The detection of autism is very important because the lack of recognition of this disorder has elevated costs for the families, health care and education providers. Diagnosis is made frequently four or five years after parents notice the first signs. The reasons for this delay are many, but a common one is the lack of recognition of key symptoms that can lead to a more complex diagnosis assessment. Another reason is that screening and diagnostic instruments are not well known by primary caregivers in health and education systems, as these professionals are the first to hear parents' concerns. Moreover the instruments are not well known because the cost of acquiring them and receiving formal training is very high. The need to make comparable assumptions of this complex disorder makes it important to use the same instruments as other countries. Growing efforts for an early recognition have been made in recent years because early intervention programs benefit children with autism. In the last decade, important advances in the design of diagnostic and screening instruments have been made. These tools have primarily been used for clinical, epidemiological or research uses. In some countries their use has become routine in schools, leading to better detection and increasing prevalence rates of autism. Misdiagnosis is not uncommon in autism. Almost 60% of children with Asperger disorder first receive an erroneous diagnosis of attention deficit disorder, oppositionistic or bipolar disorder. Autism presents with a bizarre clinical picture during the years in which many thought it was untestable. Gradual characterization of behaviors and studying different aspects of the symptomatology had led to a better comprehension and descriptions. Most authors have incorporated this knowledge to design reliable instruments. The most common behaviors explored are: protodeclarative pointing, joint attention, repetitive/ stereotyped movements and absence of characteristic symbolic play. This target behavior can be explored through the diverse rating scales and interviews. The instruments are very diverse and varing form. There are rating scales for parents to record their children symptoms and observation schedules to be completed by a clinician or trained professional for that purpose. The best approach is to combine modalities to include as much information as possible. CHAT (Checklist for Autism in Toddlers) is a brief screening instrument intended to detect autism in toddlers. The first part consists of nine questions for parents to complete, while the second part is an observation schedule with five brief age-appropriate interactions with the children. This instrument is an important antecedent of more sophisticated and expanded play observation schedules. Checklist for Autism in Toddlers Modified (CHAT M) is a modified version which consists of an expansion of the parent questionnaire by eliminating the observational section. The Childhood Autism Rating Scale (CARS) is another instrument which assesses the severity of autism. This instrument is rated by clinicians or by trained observers. CARS was designed before DSM IV criteria were published so it does not contain an algorithm to distinguish between different developmental problems. In spite of this limitation, it is the most used rating scale for autism diagnosis. The Child Behavior Checklist (CBCL/1.5-5) is a broad band rating scale which evaluates psychopathology of children between 18 months and five years old. It has a DSM oriented subscale to evaluate developmental problems such as autism or Asperger disorder. It also contains a withdrawn subscale which has proven to be useful as demonstrated by some studies done with the CBCL/4-18. This instrument also allows assessing other associated problems common in autistic children such as attention problems, depression and anxiety. The Language Developmental Survey (LDS) associated to this rating scale, gives the opportunity to screen vocabulary for the identification of language delays, which are common in children with pervasive developmental disorders. It was necessary to have more structured instruments to diagnose autism and not only for screening purposes, so in 1989 the first diagnostic interview was published. The instrument has gone through an extensive review and creative process which has led to the most important tools for diagnosing autism in adults and children. The Autism Diagnostic Interview (ADI) was published in 1989 and correlated to the ICD-10 definition of autism. The original ADI was intended primarily for research purposes, providing behavioral assessment for subjects with a chronological age of at least five years and a mental age of at least two years. The ADI explores three key domains defining autism: (1) reciprocal social interaction, (2) communication and language, and (3) repetitive, stereotyped behaviors. The Autism Diagnostic Interview Revised ADI-R is a semi/ standardizer interview shorter than the ADI, which has been developed for clinical use. It is more appropriate for younger children than the ADI. The ADI-R takes from 2 to 3 hours to administer and can be used with children as young as two years of age (with a mental age greater than 18 months). It explores information about the child functioning in the present and the past. It contains an algorithm based on DSM criteria for autistic disorder, and allows for distinguishing between autistic disorder and non autistic disorder. Pre Linguistic Autism Diagnostic Observation Schedule (ADOS-PL) is a modified version of the ADOS used to diagnose young children (under the age of six years) who are not yet using phrase speech. It is a semi-structured assessment of play, interaction, and social communication and takes about 30 minutes for a trained clinician to administer. The Autism Diagnostic Observation Schedule-Generic (ADOS-G) is a standardized play observation schedule. Through structured play materials and activities promoted by the examiner, social interactions are rated for common autistic features like joint attention, protodeclarative pointing, quality of reciprocal social interaction and symbolic play. Different modules are available from one to four, with specified criteria to match the participants' developmental and language level. It contains an algorithm related to the DSM IV domains of an Autistic Disorder or PDD-NOS. The ADI, ADI/R, ADOS PL, and ADOS G are considered the gold standards for autism diagnosis. There are important reliable instruments for diagnosing autism but extensive training is needed to obtain useful diagnostic information. Since these instruments are very recent, they have not been validated in some countries and neither their cultural bias has been investigated. It is not enough to assess autistic symptoms only for diagnostic purposes; patients need further evaluation to determine their psychosocial functioning, cognitive abilities, and language delay or deviations. The information from these assessments is very important for planning well designed interventions. Even though there is a growing interest in perfecting these modern instruments, diagnosis cannot rely exclusively on them. They are important tools to facilitate the diagnosis, but broader assessment should be pursued. It is important to validate and culturally adapt these instruments so different countries can utilize the same tools and research results can be comparable. In the future more rating scales, observation schedules and diagnostic interviews will be developed for assessing Asperger disorder, to be used in genetic studies, for assessing broad band syndromes. Better cognitive measures will be necessary to evaluate psychosocial impact. But this growing specialization will increase costs so it is important to develop briefer and more cost-effective methods to evaluate persons with autism. The availability of these tools will guarantee early diagnosis and treatment not only for research purposes but for identification in the community.

La detección del autismo en México es muy importante ya que la falta de reconocimiento de este trastorno tiene costos muy elevados para las familias y los prestadores de servicios de salud y educación. Muy a menudo el diagnóstico de autismo se realiza cuatro o cinco años después de que los padres observan los primeros síntomas. Las razones para este reconocimiento tardío son diversas; pero una de las principales es la falta de identificación de síntomas clave que obliguen a una evaluación diagnóstica en forma. Otro motivo es que en nuestro país son poco conocidos los instrumentos de tamizaje y diagnóstico por parte de los profesionistas primarios como maestros y médicos familiares, quienes son los primeros en escuchar las quejas y preocupaciones de los padres. Aun en contextos más especializados, estas herramientas son poco conocidas pues su adquisición y aplicación es un proceso complejo y costoso que a menudo debe realizar el profesionista por su cuenta. A pesar de estos inconvenientes, en años recientes se han realizado grandes esfuerzos para el reconocimiento del autismo puesto que hay evidencias de que las intervenciones tempranas mejoran el pronóstico en estos niños. En la última década se han realizado avances muy importantes en el diseño de instrumentos de diagnóstico y tamizaje, a los que se han utilizado con propósitos de investigación clínica y epidemiológica. En algunos países su uso se ha vuelto una rutina en las escuelas y se ha logrado una mayor detección de autismo por lo que se han elevado las tasas de prevalencia. Los instrumentos son muy diversos, pueden ser listas de autoinforme dirigidos a los padres para que registren los síntomas de los niños, o cédulas de observación para ser completadas por el clínico o el personal entrenado para tal propósito. Lo mejor es el uso mixto de instrumentos para obtener la mayor cantidad de información posible como es el caso del CHAT que incluye una sección de interrogatorio y otra sección de observación con actividades que el niño debe desarrollar. Este instrumento es precursor de actividades sencillas y creativas con un componente lúdico, diseñadas con el propósito de evaluar al niño preescolar. Hoy este es un importante antecedente de otros instrumentos más elaborados. El cuestionario para el autismo en niños preescolares modificado CHAT M es una versión modificada del el cuestionario para el autismo en niños preescolares (CHAT) que consiste en una expansión de la sección de interrogatorio para el padre, con un formato de autoinforme que parte de la eliminación de la sección de observación. Otros instrumentos miden la gravedad del autismo como la Escala de Evaluación de Autismo Infantil (CARS), dirigida al clínico que evalúa la intensidad del autismo. La lista de síntomas del niño de 1.5-5 (CBCL/1.5-5) es un instrumento de banda ancha que evalúa la psicopatología general en niños con edad entre 18 meses y cinco años; contiene una subescala de problemas del desarrollo que sirve como tamizaje para evaluar el autismo y el trastorno por Asperger con base en los criterios del DSM. Se han diseñado y se han perfeccionando paulatinamente varias entrevistas de diagnóstico. La entrevista de diagnóstico de autismo (ADI), la entrevista de diagnostico para el autismo revisada (ADI-R), la cédula prelingüística genérica de observación para el autismo (ADOS PL), y la cédula de observación genérica para el autismo (ADOS G) son escalas consideradas standard de oro para el diagnóstico del autismo. Conforme se han mejorado las propiedades psicométricas, de los instrumentos, éstos también se han ajustado para cubrir las necesidades de evaluación de los pacientes autistas con un amplio rango de edad, destreza verbal y cognitiva. Como resultado, podemos contar con instrumentos confiables y adecuados para una población con necesidades muy diversas; estas herramientas nos han demostrado que un constructo tan complejo y amplio como el autismo se puede medir. En este artículo se presenta una breve revisión de la evolución histórica de la clasificación acorde a los criterios del DSM y a la descripción de los principales instrumentos de diagnóstico, y los datos de su validez y confiabilidad.