Guillermo Savio* Maria Cecilia Perez-Abalo* Jose Gaya* Odelaysis Hernandez $ Eleina Mijares* *Cuban Neuroscience Center, La Habana, Cuba $ Ramon Gonzalez Coro Maternity Institute, La Habana, Cuba Key Words Multiple auditory steady state responses Objective frequency specific audiometry Screening High-risk babies Orginal Article International Journal of Audiology 2006; 45:109 /120 Test accuracy and prognostic validity of multiple auditory steady state responses for targeted hearing screening Eficiencia diagno ´ stica y valor prono ´ stico de los potenciales evocados auditivos de estado estable a mu ´ ltiples frecuencias en un programa de tamizaje auditivo para nin ˜ os en riesgo Abstract The test accuracy and prognostic validity of Multiple Auditory Steady State Responses (MSSR) and click Auditory Brainstem Responses (cABR) was compared within the context of a targeted screening protocol. A sample of 508 high-risk babies was first screened using cABR and MSSR (500 and 2000 Hz). All children (failed/ pass) were called back at three to four years of age to determine their hearing status (pure-tone audiometry). Although both methods showed an equally good test performance in the first screen (sensitivity: 100% and specificity: 92 /95%), the MSSR may have some potential advantage to identify low-frequency hearing loss. Furthermore, the confirmatory audiometry with MSSR predicted the child hearing status more accurately than the cABR. In conclusion, the MSSR can provide valuable information for the diagnosis and management of infants earlier detected by a screening protocol and further developed might be also useful as a screening test. Sumario Se comparo ´ la eficiencia diagno ´stica y el valor prono ´ stico de los potenciales evocados auditivos de estado estable a mu ´ ltiples frecuencias (MSSR, siglas en ingle ´s) con los potenciales evocados auditivos de tallo cerebral con clicks (cABR, siglas en ingle ´s) en el contexto de un programa de tamizaje auditivo de nin ˜os con factores de riesgo. Inicialmente se realizo ´ un tamizaje con cABR y MSSR (0.5 y 2 kHz) a una muestra de 508 bebe ´s de alto riesgo. Todos los nin ˜os (los que pasaron y los que fallaron) se reevaluaron a los 3 o 4 an ˜os de edad para determinar su estado auditivo (audiometrı ´a tonal). Aunque los dos me ´todos mostraron igualmente buen desempen ˜o en el primer tamizaje (sensibilidad: 100% y especificidad: 92 /95%) los MSSR pueden tener una ventaja potencial para identificar hipoacusias en tonos graves. Adema ´s la prueba audiome ´trica confirmatoria con MSSR predijo el status auditivo del nin ˜o con mas precisio ´n que los cABR. En conclusio ´n, los MSSR pueden proporcionar informacio ´n valiosa para el diagno ´stico y tratamiento de nin ˜os previamente detectados por un programa de tamizaje y en un futuro pueden ser tambie ´n u ´ tiles como prueba de tamizaje. Recent developments in techniques and instrumentation as well as strong legislative efforts, have made neonatal hearing screen- ing a reality nowadays in many countries worldwide. An important health goal has been set forth: to identify and start the appropriate intervention of children with significant hearing losses by six months of age (NIH, 1993; JCIH, 1994, 2000). To fulfill this goal, comprehensive screening programs must be organized, ensuring not only early detection, but also diagnosis, treatment and intervention of the hearing-impaired child (Mencher et al, 2001). There are several available techniques which can reliably identify hearing impairments. The most frequently used are click auditory brainstem response (cABR) (Hyde et al, 1990; Durieux-Smith et al, 1991; Galambos et al, 1994; Mason et al, 1998) automated ABR (Jacobson et al, 1990; Mason and Herrmann 1998; Dort et al, 2000; Clemens et al, 2000; Sininger et al, 2000) and otoacoustic emissions (OAE) (White et al, 1993; Mehl and Thomson 1998; Dort et al, 2000). Although these techniques have been proven efficient (Sininger et al, 2000; Norton et al, 2000) within a screening context, further improve- ments may be still beneficial. It has been shown that some specific hearing impairments can be missed by either of these screening tests (Durieux-Smith et al, 1991; Norton et al, 2000). Regardless of the technique used for screening, attention must be also paid to the subsequent diagnostic and intervention stage. As part of this stage, some form of objective audiometry would be needed to characterize the residual hearing in early infancy. The technologies used for this purpose need to be also carefully considered. There are several techniques, based on the recording of transient auditory evoked potentials to brief acoustic stimuli, which can accurately estimate the audiogram (Stapells et al, 1995; Stapells 2000). However these frequency-specific methods can be technically more difficult, time consuming, and require much expertise to accurately identify threshold responses. Also, these responses show greater variability at the lower audiometric frequencies (Gorga et al, 1988). Probably for these reasons, in ISSN 1499-2027 print/ISSN 1708-8186 online DOI: 10.1080/14992020500377980 # 2006 British Society of Audiology, International Society of Audiology, and Nordic Audiological Society Guillermo Savio, MD Cuban Neuroscience Center, Ave. 25 esq. 158 Playa, POB 6880, La Habana, Cuba. Email: guille@cneuro.edu.cu Received: October 11, 2004 Accepted: September 12, 2005