Research in Developmental Disabilities 78 (2018) 78–88

Contents lists available at ScienceDirect

Research in Developmental Disabilities

journal homepage: www.elsevier.com/locate/redevdis

“Survey of Wellbeing of Young Children (SWYC)”: how does it fit

T
for screening developmental delay in Brazilian children aged 4 to
58 months?
⁎
R.S. Moreiraa, , L.C. Magalhãesb, C.M. Siqueirac, C.R.L. Alvesc
a
Department of Health Sciences at the Federal University of Santa Catarina/Araranguá, Rod. Gov. Jorge Lacerda, 3201 Jardim das Avenidas,
Araranguá, SC, ZIP CODE: 88.906-072, Brazil
b
Department of Occupational Therapy at theFederal University of Minas Gerais, Av. Pres. Antônio Carlos, 6627 Campus, Pampulha, Belo Horizonte,
MG, ZIP CODE 31270-901, Brazil
c
Department of Pediatrics at the Federal University of Minas Gerais, Av. Prof. Alfredo Balena, 190-room 267, Belo Horizonte, MG. ZIP CODE
30130100, Brazil

A R T IC LE I N F O ABS TRA CT

Number of reviews completed is 2 Objective: To replicate the original normative study of the SWYC’s Milestones Questionnaires for
Keywords: children in Brazil. Our goals were to compare the performance of Brazilian and North American
Child development children using this screening tool and to verify the reliability and validity of the Brazilian version.
Questionnaires Study design and setting: Cross-sectional study with children aged 1–65 months and their guar-
Test validity dians, recruited in southern Brazil. Parents were interviewed using the Developmental Milestones
Data reliability questionnaire, which contains 10 questions about cognitive, motor, social, and language abilities.
Culture and primary health care Item response theory was used to examine item validity.
Results: We interviewed 415 parents. SWYC provided the most information on the children’s
development between 10 and 30 months. The performance of Brazilian and North American
children was quite similar when children are younger than 36 months old. Above 36 months,
North American children performed almost all items earlier than Brazilians. Convergent validity
was 0.73 and internal consistency 0.97.
Conclusion: The Brazilian version of the Developmental Milestones questionnaire presented ac-
ceptable measurement qualities that support the SWYĆs potential as a developmental screening
tool. As we found important differences between North American and Brazilian children in
achieving the milestones, especially among the oldest children, additional normative studies are
needed.

What is new?

This pilot study analyzed the Brazilian version of a questionnaire (Survey of Wellbeing of Young Childre – SWYC) designed to
screen developmental delay in children aged 0–5 years. The SWYC is simple, easy to apply and interpret, free of charge, and can be
used on a large scale in primary care in Brazil.

⁎
Corresponding author.
E-mail addresses: rafaela.moreira@ufsc.br (R.S. Moreira), liviacmag@gmail.com (L.C. Magalhães), cmsiqueira@hotmail.com (C.M. Siqueira),
lindgrenalves@gmail.com (C.R.L. Alves).

https://doi.org/10.1016/j.ridd.2018.05.003
Received 4 November 2017; Received in revised form 1 March 2018; Accepted 8 May 2018
Available online 21 May 2018
0891-4222/ © 2018 Elsevier Ltd. All rights reserved.
R.S. Moreira et al. Research in Developmental Disabilities 78 (2018) 78–88

Key findings

The Brazilian version of the Developmental Milestones questionnaire presented acceptable measurement qualities that support
the SWYĆs potential as a developmental screening tool. We found important differences between the North American original sample
and the Brazilian one in achieving the milestones, especially among the oldest children, indicating the need to invest in developing
local norms.

Implications

The use of developmental screening instruments in primary health care can contribute to Brazilian population-based studies and
subsidize the implementation of child health care policies. However, this was a pilot study, and additional research is needed to better
understand whether these results reveal a sociocultural pattern or are related to methodological issues.

1. Introduction

Developmental delays occur when a child does not show the expected motor, language, socioemotional, and cognitive abilities
that are required for achieving a range of personal and social competencies in his/her individual life (Dosman, Andrews, & Goulden,
2012). Data based on the prevalence of children under 5 years of age stunted and living in extreme poverty estimates that about 219
million children in low- and middle-income countries, meaning 18% of children from Latin America and the Caribbean Region, are at
risk of not reaching their potential development (Black et al., 2017). In an attempt to promote healthy development and recognize
children with delays as early as possible, the American Academy of Pediatrics recommends constant surveillance of a child’s de-
velopment (Dosman et al., 2012; Hagan, Shaw, & Duncan, 2017; Sheldrick, Merchant, & Perrin, 2011). This monitoring consists of
observing the child carefully, the use of a detailed anamnesis that values the parents’ concerns, and physical examinations conducted
by health care professionals familiar with the child’s development(Dosman et al., 2012; Hagan et al., 2017).
Developmental surveillance without the use of appropriate screening instruments fails however to identify a great number of
children with developmental problems(Aly, Taj, & Ibrahim, 2010; Fernald, Prado, Kariger, & Raikes, 2017). Standardized screening
instruments are widely recommended, because, in general, they are compact and because there is scientific evidence of their potential
to detect delays, aiding in the identification of children who need diagnostic assessment. Screening tests can identify childreńs
abnormalities with more accuracy than when decisions are based solely on professionals’ clinical judgment (Aly et al., 2010). A
systematic review indicates that less than half of the patients with developmental and behavioral problems are identified only by a
pediatricians’ clinical assessment (Sheldrick et al., 2011). Moreover, when compared to diagnostic tests, developmental screening
questionnaires are more rapid to perform, what enables their large-scale use, such as in public health care (Fernald, Kariger, Engle, &
Raikes, 2009; Sabanathan, Wills, & Gladstone, 2015).
It is important, however, to emphasize that the choice and use of screening instruments should be based on their psychometric
properties, the existence of normative data for the target population, and the accessibility of tests for professionals, including costs
and the need for training (Aly et al., 2010; Sabanathan et al., 2015). When health care professionals use screening instruments with
adequate psychometric properties, the detection of suspected delay exceeds 70%, while tests with poor measurement properties or
informal checklists only detect 30–40% of children at risk (Glascoe, 2015).
Health professionals in low-and-middle-income countries face difficulties at detecting and diagnosing developmental delays due
to the scarcity of instruments validated for their cultures (Black et al., 2017; Madaschi, Mecca, Macedo, & Paula, 2016; Saccani &
Valentini, 2012). Considering Brazil, only the “Escala do Desenvolvimento do Comportamento da Criança no primeiro ano de vida”
was created and standardized for Brazilian children aged 1–12 months, and its 64 items assess only the motor and communication
domains (Pinto, Vilanova, & Vieira, 1997). Madaschi et al. (2016) carried out a cross-cultural adaptation and validity study of the
Bayley Scales of Infant and Toddler Development-III (Bayley-III) for Brazilian children from 12 to 42 months (Madaschi et al., 2016).
However, despite being an internationally recognized developmental test, the use of the Bayley-III is still not feasible in clinical
practice in Brazil because the translation is not available for sale, the test application requires a long time, and professionals must be
trained, in addition to the need for an expensive kit with materials.
Although several screening tests widely used internationally, such as the Denver Developmental Screening Test-II (Denver-II), the
Ages & Stages Questionnaires, Third edition (ASQ-3), and the Alberta Infant Motor Scale (AIMS), have been translated for clinical use
in Brazil, only a few have established performance age norms for Brazilian children. Drachler, Marshall, and Leite (2007) carried out
a simple translation of the Denver-II for Brazilian children, without going through the steps of transcultural adaptation, re-
commending adjustments in the scoring criteria for specific items (Drachler et al., 2007).
Filgueiras et al. (2013) a cross-cultural adaptation of the ASQ-3 protocols, gathering data on 45,000 children from a major city in
Brazil. The authors analyzed the psychometric properties of the instrument and reported means and standard deviations by age range
and gender for that sample (Filgueiras, Pires, Maissonette, & Landeira-fernandez, 2013). In 2015, the same group published a review
of the adaptation process, showing improved psychometric properties. However, they did not publish cut-off points for Brazilian
children (Santana, Filgueiras, & Landeira-Fernandez, 2015). Neither the ASQ-3 manual nor the questionnaires are currently available
for purchase in Brazilian Portuguese, which makes its clinical use not feasible (Squires & Bricker, 2018).
Only the AIMS has reference curves for the assessment of the gross motor domain in Brazilian children, defining percentiles curves
for sex (Gontijo, Magalhães, & Guerra, 2014; Saccani & Valentini, 2012). The majority of screening tests are still being used with the
cut-off points of the country of origin. This may lead to the inappropriate classification of delay, as the performance on the test items

79
R.S. Moreira et al. Research in Developmental Disabilities 78 (2018) 78–88

can be influenced by economic, ethnic, and cultural factors(Fernald et al., 2017; Madaschi et al., 2016; Saccani & Valentini, 2012).
Given the need for global development screening instruments, we conducted the Brazilian cross-cultural adaptation of the Survey
of Wellbeing of Young Children (SWYC), a questionnaire for developmental surveillance created and validated in the United States of
America (Moreira, Magalhães, & Alves, 2016; Perrin, Sheldrick, Visco, & Mattern, 2016). The SWYC is organized into three sub-
sections: 1) development, 2) behavior, and 3) family risk factors (depression, conflicts between parents, abuse of illicit substances,
and food insecurity). The instrument consists of short questionnaires to be answered by parents or caretakers of children aged 1–65
months, which can be applied separately or in conjunction. The SWYC is fast and easy to apply, takes on average 10 min, has the
advantage of being free of charge, and presents evidence of its validity and reliability, making its use feasible in primary health care
context (Perrin et al., 2016). The instrument, as well as the interpretation tables, are available at the website https://www.
floatinghospital.org/The-Survey-of-Wellbeing-of-YoungChildren/Translations/Portuguese-SWYC.
In this study, we intended to replicate the normative study of the SWYC’s Milestones Questionnaires for children in Brazil. Our
goals were to compare the performance of Brazilian and North American children using this screening tool and to verify the reliability
and validity of the Brazilian version.

2. Material and methods

We conducted a cross-sectional observational study involving typical children aged 1–65 months and their guardians, recruited in
Public Health Units (PHUs) from a municipality in southern Brazil (Araranguá-SC). The present study is part of a broader project that
consists of analyzing the appropriateness of all questionnaires of the SWYC to screen developmental delay, behavioral disorders, and
family risk factors in Brazilian children up to 65 months.
All children whose legal guardian authorized the participation by signing the informed consent form were included in the study.
Guardians who had already responded to the SWYC for another child and children who presented any neuromotor, sensory, or
cognitive disorder that was previously diagnosed were excluded. Preterm children were included, considering corrected age.
The sample size was defined in order to replicate the parameters used in the creation and validation of the original instrument. We
estimated including at least 30 children for each age range of the SWYC, totaling 400 children (Sheldrick & Perrin, 2013). The
children eligible for this study were identified by community health workers in each of the 15 PHUs of the municipality. The
percentage of children to be recruited in each PHU in every age group was calculated based on the distribution of children in the
municipality. Only one researcher (a pediatric physiotherapist) was responsible for data collection, which lasted from June 2014 to
April 2015. The guardians were recruited when they came to the PHU for medical or dentistry appointments, or for vaccination of
their children.
The study consisted of individual interviews lasting about 25 min, with those responsible for the child, conducted in a private
room at the PHU. The guardians responded to a questionnaire created by the researchers, composed of 31 questions addressing the
family’s socioeconomic aspects, the child’s current and past health conditions, and resources to promote the child’s development in
the family environment. Anthropometric data were collected from the child́ s records or were obtained by the PHU nurses.
The “Economic Classification of the Brazilian Association of Research Companies (ABEP)/Criterion Brazil” was applied to esti-
mate the families’ purchasing power. ABEP contains information about the possession and quantity of durable household goods, the
existence of running water, conditions of the street where the family lives, and the educational level of the head of the household
(Kamakura & Mazzon, 2015). Considering this criterion, the Brazilian population is divided into six main socioeconomic strata: A, B1,
B2, C1, C2, and, D/E, D/E being the lowest purchasing power stratum (Kamakura & Mazzon, 2015).
Guardians answered the SWYC questionnaires to finalize the interview. The Developmental Milestones section contains 10
questions about the cognitive, motor, social, and linguistic abilities for each of the age-specific forms − 2, 4, 6, 9, 12, 15, 18, 24, 30,
36, 48, and 60 months – comprising a total of 54 items (Sheldrick & Perrin, 2013). The child́ s performance on each item is scored on a
3-point scale, on which “0” is checked when the child “not yet” perform a certain task, “1” is checked when the child performs the
task “somewhat”, and “2” is checked when the child performs the task “very much”. The total score is obtained by summing the
parents’ responses to each item, using a reference table to classify if the score obtained is above or below the cut-off point for the
respective age range (Perrin et al., 2016).
The original questionnaire shows evidence of acceptable psychometric properties and can be used both for fast screening and for
surveillance of the child’s development. The concurrent validity and accuracy of the Developmental Milestones questionnaire were
estimated based on the comparison with the ASQ-3 results. All sensitivity and specificity values were above 70%, except for the age
ranges 9 (57%), 18 (67%), and 60 (63%) months for sensibility, and age range 60 (59%) months for specificity. The concurrent
validity values were above 40% in all age ranges, except at 2 months (13%) (Sheldrick & Perrin, 2013).
In order to replicate the original normative study in the Brazilian sample, the Developmental Milestones questionnaire was
applied to the age range of the child and also to the anterior and posterior age ranges. Thus, if the child was seven months old, the
guardians responded to the items for the age range of 6 months, in addition to the questionnaires for the 4- and 9-month ranges. The
interview was interrupted when guardians reported that the child did not perform three consecutive items of the age range posterior
to the one the child belonged to.
The data were stored in electronic format, double-entried, and checked subsequently. Descriptive analysis included the dis-
tribution of frequencies for categorical variables and measures of central tendency and dispersion for continuous variables.
Quantitative data were analyzed using the R statistical program (version 3.3.0). The anthropometric data were analyzed using the
Anthro software from the World Health Organization (WHO)(WHO, 2010), and the classification of nutritional status was based on
the body mass index (BMI) according to age and sex, following the protocol of the Brazilian Ministry of Health (Brasil, 2011).

80
R.S. Moreira et al. Research in Developmental Disabilities 78 (2018) 78–88

The measurement properties of the Developmental Milestones questionnaire were analyzed according to the original North
American instrument. The gradual response model of the Item Response Theory (IRT) (Andrade, Tavares, & Valle, 2000; Sheldrick &
Perrin, 2013) was used to estimate the child’s developmental age (latent trait), which is not directly observable but was inferred from
the model defined by the pattern of responses to the 54 items of the Developmental Milestones questionnaire.
As the IRT assumes that each scale measures a single construct, factor analysis was used to assess whether the items of the SWYC
Developmental Milestones questionnaire measured a unidimensional construct. Replicating the original study, the factorial analysis
was run considering the pool of 54 items of the Developmental Milestones questionnaire. The number of dimensions was estimated by
the Acceleration Factor (AF) criterion. The convergent validity of the items was evaluated according to the criterion proposed by
Fornell and Larcker (1981). An Average Variance Extracted (AVE) of greater than 50% indicates convergent validity. Cronbach's
alpha (AC) was used to verify the internal consistency of the items. Values greater than 0.70 support the reliability of the scale
(Andrade et al., 2000).
Similar to the North American version, hand-scoring means were calculated to yield the final score based on an IRT score
converted to each child’s developmental age. In the IRT model, the parameter (a) indicates the rate at which the probability of
performing an item increases with age: the higher the number, the better the discrimination. The parameter (b1) reflects the age limit
between the answers “not yet” and “somewhat”, being represented by the median age of the responses for “not yet”. The parameter
(b2) reflects the cut-off age among the options “somewhat” and “very much”, represented by the median age of the answers for “very
much”. From the estimated probability of responding “not yet”, “somewhat”, and “very much”, the hand score was calculated as the
probability of responding “not yet” x 0 + probability of responding “somewhat” x 1 + probability of responding “very much " x 2.
The sum of the expected scores for the ten items resulted in the general hand score, with the cut-off point for each age group being
rounded up to the next integer value.
The present study is part of a broader project called ‘Child development assessment and early intervention for high-risk children
and their families in Brazil’, which was approved by the Research Ethics Committee at the Federal University of Minas Gerais – UFMG
(CAAE 29437514.1.0000.5149).

3. Results

A total of 422 parents of children aged 1–65 months representing the different ages included in the SWYC were invited to
participate. Seven children were excluded, two for presenting diagnosed neuromotor/behavioral disorders, three because their
guardians had been interviewed in a different age group of the SWYC for the same child, and two because siblings of this child had
already been recruited. A total of 415 guardians effectively participated in the study. The main respondents were mothers (92%),
followed by grandmothers (6.27%), and fathers (1.69%). The sample characterization is listed in Table 1.
The result of the factorial analysis indicated that the set of items of the Developmental Milestones questionnaire presented
adequate adjustment (Kaiser-Meyer-Olkin; KMO = 0.97). According to the AF criterion, the set of items is unidimensional, because a
single factor explained most of the variance of the items (value = 39.65), while the second factor explained considerably less.
Convergent validity (VME = 0.73) and internal consistency (AC = 0.97) were adequate for the Brazilian items version.
Table 2 shows the IRT results, presenting the parameters for each item and the children’s age in months when the percentage of
answers “very much” exceeded 25%, 50%, and 75% for each item. The second column (N) shows the number of children who
answered each item. The analysis of the parameter (a) indicates that the items that had a greater degree of discrimination were the
following: “Climbs up a ladder at a playground” (item 30), “Copies sounds that you make”(item 22), “Follows directions such as
‘Come here’ or ‘Give me the ball”' (item 24), “Talks so other people can understand him or her most of the time”(item 39), “Walks
across a room without help”, (item 23) and “Names at least five familiar objects, like ‘ball’ or ‘milk”' (item 28).
The items “Names the days of the week in the correct order”(item 54), “Prints his or her name”(item 51), and “Uses words like
‘yesterday’ and ‘tomorrow’ correctly” (item 48) were those with the highest median age for the responses “not yet” (b1) and “very
much” (b2), being considered items more difficult to master. There was no response “not yet” for the items “Makes sounds that let you
know he or she is happy or upset”(item 1), “Seems happy to see you”(item 2), “Follows a moving toy with his or her eyes”(item 3) and
“Turns head to find the person who is talking”(item 4), demonstrating that these are the easier items (Table 2).
The analysis of questions 1–21 shows that 75% of the children who already performed the items “very much” were up to 12
months old. Regarding the items 22–40, 75% of children who performed these items “very much” were up to 36 months old. For
questions 41–54, the age at which 75% of the children mastered the items was greater than 40 months (Table 2).
Graph 1 shows the information function curve for the 54 items of the Developmental Milestones questionnaire according to age.
The set of items provides more information about the children’s situation between 10 and 30 months of age, while the questionnaire
provides less information about the development of children younger than 3 months and older than 60 months.
Graphs 2 and 3 show, respectively, the comparison of the age at which 25% and 75% of the Brazilian and North American
children (Sheldrick & Perrin, 2013) master each item of the Developmental Milestones Questionnaire based on the IRT model. The
numbers in the X-axis correspond to the number of the items shown in Table 2. The performance of both groups was quite similar
when children were younger than 36 months. In this age range, Brazilian performed the following items earlier than North American
children: “Looks around when you say things like ‘Where's your bottle?’ or ‘Where's your blanket?’ (item 21), ‘Copies sounds that you
make’ (item 22), “Follows directions like ‘Come here’ or ‘Give me the ball’(item 24), ‘Walks up stairs with help’ (item 26), “Climbs up
a ladder at a playground” (item 30), “Says his or her first name when asked” (item 37) and “Washes and dries hands without help”
(item 40). Above 36 months, North American children mastered almost all items earlier than Brazilian ones. In the present study, the
estimated developmental age in items 48 and 54 were not shown in the graphs because values were too much high and were

81
R.S. Moreira et al. Research in Developmental Disabilities 78 (2018) 78–88

Table 1
Childreńs characteristics and socioeconomic and demographic profile of the families included in the sample, Araranguá, 2016.
Variables n % Median (range)

Child’s age (months) 18,06 (1–65.81)

< 12 141 33,97
12 to 24 106 25,53
25–36 68 16,38
37–48 40 9,63
49–60 41 9,87
> 61 19 4,57
Sex
Male 213 51,33 –
Preterm
< 37 weeks 40 9.64 39 (27–42)
Birth weight (grams)
< 2.500 g 32 7.71 3.275 (825–4.660)
Nutritional state (BMI/age)*
Thinness 57 13.91 –
Adequate 246 60.00
Overweight/Obese 107 26.09
Children breastfed exclusively until six months** 149 35.90 –
Mother’s age at the child’s birth (years)
≤19 65 15.66 25 (14–46)
Mother’s educational level (years) 10 (0–18)
0–4 22 5.30
5–8 128 30.84
9–11 226 54.46
12–15 39 9.40
Marital Status –
Married or stable relationship 343 82.65
Divorced or Single mothers 45 10.84
Others 27 6.51
Number of children 2 (1–6)
1 191 46.02
2 123 29.64
≥3 101 24.35
Participates in the income transfer program (Bolsa Família) –
Yes 58 13.98
Economic classification (ABEP) –
A1 + A2 5 1.20
B1 + B2 119 28.67
C1 + C2 249 60.00
D/E 42 10.13
Per capita income (American dollars) 144.44 (16.66–1.000)
≤ 123 167 40.24
123≤ to ≤ 246 187 45.06
> 246 61 14.69
Childcare –
Exclusively by the Mother 259 62.40
Shared with other relatives 156 37.59
Concerns regarding child development –
Yes 96 23.13
Childhood Education attendance –
Yes 135 32.53
Total 415 100 –

* 5 children without information; ** 74 children excluded because they had less than six months and 192 children excluded because they were not
on exclusive breastfeeding until six months, − Not applied.

considered outliers.

4. Discussion

The present study intended to replicate the process of creation and validation of the Developmental Milestones Questionnaire of
the SWYC (Sheldrick & Perrin, 2013). The unidimensionality of the items of the Brazilian version of the Developmental Milestones
questionnaire was confirmed by factor analysis, and its psychometric qualities were preserved in the Brazilian Portuguese version,
achieving acceptable convergent validity and internal consistency.
The families were recruited in Araranguá (SC), a municipality in the south of Brazil that has a Human Development Index (0.76)
similar to the Brazilian average (0.72) (BRASIL, 2010). The public health care system provides universal access in Brazil, and primary

82
R.S. Moreira et al. Research in Developmental Disabilities 78 (2018) 78–88

Table 2
Parameters of the Gradual Response Model of the Item Response Theory and age when 25%, 50% and 75% of the Brazilian children achieve SWYC’s
milestones (in months), Araranguá, 2016.
Items N IRT Parameters Age when Brazilian children achieve SWYC’s milestones
(months)

a b1 b2 25% 50% 75%

1. Makes sounds that let you know he or she is happy or upset 74 1 – 1 1 1 2

2. Seems happy to see you 74 2 – 1 0 1 1
3. Follows a moving toy with his or her eyes 74 1 – 0 0 0 0
4. Turns head to find the person who is talking 74 1 – 0 0 0 1
5. Holds head steady when being pulled up to a sitting position 74 2 2 3 2 3 4
6. Brings hands together 74 2 2 2 2 2 3
7. Laughs 74 3 1 2 2 2 3
8. Keeps head steady when held in a sitting position 74 3 1 3 3 3 4
9. Makes sounds like “ga”, “ma” and “ba” 108 2 3 4 3 4 5
10. Looks when you call his or her name 108 2 2 4 3 4 6
11. Rolls over 108 3 4 6 4 6 7
12 Passes a toy from one hand to the other 108 3 4 5 4 5 6
13. Looks for you or another caregiver when upset 107 2 3 4 3 4 5
14. Holds two objects and bangs them together 78 3 5 5 4 5 6
15. Holds up arms to be picked up 141 4 5 7 6 7 8
16. Gets into a sitting position by him or herself 139 3 7 8 6 8 10
17. Picks up food and eats it 135 4 6 7 6 7 8
18. Pulls up to standing 120 3 7 7 6 7 9
19. Plays games like “peek-a-boo” or “pat-a-cake” 106 4 6 7 6 7 8
20. Calls you “mama” or “dada” or similar name 137 3 7 9 7 9 11
21. Looks around when you say things like “Where's your bottle?" or 133 3 7 8 7 8 10
“Where's your blanket?"
22 Copies sounds that you make 126 4 8 11 9 11 13
23. Walks across a room without help 120 4 11 11 10 11 13
24. Follows directions − like “Come here” or “Give me the ball” 118 4 8 9 8 9 11
25. Runs 177 4 13 15 13 15 18
26. Walks up stairs with help 169 4 11 12 10 12 14
27. Kicks a Ball 133 3 14 16 13 16 19
28. Names at least 5 familiar objects- like ball or milk 126 4 15 15 13 15 18
29. Names at least 5 body parts- like nose, hand, or tummy 161 3 19 21 17 21 26
30. Climbs up a ladder at a playground 143 5 15 16 14 16 18
31. Uses words like “me” or “mine” 138 4 19 21 17 21 25
32. Jumps off the ground with two feet 135 4 17 20 16 20 24
33. Puts 2 or more words together − like “more water” or “go outside” 127 4 19 22 18 22 26
34. Uses words to ask for help 112 4 19 21 18 21 25
35. Names at least one color 108 3 28 30 25 30 37
36. Tries to get you to watch by saying “Look at me” 105 3 20 22 18 22 26
37. Says his or her first name when asked 104 4 20 22 18 22 26
38. Draws lines 83 3 24 29 24 29 36
39. Talks so other people can understand him or her most of the time 120 4 24 31 26 31 36
40. Washes and dries hands without help (even if you turn on the 114 3 24 26 22 26 32
water)
41. Asks questions beginning with “why” or “how” − like “Why no 110 3 31 35 28 35 43
cookie?"
42. Explains the reasons for things, like needing a sweater when it is 101 3 29 37 29 37 46
cold
43. Compares things − using words like “bigger” or “shorter” 130 3 33 39 30 39 50
44. Answers questions like “What do you do when you are cold? “or 117 3 31 35 28 35 45
"…when you are sleepy?"
45. Tells you a story from a book or tv 147 3 32 41 33 41 52
46. Draws simple shapes − like a circle or a square 145 3 35 43 35 43 54
47. Says words like “feet” for more than one foot and “men” for more 144 2 44 59 44 59 78*
than one man
48. Uses words like “yesterday” and “tomorrow” correctly 134 2 54 89* 60 89* 132*
49. Stays dry all night 126 2 36 44 33 44 59
50. Follows simple rules when playing a board game or card game 122 3 47 65* 51 65* 82*
51. Prints his or her name 118 3 68 75* 58 75* 97*
52. Draws pictures you recognize 109 3 54 69* 55 69* 87*
53. Stays in the lines when coloring 88 3 49 61* 49 61* 77*
54. Names the days of the week in the correct order 79 2 125* 180* 122* 180* 266*

Note: “a”: rate of item discrimination; “b1”: median age of responses “not yet” and “b2”: median age of responses “very much”. * Expected age based
on the IRT model.

83
R.S. Moreira et al. Research in Developmental Disabilities 78 (2018) 78–88

Graph 1. Information function curve for the 54 items of the Developmental Milestones questionnaire of the SWYC according to the children’s ages.

Graph 2. Comparison of the age when 25% of the Brazilian and North American* children perform each item of the Developmental Milestones
Questionnaire based on the IRT model.
Note: * SHELDRICK; PERRIN, 2013

health care units cover over 80% of Araranguá’s population(Brasil, 2017). The recruitment process included families from different
socioeconomic backgrounds, as we stratified the sample considering the number of families under the responsibility of each primary
health care team. The majority of the families recruited had a per capita income of less than the minimum wage, which is consistent
with official data that characterize the majority of the Araranguá households as belonging to the classes C1, C2, and D/E. In 2010, the
average number of school years of Araranguá́s population was 11 years (SEBRAE/SC, 2013), similar to the average duration of
maternal education found in the present study.

84
R.S. Moreira et al. Research in Developmental Disabilities 78 (2018) 78–88

Graph 3. Comparison of the age when 75% of the Brazilian and North American* children perform each item of the Developmental Milestones
Questionnaire based on the IRT model.
Note: * SHELDRICK; PERRIN, 2013

We excluded the main clinical conditions that would potentially impair the children’s development and, although poverty is a
known risk factor for developmental delays (Black et al., 2017), the children enrolled in this study were considered healthy. Sheldrick
& Perrin (2013) also included children and families from primary and community care units in urban and suburban settings to assure
diversity in the original study (Sheldrick & Perrin, 2013).
Just as it occurs with most developmental tests, the information function curve has shown that this questionnaire is more con-
sistent at intermediate age ranges and provides little information on the development of children under 3 and over 60 months of age.
Drachler et al. (2007) found similar results when using IRT to analyze the Denver-II in Brazilian children, showing that the scale was
more symmetric and homogeneous between 3 and 59 months of age (Drachler et al., 2007).
The analysis of the parameters of the IRT model showed that the questionnairés most discriminating items are related to language,
cognition, and gross motor domains, at ages between 12 and 36 months. Guedes et al. (2011), when investigating the validity of the
Bayley-III screener in Brazilian preterm children, also found that items included in the age range from 12 to 24 months such as
“Copies words”, “Walks alone”, or “Names three objects” were the most discriminating (Guedes, Primi, & Kopelman, 2011).
It is likely that Brazilian, in comparison to North American children, showed greater mastery on gross motor items because these
aspects are valued culturally, being required for independence and integration into society. The language, cognitive, and fine motor
items presented greater difficulty for Brazilian children, and, as expected, they appeared in the latest age ranges, supporting the
consistency of the questionnaire. These items represent skills that are almost always developed in school environments. The fact that
only a small proportion of children was attending educational programs when the parents were interviewed may have contributed to
the results presented here.
It is interesting to observe that, while in the first 36 months the performance of Brazilian and North American children was quite
similar, the differences between the samples became bigger as the children got older. Although the Developmental Milestones
Questionnaire offers a score that represents global child development, the majority of the items that North American children
mastered earlier than Brazilian children are included in the language and cognitive domains of the Bayley III and ASQ-3 scales
(Bayley, 2005; Squires, Bricker, & Potter, 2009).
A Brazilian study showed a prevalence of 50.7% of language disorders in children under 5 years of age in public schools, with
children under 5 years being 2.5 times more likely to fail in language assessments than older children (5–10 years old) (Rabelo et al.,
2015). Basílio, Puccini, Silva, and Pedromônico (2005), assessing the receptive language of 201 healthy children from the state of São
Paulo (Brazil) with the Peabody Picture Vocabulary Test, found that 44.3% presented scores below their age expectation. Only 42%

85
R.S. Moreira et al. Research in Developmental Disabilities 78 (2018) 78–88

of the children aged 2–5 years had been attending childcare centers for 6 months, and there was a positive correlation between the
Peabody test, the mothers’ educational level, and the families’ purchasing power. The authors argue that mothers with higher
educational levels contribute to the familýs income, thus offering better schools for their children, highlighting the protective and
preventive role of infant education for global developmental delay, but mainly in the language domain (Basílio et al., 2005).
Parents’ low educational level as well as lack and/or low quality of infant educational programs in Brazil might have contributed
to the poorer performance of children at the oldest ages ranges in this study. Comparing our sample with the Sheldrick and Perin
(2013) study, about 66% of the North American parents had finished high school and had some level of college education, versus only
9.4% of the mothers interviewed in the present study(Sheldrick & Perrin, 2013). In the Brazilian sample, although 54.5% of the
mothers finished high school, they studied on average for only 10 years.
Unlike the North American children, the majority of children included in this study did not attend educational programs (Rhodes
& Huston, 2012). In the present study, only 33% of children were attending childhood education, and this is not very different from
the general situation in Brazil. According to official data, in 2015, only 25,6% of the Brazilian children under 3 years of age were
attending childhood education programs(IBGE, 2017) The quality of childhood education in Brazil is also quite poor, as shown by
research funded by the Ministry of Education, in 2009, to assess the quality of childhood education in six states of the country,
involving 147 daycare centers and preschools (Campos, Esposito, Bhering, Gimenes, & Abuchaim, 2011). Using the Infant/Toddler
Environment Rating Scale-Revised Edition (ITERS) and the Early Childhood Environment Rating Scale-Revised Edition, the general
average on the ITERS was 3.4/10, showing a basic quality of daycare centers. The worst scores were found for the items activities and
program structure, with a lack of musical activities and materials like blocks and books. The authors concluded that, although the
coverage of childhood education in Brazil has expanded considerably, much effort is still necessary to assure appropriate environ-
ments to promote early child development (Campos et al., 2011).
The other side is what happens to children who do not attend childcare centers. The study “Primeiríssima Infância – creches:
necessidades e interesses de famílias e crianças” interviewed 991 caregivers of children up to 3 years of age all around the country
(FMCSV, 2017). About 30% of the children were attending childcare centers, 12% also stayed part of the day in a house other than
their home, and 57% stayed at home. Many children who stayed at home were cared for by lay woman, relatives (grandmothers,
siblings, etc.), or, alternatively, their mothers split their time between other domestic functions and childcare. The study showed that
children who stay at home have fewer opportunities of playing outside, manipulating books, drawing, and playing with adults and
other children (FMCSV, 2017).
The picture seems to be a mix of sociocultural patterns and some lack of appropriate stimulation, imputing differences in the
performance of Brazilian children when compared to the original SWYC sample. These exploratory results raise the question of
assuming or not the original cut-off for developmental delay suspicion when assessing Brazilian children. Authors investigating other
instruments found the need to establish new cut-off points for the Brazilian population (Gontijo et al., 2014; Magalhães, Fonseca,
Martins, & Dornelas, 2011; Saccani & Valentini, 2012). Magalhães et al. (2011) analyzed the performance of Brazilian preterm
children with very and extremely low weight on the Denver-II scale (Magalhães et al., 2011). In the first year of life, the Brazilian
preterms showed superior performance in some areas when compared with the normative North American sample, suggesting the
need for further studies of the validity of the test for the Brazilian population. Saccani and Valentini (2012) found that motor
acquisition in Brazilian children was usually lower than in Canadian children, using the AIMS (Saccani & Valentini, 2012). Gontijo
et al. (2014) obtained similar results, observing lower performance in Brazilian compared to Canadian children on the 5th and 10th
percentile of the AIMS (Gontijo et al., 2014). The authors of both studies concluded that the observed differences could be attributed
to sociocultural diversity and its influence on parental practices (Gontijo et al., 2014; Saccani & Valentini, 2012).
Capovilla & Capovilla (1997), when conducting the Brazilian translation and standardization of the Peabody Picture Vocabulary
Test and Language Development Survey also made changes to the cut-off points in both instruments, understanding that the ex-
pressive and receptive vocabulary is influenced by economic and social levels and by parentś educational level (Capovilla &
Capovilla, 1997). In the cognitive area, Nascimento and Figueiredo (2002), when creating the cross-cultural adaptation of the
Wechsler Intelligence Scale for Children (WISC-III), reported adjustments to the content, order, time, and granting bonuses to scores,
as well as modifications of the age ranges in the normative tables (Nascimento & Figueiredo, 2002).
A systematic review showed that of 64 translations of instruments held in different countries, only in 24 cases data was compared
to the original version, and of these, 16 cases showed evidence of the need to change the cut-off point. The authors argue that,
although changes in the rules to interpret the results are often recommended, due to linguistic and cultural differences, very few
translated instruments went through re-standardization (El-Behadli, Neger, Perrin, & Sheldrick, 2015). National studies also show the
importance of screening for early identification of developmental delay in Brazilian children (Guedes et al., 2011; Madaschi et al.,
2016; Saccani & Valentini, 2012). In the Brazilian context, the Developmental Milestones questionnaire has many advantages. As it is
included in the SWYC that offers a comprehensive approach of the developmental and behavioral disorders and also includes family
risk factors. It has no cost, is fast, and can be implemented in primary health care services, without any special training or kit of
materials. Primary health care professionals are those who can better monitor and understand the developmental trajectory of their
patients, as they maintain regular and longitudinal contact with families (Aly et al., 2010). The use of developmental screening
instruments in primary care can contribute to Brazilian population-based studies; it is therefore necessary to subsidize child health
care policies.
This study, however, points out some weakness of the SWYC, that had not yet demonstrated its accuracy compared to diagnosis
tests. The Milestones Questionnaires seem to be useful only for children between 4 and 58 months, with item discrimination not
supporting the use of the full range of the age-specific forms. Like most developmental assessments, it loses its discriminative power
at the extremes of the scale (Perrin et al., 2016; Sheldrick & Perrin, 2013). It must also be noted that our sample had size limitations

86
R.S. Moreira et al. Research in Developmental Disabilities 78 (2018) 78–88

and that further studies are necessary, with samples representative of the Brazilian diversity, to define cut-off scores and investigate
the accuracy of the instrument as a screening tool for developmental delay in Brazilian children.

5. Conclusion

The Brazilian version of the Developmental Milestones questionnaire presented acceptable measurement qualities that supports
the SWYĆs potential as a developmental screening tool. We found important differences between the North American original sample
and the Brazilian one in achieving the milestones, especially among the oldest children, which claims for broader studies to better
understand whether these results reveal a sociocultural pattern or are related to methodologic issues. Due to socioeconomic and
cultural diversity in Brazil, future studies should expand the sample and include children from other regions of the country as well as
of all social strata. It is also important to emphasize that further validity and accuracy studies of the SWYC in Brazil as well as in the
United States of America are needed, especially regarding its usefulness to screen developmental delay. As the Developmental
Milestones questionnaire is only a section of the SWYC, the use of the full instrument will provide a better estimate of overall
development.

Funding source

- FAPESC (Fundação de Amparo à Pesquisa e Inovação do Estado de Santa Catarina) [grant number 2015TR328]
- Grand Challenges Canada (grant number 582-03)

Acknowledgement

We are grateful to FAPESC and Grand Challenges Canada for funding the research and to Pró-Reitoria de Pesquisa da
Universidade Federal de Minas Gerais for supporting the publishing process.

References

Aly, Z., Taj, F., & Ibrahim, S. (2010). Missed opportunities in surveillance and screening systems to detect developmental delay: A developing country perspective.
Brain and Development, 32(2), 90–97.
De Andrade, D. F., Tavares, H. R., & Valle, R. D. C. (2000). Teoria da Resposta ao Item: Conceitos e Aplicações. Simpósio Nacional de Probabilidade e Estatística (SINAPE),
1, 109–121.
BRASIL (2010). Instituto Brasileiro de Geografia e Estatística-IBGE. [Retrieved December 23, 2015, from] http://www.cidades.ibge.gov.br.
Basílio, C. S., Puccini, R. F., da Silva, E. M. K., & Pedromônico, M. R. M. (2005). Living conditions and receptive vocabulary of children aged two to five years. Rev
Saúde Pública, 39(5), 725–730.
Bayley, N. (2005). Bayley Scales of Infant Development, 3nd edition (Bayley-III) (3rd ed.). San Antonio: Pearson.
Black, M. M., Walker, S. P., Fernald, L. C. H., Andersen, C. T., DiGirolamo, A. M., Lu, C., & Grantham-McGregor, S. (2017). Early childhood development coming of age:
Science through the life course. The Lancet, 389(10064), 77–90.
Brasil (2011). Orientações para a coleta e análise de dados antropométricos em serviços de saúde: Norma Técnica do Sistema de Vigilância Alimentar e Nutricional − SISVAN/
Básica (1 st ed.). Brasília: Secretaria de Atenção à Saúde. Departamento de atenção Básica.
Brasil (2017). e-Gestor- Informação e Gestão da Atenção Básica. [Retrieved February 20, 2018, from] https://egestorab.saude.gov.br/paginas/acessoPublico/relatorios/
relHistoricoCoberturaAB.xhtml.
Campos, M. M., Esposito, Y. lúcia, Bhering, E., Gimenes, N., & Abuchaim, B. (2011). A qualidade da educação infantil: um estudo em seis capitais Brasileiras. Cadernos
de Pesquisa, 41(142), 20–54. http://dx.doi.org/10.1590/S0100-15742011000100003.
Capovilla, F. C., & Capovilla, A. G. S. (1997). Desenvolvimento Linguistico na criança dos dois aos seis anos: Tradução e Estandardização do Peabody Picture
Vocabulary Test de Dunn & Dunn, e da Language Development Survey de Rescorla Ciência Cognitiva: Teoria. Pesquisa E Aplicação, 1(1), 353–380.
Dosman, C. F., Andrews, D., & Goulden, K. J. (2012). Evidence-based milestone ages as a framework for developmental surveillance. Paediatr Child Health, 17(10),
561–568.
Drachler, M. D. L., Marshall, T., & de C Leite, J. C. (2007). A continuous-scale measure of child development for population-based epidemiological surveys: A
preliminary study using item response theory for the Denver test. Paediatric and Perinatal Epidemiology, 21(2), 138–153.
El-Behadli, A. F., Neger, E. N., Perrin, E. C., & Sheldrick, R. C. (2015). Translations of developmental screening instruments: An evidence map of available research.
Journal of Developmental and Behavioral Pediatrics, 36(6), 471–483.
FMCSV (2017). Primeiríssima infância-Creche: necessidades e interesses de famílias e crianças (1 st ed.). [São Paulo].
Fernald, L. C. H., Kariger, P., Engle, P., & Raikes, A. (2009). Examining early child development in low-income countries: A toolkit for the assessment of children in the first five
years of life. In The World Bank1–133.
Fernald, L. C. H., Prado, E., Kariger, P., & Raikes, A. (2017). A toolkit for measuring early childhood development in low-and middle-income countries. (The world bank, ed.)
(2nd ed.). Washington: The World Bank.
Filgueiras, A., Pires, P., Maissonette, S., & Landeira-fernandez, J. (2013). Early Human Development Psychometric properties of the Brazilian-adapted version of the
Ages and Stages Questionnaire in public child daycare centers. Early Human Development, 89(8), 561–576.
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1),
39–50.
Glascoe, F. P. (2015). Evidence-Based early detection of developmental-Behavioral problems in primary care: What to expect and how to do it. Journal of Pediatric
Health Care, 29(1), 46–53.
Gontijo, A. P. B., de C. Magalhães, L., & Guerra, M. Q. F. (2014). Assessing gross motor development of Brazilian infants. Pediatric Physical Therapy, 26(1), 48–55.
Guedes, D. Z., Primi, R., & Kopelman, B. I. (2011). BINS validation – Bayley neurodevelopmental screener in Brazilian preterm children under risk conditions. Infant
Behavior and Development, 34(1), 126–135.
Hagan, J., Shaw, J., & Duncan, P. (2017). Bright futures: Guidelines for health supervision of infants, children, and adolescents [pocket guide]. (American Academy of
Pediatrics, ed.) (4th ed.). Elk Grov Village: American Academy of Pediatrics.
IBGE (2017). Síntese de Indicadores Sociais-uma análise das condoções de vida da população brasileira (1 st ed.). Rio de Janeiro: Estudos e Pesquisas.
Kamakura, W., & Mazzon, J. A. (2015). Associação Brasileira de Empresas e Pesquisa-ABEP. [Retrieved December 23, 2015, from] http://www.abep.org/.
Madaschi, V., Mecca, T. P., Macedo, E. C., & Paula, C. S. (2016). Bayley-III scales of infant and toddler development: Transcultural adaptation and psychometric
properties. Paidéia (Ribeirão Preto), 26(64), 189–197.

87
R.S. Moreira et al. Research in Developmental Disabilities 78 (2018) 78–88

Magalhães, L. D. C., Fonseca, K. L., Martins, L. D. T. B., & Dornelas, L. D. F. (2011). Desempenho de crianças pré-termo com muito baixo peso e extremo baixo peso
segundo o teste Denver-II. Revista Brasileira De Saúde Materno Infantil, 11(4), 445–453.
Moreira, R. S., de C. Magalhães, L., & Alves, C. R. L. (2016). Triagem de Atraso de Desenvolvimento e de Alterações de Comportamento: Estudo Normativo do Survey of
Wellbeing of Young Children (SWYC) no contexto brasileiro. Universidade Federal de Minas Gerais.
Nascimento, E., & Figueiredo, V. L. M. (2002). WISC-III e WAIS-III: alterações nas versões originais americanas decorrentes das adaptações para uso no Brasil.
Psicologia: Reflexão E Crítica, 15(3), 603–612. http://dx.doi.org/10.1590/S0102-79722002000300014.
Perrin, E. C., Sheldrick, C., Visco, Z., & Mattern, K. (2016). The survey of well-being of young children (SWYC) user’s manual (1.01). Boston: Center, Tufts Medical.
[Retrieved from] www.theSWYC.org.
Pinto, E. B., Vilanova, L. C. P., & Vieira, R. M. (1997). O desenvolvimento do comportamento da criança no primeiro ano de vida (1 st ed). São Paulo: Casa do Psicólogo.
Rabelo, A. T. V., Campos, F. R., Friche, C. P., Da Silva, B. S. V., Friche, A. A. D. L., Alves, C. R. L., & Goulart, L. M. H. D. F. (2015). Speech and language disorders in
children from public schools in Belo Horizonte. Revista Paulista De Pediatria, 33(4), 453–459.
Rhodes, H., & Huston, A. (2012). Building the Workforce Our Youngest Children Deserve. Social Policy Report. Society for Research in Child Development, 26(1), 1–32.
SEBRAE/SC (2013). Araranguá em Números (1 st ed.). Florianópolis: Sebrae.
Sabanathan, S., Wills, B., & Gladstone, M. (2015). Child development assessment tools in low-income and middle-income countries: How can we use them more
appropriately? Archives of Disease in Childhood, 100(5), 482–488.
Saccani, R., & Valentini, N. C. (2012). Reference curves for the Brazilian Alberta Infant Motor Scale: Percentiles for clinical description and follow-up over time. Jornal
De Pediatria, 88(1), 40–47.
Santana, C. M. T., Filgueiras, A., & Landeira-Fernandez, J. (2015). Ages &amp; Stages Questionnaire-Brazil-2011: Adjustments on an Early Childhood Development
Screening Measure. Global Pediatric Health, 2(0), 1–12.
Sheldrick, R. C., & Perrin, E. C. (2013). Evidence-based milestones for surveillance of cognitive, language, and motor development. Academic Pediatrics, 13(6),
577–586.
Sheldrick, R. C., Merchant, S., & Perrin, E. C. (2011). Identification of developmental-behavioral problems in primary care: A systematic review. Pediatrics, 128(2),
356–363.
Squires, J., & Bricker, D. (2018). Ages & stages questionnaire. [Retrieved February 14, 2018, from] http://agesandstages.com/languages/#silk-tabs-0-1.
Squires, J., Twombly, E., Bricker, D., & Potter, L. (2009). Ages & stages questionnaires, (ASQ-3) user’s guide (3rd ed.) (3rd ed.). Baltimore: Brookes Publishing.
WHO (2010). WHO Anthro for personal computers, version 3.2.2, 2011: Software for assessing growth and development of the world’s children. Geneva: WHO.

88