Abstract
We studied the feasibility of low-cost videoconferencing (using Skype) in urban community health clinics for speech, language and hearing screening of children up to six years of age. During a two-year study, screening services were provided via videoconferencing at two community clinics in an inner city area of Cleveland, Ohio. In total, 411 screenings were completed. Of these, 358 children (87%) received hearing screenings, 377 (92%) received tympanometry screening and 263 (64%) received speech and language screening only. A total of 151 children were aged three years or under (37%). The reliability of pure tone hearing screening (n = 7), DPOAE screening (n = 51) and speech-language screening (n = 10) was 100%. Typanometry screenings (n = 55) were 84% reliable. Families reported a high level of satisfaction with both the technology and with the videoconferencing. The results indicate that low-cost videoconferencing for screening of speech, language and hearing development in very young children in urban community health clinics is feasible, reliable and strongly supported by the community.
Introduction
When children are at risk of or diagnosed with speech, language or hearing disorders, early intervention substantially improves communication development. 1 Despite the positive outcome, there are many barriers to families utilizing these services in practice. These include professional shortages and the accessibility of clinic locations to families. 1–4 Children and families in urban, minority, low socioeconomic areas face additional barriers. These include insufficient health insurance, the inability to otherwise pay for medical services, difficulties with reliable transportation, scheduling conflicts and childcare/family problems. Lack of consistent access to a primary care physician further complicates obtaining a referral to speech, language and hearing services. 5
The American Speech-Language Hearing Association (ASHA) has indicated that telehealth (telepractice) should be considered when attempting to address barriers to service provision and access. 3,4 While there is a growing body of work on the use of telehealth and its efficacy for children with speech, language and hearing disorders, 6–14 the published research has focused mainly on school-age children and there is little on children under three years of age. Furthermore, there has been very little research on speech, language and hearing services provided via telehealth to very young children in urban minority communities.
The purpose of the present study was to explore:
The feasibility of low-cost videoconferencing in urban community health clinics for speech, language and hearing screening for children up to six years of age; Family satisfaction with low-cost videoconferencing for screening for speech, language and hearing developmental milestones; The reliability of speech, language and hearing screening practices.
Methods
Two community-based urban health clinics were selected. The two clinics were part of the Northeast Ohio Neighborhood Health Services, Inc (NEON) and were located in an inner city area of Cleveland, Ohio. The clinics provide primary care services for patients of all ages, regardless of insurance status. The clinicians were located at the Cleveland Hearing and Speech Center (CHSC) and provided the clinical speech, language and hearing services by telemedicine. The organizations involved in the study had a large patient base of minority, low-income families.
Recruitment
The study was approved by the appropriate ethics committee. Telehealth involved one weekly three-hour block at each of the community health clinics from September 2007 to June 2009. Participants were not recruited during the summer months. Participants were families with children up to six years of age.
The facilitator at the two clinics was a student in speech-language pathology or audiology. Each facilitator was assigned for a 15-week semester and completed an initial training session with the supervising clinicians on the protocol for recruitment and screening at the beginning of the rotation.
Equipment
Computer kiosks were set up at the two clinics, each equipped with Dell laptops with 43 cm screens, web cameras (Microsoft Life Cam VX-3000) and Skype 3.8 for Windows. The same hardware and software was used at the clinician site.
The equipment used for the audiology screening was: (1) acoustic impedance audiometer (Earscan); (2) automatic handheld otoacoustic emissions instrument (OtoRead); (3) otoscope (Welch Allyn); (4) diagnostic audiometer (MedRx Avant A2D); (5) standard typmanometer (Maico Ero-Scan Pro DP).
Screening
Once identified, participants were told the purpose of the study and the general screening protocol by the clinic facilitator. Videoconferencing began once all the family's protocol questions had been answered and they had agreed to participate.
Screening began with a brief parental interview conducted by the speech-language pathologist and audiologist at the clinician site. The purpose of the interview was to identify any parental concerns and to obtain relevant medical, family and developmental information. Families then identified the components of the screening protocol that they were most interested in receiving: speech and language, hearing or both. The complete screening procedure lasted approximately 25 min per child.
Screening protocol – hearing
The audiology screening protocol consisted of two components: screening of middle ear health (typanometry) and screening of hearing acuity. During the first year, typanometry was performed by the clinic facilitator with supervision provided via videoconferencing. On completion, the facilitator relayed the results to the audiologist for realtime interpretation. In the second year, the facilitator used the PC-based tympanometer with the clinician site audiologist accessing the results using remote desktop computing (NTRconnect) for realtime interpretation.
The second component of the audiology screening was hearing acuity. For children aged up to three years (chronological or developmental) who could not provide a reliable behavioural response, Distortion Product Otoacoustic Emissions (DPOAE) were performed. In year 1, the clinic facilitator performed the DPOAE with the results provided to the audiologist via videoconferencing for realtime interpretation. In year 2, the clinic facilitator used a PC-based DPOAE screener and the clinician site audiologist accessed the results using remote desktop computing for realtime interpretation. The DPOAE screening was conducted at 2, 3, 4 and 5 kHz at 65 dB/55 dB with a signal-to-noise ratio for pass at 6.0 dB.
Behavioural audiometry was performed in children who were able to provide a behavioural response during the testing (approximately age three years and above). In year 1, the clinic facilitator used a portable audiometer and performed the screening with videoconferencing supervision by the audiologist. In year 2, the clinic facilitator used a PC-based audiometer and performed the screening via videoconferencing with the results automatically provided for realtime interpretation. In both years, a routine audiometry response (hand raise) was employed when the child was able or conditioned play audiometry was used (chip in a bucket) if necessary. The behavioural hearing acuity screening was conducted at 1, 2 and 4 kHz at 20 dB.
In year 2, the clinic facilitator re-screened a small number of children to determine the reliability of typanometry and hearing sensitivity testing by videoconferencing compared with client-site screening. The facilitator was supervised via videoconferencing by the clinician-site audiologist during the reliability testing.
Screening protocol – speech and language
The speech and language screening protocol focused on speech sound production (articulation), verbal expressive language, auditory receptive language and play/social behaviour. The battery of screening tools used depended on the child's age and their estimated speech and language abilities, based on information obtained during the family interview.
For the youngest children, aged up to three years or with speech and language abilities estimated to lie in that age range, the Receptive-Expressive Emergent Language Test – Third Edition (REEL-3) 15 was used to screen speech sound inventory, expressive and receptive language, and play behaviour. For children aged 2.5–4 years, the Screening Kit of Language and Development (SKOLD) 16 was used to test expressive and receptive language. The SKOLD was chosen because it allows for variations of both Standard American English and Black English. In the study population, Black English was as likely to be used in the home as Standard American English. In addition, the Preschool Language Scale 4 Articulation Screener (PLS-4), 17 was used to screen the speech sound inventory of children who were given the SKOLD. If the SKOLD could not be administered, an abbreviated version of the PLS-4 was used to screen expressive and receptive language and was used for the articulation screening.
In year 2, children were chosen for screening by the clinic facilitator to determine the reliability of speech and language screening by videoconferencing compared with client-site screening. The facilitator was supervised by the speech-language pathologist via videoconferencing during the reliability check. Children who were selected for the client-site only screening were matched to the larger videoconferencing screening group by age.
Follow-up and education
After the screening, each family was immediately provided with the results. A written summary of the results was given to the family and to the referring NEON physician by the clinic facilitator. For families with a child that did not pass the screening, referral information was provided for the appropriate early intervention resources.
To provide a framework for understanding typical speech, language and hearing developmental milestones, educational materials were also provided to families following completion of the screening. Each family was given the ASHA educational brochure ‘How Does Your Child Hear and Talk?’ Families were also offered the option of viewing the ASHA educational DVD ‘Speech, Language, and Hearing Milestones: Birth to Age Five.’
Satisfaction
The satisfaction survey was a seven-item, five-point Likert scale instrument with questions that explored satisfaction with the technical aspects of videoconferencing and satisfaction with videoconferencing as a method of service delivery. Each family was given the option of completing the survey on completion of the videoconferencing visit. A verbal survey was offered if there was a problem with literacy.
Reliability
Screening reliability was determined using a comparison of pass/fail rates for all screening components between those conducted via videoconferencing compared to those conducted on-site at the NEON clinics. Reliability for the audiology components was completed using the same child, but because of the potential learning effect, the speech-language components were completed on different children. Those included in the speech-language reliability testing were matched to the videoconferencing group by age.
Results
Screening
A total of 411 audiology/speech-language pathology videoconferencing screenings were conducted during the two-year study period. Because families could select the components of the screening protocol, not all children received a full audiology and speech-language screening. Of the total, 358 (87%) received hearing screening, 377 (92%) received tympanometry screening and 263 (64%) received speech and language screening only. In year 1, 33% of the children were three years of age or under (n = 60) and in year 2, 40% of the children were three years of age or under (n = 91).
For those that participated in the complete screening protocol, the highest percentage passed the hearing screening components and the lowest percentage passed their speech and language screening (Table 1). Of the families who had a child fail one or more components of the screening, 72% scheduled an appointment for a comprehensive evaluation appointment at the CHSC and kept that appointment.
Telehealth screening pass rates
Satisfaction
A total of 160 surveys were completed during the study. Each family completed only one survey regardless of how many of their children participated. Overall, the responding families strongly agreed that the quality of the technology was good (sound and picture) and that they generally felt very comfortable in using the computer, see Table 2. Families also reported a high level of satisfaction with the use of videoconferencing as a method of service delivery for the screening of speech, language and hearing abilities, and that they would be in favour of using videoconferencing again. Families generally reported a preference for speech, language and hearing screenings to be provided via videoconferencing during their primary care visit at the community health clinic, rather than having a separate appointment at another facility with a speech-language pathologist and/or audiologist.
Survey responses for satisfaction with telehealth screening. Families were asked to respond to the survey questions on a five-point scale, from 1 = Strongly disagree to 5 = Strongly agree
Reliability
Reliability for pure tone hearing screening (n = 7), DPOAE screening (n = 51) and speech-language screening (n = 10) was 100%. Typanometry screenings (n = 55) were 84% reliable between those administered using videoconferencing versus those administered at the client site.
Discussion
The present study indicates that low-cost videoconferencing for screening speech, language and hearing development in very young children in urban community health clinics is feasible, reliable and strongly supported by the community. Families were so supportive of the approach that they indicated a preference for videoconferencing screening rather than travelling to a separate provider. Flores and Tomany-Korman 5 found that community-based interventions were effective in improving access to and utilization of health-care services for urban minority children. The model used in the present study ensured that speech, language and hearing screenings were completed during an already-scheduled appointment with a paediatrician. This approach overcame some potential barriers to service utilization, including scheduling conflicts, transportation limitations and the need to schedule childcare for an additional appointment. It also streamlined communication with the primary care physician, because the screening results were immediately conveyed and available in the child's medical record. Furthermore, the primary physician could become an active participant in moving the child and family into the appropriate early intervention services, thereby enhancing the continuity of care.
The present study provides some of the first data on the use of videoconferencing for the screening of speech, language and hearing in children up to six years of age. The results are very encouraging. The study addressed some of the limitations in telehealth research in childhood speech, language and hearing disorders because a relatively large sample size was used 18 and the problem of reliability between videoconferencing and client site results was directly addressed. 19 Moreover, the study was completed in a real clinical environment, and so is potentially generalizable.
Limitations
The present work had some limitations. For example, the results may not be generalizable to a more heterogeneous population. Also, despite the relatively large sample size, the total number of cases used to calculate reliability was small. Furthermore, reliability was not calculated for each individual test administered in the speech and language screening protocol. There was also a discrepancy between the reliability measures obtained for the typanometry component of the screening compared to the other screening measures. Finally, the study focused on the use of screening and therefore a full assessment battery was not offered.
Future directions
In order to be able to provide robust clinical recommendations for the use of videoconferencing for the early intervention screening of speech, language and hearing disorders for children aged up to six years, various matters need to be addressed. First, random selection of the experimental and control group should be used. Second, a properly-powered trial should be conducted. Third, a systematic evaluation of the screening measures is needed to identify those most appropriate for use in videoconferencing. Fourth, it will be important to compare the results from a graduate student facilitator with the results obtained from a facilitator who may not have direct training in speech-pathology and/or audiology.
Conclusion
The present study provides a foundation for improving access for urban families to early intervention speech, language and hearing services using a combination of community health clinics and videoconferencing. This service delivery method overcomes many of the barriers to health care access for urban, minority and low income communities.
Footnotes
Acknowledgements
The study was funded by the St Luke's Foundation and the Hershey Foundation in a grant to the CHSC that was clinically coordinated by Kay McNeal and Bridgid Whitford. We thank the staff at the NEON Hough and Southeast Clinics for their assistance and the families who participated in the study. We also thank Frank Ballish, Bernard Henri and Michelle Burnett (CHSC) and Lia Santiago, Matthew Brown, Adrienne Schiffman, Kris Baumann, Mallory Doup, Will Farnham, Lindsay Kodramaz, Rachel Wellman and Leah Wondrak.