Translation and usability of autism screening and diagnostic tools for Autism Spectrum Conditions in India. Rudra A1, Banerjee S2, Singhal N3, Barua M3, Mukerji S2, Chakrabarti B1,4 1 School of Psychology and Clinical Language Sciences, University of Reading, UK 2 Creating Connections, Kolkata, India 3 Action for Autism, National Centre for Autism, Delhi, India 4 Autism Research Centre, University of Cambridge, UK Banerjee S is now at University of Haifa, Haifa, Israel. [NOTE: This is the final author-version of the manuscript, the formatted and published version is available at the journal website at : http://onlinelibrary.wiley.com/doi/10.1002/aur.1404/abstract] Correspondence concerning this article should be addressed to: Dr. Bhismadev Chakrabarti School of Psychology and Clinical Language Sciences, University of Reading, Reading RG6 6AL, UK Email: [email protected] Phone: +44 118 378 5551 Fax: +44 118 378 6715 Grant sponsor: Autism Speaks Lay Abstract: Among all the major developing countries, India is conspicuous by the absence of an estimate of autism prevalence. One key reason for this absence is the the lack of availability of standardized screening and diagnostic tools (SDT) for autism in regional languages in India. To address this gap, we translated four widely-used SDT (Social Communication Disorder Checklist, Autism Spectrum Quotient, Social Communication Questionnaire, Autism Diagnostic Observation Schedule) into Hindi and Bengali, two of the main regional languages (~360 million speakers) and tested their usability. We tested these translated instruments on 170 children with and without autism, and found that scores of children with autism were significantly and reliably different from those of control children. This provides the first evidence that these SDTs are useable in Hindi and Bengali, and provides the essential toolkit for future autism epidemiological research in this region. Abstract: There is a critical need for screening and diagnostic tools (SDT) for Autism Spectrum Conditions (ASC) in regional languages in South Asia. To address this, we translated four widely used SDT (Social Communication Disorder Checklist, Autism Spectrum Quotient, Social Communication Questionnaire, Autism Diagnostic Observation Schedule) into Bengali and Hindi, two main regional languages (~360 million speakers), and tested their usability in children with and without ASC. We found a significant difference in scores between children with ASC (N=45 in Bengali, N=40 in Hindi) and typically developing children (N=43 in Bengali, N=42 in Hindi) on all SDTs. These results demonstrate that these SDT are usable in South Asia, and constitute an important resource for epidemiology research and clinical diagnosis in the region. Keywords: Screening; Global mental health; Translation; Cultural Abbreviations: ASC, SCDC, TQ, SCQ, AQ-C, ADOS, SDT ASC= Autism Spectrum Conditions SCDC= Social Communication Disorder Checklist SCQ= Social Communication Questionnaire AQ-C= Autism Quotient- Children’s version ADOS= Autism Diagnostic Observation Schedule TQ= Ten Questions SDT= Screening and Diagnostic Tools In most of the developed world, autism can be identified by the age of 3 years (Howlin & Moorf, 1997) and as early as 18 months in some cases (Baird et al., 2000; Baron-Cohen et al., 1996). Early identification of ASC is necessary for early intervention, which is associated with better outcome (Eikeseth, 2009). Early identification relies crucially on the availability of standardised and validated diagnostic tools. This is particularly relevant for developing countries such as India, where there is a widespread lack of awareness about autism, and a shortage of trained professionals. A number of screening and diagnostic tools for autism have been developed over the past decades (Charman & Gotham, 2012; Corsello et al., 2007). However, these tools are neither available nor characterised in local languages in South Asia, where the majority of the population is not proficient in English. The only related measure that is internationally used and well-characterised in this population, that is available in local languages is the WHO Ten Questions (TQ), which is a screening measure for severe childhood disability (Zaman et al., 1990). The TQ screens for major developmental disorders (e.g. epilepsy, blindness, hearing and speech problems, mental retardation and motor disability), and is not specific to autism. To address this gap in the availability of appropriate autism screening and diagnostic tools (SDT), the aim of the current study was to translate some of the widely used SDTs into two main South Asian languages, i.e. Bengali and Hindi. These two languages have the fifth largest number of speakers in the world (total number of speakers estimated over 360 million people (estimates from www.ethnologue.com). Kanner (Kanner, 1943) was the first to report that critical symptoms of autism included a lack of communicative eye contact, delayed language development, ritualistic repetitive behaviour, and a general deficit in social understanding. The DSM-IV and ICD-10 criteria for ASD refined these early observations by using the following features for diagnosis: atypicalities in socialization and communication, as well as restricted interests and repetitive behaviour (APA, 2000). These refined criteria have been used to develop SDT for ASC, which include, among others, the Social and Communication Questionnaire (SCQ) (Berument, Rutter, Lord, Pickles, & Bailey, 1999) and the Autism Diagnostic and Observation Schedule (ADOS) (Lord et al., 2000). It is generally believed that the core features of ASC are relatively universal. However, it is not known how autistic symptoms manifest themselves differently across cultures. Cultural factors can influence the definition and recognition of symptoms, as well as their treatment (Daley & Sigman, 2002). This may create a gap between cultures in how ASC is detected and managed. The first step to bridging this cultural gap is to ensure that the tools used for screening and diagnoses are available in a range of languages. This is particularly challenging for ASC, as subtle features of social- communication behaviour are captured differently by different languages. Subtle cultural differences may be missed due to language if the tools are used in English. E.g. in both Hindi and Bengali (unlike in English), there is an honorific second and third person pronoun that is used to address individuals who are senior in age or who have met for the first time. Individuals with ASC might make the distinction between first, second or third pronoun but may show deficits in choosing the appropriate second or third person pronoun. The formal you translates as ‘aap or aapni’; in comparison to ‘Tu or Tui’ for a younger person and ‘Tum’ for peers and friends. Using a diagnostic tool in English may not capture the subtleties because of a single second person pronoun (you). Additionally, different beliefs held by health professionals regarding the time of appearance of symptoms (some Indian professionals believe Indian children may speak their first word at 1.5 years of age and a delay in language should not be a point of concern until 3 years of age) may lead to differential diagnosis. Cultural differences in beliefs held by health professionals such as that mentioned above may lead to varied opinions in prioritising symptomatology for ASC (Daley & Sigman, 2002). A wide variability in the prevalence of autism is noted across the world (Elsabbagh et al., 2012). (Matson & Kozlowski, 2011) suggest that these increases maybe due to changes in diagnostic criteria over time, new assessment instruments, differences in cultural norms for identifying ASC, inaccurate diagnoses, utilization of different research methodologies to identify prevalence estimates or genetic differences between populations. Specifically within Asia, estimates vary widely across time and country (China: 0.003%- 0.17%, Japan: 0.011%-0.21% (Sun & Allison, 2010); South Korea: 2.64% (Kim et al., 2011). Sun and Allison suggest that the heterogeneity of SDTs used in Asia contributed to this observed variability. Eight screening instruments have been used for the 26 prevalence studies in Asia. Five studies in Japan have used an 18-month health checklist (HC-18) (Honda, Shimizu, Imai, & Nitto, 2005; Honda, Shimizu, Misumi, Niimi, & Ohashi, 1996; Kawamura, Takahashi, & Ishii, 2008; Sugiyama & Abe, 1989; Tanoue, Oda, Asano, & Kawashima, 1988). In China, five studies used the Chinese Autism Behaviour Scale (CABS)(Zhang & Ji, 2005), 2 used the translated version of the Autism Behaviour Checklist (ABC) (Volkmar et al., 1988), and others used a translated version of the Checklist for Autism in Toddlers (CHAT) (Baron-Cohen et al., 2000; Wong et al., 2004). The local Bryson’s Screening Scale was used in Indonesia while an Iranian study used the Childhood Symptom Inventory-4 (CSI-4) (Gadow & Sprafkin, 1994; Ghanizadeh, 2008). A recent Korean study used a translated and validated version of the ASSQ followed by ADOS and ADI to confirm diagnostic status. As is evident from the discussion above, a number of SDTs (including the AQ, SCQ, ADOS) for ASC have been translated in other languages in countries such as Brazil, Italy, China and Spain (Canal-Bedia et al., 2011; Pereira, Riesgo, & Wagner, 2008; Ruta, Mazzone, Mazzone, Wheelwright, & Baron-Cohen, 2011). However, none of these tools are available in any of the major regional languages of India. Availability of such widely