Clinical UM Guideline

This document addresses augmentative and alternative communication (AAC) devices with digitized or synthesized speech output. Digitized speech output refers to natural speech that is recorded and stored and then reproduced by the device. In contrast, synthesized speech devices translate user input into machine-generated speech and thus are not dependent on pre-stored messages. These devices are aids to improve the functional communication needs of individuals with severe speech impairment or absent speech. Associated functional disabilities may limit an individual’s ability to use alternative natural methods of communication such as writing notes, using sign language, or to manipulate a low technology augmentative communication system.

Note: For a high-level overview of this document, please see “Summary for Members and Families” below. 

Medically Necessary:

Augmentative and alternative communication devices with digitized or synthesized speech output are considered medically necessary when all of the following criteria A through C are met, and when applicable, criteria D or E are met:

1. The device has been recommended by the individual’s physician and licensed speech language pathologist who have each conducted and documented a thorough assessment which includes all of the following information:
   1. Medical diagnosis, physiological description of the underlying disorder, description of functional limitation, nature and severity of speech or communication impairment, and prognosis for improvement (or deterioration); and
   2. Therapeutic history including speech, occupational, or physical therapies as appropriate; and
   3. Documentation of the cognitive ability to utilize the selected device including, when appropriate, results of at least one validated cognitive and/or developmental test; and
   4. Documentation of the visual, auditory, language and motor ability to utilize the selected device including results of any test(s) performed; and
   5. Documentation of the specific daily functional communication needs including number of words or sounds used without a device at baseline; and
   6. Expected functional communication goals with the device; and
   7. Plan of care for the device including: anticipated training needs for the individual and caregiver(s), programming needs and planned evaluations;
      and
2. The individual has severe expressive speech impairment and alternative natural communication methods such as writing or sign language are not feasible or are inadequate for that individual’s daily functional communication needs;
   and
3. The individual has tested the device for at least 60 days, has demonstrated the ability to use the device and there is documentation of the rationale for the specific device selected which should include the following elements:
   1. Duration of device trial (number of trials and length of sessions, total duration in days); and
   2. Communication task(s) evaluated (such as initiating communication, responding to questions, making requests, effectively expressing wants, needs, and ideas, participating in conversations); and
   3. Language functions evaluated (such as making requests, initiating and responding to greetings, expressing feelings, and asking basic/functional questions); and
   4. Type and number of symbols/pictures and/or words used with each device trial; and
   5.  Extent to which individual can independently navigate the device.
4. If the individual has a degenerative disease causing the speech impairment, the communication device selected should be capable of modifications necessary to meet the individual’s anticipated needs.
5. If the individual is preliterate, the device should be capable of modifications such as spelling and text capabilities to meet the individuals anticipated learning potential.

Accessories are considered medically necessary if criteria for the base device are met and the medical necessity for each accessory is clearly documented in the formal evaluation by the speech language pathologist. For any subsequent upgrade of equipment or accessories to a previously issued device, information regarding the functional benefit to the individual of the upgrade compared to the initially provided device must be submitted to demonstrate medical necessity.

When the above criteria A through C are met, and when applicable, criteria D or E are met, specific communication software for dedicated digitized or synthesized speech generating devices is considered medically necessary.

Not Medically Necessary:

Synthesized and digitized speech generating devices are considered not medically necessary if the above criteria are not met or if they are not primarily and customarily used to serve an augmentative communication function.

The following are considered not medically necessary:

1. Devices that are not dedicated speech devices, but are devices that are capable of running software for purposes other than for speech generation, for example, devices that can also run a word processing program or perform other non-augmentative communication functions.
2. Laptop, tablet or desktop computers, smartphones or other devices which may be programmed to perform the same function as a speech generating device.

This document describes clinical studies and expert recommendations, and explains when augmentative and alternative communication (AAC) devices with digitized or synthesized speech output are clinically appropriate. The following summary does not replace the medical necessity criteria or other information in this document. The summary may not contain all of the relevant criteria or information. This summary is not medical advice. Please check with your healthcare provider for any advice about your health.

Key Information

AAC devices are tools that help people express wants, needs, feelings, and ideas with digitized or synthesized speech when their speech abilities are very limited or absent and when other communication methods such as writing or signing do not meet their daily needs. Some devices play pre-recorded messages, while others create speech from typed words, symbols, or pictures using a keyboard or other interface. These devices can help children and adults with conditions that affect speech, such as autism, cerebral palsy, stroke, Parkinson disease, amyotrophic lateral sclerosis (ALS), traumatic brain injury, or other conditions. Professional organizations support AAC use and recommend evaluation by a speech-language pathologist (SLP) and other healthcare professionals to help select the most appropriate device.

What the Studies Show

Studies have generally found that AAC devices improve communication skills. Reviews of research in children with autism and developmental disabilities found better communication, especially for making requests and participating in interactions. Research in adults with speech problems after stroke found positive or mixed results. A 2023 study found that both high-technology and low-technology AAC devices improved communication skills, with no clear advantage for one type over the other. Some studies had limitations, including small numbers of participants, differences in study methods, and limited information about long-term results. Better studies are needed to know if some AAC approaches improve health more than others.

When is Augmentative and Alternative Communication with Digitized or Synthesized Speech Clinically Appropriate?

AAC devices with digitized or synthesized speech may be appropriate in these situations:

• A physician and a licensed speech-language pathologist have completed and documented a detailed evaluation; and
• The evaluation includes the medical condition, communication problems, functional limitations, communication goals, and a plan for training and follow-up; and
• The person has the cognitive, visual, hearing, language, and motor abilities needed to use the device; and
• The person has severe difficulty expressing speech, and writing, signing, or other natural communication methods are not possible or do not adequately meet daily communication needs; and
• The person has completed a device trial lasting at least 60 days and has shown the ability to use the device; and
• Documentation explains why the selected device is the best choice based on communication tasks, language functions, symbols or words used, and the person's ability to navigate the device.

Additional considerations include:

• For progressive conditions that may worsen over time, the device should be able to be modified to meet future communication needs; or
• For people who are not yet able to read or write, the device should support future learning needs, such as spelling or text-based communication features.

Accessories and communication software may also be appropriate when the base device meets the criteria above and there is documentation showing why the accessory or software is needed.

When is this not Clinically Appropriate?

AAC devices with digitized or synthesized speech are not clinically appropriate when the criteria above are not met.

The following are also not clinically appropriate:

• Devices that are not mainly designed for augmentative communication and are commonly used for other purposes; or
• Laptop computers, tablet computers, desktop computers, smartphones, or similar devices that can perform many functions beyond speech generation.

These devices are not considered clinically appropriate because they do not meet the requirements for dedicated AAC speech-generating devices (SGDs). Studies discussed in the document focus on AAC systems designed to support communication needs. Better studies are needed to know if broader multipurpose devices improve health when used in place of dedicated AAC devices.

(Return to Description)

The following codes for treatments and procedures applicable to this document are included below for informational purposes. Inclusion or exclusion of a procedure, diagnosis or device code(s) does not constitute or imply member coverage or provider reimbursement policy. Please refer to the member's contract benefits in effect at the time of service to determine coverage or non-coverage of these services as it applies to an individual member.

When services may be Medically Necessary when criteria are met:

When services are Not Medically Necessary:
For the procedure codes listed above when criteria are not met or for situations designated in the Clinical Indications section as not medically necessary.

Summary

Speech-generating devices (SGDs), including synthesized and digitized technologies, are designed to enable individuals with severe or absent speech to meet their communication needs, especially when natural methods like writing or signing are not feasible. Digitized SGDs use pre-recorded messages, while synthesized SGDs allow users to create messages independently. Evidence supporting SGDs largely comes from small studies and systematic reviews of these studies. Studies have shown positive communication outcomes for children with autism and individuals with intellectual disabilities and mixed results for adults with post-stroke aphasia. A 2023 randomized controlled trial (RCT) found both high- and low-tech augmentative and alternative communication (AAC) devices significantly improved communication in children with autism and intellectual disability, with no clear advantage of high-tech over low-tech options.

Discussion

According to the International Society for Augmentative and Alternative Communication (ISAAC):

AAC is a set of tools and strategies that an individual uses to solve everyday communicative challenges. Communication can take many forms such as: speech, a shared glance, text, gestures, facial expressions, touch, sign language, symbols, pictures, speech-generating devices, etc. Everyone uses multiple forms of communication, based upon the context and our communication partner. Effective communication occurs when the intent and meaning of one individual is understood by another person. The form is less important than the successful understanding of the message.

Speech aids such as SGDs can provide individuals with severe speech impairment or absent speech the ability to meet their functional communication needs. Etiologies of speech impairment in children may include cerebral palsy, intellectual/developmental disorder, autism-like disorders and other genetic or speech disorders. Etiologies in adults may include stroke, traumatic brain injury, amyotrophic lateral sclerosis (ALS), Parkinson disease and head and neck cancer among others. There may be associated functional disabilities that also limit the individual’s ability to use alternative natural methods of communication such as writing notes, using sign language, or to manipulate a low technology augmentative communication system.

Digitized SGCs, sometimes referred to as devices with “whole message” speech output, use words or phrases that have been recorded by an individual other than the SGD user for playback upon command of the SGD user. Synthesized speech is a technology that translates user’s input into device-generated speech using algorithms representing linguistic rules. Users of synthesized SGDs are not limited to pre-recorded messages but instead can independently create messages as their communication needs dictate. Some SGDs require a message formulation by spelling, and access by physical contact with a keyboard, touch screen, or other display containing letters. Speech generating software programs enable a laptop, tablet, desktop computer or mobile device to function as SGDs.

A 2018 meta-analysis by Muharib and Alzrayer evaluated studies on use of high-technology SGDs in children with autism spectrum disorder (ASD). The authors identified 20 single-case studies with a total of 54 young children with ASDs and assessed the effectiveness of high-tech SGDs in improving communication skills. In nearly all of the studies, the SGDs were applications used on an iPad or iPod. In 13 out of 18 studies, use of SGDs had a strong effect in teaching verbal behaviors and in another 4 studies, the SGDs were moderately effective in improving verbal skills such as requesting (manding), labeling (tacting), and conversational responses (intraverbals). Effects on vocal speech production were limited. The quality of the evidence was mixed, with only 2 high-quality studies and 16 rated as acceptable, resulting in a moderate level of evidence supporting SGDs as an intervention. Limitations included variability in study quality, limited assessment of generalization and social validity, and a predominance of highly structured teaching settings rather than natural environments.

In a systematic review, Therrien (2025) evaluated 20 intervention studies examining the use of SGDs as part of AAC for autistic preschoolers, focusing on expressive communication outcomes. Overall, 55% of the studies demonstrated moderate to strong effects, indicating that SGDs can effectively support communication development, particularly for requesting behaviors, which was the most commonly targeted function. Interventions were typically multicomponent and relied heavily on strategies such as prompting and modeling. There was limited focus on broader communication functions beyond requesting. Notably, fewer than half of the studies demonstrated strong ecological validity and only a minority assessed social validity, highlighting concerns about real-world applicability.

A 2017 meta-analysis by Ganz and colleagues evaluated 24 single-case studies examining the use of high-technology AAC, including SGDs and mobile applications, for individuals with intellectual and developmental disabilities and complex communication needs. Studies differed in the interventions they evaluated and the outcome variables they measured. All of the interventions provided statistically significant benefits, compared with baseline, and the overall pooled effect size was 0.70 (95% confidence interval [CI]; 0.63 to 0.77). Limitations include small sample sizes, limiting the ability to draw firm conclusions for certain variables and the analysis included only single-case designs, as effect sizes could not be combined with group studies, which restricts generalizability. Different disability populations were grouped together, making it difficult to determine whether outcomes differ by specific diagnoses (for example, autism compared to other developmental disabilities). Additionally, the study did not assess maintenance or generalization outcomes, focusing only on acquisition, and noted that available data in the literature were insufficient for such analyses. Lastly, the inclusion of only peer-reviewed studies introduces potential publication bias, as studies with negative findings may be underrepresented.

A systematic review of studies on high-technology AAC devices for adults with post-stroke aphasia was published by Russo and colleagues in 2017. The review included 30 publications and included a total of 250 individuals with acquired non-progressive post-stroke aphasia. Study sample sizes ranged from 1 to 10. AAC included desktop computer software (n=11), portable computer software (n=9), dedicated AAC devices (n=6) and mobile devices with communication apps (n=4). Results were largely positive or mixed, indicating that high-tech AAC can enhance communication abilities, particularly for everyday functions such as conversation, storytelling, and written communication. However, the evidence base was highly heterogeneous, with small sample sizes, variable participant characteristics, inconsistent outcome measures, and limited reporting of key factors such as device acceptance and caregiver involvement. Overall, while high-tech AAC shows promise as a compensatory strategy to improve communication and social participation, the review highlights the need for more rigorous, standardized research designs and outcome reporting to better guide clinical implementation and decision-making.

A systematic quality review evaluated 23 studies (including 18 multiple-baseline/multiple-probe experiments across 17 studies) examining high-tech AAC interventions for individuals with ASD or intellectual disability and complex communication needs (Morin, 2018). Overall, the evidence met established criteria for an evidence-based practice, with 12 experiments demonstrating strong or moderate evidence across a total sample of 33 participants, exceeding methodological thresholds for replication and sample size. Most studies showed improvements in social-communication skills, particularly requesting, although only one experiment demonstrated strong evidence, with the remainder largely showing moderate or weak effects based on visual analysis of single-case data. Importantly, comparison (alternating treatment) designs indicated that high-tech AAC was not consistently superior to low-tech AAC, suggesting comparable effectiveness across modalities. Despite generally positive findings, methodological limitations were common, including inconsistent reporting of generalization, maintenance, procedural integrity, and social validity, highlighting the need for more rigorous and comprehensive study designs.

A systematic review evaluated 29 studies examining AAC interventions in children aged 0-6 years with diverse developmental and communication needs (Leonet, 2022). Overall, findings demonstrated that AAC interventions, across aided (low- and high-tech) and unaided modalities, consistently improved multiple domains of communication, including expressive and receptive language, functional communication behaviors, interaction skills, and symbol use. Despite considerable heterogeneity in study designs, participant characteristics, and intervention approaches, positive outcomes were reported across nearly all studies, supporting early introduction of AAC. Methodological limitations were common (for example, inconsistent baseline reporting, limited replication and generalization data), and no single “best” intervention approach emerged, highlighting the importance of individualized, multimodal AAC strategies and further high-quality research to guide practice.

A 2023 meta-analysis by Pak evaluated 19 single-case studies involving 66 young children with complex communication needs to compare the effectiveness of SGDs with other AAC modes, including picture exchange and manual sign language. Overall, findings indicated that SGDs and picture exchange systems were similarly effective for teaching requesting behaviors, with no significant differences in performance across these modalities (m=15 studies, n=49; log response ratio [LRR], −0.20, p=.46). SGDs were more effective than manual sign language (m=7 studies, n=20; LRR, 2.38, p<.001). Results also suggested that individual preference influenced outcomes, as individuals tended to learn more efficiently with their preferred system. Most studies focused narrowly on simple requesting tasks in structured settings, limiting conclusions about broader communication skills or real-world use. The evidence supports the effectiveness of aided AAC systems overall, while emphasizing the importance of individualized selection and the need for further research on diverse communication functions, participant characteristics, and long-term outcomes.

In 2023, Gilroy and colleagues published an RCT evaluating AAC use in school-age children with co-occurring autism and intellectual disability. Children were randomized at the classroom level to either a high-technology tablet-based speech-generating AAC device or low-technology AAC using laminated picture cards. In addition, there was a waitlist-control group that received no intervention. Analysis was done using linear mixed-effects modeling using data after 3 months of the intervention. Modeling found that the children who received any AAC device had greater acquisition of communication skills than those in the waitlist-control group. Modeling did not find a significant effect of high-technology when compared to low-technology devices in terms of communication skill acquisition.

In 2025, Desch published an American Academy of Pediatrics (AAP) clinical report that provides guidance on the evaluation, prescription, funding, and implementation of AAC technologies for children with complex communication needs. The report notes that data from the 2021-2022 National Survey of Children’s Health found that nearly 10% of children aged 3-17 years had speech or language disorders. It notes that approximately 30% of children with ASD without severe intellectual disability have complex communication needs, and that 60-85% of children with cerebral palsy experience significant communication impairments. The report reviews the spectrum of AAC options, including low-tech systems (for example, communication boards and Picture Exchange Communication System [PECS]), mid-tech devices, and high-tech SGDs. The AAP emphasizes that there are no prerequisite skills required before a child can be evaluated for AAC and that decades of research support AAC as an effective intervention that promotes communication and language development rather than hindering speech acquisition. The speech-language pathologist (SLP), using their clinical expertise and experience, identifies the device or devices most likely to meet the individual's communication needs, with input from other members of the interdisciplinary team as appropriate. Preferably, the individual should participate in a trial period using a loaned or rented version of the recommended device before a final purchase decision is made. A trial period of approximately 1 to 2 months allows the individual, family members, educators, and caregivers to receive training and evaluate the device's effectiveness across various communication settings, while identifying the system's strengths, limitations, and overall suitability for the individual's needs. The report cites systematic reviews demonstrating that AAC use generally improves communication outcomes, vocabulary development, and in some cases receptive language and social interaction skills. It also cites evidence that up to 75% of assistive devices may ultimately be abandoned or unsuccessfully used, underscoring the importance of appropriate assessment, training, and follow-up. After implementation, ongoing monitoring, training, and guidance are recommended to support communication and literacy development. For school-aged children, federal regulations require formal re-evaluations every 3 years, unless an earlier review is requested by the family or educational team. Reassessment of the AAC system should be considered when new behavioral concerns or changes in communication needs are identified. The authors recommend a multidisciplinary evaluation process led primarily by a SLP, with involvement from physicians, therapists, educators, assistive technology specialists, and families to ensure appropriate device selection, training, and long-term support.

Digitalized speech: Speech that has been converted from an analog sound wave into a digital format that can be stored, processed, transmitted, or reproduced by electronic devices.

Laryngectomy: Partial or complete surgical removal of the larynx (voice box).

Speech disorder: A condition affecting the ability to produce normal speech may affect articulation (phonetic or phonological disorders); fluency (stuttering or cluttering); and/or voice (tone, pitch, volume, or speed) resulting in difficulty communicating through spoken language; most speech disorders have their roots in the muscles of the mouth and/or mouth movements.

Speech language pathologist (Speech Therapist): A licensed healthcare professional who evaluates, diagnoses, and treats disorders related to speech, language, voice, fluency, cognitive-communication, and swallowing.

Synthesized speech: Artificially generated speech produced by a computer or electronic device using text-to-speech (TTS) technology or other speech synthesis methods, rather than being recorded from a human speaker.

Peer Reviewed Publications:

1. Desch LW, Hobart-Porter LJ, Houtrow A; Council on Children With Disabilities. Prescribing assistive technology: focus on children with complex communication needs: clinical report. Pediatrics. 2025; 156(1):e2025072216.
2. Ganz JB, Morin KL, Foster MJ, et al. High-technology augmentative and alternative communication for individuals with intellectual and developmental disabilities and complex communication needs: a meta-analysis. Augment Altern Comm. 2017; 33:224-238.
3. Gilroy SP, McCleery JP, Leader G. A delayed intervention start randomized controlled trial of high- and low-tech communication training approaches for school-age autistic children with co-occurring intellectual disability. J Appl Behav Anal. 2023; 56(3):593-606.
4. Leonet O, Orcasitas-Vicandi M, Langarika-Rocafort A, et al. A systematic review of augmentative and alternative communication interventions for children aged from 0 to 6 Years. Lang Speech Hear Serv Sch. 2022; 53(3):894-920.
5. Morin KL, Ganz JB, Gregori EV, et al. A systematic quality review of high-tech AAC interventions as an evidence-based practice. Augment Altern Commun. 2018; 34(2):104-117.
6. Muharib R, Alzrayer NM. The use of high-tech speech-generating devices as an evidence-based practice for children with autism spectrum disorders: a meta-analysis. Review Journal of Autism and Developmental Disorders 2018; 5(1):43-57.
7. Pak NS, Bailey KM, Ledford JR et al. Comparing interventions with speech-generating devices and other augmentative and alternative communication modes: a meta-analysis. Am J Speech Lang Pathol. 2023; 32(2):786-802.
8. Russo MJ, Prodan V, Meda NN, et al. High-technology augmentative communication for adults with post-stroke aphasia: a systematic review. Expert Rev Med Dev 2017; 14:355-370.
9. Therrien MCS, Whalon K, Nunes D, et al. A Systematic review of AAC interventions using speech generating devices for autistic preschoolers. Augment Altern Commun. 2025; 41(2):114-128.

Government Agency, Medical Society, and Other Authoritative Publications:

1. Centers for Medicare and Medicaid Services. National Coverage Determination: Electronic Speech Aids. NCD #50.2. Available at: https://www.cms.gov/medicare-coverage-database/details/ncd-details.aspx?NCDId=237&ncdver=1&SearchType=Advanced&CoverageSelection=National&NCSelection=NCA%7CCAL%7CNCD%7CMEDCAC%7CTA%7CMCD&KeyWord=radiation&KeyWordLookUp=Doc&KeyWordSearchType=Exact&kq=true&bc=IAAAACAAAAAA&. Accessed on August 13, 2026.
2. Centers for Medicare and Medicaid Services. National Coverage Determination: Speech Generating Devices. NCD #50.1. Effective July 29, 2015. Available at: https://www.cms.gov/medicare-coverage-database/view/ncd.aspx?ncdid=274&ncdver=2&bc=AAAAEAAAAQAAAAAA&=. Accessed on  August 13, 2026.
3. International Society for Augmentative and Alternative Communication (ISAAC). About AAC. Available at: https://isaac-online.org/english/about-aac/. Accessed on August 13, 2026

1. National Institute on Deafness and Other Communication Disorders (NIDCD). Assistive devices for people with hearing, voice, speech or language disorders. November 2019. Available at: http://www.nidcd.nih.gov/health/hearing/Pages/Assistive-Devices.aspx?nav=update. Accessed on August 13, 2026.

Digital Speech
Speech Impairment
Synthesized Speech

Federal and State law, as well as contract language, and Medical Policy take precedence over Clinical UM Guidelines. We reserve the right to review and update Clinical UM Guidelines periodically. Clinical guidelines approved by the Medical Policy & Technology Assessment Committee are available for general adoption by plans or lines of business for consistent review of the medical necessity of services related to the clinical guideline when the plan performs utilization review for the subject. Due to variances in utilization patterns, each plan may choose whether to adopt a particular Clinical UM Guideline. To determine if review is required for this Clinical UM Guideline, please contact the customer service number on the member's card.

Alternatively, commercial or FEP plans or lines of business which determine there is not a need to adopt the guideline to review services generally across all providers delivering services to Plan’s or line of business’s members may instead use the clinical guideline for provider education and/or to review the medical necessity of services for any provider who has been notified that his/her/its claims will be reviewed for medical necessity due to billing practices or claims that are not consistent with other providers, in terms of frequency or in some other manner.

No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, photocopying, or otherwise, without permission from the health plan.

© CPT Only - American Medical Association