70% (confidence interval [CI] 95% 34.75-93.33) and specificity was 75% (CI 95% 58.8-87.31); in evaluation 4, sensitivity was 100% (CI 95% 29.4-100) and specificity was 78.72% (CI 95% 64.34-89.3), both of which were significant (p = 0.007 and p = 0.001, respectively). A sensitivity of 57.1% (CI 95% 18.41-90.1) and a specificity of 74.4% (CI 95% 58.83-86.48) were obtained in evaluation 2 and a sensitivity of 62.5% (CI 95% 24.49-91.48) and a specificity of 71.4% (CI 95% 55.42-84.28) were obtained in evaluation 3, without being significant (p = 0.091 and p = 0.063, respectively) (Table 5).

Discussion

Neurodevelopmental disorders are a major problem in developing countries, affecting child morbidity and public health. Their assessment at the point of care is essential but is not always possible due to factors such as the complexity of existing screening tests and, more recently, the isolation caused by the pandemic. Therefore, simpler tests are needed. During our study, we confirmed that the screening tool KARVI scale is as sensitive as the EDI test for timely detection of suspected neurodevelopmental delay in children under 1 year of age, as we obtained sensitivity percentages > 70% in scores 1 and 4, and specificity percentages > 70% in all 4 scores. However, we recognize that our study did not reach the ideal sample size for this type of scale, resulting in an inability to perform psychometric analysis and obtain wide CIs adequately. In addition, although Pearson's χ² test was performed with significant results in evaluations 1 and 4, given that it is very sensitive to the sample size, we could face errors in its interpretation, overestimating the test's usefulness. Therefore, it is recommended that the study should be reproduced in a larger population.

Among our limitations are the age of the patients, since only patients from 0 to 12 months were included; the recruitment process, referred by physicians from the metropolitan area of Monterrey; the follow-up of the patients, because it was only carried out for 4 months; the interpretation of the results, since the EDI test has three groups, while KARVI scale has four, which made it difficult to compare the results. Lastly, the application, since the same person performed both tests, which could lead to bias when knowing the result of the other test. In addition, it is necessary to consider the limitations resulting from the pandemic since the isolation prevented to conduct a face-to-face assessment of the patients, which affected the size of the sample and made it difficult to find caregivers willing to conduct sessions through remote access to electronic media and privacy issues, as well as communication problems and lack of understanding of the items by the child's caregivers.

In the literature, we found several validation studies of neurodevelopmental screening tests, highlighting the EDI test conducted in 2013 by Rizzoli et al. and the VANEDELA test conducted in 2011-2012 by Sanchez et al. their sample sizes and distribution by gender are shown in Table 6¹⁷,¹⁹.

Our results were compared with those obtained with the modified version of the EDI test in children under 16 months of age during its validation process. In conclusion, KARVI is more specific than EDI, i.e., it is better at obtaining a negative result in healthy patients (Table 7). Despite the conditions, we had a diverse sample in terms of age and sex, representative of the general population; however, it is suggested to expand the sample to obtain more significant results.

The KARVI scale has the elements to be an effective screening test to detect suspected neurodevelopmental delay since it has adequate sensitivity and specificity (> 70%). In addition, it does not require special materials, caregivers can use it without prior training, and its use requires less time than other tests. However, we faced some limitations in the present study, such as the sample size. therefore, a study with a larger number of patients and recruitment sites should be conducted. Both tests should be administered in person, and more personnel should be available to administer the tests separately and minimize bias. Finally, the results are satisfactory under the circumstances in which this study was conducted. We will continue the validation process and implement the electronic scale project.