Evidence of loneliness, it is difficult to save autistic children

Many parents of children with autism have spent a lot of effort and have persisted in intervening for their children for many years to improve their children's mental conditions. Some people have good results, while others have average results. Everything seems to be arranged by fate. They don't know that all these therapies have not yet been supported by rigorous clinical evidence.

Compiled by Xiaoye

Twenty years ago, Andrew Whitehouse was a speech therapist. He couldn’t remember how many special little patients he had treated, but they all had a common name: Star Children. Because the diseases these children suffered from were related to brain development, they were not sociable and had difficulty communicating, as if they were trapped on a lonely planet. After years of popularization of science, most people have now heard of this disease called "autism", which is professionally called autism spectrum disorder. According to statistics from the World Health Organization, approximately one in every 100 children in the world suffers from autism [1].

Whitehouse found that many worried parents were spending a fortune trying various treatments, some of which had not even been scientifically proven. Some families scrimped and saved money just to get their children machines that claimed to “rewire the brain”[2]. In reality, the so-called machines did not really work except beeping.

Today, Whitehouse, who has extensive experience in diagnosis and treatment, is a professor of autism research at the University of Western Australia and director of the autism research program at the Telethon Kids Institute. Over the years, Whitehouse has led his team to conduct randomized controlled clinical trials, delving into the field of autism from the perspectives of patients and their families, scientific research, and medical practice, because he has always had a heavy problem in his mind: in this field, especially in the field of autism intervention research, the clinical evidence standards accepted and even expected by scientists, doctors, and medical practice system developers are much lower than those in other research fields. This problem involves a wide range of issues. Not only do families have to bear the risk of the patient's life safety, but there are also more emotional problems, economic risks, and opportunity costs [2].

To this end, he led his team to start by summarizing previous academic literature, examining the research on autism interventions that are widely used in clinical practice, and examining the reliability of the scientific evidence. In November 2020, the team published the report "Interventions for children on the autism spectrum: A synthesis of research evidence" [3], which concluded that there is very little solid scientific evidence to support the autism interventions used in clinical practice. This negative conclusion instantly set off a wave of controversy in the academic community, and a large number of clinicians, treatment providers, and professional organizations took action within a few weeks, threatening to sue Whitehouse, complaining to his employer, and even harassing Whitehouse's family and threatening their personal safety. Whitehouse never thought that a scientific research report would put himself, his family, and his colleagues in danger.

However, it is not only Whitehouse's team that came to this conclusion. In 2020, the Psychological Bulletin published a meta-analysis of intervention therapies for children with autism aged 0-8 years old (referred to as the AIM project) [4]. Micheal Sandbank, an assistant professor of special education at the University of Texas, led a team to track the seven most popular and allegedly effective early intervention therapies for autism. The results showed that if the research quality indicators were not considered, only three therapies (behavioral intervention, developmental intervention, and spontaneous developmental behavioral intervention (NDBI)) showed significant positive effects. However, if the research standards are strictly considered - whether it is a randomized controlled trial and whether there is detection bias in the research results - then all therapies are ineffective (statistically insignificant). However, Sandbank himself pointed out that although the analysis results are not ideal, this does not mean that we should abandon all these intervention methods [5].

In fact, both reports can be attributed to the same key question: Why is there such a lack of high-quality evidence for autism intervention research?

The development of autism intervention science can be traced back to the 1970s and 1980s, when some groundbreaking clinical studies had small sample sizes and statistical loopholes. Here we have to mention the pioneer of behavior analysis and intervention in the field of autism: Norwegian-American psychologist Ole Ivar Løvaas.

O. Ivar Løvaas in the 1980s. Source: Lovaas Institute/UCLA

Since the mid-1960s, Løvaas has been trying to use Applied Behavior Analysis (ABA) to establish intervention therapy. Through one-on-one language teaching for autistic children, he broke other core barriers of autism and developed abilities such as play and self-help[6]. In 1987, Løvaas published a groundbreaking study titled "Behavioral Treatment and Normal Intellectual Functioning in Young Children with Autism"[7]. The article pointed out that after early and intensive ABA intervention, 47% of the children tested (n=19) successfully recovered normal intelligence and educational functions, which were no different from healthy peers. Among the autistic children in the control group (n=40), only 2% recovered to normal levels. Since then, people have begun to believe that autistic children can eventually live and study independently through teaching. Løvaas is also known as the "Father of ABA". He actively promoted ABA therapy and inspired many students and colleagues to engage in autism intervention research. Later, many new therapies were based on ABA intervention.

It can be said that Løvaas's ABA research has the innovative "quasi-experimental" characteristics, but it also has certain limitations. One is the small sample size; in addition, during the research process, the children were not randomly grouped. His approach was based on the background of the time: psychological research at the same time as Løvaas mainly followed the design principle of "single-case", that is, case studies lacked a formal control group, and the subjects themselves were both the experimental group and the control group.

In fact, by the 1980s, the current gold standard for clinical research in medicine, the randomized controlled trial, had been widely accepted, but autism intervention research that combined psychology and medicine still had a hard time keeping up with the pace of other disciplines. Jonathan Green, professor of child and adolescent psychiatry at the University of Manchester, UK, pointed out that from the beginning, some researchers believed that RCTs were neither ethical nor feasible in the face of such a complex autism. This mentality contributed to the default attitude of everyone in the field towards low-standard research evidence [8].

In the AIM project led by Sandbank mentioned above, less than one-third of the ABA intervention therapy studies statistically analyzed by researchers used a randomized controlled design, not to mention that most of the reports read by American clinicians are case studies. In 2021, the National Autism Evidence and Practice Exchange Center (NCAEP) reported that [9] after a comprehensive analysis of 972 articles, it was found that 28 medical practices were evidence-based, including many behavioral intervention therapies, but 85% of the studies were case studies. In addition, the 2015 National Standards Report (NSP) of the United States confirmed 14 effective intervention therapies for children, adolescents, and young people with autism [10]. Similarly, 73% of the studies cited were case studies.

It is true that in the development of modern psychology, case study methods help to observe and study a certain or a small group of subjects in depth, so as to discover the causes of certain behaviors, symptoms and psychological phenomena, observe their clinical development and changes, and help researchers establish theories. However, in clinical treatment research, the shortcomings of case studies are also quite obvious. First, due to the small number of cases, if the research results are to be generalized to a more general conclusion, caution must be exercised. Secondly, case studies may not reveal causality. Finally, case studies are not suitable for tracking long-term development changes, which is precisely the focus of intensive intervention therapy. The so-called intensive intervention therapy is a traditional treatment for children with autism aged 0 to 9 years old. It takes 25 to 40 hours a week, from one-on-one classroom learning to daily life, and intervention mechanisms are adopted at all times. The whole process lasts for two to three years. Despite the huge cost of manpower and time, the rehabilitation effect is not immediate and requires long-term follow-up. However, the academic community currently lacks high-quality research to prove its true effect. Sandbank points out that while case studies can help researchers test changes in specific treatment approaches (such as regular classroom learning in schools), "we cannot make general treatment recommendations based on case studies."[8]

It is well known that unvalidated therapies may cause serious and substantial harm, so modern medical clinical trials place great emphasis on whether the validation method is reliable [11]. Whitehouse believes that maintaining the current status of low-quality evidence will push the development of autism intervention therapies to the brink of danger, which is also the original intention of his active promotion of randomized controlled studies in the field.

In addition to conflicts in research theories and methods, intertwined conflicts of interest are also a major hidden crisis in the field of autism intervention. It is this powerful force that is stubbornly suppressing the improvement of data standards.

In the United States, autism treatment has become a multibillion-dollar industry, thanks in large part to national insurance mandates and financial companies that subsidize ABA intervention providers. The commercialization of autism treatments has increased access to care, but it may also undermine the importance of high-quality research evidence. After all, private industry seeks profit, and in the game between profit and good medical practice, profit always wins. [8]

In addition to the problems caused by commercialization, researchers involved in research evidence assessment may have multiple roles, and the positions of the relevant scientific research institutions behind them cannot be ignored. Conflicts of interest may hinder the fairness of the evaluation. For example, the editorial board members of the Journal of Applied Behavior Analysis, which publishes ABA research, are mostly board-certified analysts who have received ABA qualification training. Therefore, they naturally tend to support ABA intervention, and the evaluation of the paper may be biased.

In addition, many ABA analysts also participated in the writing of the national standard report, and it was logical to include ABA therapy in the standard list of "established intervention therapies." At the same time, the May Institute, a nonprofit organization that provides ABA intervention therapy services throughout the United States, is also one of the funders of the national standard report. Although the contributions of analysts and institutes are listed in the report, the hidden conflicts of interest are not mentioned at all. After all, it was not a routine operation to strictly inquire about conflicts of interest in the evaluation agenda at that time. Although it does not mean that the practices of ABA analysts must be excluded, it shows that conflicts of interest can allow others to examine research evidence more fairly.

Conflicts of interest also involve researchers’ dual roles: someone who reviews autism intervention studies may also be the creator of a treatment. However, this overlapping role is rarely mentioned in publicly published research. Connie Kasari, a professor of human development and psychology at the University of California, San Diego, commented: “Researchers often have little motivation to step out of their own ‘islands’ and are unwilling to independently verify or collaborate with others to jointly verify interventions. It all comes down to money.”[8]

Future Development Trends of Autism Treatment The field of autism has been developed for less than a century, but it is necessary to go beyond the concepts and methods inherited from history. Many scientists are committed to promoting the development of this field in a more multidimensional direction, and everyone is still optimistic about the future.

First, regarding the issue of experimental methodology, autism research has been turning to randomized controlled trials. According to statistics, the number of randomized controlled trials has increased from 2 in 2000 to 48 in 2018, most of which appeared after 2010. Only 12.5% ​​of the studies had a risk of bias and the risk was low. [12]

Second, autism intervention research needs to go beyond validating the operation of a single intervention therapy within a control group and instead compare multiple intervention therapies. The ultimate goal is to enable physicians to list the pros and cons of different therapies to patient families so that they can make informed choices that benefit their patients. For example, a study published in the Journal of the American Academy of Child & Adolescent Psychiatry in 2021 compared the efficacy of ABA intervention therapy and the Early Intervention Denver Model (ESDM) [13]. The latter is mainly aimed at children with autism aged 1 to 3 years old and can be extended to children aged 4 to 5 years old. It uses games as an intervention framework, focuses on the establishment of emotional interaction and social motivation, and emphasizes teaching and family intervention models in natural settings [14]. The results showed that the two were equally effective. In fact, if more studies like this could appear, it would help professionals to promptly understand which intervention therapies can achieve the best results with the least time and economic cost. However, there are still too few such studies.

In addition to comparing the efficacy of multiple therapies, the order of application of different intervention therapies is also worth discussing. Connie Kasari, a professor of human development and psychology at the University of California, Los Angeles (UCLA), led a team to develop the game-based natural intervention therapy JASPER. She compared the effects of applying JASPER before and after ABA intervention. For some children, receiving structured ABA intervention methods first will have better results, while for others, the opposite is true. Such sequential multiple allocation randomized trials will help develop personalized treatment strategies for individuals.

For researchers, Sandbank recommends that they break away from the traditional model of validating their own interventions through individual cases. The above examples have shown that experimental protocols can be designed from multiple angles to prioritize the independent replication of intervention effects. Although the results of such studies may not be as ideal as those of the original studies, "we must not be afraid of any findings." [8] At the same time, researchers should always remember that they have an obligation to carry out high-quality scientific research, rather than shifting the responsibility to other parties.

Finally, if autism intervention research is to advance, scientific regulation also needs to change from top to bottom. Many autism journals need to tighten their standards for publishing papers, and more funding is needed to encourage researchers to conduct more expensive and complex trials, such as the sequential multiple allocation randomized trial designed by Kasari mentioned above.

The two reports mentioned at the beginning of this article (one from the Whitehouse and one from the Sandbank teams) have brought the underlying problems in the field of autism science to the public’s attention. Although scientific research is still struggling to catch up with medical practice, and “the field as a whole is only beginning to obtain high-quality data”[8], only by doing something can we save more “star children”.

References

[1] https://www.who.int/zh/news-room/fact-sheets/detail/autism-spectrum-disorders

[2] https://www.spectrumnews.org/opinion/beyond-the-bench-a-conversation-with-andrew-whitehouse/

[3] https://www.autismcrc.com.au/interventions-evidence

[4] https://doi.apa.org/doiLanding?doi=10.1037%2Fbul0000215

[5] https://www.spectrumnews.org/news/studies-find-thin-evidence-for-early-autism-therapies/

[6] http://thelovaascenter.com/about-us/dr-ivar-lovaas/

[7] https://doi.apa.org/doiLanding?doi=10.1037%2F0022-006X.55.1.3

[8] https://www.the-scientist.com/news-opinion/why-autism-therapies-have-an-evidence-problem-69916

[9] https://link.springer.com/article/10.1007/s10803-020-04844-2

[10] https://nationalautismcenter.org/national-standards-project/phase-2/significant-findings/

[11] https://blog.sciencenet.cn/blog-279293-1254793.html

[12] https://acamh.onlinelibrary.wiley.com/doi/10.1111/jcpp.12828

[13] https://linkinghub.elsevier.com/retrieve/pii/S0890856720313502

[14] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7390010/#:~:text=%E6%97%A9%E6%9C%9F%E4%BB%8B%E5%85%A5%E4%B8%B9%E4%BD%9B%E6 %A8%A1%E5%BC%8F%EF%BC%88Early,%E5%84%BF%E7%AB%A5%E7%9A%84%E7%BB%BC%E5%90%88%E5%B9%B2%E9%A2%84%E4%BD%93%E7%B3%BB%E3%80%82

Special Tips

1. Go to the "Featured Column" at the bottom of the menu of the "Fanpu" WeChat public account to read a series of popular science articles on different topics.

2. Fanpu provides a function to search articles by month. Follow the official account and reply with the four-digit year + month, such as "1903", to get the article index for March 2019, and so on.

Copyright statement: Personal forwarding is welcome. Any form of media or organization is not allowed to reprint or excerpt without authorization. For reprint authorization, please contact the backstage of the "Fanpu" WeChat public account.