1. Bronwyn Hemsley1, Ralf Schlosser2,3,4 Mark Mostert,5 Scott Lilienfeld,6 Samuel Odom,7
James Todd,8 David Trembath,9 Howard Shane 2,10,11
Russel Lang,12 & Seraphina Fong1
4
2
A systematic review of the Rapid Prompting
Method: Findings and Implications for
Speech Language Therapists worldwide
3
1 5
6
7
8
9
10
11
12
Paper Presented at #CPLOL18 10th European Congress of Speech and Language
Therapy, Estoril, Portugal on 12th May 2018
5. Video: Example of RPM
5#CPLOL18
Table provides a good frame of reference square to show board is being moved.
Removed for slideshare
6. According to Soma Mukhopadhyay,
as distilled from her Halo-Soma website and books,
Rapid Prompting Method (RPM) is:
• A brain-based teaching and assistive method
designed to establish and promote pointing-
based textual communication in people who
cannot speak, usually autism.
• Involves a series of “teach-ask” trials of
graduated difficulty, starting with the pupil
choosing a correct answer from two written
options, progressing through composing
responses by pointing to printed letters on a
card.
• Linguistic competence is presumed even if
not evident, and learning materials are
presented verbally and textually at age level.
6
Image credit: Halo website
#CPLOL18
7. RPM theory:
An ‘unusual’ way of looking at autism
• The Sensory Integration Theory of A. Jean
Ayers is foundational; teachers are to assess
and use the “open sensory channel.”
• Some problems of autism result from a failure
of the child to move from earlier to later
Piagetian stages.
• Teaching does not require explicit rewards as
these are built into success.
• Autism is fundamentally a “failure of
initiation.” Thus, prompts are a necessary and
probably permanent feature of RPM
assistance.
• Self-stimulatory and other common features
of autism may be used in assessing the
“open sensory channel,” but usually can be
otherwise ignored.
7
Images credit: Halo website
#CPLOL18
8. Why the urgency to
systematically review RPM?
1. Anecdotal reports that it is growing in popularity in US and Ireland
2. Appearance in Ireland senate (Senator Rónán Mullen) (2016) and Irish Media
(Jan 2017)
3. Front page article, Washington Post (March 2017)
4. YouTube - search ‘Rapid Prompting’ yields about 1210 results (Nov 2017)
5. Appearance in occupational therapy open access literature 2017
6. Parents have a right to the highest level of evidence available in making
decisions
8
#CPLOL18
9. Previous Review of RPM
(Deacy et al., 2016)
Three studies were included and
“findings were summarised into
four categories. Findings appear
to indicate an association
between RPM and a decrease in
repetitive behaviours.” (p. 92)
“Further research in the area of
RPM is warranted in order for it
to be considered as a qualifying
evidence-based practice” (p. 92)
9
Deacy, E., Jennings, F., & O'Hallorhan, A. (2016). Rapid Prompting Method (RPM): A
suitable intervention for students with ASD?. Reach, 29(2), 92-100.
#CPLOL18
10. Gernsbacher (2004)
Retrospective case study of
1 child.
“The RPM core element of
‘repetitive prompting’ is not
referred to in this study, so it
is unclear if the method used
is closer to ‘facilitated
communication’ which has
been discredited in the
literature” (p. 94). 10
Solomon (2006)
Descriptive-ethnographic
study, Compared
practice of RPM at HALO
vs. at home.
Chen et al. (2012)
Descriptive study of
RPM and repetitive
behavior.
Deacy, E., Jennings, F., & O'Hallorhan, A. (2016). Rapid
Prompting Method (RPM): A suitable intervention for students
with ASD?. Reach, 29(2), 92-100.
Previous Review of RPM
(Deacy et al., 2016)
11. Not all reviews are equal
Commentary on Deacy et al., review
• Numerous methodological flaws identified in Deacy et al., 2016
• Conflicts of interests identified for its included studies
11#CPLOL18
12. Aim
To determine whether RPM is effective for
individuals with ASD in terms of:
(1) improving body motor skills, speech
skills, communication skills, and
communication-related cognitive skills,
and/or
(2) reducing challenging behaviors
12#CPLOL18
13. Aim in PICOS Format
P (Person/Population) => Individuals with ASD
I (Intervention) => RPM
C (Comparison) => None required, any allowed
O (Outcomes) => Body motor skills, speech skills,
communication skills,
communication-related cognitive,
challenging behaviors
S (Study Design) => Single-case experimental
designs or group experimental
designs capable of showing
a functional relation between
RPM and outcomes
13#CPLOL18
16. Inclusion Criteria
• Participants with a documented diagnosis of ASD
• Design that establishes a functional relationship
between RPM and outcomes
If qualifying group design:
(a) if only group level data were provided, all participants met ASD
criteria
(b) if only some participants met ASD criteria, subgroup analyses
were provided
If qualifying single-case experimental design:
At least one participant met ASD criteria and data are provided for
that participant
• RPM was the intervention studied
• Published in a peer-reviewed journal, as a master’s
thesis, as a dissertation, or full-length unpublished
manuscript
16#CPLOL18
17. Literature Search
17
• Bibliographic databases
• Cochrane Library
• ComDisDome (ProQuest)
• Cumulative Index of Nursing & Allied Health
Literatures (CINAHL, EBSCOhost)
• Education Journals (ProQuest)
• Education Research Complete (EBSCO
Host)
• Education Resources Information
Clearinghouse (ERIC, ProQuest)
• Google Scholar
• Linguistics and Language Behavior Abstracts
(LLBA, ProQuest),
• Medline (via Pubmed)
• ProQuest Dissertations and Theses Global
• PsycINFO (ProQuest)
• Psychology and Behavioral Science
Collection (ProQuest)
• Web of Science
#CPLOL18
18. Literature Search
18
• Publisher-specific databases
• Sage Journals Premier
• ScienceDirect Elsewier
• SpringerLink Journals
• Taylor & Francis Online
• Trial Registers
• EU Clinical Trial Register
• Australian New Zealand Clinical Trial Registry
• Clinical Trials
• ClinicalTrials.gov
• ClincalTrialSearch.org
• Ancestry Searches
• of included and relevant other studies, reviews, book chapters, and books
• Forward citation searches
• Of included and other studies/papers that referenced RPM
• Contact authors of select published peer-reviewed articles
19. Literature Search
19
Lessons from Preliminary Search
• “RPM” not a viable acronym to search
• Author search only viable if the database
permitted a search of the full name, i.e., last
name (“Mukhopadhyay”) with first initial (“S”)
• Not necessary to combine “Rapid Prompting
Method” with an autism-related keyword
Search Strategy
• “Rapid Prompting Method” as a free text
word in title, abstract, and/or full text
• “Informative pointing” as a free text word in
title, abstract and/or full text
#CPLOL18
20. Inclusion Reliability
• First author (Schlosser) screened search results by title and
abstract
• All documents (100%) submitted to the inclusion checklist were
evaluated by the first and second author (Hemsley)
independently from each other
• Percent agreement on Inclusion Decision (“in or out”) = 100%
• Percent agreement on item-specific ratings before consensus
building: 91.18%
• Percent agreement on item-specific ratings after consensus
building: 100%
20#CPLOL18
23. Reasons for Exclusion
of the 6 full texts
Study – Full-text Review Checklist item/s
not met
Exclusion notes
Chen et al. (2014) # 2 Descriptive study
Cleveland (2008) # 5 Although Mukhopadhyay’s work is reviewed in the literature
section, the procedure described (p. 11) has components of
RPM but is not specifically RPM as described by
Mukhopadyay.
Gernsbacher (2004) # 2, # 5 Descriptive case study (# 2);
The mother had met with Mukhopadhyay and as a result
adapted some of her modalities. For example, she tried
handwriting as a form of AAC. It is stated she wouldn’t go
through the procedures at great length as described by the
developer (# 5).
Solomon (2006) # 2, # 5 Ethnographic study (# 2);
Even though RPM was studied it was examined as a
practice rather than an intervention – comparing the
practices of RPM at HALO versus by parents elsewhere (#5)
Van Acker (2016) # 1, # 2, # 5 Discussion paper without participants or design (#1, # 2);
RPM was mentioned in the text but was not studied as an
intervention (#5)
Van der Meer (2012) # 5 RPM was mentioned in the text but not studied as an
intervention (#5)
23
24. Result
There are NO studies establishing a
relationship between RPM and:
1. Improvements in body motor skills,
speech skills, communication skills,
and communication-related cognitive
skills, and/or
2. Reduction in challenging behaviors
24#CPLOL18
26. Empty Review
• There is no evidence that RPM results in
the claims purported by its developer and
proponents
• RPM is an unsupported technique
26#CPLOL18
27. Evaluating the Proclaimed
Underlying “Theory of Action”
A neurological examination of RPM suggests:
• Mukhopadhyay’s statements about mirror neurons, brain
anatomy, and other aspects of neurology bear little relation
to actual neurology.
• In addition to its reliance on Ayers’s doubtful theories of
sensory integration, RPM provides no specific or reliable
method to assess or employ the “open sensory channel.”
• The strong recommendation to teach from the right in order
to engage the left side of the brain reveals a profound
reliance on pop psychology rather than actual neurology.
27
Todd, J.T. (2015). Old horses in new stables: Rapid
prompting, facilitated communication, science, ethics,
and the history of magic. In R. M. Foxx & J. A. Mulick
(Eds.). Controversial Therapies for Developmental
Disabilities: Fad, Fashion, and Science in Professional
Practice (2nd ed.). Mahwah, NJ: Routledge
Hickok, G. (2014). The myth of mirror
neurons: The real neuroscience of
communication and cognition. New York:
WW Norton & Company
28. Evaluating the Proclaimed
Underlying
“Theory of Action”
A learning theory examination of RPM suggests:
• The pointing-based communication is actually simulated,
authored by the aide rather than the pupil.
• The “teach-ask” teaching trials contain no real linguistic
instruction, and should not be expected to establish genuine
expression.
• Rather, the trials are an errorless program by which the pupil is
taught to respond to increasingly subtle cues given by the aide
via letterboard movements, gestures, and verbal prompts.
• The aide gradually loses awareness of giving the cues to the
point doing so is likely to become unconscious.
28
Todd, J.T. (2015). Old horses in new stables: Rapid prompting, facilitated communication,
science, ethics, and the history of magic. In R. M. Foxx & J. A. Mulick (Eds.). Controversial
Therapies for Developmental Disabilities: Fad, Fashion, and Science in Professional
Practice (2nd ed.). Mahwah, NJ: Routledge
29. Proposed consumers of RPM most likely can initiate
movement. Watch them:
• Cross mid-line
• Independently point – occasionally require modeling
• Often point independently to common items represented by a
symbol, but not to corresponding words or letters
• Manipulate iPad / smart phone (point, swipe, etc.)
• Grasp and release small objects
• Accurately place small objects on targets
• Eat independently
• Handle small manipulatives (e.g., string beads, operate tv
remote, open containers, complete puzzles) 29
Evaluating the Proclaimed
Underlying
“Theory of Action”Inability to initiate movement is a (convenient) red
herring
Evaluating the Proclaimed
Underlying
“Theory of Action”
30. Given state-of-the-art
contemporary technology, almost
no person is too physically
disabled to communicate using
language -- If they have psycho-
linguistic competence.
30
Just to be clear …
31. Facilitated Communication (FC)
& RPM
FC’s mechanism of action
• Ideomotor Effect
• Expectancy Bias
• Confirmation Bias
FC by any other name…
• Supported Typing
• Assisted Typing
• Inventive Spelling
Despite minor procedural variations, the
above interventions share the same
mechanism of action and are thus rightfully
grouped together.
RPM is in the (expanding) group of
facilitated communication methods.
31
Same
action
RPM
FC
Supported
typing
Assisted
Typing
#CPLOL18
32. RPM & FC:
Procedural Differences and Similarities
RPM has a “Teaching” phase
before letter board introduction.
• intentional, acknowledged,
prompting
• letter board is held by the facilitator
while the facilitated person is not
touched.
In RPM, there is no meaningful
change to FC’s mechanism of
action.
32
Same
action
RPM
FC
#CPLOL18
33. Abuse and RPM:
Not just sexual exploitation
• Educational mistreatment
– wasted time not learning useful / functional skills
• Parental exploitation
– unrealistic expectations planted by FC & RPM
evangelists
• Child ‘mishandling’ of identity
– loss of own voice and actual person
• Professional manipulation
– of those teachers/SLPs who unwittingly become
believers
33#CPLOL18
34. RPM is Pseudoscientific
FC and RPM are “untestable”
• proponents say that research breaks the trust between facilitator
and facilitated; or research is said to be unnecessary
Evidence that does not support pre-drawn conclusions is
dismissed (despite sufficiently rigorous research procedures),
and supportive “evidence” is accepted
FC and RPM are largely supported by anecdotal accounts
Inconsistent with well established replicated findings regarding
the communication and language development in people with
ASD/ID
Extraordinary claims
• unexpected literacy
Reliance on “vaguely described holistic framework”
• the mechanism of action suggested by proponents is unclear and
lacks parsimony
34
Finn, Bothe, & Bramlett’s (2005) Criteria for distinguishing Science from Pseudoscience*.
* Adapted from Lilienfeld, Lynn, and Lohr (2003, revised edition 2014) “Science and pseudoscience in
clinical psychology.” New York, Guilford.
36. Our audience is now …
New to
RPM
Uncertain
either way
Evidence
Aware
RPM
Advocate
36#CPLOL18
37. Code of Ethics
Each association’s code of ethics is applicable in
terms of avoiding harms and making ethical
decisions in relation to treatments.
37
Ethics
Aware
#CPLOL18
38. 1. Refrain from using RPM
or supporting RPM
Use evidence-based AAC interventions and
ABA
• Wong, C., Odom, S. L., et al. (2015). Evidence‐ Based
Practices for Children, Youth, and Young Adults with Autism
Spectrum Disorder: A Comprehensive Review. Journal of
Autism and Developmental Disorders, 1‐16.
• National Autism Center (NAC) (2015). Findings and
conclusions: National Standards Project, Phase 2.
http://www.nationalautismcenter.org/national-standards-
project/results-reports/
Using RPM or UNCERTAIN?
Ask us about testing authorship. 38
Uncertain
either
way?
#CPLOL18
39. • Promoting Rapid Prompting Method (e.g., in social
media) could contravene the professional’s code of
ethics, and is irresponsible.
• Speech language pathologists should not promote or
support the use of RPM.
• Refer to your professional association if uncertain how to
respond to authority figures pressing you to use RPM.
• Families, teachers, therapists and other decision-makers
(e.g., funding bodies, administration of schools) should
be advised of the findings of this review, as per ‘duty of
care’ ‘duty to warn’ (professional responsibility).
• Ask the authors of this paper for updates (see last slide)
39
2. Be Aware & Advise
#CPLOL18
40. 3. Read studies CLOSELY
❓Participants with a documented diagnosis of ASD
❓Design that establishes a functional relationship
between RPM and outcomes
❓If qualifying group design:
❓(a) if only group level data were provided, all participants met ASD
criterion
❓(b) if only some participants met ASD criterion, subgroup analyses
were provided
❓If qualifying single-case experimental design:
❓At least one participant met ASD criterion and data are provided for
that participant
❓RPM was the intervention studied
❓Published in a peer-reviewed journal, as a master’s
thesis, as a dissertation, or full-length unpublished
manuscript
40#CPLOL18
41. 4. Be an evidence champion
#ResNetSLT #SLPeeps @WeSpeechies #WeSpeechies
#AugComm
41
Evidence
Champion
42. Is further research needed?
Testing of authorship
“The paramount question to be addressed by researchers
investigating RPM should pertain to authorship of the
messages … evidence is needed regarding the authenticity of
messages generated by people with ASD who use a
letterboard held by facilitators” (Lang et al., 2014, p. 46).
42
Lang, R. Harbison Tostanoski, A, Travers, J., Todd, J., Todd,
J., & Shane, H. (2014). The only study investigating the rapid
prompting method has serious methodological flaws but data
suggest the most likely outcome is prompt dependency.
Evidence- Based Communication Assessment and
Intervention, 8, 40–48.
Although Hemsley (2016) contended that further primary research was not
warranted, would research on ‘authorship’ in RPM protect the interests of person
with autism?
#CPLOL18
43. Conclusion
43
Rapid Prompting Method has no evidence, is not
supported, is pseudoscientific, and - because of the
influence of facilitators, loss of the human right of
communication, and loss of opportunity to treatments that
work - is harmful to people with autism who are exposed
to it.
Rapid Prompting Method should not be used in the
treatment of communication or behavior for people with
autism.
#CPLOL18
44. Ask the Authors for Autism
Research & Evidence Updates
• Prof Bronwyn Hemsley Bronwyn.Hemsley@uts.edu.au
• Prof Ralf Schlosser R.Schlosser@northeastern.edu
44
Notes de l'éditeur
This paper considers any audience might have people in one of these camps, or moving between them from time to time as new information about RPM comes in, or as they are exposed to ‘miraculous’ stories, or policies at work encouraging them to use RPM. This presentation is for all of you, in the hope that together we can improve access to evidence-based interventions for people with autism and their families.
Pause 1 second
First, an introduction to RPM.
The HALO website headline is ‘A New Way of Looking at Autism’. It is rather an unusual way of looking at RPM.
In this review, the inclusion criteria were not defined.
Gernsbacher (2004)
Point out that there were no data presented
Even though they noted that the core element of RPM was not mentioned, they still included it – not logical and inconsistent
Solomon (2006)
Included even though it was more about the practice or implementation of RPM in two settings by two types of communication partners rather than RPM as intervention
Hemsley (2016) identified numerous methodological issues in her structured appraisal commentary on this review. She also identified previously undisclosed conflicts of interests present for some of the included studies. Clearly, a more rigorous effort is needed.
Pause 1 second
Participants with a documented diagnosis of ASD:
-NOTE: of any chronological age
includes Autism Prototype Disorder, Childhood Disintegrative Disorder, Asperger Syndrome, Pervasive Developmental Disabilities- Not Otherwise Specified [PDD-NOS]).
-Individuals with additional diagnoses (e.g., blindness/visual impairments/intellectual disability) qualify as long as they do carry any of the above ASD diagnoses as well.
-EXCLUSION: Individuals described as “autistic-like” or individuals whose diagnosis is not confirmed by accepted means of evaluation were excluded.
Design that establishes a functional relationship between RPM and outcomes
single-case experimental designs (e.g., ABA, ABAB, ATD, AATD, MBD, MPD, PTD)
group experimental designs (e.g., RCT).
EXCLUSION: Pre-experimental designs such as AB designs or Single Group Pre-Post designs were excluded.
RPM was the intervention studied:
specific reference/s made to writings of its developer (e.g., Mukhopadhyay S. [2008]. Understanding autism through rapid prompting method. Denver, CO: Outskirts Press) or any subsequent guides available for purchase on the website of Helping Autism Through Learning and Outreach (HALO): http://www.halo-soma.org/store.php?sess_id=7ad457a169dd656ce33c6d8a7c0d7f68
Peer-reviewed journal:
In English only
We consulted the following types of sources: (a) bibliographic databases, (b) publisher-specific databases, and (c) trial registers
In terms of search methods, we conduced (a) ancestry searches, (b) forward citation searches (on Web of Science), and (c) author contacts
I'd like to start the discussion by sharing our descriptive findings.
Based on the findings of this review it is safe to say that RPM is an unsupported technique. But we are going to take this a step further in the remaining discussion to challenge RPMs underlying assumptions
The findings of the review also confirm the conclusions of the earlier review by Deacey (2016).
As with any intervention whether medical, psychological or educational; it is important to understand the “mechanism of action”. That is, to understand exactly how the intervention’s components cause any observed change in behavior, health or wellbeing of the recipient. There are several benefits to a thorough understanding of an intervention’s mechanism of action. For example, with interventions that provide actual benefit, the knowledge may be useful in correcting treatment failures, training others to implement the procedures, and in designing more effective interventions. Further, interventions may be classified by their mechanisms of action. For example, interventions that involve programmed reinforcement contingencies and/or antecedent environment manipulations are typically classified as “behavioral”.
For interventions that cause harm, an understanding of the mechanism of action is useful in recognizing and avoiding rebranded but equally harmful variants of the intervention approach. Comparable to changing the flavor of cough syrup, changes that do not influence the mechanism of action have no meaningful effect- other than possibly altering the probability of consumption for some consumers.
FC has been rebranded a number of times, For example, being referred to as supported typing and inventive spelling. In some cases, these variants include more than a name change and involve the addition or deletion of inconsequential procedural aspects (nothing more than a change in flavor). Research has identified several mechanisms of action involved in FC. The ideomotor effect, expectancy bias, and confirmation bias are among the primary mechanisms responsible FC’s powerful illusion and any intervention that relies on these mechanisms can be accurately categorized within the same class of pseudoscientific interventions as FC.
Rapid Prompting involves a number of procedural differences, including but not necessarily limited to: (a) a teaching phase that occurs before the full letter board is introduced; (b) acknowledged prompts that are purportedly based on the sensory products of stereotyped behaviors. For example, a person who makes repetitive vocalizations might be said to require auditory prompts to compete with audible input of the scream; and (c) the person being facilitated is not touched by the facilitator. Instead, the facilitator is holding the letter board which is touched by (or touched to) the facilitated.
[RALF: might be a good spot for video clip of RPM in action]
Nothing about these differences can discount the influence of expectancy bias or confirmation bias and it is irrelevant whether or not the facilitator moves the person’s hand or moves the board that contact’s the person’s hand. These differences are unimportant and suggest that RPM belongs in the same class of interventions with FC and the others.
FC and RPM both meet many of Finn et al’s criteria for identifying pseudoscientific communication interventions. For example…
[RALF: I didn’t add too many notes, here- slide points seemed sufficient. Let me know if you need more on any of this stuff.
I'd like to start the discussion by sharing our descriptive findings.
Our audience might have moved over the course of this session, from being ‘new’ to RPM, to now being ‘evidence aware’.
All in the audience, regardless of their own personal opinion of RPM, must be ethics aware.
For those who are uncertain about RPM, err on the side of safety, and choose to use evidence-based AAC interventions and ABA.
If at all tempted with or already using RPM, test authorship. How do you do that?
Howard,
Can you speak to the testing of authorship for those who wish to explore the use of RPM, despite the lack of evidence in its support.
Participants with a documented diagnosis of ASD:
-NOTE: of any chronological age
includes Autism Prototype Disorder, Childhood Disintegrative Disorder, Asperger Syndrome, Pervasive Developmental Disabilities- Not Otherwise Specified [PDD-NOS]).
-Individuals with additional diagnoses (e.g., blindness/visual impairments/intellectual disability) qualify as long as they do carry any of the above ASD diagnoses as well.
-EXCLUSION: Individuals described as “autistic-like” or individuals whose diagnosis is not confirmed by accepted means of evaluation were excluded.
Design that establishes a functional relationship between RPM and outcomes
single-case experimental designs (e.g., ABA, ABAB, ATD, AATD, MBD, MPD, PTD)
group experimental designs (e.g., RCT).
EXCLUSION: Pre-experimental designs such as AB designs or Single Group Pre-Post designs were excluded.
RPM was the intervention studied:
specific reference/s made to writings of its developer (e.g., Mukhopadhyay S. [2008]. Understanding autism through rapid prompting method. Denver, CO: Outskirts Press) or any subsequent guides available for purchase on the website of Helping Autism Through Learning and Outreach (HALO): http://www.halo-soma.org/store.php?sess_id=7ad457a169dd656ce33c6d8a7c0d7f68
Peer-reviewed journal:
In English only
The onus is on the developer
At the beginning of the presentation we asked whether the emperor has any clothes. Based on our systematic review, documented similarity to FC, characteristics consistent with pseudoscientific intervention, and questionable underlying theory of action, we conclude that the emperor is naked.