Validity of the Greek Knowledge About Childhood Autism Among Health Workers (KCAHW) Questionnaire

Abstract

The study investigates the reliability and validity of the Greek version of the knowledge about childhood autism among health workers (KCAHQ) questionnaire. A total of 541 allied healthcare participants (n1 = 471 students; n2 = 70 professionals) provided their socio-demographic variables and completed the KCAHQ. Analysis showed overall floor-ceiling effects lower than 15% and skewness-kurtosis values between ±2. The internal consistency was good (Kuder–Richardson 21 = 0.80) and all domains were significantly correlated with each other, with Spearman’s coefficients ranging from 0.26 to 0.57. The split-half reliability was satisfactory, with the Spearman–Brown and lambda 4 coefficients equal to 0.618 and 0.613, respectively. All the goodness of fit indices generated by confirmatory factor analysis were acceptable (CFI = 0.88; TLI = 0.861, RMSEA = 0.052). Being a professional, having personal contact with an individual with autism, and having received autism-specific training significantly increased scores on the KCAHW. Overall, the Greek KCAHQ seems reliable and valid; however, domain 4 presents insufficient internal consistency.

1. Introduction

Autism Spectrum Disorder (ASD) is a lifelong neurodevelopmental syndrome that is characterized by social communication deficits in different contexts, stereotyped, perseverative behaviors, adherence to routines, and restricted interests [1].

While autism was a rare condition in the early 1970s, later epidemiological studies have steadily reported an increase in the prevalence of the autistic condition, indicating a mean worldwide prevalence estimate between 0.72% and 1% [2,3]. Low-income countries present lower prevalence rates than middle-income or high-income countries, while for several low or middle-income regions of the world, the prevalence estimates of ASD are still missing [3,4]. These differences in the prevalence of ASD possibly reflect the accessibility to diagnostic and support services offered by the healthcare system of each country in the affected population [2]. In Greece, which is a middle-income European country, epidemiological data for children under 17 show similar prevalence estimates (between 0.94% and 1.15%) with those documented globally [5,6], and even though significant differences are noted between the 13 different regions of the country, these are not associated with the difficulty of access to diagnostic services [6].

The dramatic growth in the global prevalence of ASD since the late 1900s mirrors multiple environmental, biological, and social factors, but it is not caused by a true increase in the prevalence of autism [2,3]. It is rather associated with changes in the diagnostic criteria in psychiatric classification systems; the design and standardization of specific screening and diagnostic tools for ASD; public awareness; higher levels of knowledge and education of healthcare workers, which permits the identification of more mild cases of autism; easier access to diagnostic services; and the more accurate identification and diagnosis of the autistic condition, especially when it coexists with other neurodevelopmental disorders [2,3]. Finally, it is also attributed to improvements in methodology and the quality of research in this field [3].

1.1. The Role of Health Professionals in Early Identification and Diagnosis of Autism

Despite increased global awareness and knowledge of autism, a significant gap still exists between the time parents first express concerns about their child’s development and the official diagnosis of ASD [7]. Most children with ASD are diagnosed at 3 to 4 years old, even though parents often notice the first signs of abnormal behavior as early as 6 months [7,8]. Howling and Moore [9] documented that the period leading up to the initial diagnosis is not only stressful for families, but also a minefield. Less than 10% of the children reported to a general practitioner, health visitor, or pediatrician for atypical behavior received an ASD diagnosis during their first referral. Despite this 59.6% still missed a formal diagnosis in their second referral, even though 42.3% of the families consulted three or more different professionals in the second assessment, and over a quarter were referred again to another specialist. Moreover, parents often consult allied health professionals such as speech-language pathologists (SLPs), psychologists, occupational therapists, or physical therapists before visiting a physician. These professionals may be the first to observe atypical communicative, social, kinetic, and/or sensory behaviors in the child, making them crucial in recognizing autistic traits and initiating further assessment and diagnosis [10].

Furthermore, children with autistic syndrome present high comorbidity with organic and neurological disorders, such as epilepsy, chronic constipation, and infections. They typically experience chronic pain and lower levels of general health, leading to frequent short hospital stays [11,12]. During these stays, general doctors, practitioners, and pediatric nurses must manage possible sensory and emotional crises of undiagnosed children, address their organic needs, and provide consultation to primary caregivers about their child [12,13,14]. These consultations often provide an opportunity for caregivers to express their concerns about their child’s developmental and behavioral issues. However, these concerns are frequently downplayed or dismissed by healthcare professionals in primary care settings [11]. To make matters worse, Dosreis and colleagues [15] found that only a small percentage of licensed pediatricians (8%) systematically screen children for autism, despite recommendations for a first-level assessment for ASD in all children referred to physicians at 18 and 24 months [16,17]. The reported reasons for this omission included a lack of time, unfamiliarity with screening tools, limited staffing, and inadequate reimbursement [10,15].

1.2. Knowledge of Healthcare Professionals About ASD

A recent review of healthcare professionals’ knowledge of ASD revealed only moderate levels of autism knowledge and lack of training [14]. The professional background seems to significantly impact knowledge about ASD, even though the lack of standardized golden standard tool(s) hinders direct comparison(s) between different studies [14]. Among the healthcare professionals, primary healthcare caregivers (general practitioners, physicians, nurses, family practitioners, pediatricians) exhibited lower levels of knowledge about ASD [14] than children and adolescent mental health specialists (CAMH), and allied healthcare workers (SLPs, occupational therapists, physical therapists) [10,18]. Generally, pediatric providers’ knowledge was higher among those who had personal experience with autism (e.g., having autistic friends or family) or had participated in autism-specific education or training programs. Allied healthcare professionals reported attending more continuing education training about ASD than physicians [10] a fact that possibly explains the difference in ASD knowledge between these healthcare professionals. Finally, the findings indicated that knowledge about ASD improved with years of experience working with autistic clients [14].

1.3. The Implications of a Late Diagnosis of ASD

A late diagnosis of ASD has a significant long-term influence on the development of individuals with ASD, as well as on levels of parental stress and acceptance [9]. To begin with, poor knowledge and training among healthcare professionals impedes the early identification and accurate diagnosis of autism, leading to delayed access to early intervention services for toddlers with ASD. This delay is critical because early intervention programs are most effective when implemented before the age of three, resulting in better outcomes later in life, including better cognitive, verbal, and adaptive skills in school [7,18,19]. Moreover, the period of first diagnosis is a stressful period for the whole family with parents having feelings of despair and self-blame which often lead to the deterioration of their mental health and inability to perform correctly their parental role [19]. The prolonged period of the first diagnosis and an unclear first diagnosis has been related to feelings of dissatisfaction which negatively affect parents’ views of healthcare professionals and of services provided [9]. This loss of trust in diagnostic and supportive services may negatively impact the acceptance of the ASD diagnosis [9]. Finally, insufficient knowledge and misconceptions about ASD from pediatric professionals degrade the quality of the provided family consultation services [18].

The above findings highlight two key points: first, the need for all pediatric providers to be familiar with ASD screening instruments as recommended by the American Academy of Pediatrics, to ensure accurate and timely identification of autism, which leads to cost-effective diagnosis [17]. Second, they underscore the importance of specialized training for healthcare professionals to deliver appropriate services to children with autism and their families [17].

1.4. Autism Knowledge Measures

Although knowledge of healthcare professionals about autism is crucial for delivering effective services to children with ASD and their families, it has only recently gained attention in terms of designing proper instruments to measure this knowledge. An international review identified 44 distinct measurements, most of which are not standardized, but are used in independent surveys for specific studies [20]. Among these measures, few can be considered well-established assessments. Only a small number have been used by two independent research groups in published studies and are noted for acceptable psychometric properties, such as validity and reliability [20,21]. Regarding cross-cultural adaptations of standardized autism knowledge measures, while several studies reported formal translation and back-translation procedures, very few detailed the full processes involved in cross-cultural adaptation [20]. The KCAHW questionnaire [22] is one of the most commonly used standardized measures, presenting good psychometric properties [14,20]. It has been used to assess knowledge about ASD in different samples of healthcare professionals (general practitioners, physicians, nurses, midwives, psychiatrists, CAMH) and healthcare students (psychology students, nursing students, medicine students) in different countries (Nigeria, Pakistan, Italy, Sri Lanka, Turkey, Ghana, Saudi Arabia, India) [13,14]. The original tool, which is a 19-item scale, demonstrates excellent internal consistency with a Cronbach alpha value equal to 0.97 and an overall test-retest reliability equal to 0.99 [22]. It has been adapted and validated only for Turkey in two independent studies, showing acceptable internal consistency with Kuder–Richardson, with 21 and 20 coefficient values equal to 0.683 and 0.70, respectively [23,24]. Furthermore, it demonstrated good split-half reliability with a Guttman split-half value equal to 0.84 [23] and an adequate test-retest reliability with an intraclass correlation coefficient value equal to 0.83 [23]. In Gürbüz-Özgür et al.’s [23] study, domain 4 presented insufficient internal consistency, while Ozdemir et al. [24] reported acceptable psychometric properties for a shorter version of the scale, comprising only 12 items. Finally, according to one of the Turkish validation studies, factors positively affecting performance on the KCAHW questionnaire were previous ASD-specific training, contact with someone diagnosed with autism, mental health clinic experience, and working experience with ASD individuals [23].

1.5. The Present Study

Healthcare professionals’ knowledge of ASD significantly influences the level of supportive services provided to autistic people and their families, thereby impacting the quality of their lives [10,17,18,19]. For this reason, it is vital to assess healthcare professionals’ knowledge using well-designed instruments that present adequate psychometric properties. Relevant research suggests that the best practice involves identifying specific measures as the gold standard for ASD knowledge and examining their psychometric qualities in replication studies in different samples and different cultural environments. This approach will ensure the construct equivalence of tools in different cultural contexts and will determine the possible adjustments required [20,25]. Hence, the current cross-sectional study aims to investigate the validity and reliability of the Greek version of the KCAHW questionnaire. Even though this tool has been used in different samples and different countries, most previous studies used small samples, and no psychometric information is reported except for the Turkish adaptation of the tool.

2. Materials and Methods

2.1. Participants

A conveniently large sample was used, including 471 undergraduate healthcare students attending the Departments of Nursing and Speech-Language Therapy, and 70 professional nurses and SLPs with at least two years of clinical experience. SLPs are often the allied healthcare professionals who most commonly serve children with ASD and the first professionals who evaluate a child at risk of ASD, since speech delay is the most frequent reason for referral [9,18]. SLPs present adequate knowledge about the core diagnostic features of ASD, even though they still show some knowledge gaps, and they receive specific coursework and training about autism [10,18]. On the other hand, although nurses are the first to adapt to the organic needs of a child with ASD in a primary healthcare setting and to provide consultation to parents, they demonstrate low levels of knowledge according to relevant research, and they lack specific training about ASD [13,14,22]. The comparison of performance between undergraduate students and professionals on the KCAHW questionnaire, as well as between those two healthcare professions, will provide information about the sensitivity of the questionnaire. The mean age of participants (n = 326) who reported their age was 25.6 years (SD = 8.8). More than 80% of participants in both samples (students and professionals) were females, mirroring the actual ratio of female gender in these healthcare professions. Most of the professional examinees (51.4%) had less than 10 years’ clinical experience. Regarding special education about the autistic condition, 24.3% of the professionals had attended special undergraduate courses, 14.3% had attended special postgraduate courses, 2.9% had completed an annual specialization, and most of them (58.5%) had attended special seminars. The demographic characteristics of participants are presented in the following table (

Table 1. Sociodemographic characteristics of participants.

Variable		N	%
Gender	Male	75	13.8
	Female	438	81.0
	N/A	28	5.2
Specialty	Speech and language therapy	295	54.5
	Nursing	246	45.5
Type	Professional	70	12.9
	Student	471	87.1
Working experience οf participants	0–5 years	18	25.7
	6–10 years	18	25.7
	years	13	18.6
	>20 years	21	30.0
ASD-specific training	Bachelor courses	17	24.3
	Postgraduate courses	10	14.3
	Annual specialization	2	2.9
	Seminars	41	58.5
Relationship with a person in the autistic spectrum	No	421	77.8
	I have a diagnosed relative/friend	95	17.6
	I have a diagnosed family member	12	2.2
	I am diagnosed	1	0.2
	N/A	12	2.2
Total		541	100.0

N/A: Not reported.

).

2.2. KCAHW Questionnaire

The KCAHW questionnaire [22] is a 19-item scale evaluating knowledge about ASD disorder. The items of the scale are divided into four domains. The first domain includes eight items, examining knowledge about deficits in social interaction that present in autistic individuals. The second domain only includes one item that addresses deficits in communication and language development, which are part of the autistic phenotype. The third domain contains four items evaluating behavioral impairments, such as compulsive, obsessive, and repetitive or stereotyped patterns of behaviors, the existence of which constitute the diagnostic criterion for ASD disorder. Finally, the fourth domain includes six questions evaluating knowledge about the onset and the nature of the disorder and its comorbidity with other psychiatric conditions. Based on previous research [23], we suggest that, henceforth, the above domains are named as “Social interaction-D1”, “Communication-D2”, “Behavioral pattern-D3” and “Characteristics of the disorder-D4”. The possible answers for the first eighteen items of the instrument are “Yes”, “No” and “Don’t know”, while for the last item, the possible answers are “Neonatal age”, “Infancy” and “Childhood”. Each correct answer takes one point, while incorrect and don’t know answers are scored with zero points. Therefore, the total maximum score is 19 points and the minimum score is zero points, respectively.

Translation and Adaptation of the KCAHW Questionnaire

For the preparation of the Greek version of the KCAHW questionnaire, a forward and backward translation was performed, as indicated by the World Health Organization [26]. Initially, the KCAHW was translated independently into Greek by two members of the research team, who are proficient bilingual speakers of Greek and English. Then, the two translated and adapted versions were compared and discussed, and a reconciled Greek version of the questionnaire was produced. Accordingly, the reconciled Greek translation was translated back into English by an independent bilingual professional translator, and the back translation was reviewed by the team members, establishing the pre-final adapted version of the questionnaire. Finally, the comprehensibility of the items was evaluated by asking ten students to review the questions of the questionnaire and minor modifications were made, leading to the final Greek version of the questionnaire.

2.3. Data Collection

Prior the beginning of the study, the research protocol was approved by the Ethics Committee of the University of Ioannina (ΡΝ: 49627/29-09-2022). All participants were informed of the purpose of the study, and they provided their written consent. Participation in the study was voluntary. The undergraduate students were informed about the research project during regular courses, while professionals were recruited via personal contacts through social media platforms such as Facebook and personal email. For the undergraduate students, the completion of the KCAHW questionnaire was part of a broader project about knowledge and attitudes toward ASD peers. The KCAHW questionnaire was either distributed online via Google Forms or it was paper-and-pencil completed. Questionnaires that were not fully answered were excluded from the analysis.

2.4. Statistical Analysis

Categorical variables are presented as absolute and relative (%) frequencies. Quantitative variables are presented as mean ± SD. The normality of distribution was evaluated with the Shapiro–Wilk test. The Student’s t-test or Mann-Whitney test were used for the comparison of means between different groups. The floor and ceiling effects, as well as the skewness and kurtosis of raw scores of each domain of the questionnaire, were evaluated. To investigate the internal consistency and reliability of the questionnaire, the Kuder–Richardson 21 coefficient was calculated for each domain and overall and the split-half reliability was measured using the Spearman–Brown correlation coefficient and the Guttman Split-Half value (lambda 4). Moreover, Spearman’s correlation coefficient was implemented to evaluate the linear relationship between different KCAHW domains. Additionally, to investigate the construct validity of the Greek version of the KCAHW questionnaire, confirmatory factor analysis (CFA) was implemented. To evaluate the goodness of fit of the model, the comparative fit index (CFI), the Tucker–Lewis index (TLI), and the root mean square error of approximation (RMSEA) among the fit statistics generated by CFA were used. Specifically, values of the RMSEA index lower than 0.08 are considered acceptable, whereas suggested acceptable values for the CFI and the TLI indices are equal, or larger than 0.90. A two-tailed p-value < 0.05 was considered statistically significant. Statistical analysis was implemented using ΙΒΜ SPSS v. 29 [27].

3. Results

3.1. Distribution Characteristic of the KCAHW Questionnaire

The floor and ceiling effects and skewness-kurtosis values of the four domains of the questionnaire are presented in

Table 2. The floor-ceiling and skewness-kurtosis values of the domains of KCAHW† questionnaire.

Scale Domains	Floor Effect %	Ceiling Effect %	Skewness	Kurtosis
Social interaction-D1	3.1	17.6	−0.388	−0.834
Communication-D2 *	24.0	76.0	−1.219	−0.516
Behavioral pattern-D3	3.5	55.1	−1.198	0.327
Characteristics of the disorder-D4	2.2	9.2	−0.262	−0.499
Overall	0.4	3.1	−0.422	−0.814

* Domain 2 consists of only one item.

. Overall, the floor-ceiling effects are below 15%, as expected. In “Communication-D2”, which consists of only one item, floor-ceiling effects are high, 24% and 76% respectively. Also, in “Behavioral pattern-D3”, the ceiling effect is high (55.1%). Moreover, the skewness-kurtosis values are excellent and between ±2.

3.2. Internal Consistency and Reliability

An overall Kuder–Richardson 21 (KR-21) value equal to 0.80 was obtained, indicating good internal validity [28]. KR-21 values for each of the domains were as follows: Social interaction-D1 (KR-21 = 0.73); Communication-D2 (KR-21 = na); Behavioral pattern-D3 (KR-21 = 0.66); and Characteristics of the disorder-D4 (KR-21 = 0.40). The KR-21 value, depicting the internal consistency for “Characteristics of the disorder-D4” was under the acceptable limits. To further examine the reliability of the questionnaire, we investigated the correlation of each item with its domain and the other domains of the scale. The questionnaire’s success rate was 100%. All items correlated higher with their domain compared to other domains. Item correlation coefficients were higher than 0.30 and ranged from 0.31 to 0.76. Additionally, we examined how the KR-21 value would change if each individual item was removed from the scale. For all items except item 19, removing each item resulted in a lower KR-21 value for that specific domain (

Table 3. Item-domain correlations, internal consistency †, and scale success ‡.

Items	Social Interaction-D1 r (KR-21 When the Item Is Deleted)	Communication-D2 r (KR-21 When the Item Is Deleted)	Behavioral Pattern-D3 r (KR-21 When the Item Is Deleted)	Characteristics of the Disorder-D4 r (KR-21 When the Item Is Deleted)
Item 1	0.49 (0.73)	0.18	0.36	0.22
Item 2	0.65 (0.69)	0.21	0.37	0.28
Item 3	0.59 (0.71)	0.23	0.26	0.23
Item 4	0.66 (0.69)	0.26	0.37	0.37
Item 5	0.50 (0.73)	0.19	0.22	0.18
Item 6	0.61 (0.71)	0.22	0.31	0.22
Item 7	0.70 (0.67)	0.28	0.38	0.32
Item 8	0.67 (0.68)	0.26	0.46	0.38
Item 9	0.38	1.00 (na)	0.34	0.26
Item 10	0.28	0.27	0.62 (0.65)	0.2
Item 11	0.47	0.26	0.75 (0.62)	0.41
Item 12	0.38	0.28	0.76 (0.54)	0.3
Item 13	0.44	0.2	0.76 (0.54)	0.37
Item 14	0.21	0.03	0.26	0.49 (0.36)
Item 15	0.31	0.1	0.29	0.61 (0.26)
Item 16	0.39	0.24	0.44	0.6 (0.28)
Item 17	0.31	0.29	0.26	0.59 (0.28)
Item 18	0.29	0.21	0.28	0.59 (0.35)
Item 19	−0.04	−0.04	−0.03	0.31 (0.49)
KR-21 (Overall)	0.73	(na)	0.66	0.40

na: not applicable. †: Measured with Kuder–Richardson 21. ‡: A percentage of the items in one domain gives a high correlation coefficient with their domain, not with the other domains.

).

Furthermore, all domains were positively and significantly correlated with each other, and the Spearman’s correlation coefficients ranged from 0.26 to 0.57. The results are presented in

Table 4. Spearman’s correlation coefficient between the four domains of the KCAHW.

Scale Domains	Social Interaction-D1	Communication-D2 *	Behavioral Pattern-D3	Characteristics of the Disorder-D4	Overall
Social interaction-D1	-	0.38 **	0.57 **	0.46 **	0.89 **
Communication-D2 *		-	0.32 **	0.26 **	0.48 **
Behavioral pattern-D3			-	0.48 **	0.75 **
Characteristics of the disorder-D4				-	0.76 **
Overall					-

* Domain 2 consists of only one item; ** p-value < 0.05.

.

Finally, the Spearman–Brown correlation coefficient was equal to 0.618, while the split-half reliability analysis demonstrated that the Guttman split-half value (lambda 4) was 0.613 and was considered as satisfactory [29].

3.3. Construct Validity of the KCAHW

The goodness of fit of the CFA is presented in

Table 5. Confirmatory factor analysis indices.

Fit Indices	Values
CFI †	0.880
TLI ‡	0.861
RMSEA †‡	0.052

†: Comparative fit index; ‡: Tucker–Lewis index; †‡: root mean square error of approximation.

and Figure 1. The CFI and the TLI indices were equal to 0.88 and 0.861, respectively. Both indices were just below 0.90 and their values are considered acceptable, indicating a good model fit. Finally, the RMSEA index was equal to 0.052, showing a moderate fit. Overall, the model presented an adequate fit to the data.

3.4. Criterion Validity

To evaluate the criterion validity (sensitivity) of the questionnaire, we investigated the performance of different known groups of participants. The results related to the variables used in the testing of criterion validity are presented in

Table 6. Known groups and criterion validity results of KCAHW questionnaire.

		Social Interaction-D1		Communication-D2		Behavioral Pattern-D3		Characteristics of the Disorder-D4		Overall Score
		Mean (SD)	p-Value	Mean (SD)	p-Value	Mean (SD)	p-Value	Mean (SD)	p-Value	Mean (SD)	p-Value
Occupation	Nursing professionals	4.09 (2.86)	<0.001	0.53 (0.51)	<0.001	2.22 (1.52)	<0.001	3.44 (1.16)	<0.001	10.28 (4.62)	<0.001
	SLT professionals	6.79 (2.03)		0.92 (0.27)		3.95 (0.23)		4.37 (1.24)		16.03 (2.53)
	SLT students	6.13 (1.64)		0.87 (0.34)		3.72 (0.64)		4.23 (1.17)		14.94 (2.71)
	Nursing students	3.46 (1.86)		0.64 (0.48)		2.45 (1.20)		2.57 (133)		9.12 (3.25)
Contact with ASD individuals	Students
	No	6.16 (1.59)	0.654	0.87 (0.33)	0.637	3.72 (0.66)	0.514	4.22 (1.21)	0.535	14.97(2.76)	0.948
	I have a diagnosed relative/friend/family member	6.19 (1.78)		0.85 (0.36)		3.70 (0.61)		4.6 (1.00)		15.09 (2.54)
	Professionals
	No	4.95 (2.97)	0.016	0.65 (0.48)	0.028	2.84 (1.51)	0.018	3.74 (1.05)	0.051	12.19 (5.04)	0.013
	I have a diagnosed relative/friend/family member	6.52 (2.15)		0.89 (0.32)		3.67 (0.83)		4.26 (1.56)		15.33 (3.03)
Working experience	0–5 years	6.11 (2.43)	<0.001	1.00 (0.00)	<0.001	3.83 (0.38)	<0.001	4.67 (1.33)	0.007	15.61 (3.15)	<0.001
	6–10 years	7.06 (1.73)		0.89 (0.32)		3.83 (0.71)		4.06 (1.61)		15.83 (2.55)
	≥11 years	4.47 (2.99)		0.53 (0.51)		2.44 (1.56)		3.50 (1.16)		10.94 (4.90)
ASD-specific Training	Bachelor courses	5.59 (2.83)	0.059	0.71 (0.47)	0.866	2.76 (1.48)	0.025	4.00 (1.58)	0.768	13.06 (5.03)	0.090
	Postgraduate courses	7.20 (1.87)		0.80 (0.42)		4.00 (0.00)		4.10 (1.10)		16.10 (2.77)
	Annual specialization/seminars	5.16 (2.84)		0.74 (0.44)		3.12 (1.38)		3.88 (1.22)		12.91 (4.66)

Mann–Whitney or Kruskal–Wallis test.

.

As presented in Table 6, nursing students had significantly lower total knowledge scores than speech-language pathology students (p < 0.001), while both student groups exhibited significantly lower performances on the scale than their relevant professional group (p < 0.001). The above pattern of performance was observed in all domains of the questionnaire. Furthermore, there was a significant difference in all domains of KCAHW (Social interaction-D1: p = 0.016; Communication-D2: p = 0.028; Behavioral pattern-D3: p = 0.018; Characteristics of the disorder-D4: p = 0.05) and total knowledge scores (p = 0.013) between professionals who had personal contact with a person with ASD and those who did not. There was also a significant advantage in performance on all domains (p < 0.001) and on total scores (p < 0.001) of the KCAHW questionnaire for younger professionals with fewer years of working experience over older professionals who had more than 11 years of working experience. Finally, those who had received ASD-specific training had a significantly higher performance on “Behavioral pattern-D3” of the scale (p = 0.025), but this difference did not reach significant levels for the total scores of KCAHW.

4. Discussion

Even though allied healthcare professionals are widely viewed as authorities on child development, research findings suggest that their overall knowledge about ASD is moderate. They generally recognize the core characteristics of the disorder, but exhibit knowledge gaps and misconceptions about the aitiology, comorbid conditions, and specific interventions [14,18]. Poor knowledge of allied healthcare professionals significantly impacts the delivered services and the quality of life of autistic children and their families [10,17]. Despite the importance of measuring ASD knowledge to develop expertise in professional populations, there is currently no gold standard tool for assessing this knowledge [14,20]. Most researchers have used structured surveys designed for the scope of their study, a fact that inhibits both the examination of the psychometric properties of any one measure and the comparison of the findings across the studies [20]. Hence, recent research strongly recommends the rigorous evaluation of the psychometric properties of known ASD knowledge instruments in research replications, providing adequate evidence for characterizing an instrument as a “well-established assessment” [20,21]. In line with the above, the present study aimed to investigate the psychometric properties of the Greek version of the KCAHW questionnaire, which is the most used standardized questionnaire for assessing knowledge about the autistic condition [14].

4.1. Reliability of KCAHW

Regarding the reliability of the questionnaire, the distributional characteristics of the Greek version showed an overall floor and ceiling effect lower than 15%, and excellent skewness-kurtosis values between ±2. These findings are better than those presented for the Turkish version of the tool [23]. However, in “Communication-D2”, which consists only of one item, the floor-ceiling effects were higher than 15%, and in “Behavioral pattern-D3”, the ceiling effect was significantly high and equal to 55.1%. As it concerns the internal validity of the scale, statistical analysis revealed an overall KR-21 coefficient equal to 0.80, which is lower than the overall Cronbach alpha value of 0.97 reported in the original standardization of the KCAHW [22] and KR-21 values equal to 0.73, 0.66, and 0.40 for domains 1, 3, and 4, respectively. The KR-21 values for the Greek version of the tool are better than those of the Turkish version of the tool. Nevertheless, in both studies, the internal consistency for “Characteristics of the disorder-D4” was under the acceptable limits (<60%) [23]. Furthermore, all items of the scale presented higher correlation coefficient values for their domain than with others, while for all items except item 19, the overall KR-21 values were greater than the “if item deleted KR-21 values”. These results are in line with the findings of the Gürbüz-Özgür et al. study [23]. In this Turkish study, the “if item deleted KR-21 values” were higher for two items of the scale, specifically questions 16 and 19 [23]. Moreover, Spearman correlation coefficients revealed overall significant strong to very strong correlations between the four domains of the questionnaire, indicating better values than those presented in the Turkish adaptation of the questionnaire [23]. Finally, the split-half reliability analysis demonstrated a Spearman–Brown correlation coefficient equal to 0.618 and a Guttman split-half value (lambda 4) equal to 0.613, which are within the acceptable limits; however, the Greek version’s lambda 4 value is significantly lower than the lambda 4 value reported in the shorter Turkish version of the scale, which only included 12 items of the KCAHW from the domains 1, 2 and 3 [24].

4.2. Validity of KCAHW

The Greek version of the KCAHW questionnaire presents acceptable construct validity since all three indices of the CFA, specifically the CFI (0.88), TLI (0.861), and RMSEA (0.052), present acceptable values, indicating the goodness of the fit of the model. Our results reinforce the previous findings of a Turkish adaptation of the questionnaire [23], supporting the construct validity of the scale. Furthermore, the findings of criterion validity also enhance evidence about the validity of the tool, since there were significant differences (p < 0.001) in the total scores between students and professional participants, with professional participants showing a higher performance on the scale. Moreover, there was a significant difference in the mean total scores of KCAHW between SLPs (mean: 16.03; SD: 2.53) and nurses (mean: 10.28; SD: 4.62) in favor of SLPS. In our study, being a professional SLP, having personal contact with an individual with ASD, and having received ASD-specific training significantly affected positively performance on the KCAHW. The related literature shows that the mean total scores on the KCAHW questionnaire for allied healthcare professionals range from 9.01 to 13.5 out of a maximum of 19, indicating large variations in knowledge between primary healthcare providers [14]. Studies [13,30,31] examining knowledge of ASD in nurses with the KCAHW reported mean total scores between 8.79 and 12.06, which were similar to those reported to the original standardization of the tool [22], and our findings add to this body of research. SLPs correctly answered approximately 84% of questions (16 out of 19), indicating a fairly good level of knowledge. These results agree with other studies showing that SLPs exhibit acceptable levels of knowledge about autism [18,32]. Moreover, our data mirror findings supporting that personal contact with an ASD person positively impacts autism awareness knowledge [14] and are analogous to those referred to in the Turkish version of the KCAHW [23]. Finally, younger professionals with fewer years of working experience and those with postgraduate ASD-specific training had better performance on the KCAHW than older professionals with more than 11 years of working experience and those who had received bachelor courses or seminar training on ASD. The above is in line with previous findings, possibly reflecting an increase in the amount of specific training that health workers receive relating to ASD [14,33].

5. Conclusions

Based on the results of the current study, we can conclude that the Greek version of the KCAHW questionnaire is a valid and reliable tool for evaluating knowledge about ASD. A previous validity study in Turkish provides similar findings, underscoring the ability of the scale to present acceptable psychometric properties in different samples and across multiple cultural contexts [22]. However, both aforementioned studies reported unacceptable internal consistency for “Characteristics of the disorder-D4”, which includes six questions investigating knowledge about the onset and the comorbidity of autism with other psychiatric conditions. The original standardization of the scale included a rather restricted sample of participants (50 psychiatric nurses) and did not examine separately the internal consistency of each domain of the questionnaire [22]. The findings underline the need for a detailed evaluation of problematic items from a diverse panel of autism experts and/or potential respondents to improve their clarity and the construct validity of “Characteristics of the disorder-D4”, as well as to ameliorate the psychometric soundness of the scale [20]. Nonetheless, Atun-Einy and Ben-Sasson [18] in their study revealed that allied healthcare clinicians (SLPs, occupational therapists, physical therapists) presented scattered knowledge in identifying comorbidities, such as epilepsy and intellectual disability, as associated with autism. The advantages of the study are the inclusion of a large and divergent sample of participants (undergraduate students and professionals) and the detailed presentation of the recommended cross-linguistic processes of the Greek adaptation to address previous issues of transadaptation of ASD knowledge scales. In future research, the problematic items must be revised or deleted in response to expert panels’ feedback and psychometric properties must be reevaluated, while the adaptation of the KCAHW questionnaire in other language/s and its administration in different clinical populations will provide more evidence about its psychometric abilities. The standardization of valid and reliable ASD knowledge scales will be beneficial for the evaluation of the impact of ASD training professionals’ programs on healthcare workers.