Outcome measures for children with speech sound disorder: an umbrella review

Abstract

Objective

Speech sound disorder (SSD) describes a ‘persistent difficulty with speech sound production that interferes with speech intelligibility or prevents verbal communication’. There is a need to establish which care pathways are most effective and efficient for children with SSD. Comparison of care pathways requires clearly defined, evidence-based, interventions and agreement on how to measure the outcomes. At present, no definitive list of assessments, interventions or outcomes exists. The objective of this umbrella review paper is to provide a rigorous and detailed list of assessments, interventions and outcomes which target SSD in children.

Design

In December 2022, a systematic search of Ovid Medline, OVID Embase, CINAHL, PsycInfo and Cochrane and a number of grey literature platforms were undertaken. 18 reviews were included, and subsequently 415 primary research articles were assessed for data related to assessments, interventions or outcomes. The AMSTAR (Assessing the Methodological Quality of Systematic Reviews) framework was used to assess the quality of the retained reviews.

Setting

Reviews were retained which took place in any setting.

Participants

The population is children of any age with a diagnosis of SSD of unknown origin.

Primary and secondary outcome measures

Reviews reporting outcomes, assessment and interventions for children with SSD.

Results

Extraction and analysis identified 37 assessments, 46 interventions and 30 outcome measures used in research reporting of SSD. Not all of the listed outcomes were linked to specific outcome measurement tools, but these were measurable through the use of one or more of the assessments extracted from the retained reviews.

Conclusions

The findings of this review will be used to develop a Core Outcome Set for children with SSD. The findings are part of a rigorous process essential for advancing healthcare research and practice in the specific area of speech and language therapy for children with SSD.

PROSPERO registration number

CRD42022316284.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The umbrella review is designed to collate existing systematic reviews.

• This umbrella review follows the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols guidelines.

• Critical appraisal of included reviews was undertaken using the AMSTAR.

• Appraisal of the 415 individual primary research papers was not undertaken.

Introduction

Speech sound disorder (SSD) is a broad category of speech disorders that involve difficulties in producing and/or using speech sounds correctly.¹ SSD can be broadly categorised into two main types: articulation/motor speech disorders and phonological disorders. Articulation/motor speech disorders refer to difficulties in the movements required for producing speech sounds. Phonological disorders, on the other hand, involve difficulties with the rules that govern the use of speech sounds in a particular language.

The reported prevalence of SSD varies depending on the population studied and the definition of SSD used. However, there is agreement that SSD is a common communication disorder in children, affecting approximately 8%–9% of preschool-aged children and 2%–3% of school-aged children.^2–5 SSD is more common in boys than girls and may be more prevalent in certain populations, such as children with a history of ear infections or language delays.

The underlying causes of SSD are complex and multifactorial. Several factors have been identified as contributing to the development of SSD, including genetic, neurological, environmental and psychosocial factors. However, it is a minority of children with SSD that has a clear aetiology (eg, cerebral palsy, cleft palate +/− cleft lip and hearing impairment, genetic causes, childhood apraxia of speech). In most cases, SSD has no identifiable cause, and the evidence for intervention in this particularly heterogeneous group is limited.

SSD can have significant negative consequences for a child’s communication and academic development. If left untreated, SSD can lead to a range of negative sequelae, including social and emotional problems, reduced quality of life and academic difficulties.^6–12 The latter are a common consequence of SSD, as affected children may struggle to acquire reading and writing skills, leading to poor academic performance and reduced educational attainment.¹³ This may be exacerbated for those children with co-occurring developmental language disorder,¹⁴ which can affect a child’s ability to understand and follow instructions in the classroom, further impacting their academic performance.

In the UK, publicly funded, free at the point of access National Health Service (NHS) speech and language therapy is provided to children with SSD via a range of care pathways, typically defined by resource constraints, rather than robust evidence.¹⁵ Vanhaecht et al defined a care pathway as “a complex intervention for the mutual decision-making and organisation of care processes for a well-defined group of patients during a well-defined period”.¹⁶ Care pathways aim to improve care, outcomes and patient satisfaction while also optimising the use of resources.

There is a need to establish which care pathways are most effective and efficient for children with SSD. Comparison of care pathways requires both clearly defined, evidence-based interventions and agreement on how best to measure the outcomes of these interventions for children with SSD. However, a review of existing NHS case notes of children treated for SSD was found to be too incomplete to compare pathways.¹⁷ Morgan et al suggest that there is a need for agreement on a national UK-wide core outcome set for SSD.¹⁷ A national consultation in 2018 identified the need to collect consistent data and recommended that NHS England support providers to collect data on the quality and outcomes of interventions (recommendation 4.5, p29).¹⁸

Evidence from systematic reviews and trials has shown that intervention is effective for the majority of children with SSD and that these children do not make progress without intervention.^{19 20} However, a challenge with studies that have investigated the impact of interventions is that they have typically employed intervention protocols which are intense and difficult to replicate in NHS speech and language therapy services because of resource constraints and service variations.^{15 21 22} Importantly, unlike research studies, clinical intervention takes place within care pathways which vary in terms of timing of intervention (eg, preschool, school age), agent of intervention (eg, speech and language therapists (SLTs), speech and language therapy assistants, teaching assistants, parents), dosage (number, frequency and duration of sessions) and involvement of parents or education staff, as well as the assessments and outcome measures used.

To determine which care pathways are most effective for children with SSD, there is a need to compare outcomes across NHS speech and language therapy services. Given the variation observed in how outcomes are measured in routine clinical care,¹⁷ the first step is to determine what outcomes are important for service users and clinicians. Functional goals such as independence and improved social interaction have been identified as of greatest importance to parents,¹⁵ while children have indicated that improved speech, schoolwork and skill at sports as well as making friends are their most important goals.^{23 24} Preferred outcomes for preschool children with SSD among SLTs have been identified as intelligibility, social interaction and participation.²⁵

It is important to understand how these outcomes can relate to systems of classification, which might be used by healthcare professionals. The International Classification of Functioning, Disability and Health (ICF) is a framework developed by the WHO to provide a comprehensive understanding of health and health-related issues, which can reflect some of the preferred outcomes outlined above. The principles and structure of the ICF were used in the development of the Speech Outcome Reporting Taxonomy (SORT).²⁶ SORT is a tool used to document and standardise the assessment of speech outcomes in clinical and research settings. It provides a way to link clinical assessment and intervention with the ICF, making it easier to track, compare and report changes in speech and communication abilities over time. However, even with development of this type of system, clinicians and researchers still do not have a defined and agreed set of measures mapped to the facets of the framework.

However, frameworks can help establish what outcomes are important, and the next step is then to determine how these should be measured. Systematic reviews are a useful source of evidence for interventions for SSD and are increasing in frequency in published literature. They provide a rigorous and transparent knowledge base for translating clinical research into decisions and as such are ‘go to’ documents to advise healthcare service construction and evaluation. These are also a useful potential source for considering the range of assessments which are available and could be used to measure outcomes in any investigation comparing effectiveness of different care pathways. Given the number of existing systematic reviews in this field, an overarching review which captures the relevant information across multiple reviews is needed. An umbrella review is a type of systematic review that synthesises and evaluates the findings of multiple scoping reviews, systematic reviews and meta-analyses on a specific topic.^{27 28} Unlike traditional systematic reviews that focus on a single research question, umbrella reviews aim to provide a comprehensive overview of the existing evidence on a particular topic, including the quality and consistency of the evidence, gaps in the literature and areas that require further investigation.

The objective of this umbrella review was to collate the tools used for assessment and outcome measurement with children with SSD in speech and language therapy. To consider the range of possible interventions which might be used in care pathways, a secondary objective was to determine what interventions for SSD are described and defined in the reviews.

Methods and analysis

The review was conducted in accordance with the Joanna Briggs Institute (JBI) methodology for umbrella reviews,²⁹ with the addition of undertaking quality appraisal using the AMSTAR tool.³⁰ It was registered with PROSPERO (CRD42022316284). The full protocol was published in 2023 in BMJ Open,³¹ including the OVID Medline search strategy as an example.

Eligibility criteria

In line with the JBI guidance, the eligibility for inclusion in the review was undertaken using the concepts of population, phenomena of interest and context of data.²⁹ As this is an umbrella review, the only papers retained for inclusion were peer-reviewed published reviews. This included any type of review, for example, systematic reviews of effectiveness, mixed methods, qualitative and scoping reviews.²⁸

Population

The population is children of any age with a diagnosis of SSD of unknown origin. Studies were not excluded if the interventions included additional therapy targets (eg, for receptive language). Children whose speech sound needs were associated with a biomedical condition with a known association with communication, such as sensorineural deafness, autistic spectrum condition or cleft palate and neurological conditions (eg, cerebral palsy) affecting speech output, were excluded.

Phenomena of interest (concept)

To be included, reviews must have assessed children with SSD or the outcomes of intervention for children with SSD. This included articulation disorders, childhood apraxia of speech (formerly known as developmental verbal dyspraxia in the UK) or phonological disorders/delay. It excludes children with a known cause for their SSD, such as those with identified genetic or chromosomal anomalies, and congenital or acquired neurological conditions, often associated with childhood dysarthria.

Context

The context for included reviews was left open in that we considered reviews that retained studies which took place in any setting (eg, home, clinic, school) and geographical location (including outside of the UK).

Information sources

As the aim of this umbrella review was to provide a long list of assessments, outcomes and outcome tools (measures) used in the evaluation of SSD in children, it did not exclude relevant studies on account of their review methodology. The complete search was undertaken in December 2022 using Ovid Medline, OVID Embase, CINAHL, PsycInfo and Cochrane. These databases were selected to cover a broad range of journals pertaining to medicine, psychology (including child development) and the allied health professions.

In addition to these standard journal databases, other platforms were interrogated including Campbell Collaboration, COSMIN, Figshare, JBI, OSF, PROSPERO and Speechbite. Due to a limitation in resources, included studies were those published in English. To include literature relevant to current speech and language therapy practice, the search had a minimum publication year of 2010 (1 January 2010).

Search strategy

Following JBI protocol development guidance, an initial limited search of two databases was conducted prior to the full search being carried out.²⁹ A set of key terms was developed by the first author (SH), in consultation with coauthors who are subject experts with significant postdoctoral research experience in the area (JC, HS and YW). These terms were used for the initial limited search of Ovid Medline and Ovid Embase to identify articles on the topic. With the support of a clinical librarian, the text words contained in the articles and abstracts of relevant articles and the index terms used to describe the articles were used to develop a full-search strategy for Medline. Box 1 presents the full-search strategy for Medline, and online supplemental material 1 contains all search strings for the other databases. This search strategy was adapted for each selected database as appropriate. The reference list of each of the included sources of evidence was screened for additional studies.

Box 1. Full search strategy for Medline.

1. (child* or youth* or boy* or girl* or juvenil* or teenage* or adolescen* or “young person*” or “young people*” or toddler* or infan* or baby or babies).mp.

2. Child/ or Adolescent/ or Infant/ or Infant, Newborn/

3. 1 or 2

4. (phon* or speech or speech disorder* or speech impairment* or speech sound disorder* or speech sound difficult* or speech-sound* or speech retard* or speech delay* or speech disabilit* or speech handicap* or speech problem* or childhood apraxia of speech or apraxia of speech or developmental verbal dyspraxia or verbal dyspraxia or dyspraxia or articulat*).ti,ab.

5. exp Speech Sound Disorder/

6. 4 or 5

7. (“clinical service*” or “therap* service*” or NHS or “social care” or “social service*” or school* or education* or nurser* or “early year*” or preschool* or pre-school* or college* or universit*).mp.

8. Schools/ or Universities/ or Nurseries, Infant/ or Child, Preschool/ or Social Support/

9. 7 or 8

10. (exp META-ANALYSIS AS TOPIC/ or (“meta analy*” or “metaanaly*”).ti,ab. or META-ANALYSIS/ or (systematic adj1 (review*1one or overview*1)).ti,ab. or exp REVIEW LITERATURE AS TOPIC/ or (cochrane or embase or psychlit or psyclit or psychinfo or psycinfo or cinahl or cinhal or “science citation index” or bids or cancerlit).ab. or (“reference list*” or bibliograph* or hand-search* or “relevant journals” or “manual search*“).ab. or ((“selection criter*” or “data extraction”).ab. and exp REVIEW/)) not ((ANIMALS/ not (ANIMALS/ and exp HUMANS/)) and (COMMENT/ or LETTER/ or EDITORIAL/ or (letter* or comment*1one or editorial*1).ti,ab.))

11. 3 and 6 and 9 and 10

12. limit 11 to (english language and yr=“2010 -Current”)

Study/source of evidence selection

All identified citations were collated and uploaded into EndNote and duplicates removed. Inclusion and exclusion criteria are presented in table 1. The remaining citations were then downloaded and entered into the online review management software, Rayyan.ai.³² Two reviewers (SH and SB) independently excluded studies which were clearly unrelated to the population and concept of the umbrella review from their title. The reviewers achieved 100% agreement in this process and then independently reviewed all the remaining abstracts against the stated inclusion criteria. Again, 100% consensus was achieved. Once all abstracts had been reviewed, potentially relevant sources for full-text review were retrieved. The same reviewers examined all remaining papers independently at full-text level with regular consensus meetings. Reasons for the exclusion of sources at full text were recorded and are reported in table 1. This included three articles whose content was included in a larger report, which had already been retained in the umbrella review.

Table 1.

Inclusion and exclusion criteria of reviews and reports

Inclusion criteria	Exclusion criteria
Children of any age Children with SSD of unknown origin including: Childhood apraxia of speech/ developmental verbal dyspraxia Articulation disorders Phonological disorders of all types	Children with SSDs associated with a biomedical condition, for example: SSD associated with cleft palate +/− lip Cerebral palsy Traumatic brain injury Reviews not written in English Reviews that report outcomes for adults Reviews of studies with no reported assessments or outcomes from interventions for SSD

SSD, speech sound disorder.

Figure 1 shows other databases searched by SH using keywords (speech sound disorder AND review) including the National Grey Literature Collection (https://allcatsrgrey.org.uk), EThOS (https://ethos.bl.uk) and pre-printed servers MedRxiv (www.medrxiv.org) and PsyArXiv (https://psyarxiv.com).

Figure 1.

Preferred Reporting Items for Systematic Review and Meta-Analyses flow chart. JBI, Joanna Briggs Institute.

The reasons that studies were excluded at full text are given in online supplemental material 2. Following the final selection and retention of review articles, critical appraisal was undertaken using the AMSTAR tool.³⁰ This tool was selected as it is designed to critically appraise systematic reviews that include randomised or non-randomised studies of healthcare interventions or both. Two reviewers (SH and SB) individually appraised each study, with consensus meetings to confirm ratings. Shea et al strongly recommend that individual item ratings from the critical appraisal are not combined to create an overall score.³⁰ They propose a ‘confidence in the results’ rating. When this confidence rating was applied, one study was rated ‘moderate’,³³ five were rated ‘low’,^{15 34–37} and the remaining studies were rated ‘critically low’, even when excluding questions specifically related to meta-analysis.^{26 38–48}

Data extraction

Data from the retained reviews were identified using a researcher-developed extraction form. This form was adapted from guidance provided by the JBI Reviewer’s Manual to meet the specific requirements of the proposed review.²⁸

Three of the authors (JC, HS and SB) discussed the long list of outcomes and determined which were outcomes and which were outcome measurement tools. In cases where general terms were used, more specific wording was agreed. The same authors then used the ICF and specific areas of speech development to assign outcome domains to each of the reported outcomes and outcome measures. An outcome was defined as the ultimate or long-term goal of one or more episodes of intervention for a child with SSD. An outcome measure was defined as a tool for measuring a specific outcome (ie, from baseline assessment or an intervention outcome).

Results

Characteristics of retained review studies

16 reviews and two reports were retained for inclusion in this umbrella review (figure 1). Of these, one was a narrative review,³⁸ two were scoping reviews,^{36 40} and the rest were systematic reviews. Ratings on individual AMSTAR items are provided in online supplemental material 3.

During data extraction, it became clear that insufficient detail was reported within the retained reviews. In particular, limited data regarding assessments, interventions and outcome tools (measures) were provided within the retained review manuscripts. Therefore, all papers retained within the reviews were identified and collated. Duplicates were then identified and removed. Papers published prior to 2000 were then removed, and those remaining were checked against the inclusion criteria outlined for retention in this umbrella review. This process is shown in figure 2. Figure 2 shows the number of papers retained from within the reviews published between 2000 and 2022. The articles are spread by date, with the majority of papers published between 2008 and 2015.

Figure 2.

Follow-on Preferred Reporting Items for Systematic Review and Meta-Analyses flow chart.

The 415 articles were sourced and basic data regarding assessments, interventions, outcomes and outcome tools used were extracted and compiled. Figure 3 shows the number of articles by year of publication of 415 relevant papers retained within the reviews published between 2000 and 2022 and the year of publication of the retained reviews.

Figure 3.

Number of articles by year of publication of 415 relevant papers retained within the reviews published between 2000 and 2022 and the year of publication of the retained reviews.

Following the raw extraction, SH and SB screened for duplicates and checked eligibility of data where they were not familiar with it. Another member of the team (JC) resolved any disagreements through discussion. Consensus agreement for assessments and interventions was achieved through discussion between SB, HS and JC and by application of inclusion and exclusion criteria (table 2) developed through agreement between SB, HS and JC.

Table 2.

Inclusion and exclusion criteria for screening of assessments and interventions

	Inclusion criteria	Exclusion criteria
Assessment	Named, published assessment Available in English language SSD-specific assessment (including phonological awareness) Standardised assessments	Unpublished assessment (ie, due to unavailability and unreliability) Data analysis tools for analysing primary assessment results Unavailable in English language Assessments of language, syntax or morphology General development assessments Cognitive assessments Assessments for populations other than SSD of unknown origin (eg, children with hearing impairment or autism)
Intervention	Published evidence available for the intervention or approach	Lack of published evidence for the intervention approach

SSD, speech sound disorder.

Domains of SSD assessed by outcome measures

Table 3 presents the outcomes and outcome measurement tools that were retained in this review. The outcome measures are mapped, but not all measures extracted were linked to specific measurement tools in the retained reviews. Where this is the case, the cell is left blank. Broad outcome domains relating to the ICF and speech development are indicated in each case to highlight the spread of reported outcomes and outcome measurement tools across these domains.

Table 3.

Outcomes and outcome measurement tools

Outcome	Outcome measurement tool	Outcome domain
Improved language	Bristol Language Development Scale (BLADES)*56	Language
Improved vocabulary	British Picture Vocabulary Scales 2 (BPVS)*57	Language
Improved quality of life	Family Adaptability and Cohesion Evaluation Scales (FACES III)*58	Quality of Life
Improved communicative activity and participation	Focus on the Outcomes of Communication Under Six (FOCUS)*59	Communicative activity and participation
Increased speech intelligibility	Functional Communication Measure (FCM) for speech intelligibility*60	Speech
Increased speech intelligibility	intelligibility outcomes by peer group listeners61	Speech
Generalisation to a new context	–	Speech
Generalisation across linguistic units	–	Speech
Generalisation across word positions	–	Speech
Generalisation of known sounds	–	Speech
Generalisation of the intervention target to other response contexts (eg, non-treatment words, other word positions, conversational speech)	–	Speech
Generalisation related to the target (eg, generalisation to other phonemes within and across sound classes)	–	Speech
Increased accuracy of target	–	Speech
Increased confidence when talking	Kiddy-Communication and Attitude Test (KiddyCAT)*62	Communicative activity and participation
Increased mean length of utterance (MLU)	–	Language
Improved oromotor skills	Movements in Context and Sequenced Oral Movements tasks**63	Oromotor
Improved oromotor skills	Peabody Developmental Motor Scales 2 (PDMS-2)**64	Oromotor
Increase in number of phonemes	–	Speech
Parent report on increased structural complexity	–	Speech complexity
Parent report on increased phrase complexity	–	Language
Increase in percentage of child utterance attempts that are fully intelligible from language sample	–	Speech
Increased accuracy measured by Percentage Consonants Correct	–	Speech
Increase in egressive output	–	Speech
Increase in phonological awareness	–	Phonological awareness
Decrease in phonological variability	Phonological Variability Test	Speech
Increase in percentage of intelligible utterances	–	Speech
Decrease in proportion of errors	–	Speech
Increase in percentage of phonemes correct	–	Speech
Increase in production of target sounds	–	Speech
Increase in percentage vowels correct	–	Speech
Increase in percentage of words correct	–	Speech
Increased stimulability	Scaffolding Scale of Stimulability (SSS)*65	Speech

Measures supported by normative data are indicated by*. Those that have been validated are indicated by**.

Areas assessed by identified assessment tools

A total of 37 published assessments were identified which could be used to provide speech data for outcome measurement. Many focus on measurement of specific skills required for speech development while others were comprehensive test batteries (table 4).

Table 4.

Assessments identified

Assessment	Aspect of speech assessed
Individual Growth Development Indicator: Rhyming**66	Phonology
Arizona Articulation Proficiency Scale** (AAPS)67	Articulation, motor
Arizona Articulation Proficiency Scale**—Revised (AAPS-R)68	Articulation, motor
Hodson Assessment of Phonological Patterns—third edition** (HAPP-3)69	Phonology
Bankson-Bernthal Test of Phonology** (BBTOP)70	Phonology
Children’s Test of Nonword Repetition**71	Articulation, motor
Comprehensive Test of Phonological Processing – second edition (CTOPP-2)**72	Phonology
Computer-Based Phonological Awareness Assessment*73	Phonology
Computerized Articulation and Phonology Evaluation System (CAPES)**74	Articulation, motor
Denver Articulation Screening Exam**75	Articulation, motor
Diagnostic Evaluation of Articulation and Phonology (DEAP)**49	Articulation, motor, phonology
Edinburgh Articulation Test (EAT)**76	Articulation, motor
School Speech Questionnaire77	Communicative participation
Goldman-Fristoe Test of Articulation** (GFTA)50	Articulation, motor
Glaspey Dynamic Assessment of Phonology** (GDAP)78	Phonology
Grammar and Phonology Screening (GAPS)**79	Phonology
Preschool and Primary Inventory of Phonological Awareness (PIPA)**80	Phonology
McDonald Screening Deep Test of Articulation**81	Articulation, motor
Oral Speech Mechanism Screen Examination (OSMSE)**82	Articulation, motor
Phonological Abilities Test (PAT)*83	Phonology
Phonological Assessment Battery (PhAB)**84	Phonology
Phonological Assessment of Child Speech (PACS)**85	Phonology
Phonological Awareness Literacy Screening—PreK (PALS-PreK)**86	Phonology
Phonological Awareness Literacy Screening—PreK Pre-Reading**86	Phonology
Phonological Awareness Test**87	Phonology
Phonological Knowledge Protocol (PKP)*88	Phonology
Phonological Variability Test89	Phonology
Scaffolding Scale of Stimulability (SSS)*90	Articulation, motor
Sutherland Phonological Awareness Test—Revised (SPAT-R)**91	Phonology
Test of Phonological Awareness–Second Edition: Plus Test of Preschool Early Literacy (TOPA)**92	Phonology
Syllable Repetition Task (SRT)*93	Articulatory, phonetic and motor based assessments
Templin-Darley Articulation Screening Test**94	Articulatory, phonetic and motor based assessments
Test of Polysyllables95	Articulatory, phonetic and motor based assessments
Verbal Motor Production Assessment for Children**96	Articulatory, phonetic and motor based assessments
Word Complexity Measure97	Phonological interventions: complexity approaches

Assessments supported by normative data are indicated by *. Those that have been validated are indicated by **.

Interventions for SSD

A total of 46 interventions were retained in the review (box 2). These included national and international published clinical interventions spanning all domains of speech sound development.

Box 2. Interventions identified.

Intervention

• Articulation with facilitative vowel contexts⁹⁸

• Auditory bombardment/stimulation⁹⁹

• Broad target recasts¹⁰⁰

• Complexity approach¹⁰¹

• Contrast word procedures (min or max pairs)^{102 103}

• Core vocabulary¹⁰⁴

• Cycles therapy¹⁰⁵

• Drill play¹⁰⁶

• Electropalatography¹⁰⁷

• Focused stimulation¹⁰⁸

• FONEMZ: a multimodal approach¹⁰⁹

• Integral Stimulation/Dynamic Temporal and Tactile Cueing¹¹⁰

• Integrated Phonological Awareness Intervention¹¹¹

• Intraoral stimulation¹¹²

• Maximal oppositions contrast (maximal pairs)¹⁰²

• Maximal/empty sets¹⁰²

• Melodic intonation therapy¹¹³

• Metaphon programme¹¹⁴

• Minimal oppositions contrast (minimal pairs)¹¹⁵

• Minimal or near-minimal contrasts¹¹⁵

• Modified core vocabulary treatment¹¹⁶

• Modified cycles approach¹¹⁷

• Morphosyntax intervention¹¹⁸

• Motor speech treatment protocol¹¹⁹

• Multiple oppositions approach¹²⁰

• Multiple oppositions approach¹²⁰

• Naturalistic intervention for speech intelligibility¹²¹

• Nonlinear phonological intervention¹²²

• Non-speech oromotor intervention³³

• Nuffield Centre Dyspraxia Programme¹²³

• Parents and Children Together therapy¹²⁴

• Phonological Stimulation Program¹²⁵

• PROMPT (Prompts for Restructuring Oral Muscular Phonetic Targets) system therapy (targeting articulation)¹²⁶

• Rapid Syllable Transition Treatment¹²⁷

• Rate control therapy¹²⁸

• Sound Contrasts in Phonology software program¹²⁹

• Speech perception (SAILS - Speech Assessment and Interactive Learning System)¹³⁰

• Speech perception training¹³¹

• Stimulability (STP - Stimulability Training Protocol)¹³²

• Teaching prosodic patterns¹³³

• Touch cue method¹³⁴

• Traditional articulation therapy¹³⁵

• Traditional multiple phonemic approach¹³⁶

• Ultrasound visual biofeedback¹³⁷

• Vocal imitation training¹³⁸

• Vowel-targeted intervention¹³⁹

• Whole language therapy¹⁴⁰

Discussion

The current study sought to summarise previous reviews of outcome measures, assessments and interventions for SSD of unknown origin using an umbrella review methodology. Using a previously published umbrella review protocol,³¹ we identified 18 reviews. The critical appraisal undertaken on the reviews found that the majority (12) of the studies were rated as ‘critically low’. The aim of the umbrella review was to provide a rigorous and detailed list of assessments, interventions or outcomes, as such the quality of review did not impact this collation, so weighting has not been assigned to any of the 18 retained reviews. In order to identify specific assessments, interventions and outcomes, we needed to take the additional step of retrieving the primary sources within the review papers. From these individual studies, we identified 37 assessments, 46 interventions and 30 outcome measures. Although not all of the listed outcomes were linked to specific outcome measurement tools in the retained studies, for the most part these are measurable by one or more of the assessments listed in this review. For example, increase in percentage of phonemes correct could be derived from any of the tools which assess spoken output, such as single-word naming tests like the Diagnostic Evaluation of Articulation and Phonology and Goldman-Fristoe test of articulation.^{49 50}

It is interesting to note that there are more assessments than outcome measures identified in this umbrella review. In intervention research, it is common that assessments are used as diagnostic or screening tools to check suitability for the therapy being investigated. These measures may therefore only be completed at baseline and not at the end of any intervention and as such do not serve as outcome tools. The outcome measures are typically used to assess the impact or broader outcomes of the interventions. These measures may include standardised test scores or other indicators of change. The choice of outcome measures also frequently depends on the research goals, the scope of the intervention and the timeframe for assessing its impact.

Classifying SSD outcomes

The identified assessments include diagnostic and outcome measurement tools which as well as measures of communicative participation⁵¹ cover a full range of measurement tools that look at treatment needs or outcomes. The SORT²⁶ is a tool designed to support the classification of outcome and experiences for children with SSD. The SORT consists of eight domains ranging from clinical treatment data (Domain 1) through to levels of generalisation (Domains 2–4), intelligibility (Domain 5), activity and participation (Domain 6), quality of life and well-being (Domain 7) and the impact of SSD on other people in the child’s life (Domain 8). The outcomes identified in this current umbrella review map across all eight of the SORT domains, with the 25 of the 30 identified outcomes assigned to Domains 1, 2, 3, 5 and 6. The outcome ‘generalisation to a new context’ (table 3) does not map to the SORT because of the broad reference of this outcome and the availability of more specific generalisation outcomes within the retained data. Outcomes for quality of life, and the impact of SSD on other people in the child’s life, were also not measured in the retained papers.

The differing outcome tools and outcomes listed here may in part reflect differences in the way SSD is classified. There are several systems for the classification of SSD. Three clinically commonly used paediatric-specific SSD classification systems were reviewed and critically evaluated by Waring and Knight.⁵² These systems are the Speech Disorder Classification System (SDCS)³; the Differential Diagnosis System (DDS)⁵³ and the Psycholinguistic Framework.⁵⁴ The DDS and the SDCS are the two that are most commonly utilised globally.⁵⁵ The DDS incorporates the subtype labels of phonological delay; consistent atypical phonological disorder; inconsistent phonological disorder; articulation disorder and childhood apraxia of speech (also known as developmental verbal dyspraxia), based on the features of children’s surface-level speech presentation. The SDCS on the other hand is an aetiology-based system that includes the terms: speech delay-genetic; speech delay-otitis media with effusion; speech delay-developmental psychosocial involvement. The motor speech disorder ‘dysarthria’ is absent from the DDS. This is because the DDS is specific to SSD of unknown origin.

An example of how the differences in classification of SSD lead to differences in outcome tools and outcomes is provided by the DDS. A child with inconsistent phonological disorder may receive interventions to improve consistency, rather than correctness of phonemes. In this case, the outcome and outcome tool used will differ from children where the focus is on improving the production of specific consonants.

The mapping of the findings of this umbrella review to one or more of these classification systems for SSD will enable SLTs to select appropriate tools for their practice context.

Towards a core outcome set for SSD

With the potential to map the assessments and outcomes to a framework or system of classification, the speech and language therapy profession can start the conversation around what outcomes are needed to cover all elements/aspects of the framework. Although the measures identified in this umbrella review are the first step in drawing together a list of outcomes which could be used in a core outcome set, it is a crucial starting point and provides the data that are needed to drive follow-up work.

Developing a core outcome set for SSD can help standardise the reporting of outcomes in research studies, making it easier to synthesise findings and assess the overall effectiveness of interventions for this condition. It can also ensure that the outcomes considered most important by patients and healthcare providers are consistently measured and reported.

Strengths and limitations of this review

Umbrella reviews offer several strengths, such as efficiency, comprehensive synthesis, statistical rigour, identification of discrepancies and identification of research gaps. However, they also face challenges, including heterogeneity of studies, quality assessment complexities, publication bias, potential duplications and limited control over methodological choices. In the current review, we have mitigated some of these limitations by extracting retained articles from the individual reviews and deduplicating their representation. We have also critically appraised the reviews using the AMSTAR, although it must be acknowledged that we have also not undertaken an appraisal of all the 415 papers from which we extracted data. Understanding these strengths and weaknesses is crucial for researchers, policymakers and practitioners to interpret and apply the findings of umbrella reviews effectively in evidence-based decision-making processes. Future research should focus on addressing these limitations to further enhance the utility of umbrella reviews as a valuable tool for evidence synthesis.

Further research

The umbrella review reported herein will be used to develop a speech and language therapy core outcome set for children with SSD. However, this requires a rigorous and collaborative process aimed to provide consistent and quality outcomes data, enhance patient-centred care and facilitate evidence-based decision-making. The findings are part of the process essential for advancing healthcare research and practice in the specific area of speech and language therapy for children with SSD.

Data availability statement

No data are available.

Ethics statements

Patient consent for publication

Not applicable.