Ethical Issues in ABA-based Service Provision for Autism in Limited-Resource Contexts: A Case Example of the People’s Republic of China

Abstract

Recommendations for intervention for young children with autism spectrum disorders (ASD) focus on early, intensive, and often individual intervention based on methods of applied behavior analysis. In much of the world, there are few options for early intervention. This article examines this question: in a context where intensive and high staff-to-student ratio intervention is not possible, how can provision of high-quality evidence-based intervention be ensured? We consider the case of China, where intervention for young children with autism is provided at diverse educational organizations, and funding limitations impact teacher-to-student ratio. Due to challenges, rigorous use of evidence-based methods often lags behind best practices. This article presents an ethical analysis of the choices educators face and research-based recommendations consistent with the ethical analysis. Given the current context, we recommend a socially valid approach of systematically using group instruction based on ABA principles to increase the effectiveness and intensity of each aspect of intervention programs. While focused on the case of China, recommendations and analysis have implications for other settings with limited resources.

The number of children with autism spectrum disorders (ASD) worldwide is increasing, with a global prevalence rate of 1 in 160 (WHO, 2021) and up to 1 in 100 prevalence in China (Sun et al., 2019). Globally, much attention is paid to the importance of early intervention and education for this population. Evidence-based intervention methods have been established, many based on Applied Behavior Analysis (ABA) (National Autism Center, 2015; Steinbrenner et al., 2020), and provision of numerous hours of specialized intervention is recommended (Pellecchia et al., 2019; Sundberg, 2014). While much of this literature has been based in English-speaking, Western countries, there is a growing body of research in Chinese and English regarding evidence-based strategies based on ABA in China (e.g., Liao et al., 2018; Liu & Zhao, 2019; Wang, 2021). Moreover, the needs of children with autism in China are similar to those in other countries, including social skills, communication, self-care, etc., which are addressed by effective intervention.

However, in much of the world, there are few options for education of young children with ASD. If intensive, high staff-to-student ratio intervention is impossible, how can high-quality evidence-based intervention be provided? How can evidence-based practices be implemented in an ethical manner when intensity and staff-to-student ratio do not match expert recommendations, due to limited resources? We focus our discussion on China, where the authors are most familiar with autism services and the social situation, being from China and/or having done many years of work in the field of intervention for young children with autism there. Given the current context, in this article we recommend systematically using group instruction based on ABA principles to increase the effectiveness and intensity of each aspect of intervention programs. While we focus here on China, we believe that our discussion may also have implications for other settings where resources are limited.

Ethical Concerns

Regarding interventions for children with ASD in areas with limited resources, we begin with an analogy. The purpose of analogies is to improve our understanding of something that is less well-understood by comparing it to something different that is better understood. The analogs are usually quite different from each other in many respects, but share some important similarities that are the foundation of the analogy. In our analogy, the important similarities are the fact of limited resources and the pressing need to find an allocation of those resources that is fair. The less familiar analog is allocation of ABA resources in China, while the more thoroughly discussed analog is the allocation of limited medical resources in an epidemic. Other analogies are possible (e.g., education, public safety, housing), but the most thoroughly explored area of resources allocation concerns medical care.

As an analogy, consider that an epidemic occurs somewhere with limited medical resources, infecting 10,000 people. The disease is not lethal, but if not treated, it will significantly impact and potentially reduce most people’s quality of life. There is only one local hospital, which has a limited supply of the most effective medicine. If administered at its fully effective recommended dose, only 20% of infected patients would receive treatment, and most would show significant long-term benefits. If administered at the minimum dosage considered potentially effective, 40% would receive the treatment. Although many of these patients would benefit less than if they had gotten the full recommended dose, most of them would benefit significantly in improved quality of life. If almost everyone who is infected were given an equal dose of medicine, some people might show some benefit, but the vast majority would show no lasting quality of life improvement (Table 1).

Table 1.

Allocation of Limited Medical Resources

Dose	% Treated	Positive Outcomes	Negative Outcomes
Full Strength	20%	1600 significantly improved	400 treated unsuccessfully; 8000 not treated
Minimal Dose	40%	2800 significantly improved	1200 treated unsuccessfully; 6000 not treated
Equal Dose	100%	400 significantly improved	9600 treated unsuccessfully 0 not treated

In this highly simplified analogy, used solely to demonstrate challenges of dosage distribution in a context with limited resources, it seems reasonable to allocate medical resources in an unequal manner (perhaps giving others a less effective drug), presuming this is done by some non-discriminatory allocation method. This is hardly surprising. If two people each needed a heart transplant and there was only one heart available, no one would advocate giving each person half a heart. Some medical treatments, whether drugs or other therapies, are similarly ineffective if done in half-measures. Thus far, this is a fairly traditional top-down resource allocation problem.

One might object that our argument assumes consequentialism as a moral framework, as opposed to deontology or virtue ethics. This objection falls into a common misunderstanding that deontology and virtue ethics do not care about consequences. In fact, deontological theories include duties to bring about good consequences by benefitting others (Locke, 1980; Kant, 1964). Similarly, virtue ethics will invariably include the virtue of beneficence prescribing that people bring about good consequences for others (Hume, 1998; Aristotle, 2009). These theories differ from consequentialism in that they assert that sometimes other factors override the prescription to maximize the benefit to others, but unless some conflicting and overriding duty or virtue is present, all major moral theories agree that we should bring about the best consequences. In our analogy, no such conflicting and overriding duties or virtues exist; therefore, all three predominant types of moral theory will concur that promoting the greater good here is morally appropriate. In the real-world situation in China, it is possible that specific limited situations would have conflicting and overriding duties or virtues. For example, if an autism program promised to provide services of a specific type in order to receive funding from a donor, this could present them with a conflicting duty. Or, if the most effective ABA therapy violated some religious code that would generate a moral conflict. However, these special circumstances need to be specified to call into question and override the general moral presumption to promote the best consequences. Therefore, though we encourage bringing about the best consequences, our conclusions do not require accepting a consequentialist moral framework.

Returning to the analogy, suppose that very few people in this area can afford to pay for the full (or even minimally effective) treatment, though almost everyone could find a way to pay for the equal dose treatment, and there is little to no medical insurance provided by the state and minimal funding to support the hospital acquiring additional resources to treat this epidemic. The hospital cannot afford to give away the medication. Under these circumstances, administering the medicine to those who can afford to pay for at least a minimum effective dose, and perhaps even rationing the medicine by refusing to administer full doses, is an ethically defensible policy decision for the hospital because it is likely to do the most good without violating any overriding duties or virtues. Alternatively, one could also reasonably argue that the hospital would be ethically justified in providing full doses to those who could afford them as long as supplies last. The former is a typical consequentialist conclusion, while the latter might be justified by certain deontological bioethical approaches prioritizing the duty of doctors to patients under their care over potential patients (Garbutt & Davies, 2011). In contrast, equal doses for all seems ethically irresponsible. An appeal to egalitarianism in its defense falls flat, as the heart transplant example illustrates. In other words, in contexts with limited resources, it is sometimes ethically preferable to distribute resources unequally, and may also be ethically appropriate to distribute resources such that those who are treated do not receive a maximally effective treatment.

There is a temptation to take a wide scope view of this situation, invoking the distinction between wide scope resource allocation (i.e., dividing national resources among competing projects) and narrow scope resource allocation (i.e., choosing which patients receive what amount of care) (Daniels, 2006; Menzel, 2007). One might try to resolve this problem by claiming that somebody (i.e., the government) should provide more funding for healthcare generally, or to this particular kind of care, giving our hypothetical hospital more resources. However, we bracket this response by assuming that the government’s wide scope resource allocation either is justified as it stands or is not within one’s power to change, thereby taking wide scope solutions off the ethical table.

In this paper, we are concerned with the analogous situation of intervention services for children with ASD in China. The situation in China does not concern a disease; the analogy regards the distribution of limited resources. In China, intervention for young children with autism and other developmental disabilities is not a fully developed system (Sun et al., 2013). Intervention for young children (typically below age 8) is provided at diverse educational organizations, some private, some funded partly by the government, and some in hospitals (e.g., Chen & Cheng, 2020; McCabe, 2008a, 2013; Sun et al., 2013). Even those with government support also rely on student tuition for their operations, and funding limitations impact teacher-to-student ratio. Recently, the number of organizations providing short-term intervention has grown, but challenges remain due to lack of training, resources, and qualified providers, limiting hours and instruction availability (Guo, 2006; McCabe, 2013; McCabe & Deng, 2018). Thus, there are questions about current intervention models and whether some programs may not efficiently and effectively achieve their goals of benefiting children, perhaps causing them to miss the opportunity for more useful interventions. This article provides recommendations on ethical ways to allocate resources to improve intervention services for children with ASD in China and, by extension, other areas with limited resources.

Background

Since Lovaas (1987) reported on the significant improvement of almost 50% of children with ASD who received 40 hours/week of Discrete Trial Teaching (DTT), an intervention based on the principles of (ABA), strong interest in ABA-based services has grown, including in China. As early as 1998, ABA and DTT were introduced to China via materials and training, and Chinese teachers and families felt that they had finally found a framework for teaching children with ASD that was systematic, understandable, and beneficial. Research describes the background and use of these methods in China (Guo, 2006; Liao et al., 2018; McCabe, 2008a, 2013; McCabe & Tian, 2001; Wang, 2008).

Background: What is Meant by “Applied Behavior Analysis”?

There are common misconceptions regarding ABA, including equating ABA to “table time” and discrete trial teaching (Boutot & Hume, 2012). In fact, ABA is “not a specific set of interventions for students with ASD; rather, ABA is a scientific approach to teaching and learning” (Trump et al. 2018, p. 383). Many evidence-based practices based on ABA principles can be and are commonly used in special education settings for academic and behavior interventions (Trump et al., 2018). Indeed, it is important to use ABA-based methods to teach students to succeed in educational settings beyond one-to-one settings (i.e. Charania et al., 2010). It is also important to note that in our consideration of best practice using ABA-based intervention in China, we follow the beliefs of autistic adult and parent, Eileen Lamb (2019), who writes about her son who also has autism. She states,

You can help someone without changing who they are. Charlie [her son] isn’t in ABA therapy because we want him to be normal. We want him to be safe, independent, and to learn to communicate, to decrease his, and yes, our frustration. We want to fade away the less functional and dangerous behaviors, like playing with the cats’ litter box, swallowing rocks, and running in the street, to give him a better shot at life. I’m not trying to “fix” Charlie’s autism. (Lamb, 2019)

Lamb also points out that in effective ABA services, there is no “one size fits all” and this is important to note regarding individuals as well as regarding contextual fit.

Research on Intensity of ABA Intervention

Regardless of specific ABA-based practices used, effective intervention to promote communication, independence, and functional behaviors emphasizes individualized instruction. For young children, this often includes a significant individual or one-to-one (1:1) component. One intervention model for young children is early intensive behavioral intervention (EIBI), which Reichow (2012) describes as lasting two or more years, and comprising “comprehensive programming for upwards of 40 h per week”, initially emphasizing one-to-one discrete trial teaching (p. 512).

Intensity or amount of intervention is commonly emphasized when examining ABA intervention, including “policy and practice recommendations that young children with ASD receive many EIBI treatment hours” (Pellecchia et al., 2019, p. 1075). Many studies have examined the impact of intensity (or “dosage”) of ABA-based intervention, some finding that intervention with more hours (average 20–40 hours per week) leads to more significant improvement (Dawson et al., 2010; Eikeseth et al., 2002, 2007; Linstead et al., 2017; Lovaas, 1987), and others finding that fewer hours of intervention also lead to gains (Lotfizadeh et al., 2020; Vietze & Lax, 2020; Waddington et al., 2020). Pellecchia et al. (2019) note that Behavior Analyst Certification Board (BACB) guidelines indicate the importance of considering individual needs but include a minimum recommended 10 hours of treatment, “with a strong recommendation for a higher dose” (p. 1076). However, there are reasons in various contexts that lower dosage ABA interventions may be considered, including “cost limitations, scheduling difficulties, [and] availability of competent professionals” (Lotfizadeh et al., 2020, p. 94).

There are limitations to looking only at the duration of an intervention or overall number of hours when considering intensity. Though Rivard et al. (2019) note that in the literature, intensity is generally regarded as how many hours per week of intervention are provided directly, Warren et al. (2007) argue that simply looking at hours or duration does not provide sufficient information about other aspects of intensity, such as frequency of intervention, or “density of teaching episodes” (p. 71). In other words, it is also important to consider intensity as related to “how many opportunities to learn in a single session, in a certain number of sessions per week or month, over a certain number of weeks or months” (Warren et al., 2007, cited in Yoder & Woynaroski, 2015, p. 154).

Ethically Allocating Resources in Specific Contexts

Although it is clear that interventions based on ABA for children with autism are among accepted evidence-based practices (e.g., National Autism Center, 2015), and research indicates that more hours may have more impact than fewer hours of intervention (e.g., Linstead et al., 2017; Lovaas, 1987; Reed et al., 2007), there is an issue of access to resources and services in China, which raises the need for identifying effective methods that do not require so many hours of professional-provided intervention (Waddington et al., 2020, p. 2). Regarding the appropriate dosage of intervention, Pellecchia et al. (2019) argue that instead of thinking “more is better,” we should consider many factors, “including service access, family functioning, cultural and developmental appropriateness of intervention, and the role of other treatment modalities” (pp. 1076–1077). As in our analogy, it is essential to identify strategies to allocate and stretch available resources to do the most good, including considering treatment variations.

To extend the medical analogy, suppose that the limited drug supply was initially envisioned as only being administered intravenously, and that is what the hospital planned to do. What if someone showed that the same drug could be administered as an aerosol and that each effective aerosol dose used significantly less of the drug? Changing the mode of the drug’s delivery could allow many more people to be effectively treated, perhaps dramatically improving outcomes for the overall target population. This would be uncontroversial if the aerosol delivery were as effective as intravenous delivery for all. The point here is that even if the aerosol delivery of the drug were modestly less effective for individuals than the intravenous delivery, from a public health perspective, the ethically preferable choice is still to deliver the drug as an aerosol because they can treat dramatically more people this way using a dosage that is almost as effective as the intravenous delivery. Likewise, to best serve children with autism, it is important to consider “the effectiveness of interventions delivered according to parameters that support wider-scale implementation and better meet existing demand (Peters-Scheffer et al., 2013)” (Rivard et al., 2019, p. 493).

One might argue that the public health perspective is not the only relevant ethical perspective and that things look different from the perspective of the individual provider who has specific individuals under their care. We acknowledge that there is an ethically defensible (though not uncontroversial) position that individual providers have an overriding obligation to provide the best possible care to their clients, even at the cost of broader public health. Our point is not to argue that individual providers are morally required to adopt our suggested methods of group instruction (instead of offering full intensity 1:1 treatments). Our point is rather that both approaches are morally defensible (in contrast to equal but ineffective doses). Of course, if you are not a provider with individuals currently in your care (e.g., if you are considering how to set up a new treatment facility or what government incentives to create for setting up facilities), then there is a compelling reason to take the public health perspective.

Autism Intervention Resources in the Context of China

In many countries, intervention for autism is provided as part of the health or education system. While this does not guarantee equal treatment for all, it makes treatment accessible to more families than in a system with more limited resources. In China, there is limited funding for intervention for young children, mainly provided through reimbursement for participation in a limited number of programs and centers, which are not connected to the public education system (Li, 2019; State Council, 2015). Since intervention for children with autism began at a state-run hospital in the early 1990s, followed by the first private program for children with autism (McCabe, 2008a; McCabe & Tian, 2001), similar private and hospital programs have been established throughout the country, with increasingly rapid growth in numbers since the early 2000s (Shieh & Deng, 2011; McCabe, 2013; McCabe & Deng, 2018).

Regarding ABA-based intervention in China, there is a wide range of intervention programs, with over 10,000 intervention programs by 2020 in one estimation (China Disabled Persons’ Federation, 2021). Research shows that though intervention has developed rapidly, and there are substantial efforts to serve children, often there is not enough training for professionals, nor sufficient instructional resources (Liao et al., 2018; McCabe & Deng, 2018).

In part because there are not enough teachers, parents have played a central role in the development of intervention services for children with autism in China, from founding some of the earliest programs to attending intervention programs with their children, either to be the recipient of training or to accompany their child and provide a range of logistical and behavioral support (Feng & Wei, 2018; Liu et al., 2020; McCabe & Tian, 2001; McCabe, 2008b; Zou et al., 2008). Indeed, to understand service provision in China, it is important to highlight that in addition to children, there is frequently an equal number of parents accompanying them in and out of the classroom.

As most of these intervention programs survive solely or largely from tuition (McCabe & Deng, 2018), the cost is often based on staff-to-student ratios, and thus it is more costly to provide individual instruction. At many programs, children receive between 30 and 60 minutes per day of 1:1 instruction, or 2.5–5 hours per week (Liao et al., 2018; McCabe, 2013), with the rest of the time spent attending classes with 8–10 other children and their parents, who typically accompany them (Feng & Wei, 2018), in music, arts and crafts, exercise, etc. In many programs, group classes are staffed by one teacher, occasionally with an assistant teacher also. Parents observe and assist their children to participate, providing behavioral support as well.

As the field of services for children with autism developed, training and information about effective ABA-based methods also became available in China by the early 2000s (Hou, 1995; Zou et al., 2008), some of which emphasized that effective instruction should include discrete trial instruction for close to 30 hours per week (Chen et al., 2003). However, limited resources make this impractical. The number of children with autism has grown faster than the capacity of programs to provide evidence-based intervention by qualified practitioners.

Thus, a significant problem is that those seeking treatment to improve their children’s quality of life do not always get the effective treatment their children need and are being promised. There are several possible reasons. It may be because the service provider wants to deliver high-quality instruction but has not received training in evidence-based practices. It may also be because the intervention is being provided with a significant lack of treatment fidelity, which could lead to the child finding intervention sessions aversive, thus impacting the effectiveness and implementation of treatment. Finally, perhaps the child does not mind the treatment, but it does not adequately help the child and parents achieve their goals, causing parents to become frustrated and withdraw from treatment. There is an opportunity cost for families who are not pursuing better and more effective treatment in any of these cases.

Currently, another long-term problem involves programs that use ABA-based intervention but cannot provide an adequate intensity, which will likely be ineffective. Sub-optimal outcomes associated with ABA treatment causes a public perception of ABA as an ineffective intervention. This could result in a long-term negative consequence of many other people in the future being unwilling to try an effective intervention (ABA) even when offered by a highly effective program (McCabe, 2013).

Many individuals in China have excellent educational skills or potential to become skilled teachers, but social factors limit intensive intervention as commonly seen in the literature. With not enough teachers, parents often provide too much support to children, who are too often passive recipients of watching a video or listening to the instructor. Thus, we advocate for a new way of thinking about “intensity” and effective practice, which focuses more on evidence-based practices in group instruction, to serve more students effectively.

Recommendations: Exploring an Alternative Model of Service Provision

Given the current context, what is the solution to provide efficient evidence-based services to a larger number of children with ASD? Our recommendation is to systematically use group instruction based on ABA principles to multiply the hours of evidence-based intervention each student receives, and thus increase the effectiveness and intensity of each aspect of intervention programs. The effectiveness of ABA-based intervention in group settings has been demonstrated (e.g., Leaf et al., 2013; Taubman et al., 2001). This necessarily involves ensuring treatment fidelity whether in 1:1 or group settings, an important part of intensity (Codding & Lane, 2015). Importantly, switching from 1:1 instruction to group instruction may be a new way of service delivery, but it will reach more children.

When the government does not adequately fund intervention, and without enough qualified professionals, accessing affordable intensive intervention programs is impossible for many (Waddington et al., 2020). Considering alternative treatment models for children with autism to receive evidence-based intervention is essential. Practices and policies that emphasize the need for a minimum number of hours in fact may “perpetuate service disparities for families who, for myriad reasons, cannot access the recommended dose” (Pellecchia et al. 2019 p. 1077).

Alternative practices thus should be considered, such as lower-intensity intervention (Waddington et al., 2020) or, for example, providing DTT in groups of two or more (e.g., Leaf et al., 2011; Leaf et al., 2013; Taubman et al., 2001). Special education in China is heavily group-oriented, whether in inclusive settings where little support is provided (Liu et al., 2020; Qu, 2021), or at intervention programs where students spend most of their day in group instruction with parents accompanying them (Feng & Wei, 2018; Yan et al., 2020). Given this context, our proposal to focus on group instruction has strong social validity, targets socially significant goals, uses socially appropriate methods, and intends to bring positive changes to each individual (Wolf, 1978).

It is important that results from research settings generalize into more natural settings (Rivard et al., 2019), and we should consider contextual variables related to each client and setting to ensure that interventions work in practice (Rehfeldt, 2011). Rehfeldt’s argument focuses on moving results from clinic to natural setting; we extend this argument about context to consider extending results of research from one national context to another. Scholars emphasize “the importance of accounting for socio-cultural barriers” related to “intervention uptake, engagement, and efficacy” (Pellecchia et al., 2019 p. 1076). Implementation of evidence-based practices should be relevant to the local setting. While focusing heavily on group instruction may be unnecessary in a resource-rich nation or region where 1:1 services are widely available, this does not call into question that ABA-based group instruction is the most ethically defensible approach in a very different context.

ABA-based intervention is not synonymous with any specific intervention method and effective practices in group settings have been researched. If human and financial resources do not allow children to receive extensive 1:1 intervention, teachers will need to find a way to use evidence-based practices in group classes effectively. For example, while taking data on student behavior may be easier in a 1:1 setting, there are effective practices for measuring behavior and collecting data in groups such as, for example, PLACHECK, a modified method of time sampling used in group settings (Doke & Risley, 1972; Spangler & Marshall, 1983). More recently, Taubman et al. (2001) examined the effectiveness of group DTT in a preschool classroom of eight children with disabilities, which was more efficient than 1:1 instruction and provided more opportunities for interaction and observational learning. Leaf et al. (2013) examined DTT in a group format; in this comparison study, they found it equally effective to 1:1 DTT. In the study, group DTT involved one teacher working with multiple students, implementing both simultaneous trials with choral responding and sequential trials where students responded one at a time. Working in groups allowed for students to access more discrete trial intervention and thus was beneficial both due to efficiency and because it was more similar to a naturalistic environment (Taubman et al., 2001) and “more closely represents instruction that can be found in general education classrooms” (Leaf et al. p. 83).

Too often, whole-class teaching, a common instructional model at intervention programs in China, tends toward passively receiving information (Stephenson, 2006). However, for children with significant disabilities, including ASD, evidence-based practices must include “repeated presentations of instructional stimuli, high levels of student responding, systematic prompt fading procedures, and consistent feedback” (Pennington & Courtade, 2015, p. 40). Teachers also need to effectively keep students with ASD actively engaged and learning in group settings. Research on the importance of students’ active engagement during whole-class instruction emphasizes increasing opportunities to respond (OTR) (Adamson & Lewis, 2017; Haydon et al., 2012; Pennington & Courtade, 2015; Randolph, 2007; Thompson et al., 2019). Increased OTR, whether verbal, gestural, or written, allow teachers to observe and assess children’s understanding and progress on specific goals, provide immediate feedback, take data, and adjust instruction (Adamson & Lewis, 2017; Haydon et al., 2012). As some authors argue, it is a high rate of opportunities to learn that defines intensity (Warren et al., 2007; Yoder & Woynaroski, 2015), and this is also relevant and feasible in group settings (e.g., Leaf et al., 2013; Taubman et al., 2001).

ABA-based practices stress the core principle of reinforcement and require ongoing data collection, and whole-class instruction should not be an exception. To make whole-group instruction more effective, and to increase opportunities to receive feedback and reinforcement, responding in unison can be utilized, including using response cards or gestures so that all children, even those who are non-verbal, can actively respond at the same time (Common et al., 2020; Heward et al., 1996; Owiny et al., 2018). Moreover, children can be taught skills that are useful for responding in group settings, such as appropriate hand-raising (Charania et al., 2010).

Currently, in intervention programs around China, a significant portion of the day is spent in group instruction, with only 30–60 minutes out of 6–7 hour days in 1:1 settings where ABA principles are emphasized. Many of these group instruction classes are taught by teachers who are also the ABA therapists providing 1:1 instruction. Thus, there are at least two potential ways to increase ABA intensity: first, making more 1:1 instruction available for students and second, using additional existing group class time to provide instruction to the group based on ABA principles. The first requires a large increase in qualified practitioners, increasing costs. The second option is more contextually appropriate and feasible. For example, if therapist A provides eight students with one 30-minute ABA session each day, this will take four hours. This same therapist could (if properly trained) provide group instruction based on ABA principles, with high rates of opportunities to respond (OTR) to all eight students for four hours. This would dramatically increase the dose of ABA-based intervention for all. Even if each hour of ABA instruction is somewhat less effective than one-to-one therapy, the up to eightfold increase in the hours of ABA-based instruction will more than compensate for this, assuming instruction is provided with high treatment fidelity. Teachers could ensure increased OTRs and opportunity for reinforcement and data collection in a group setting in part, in many cases, by training parents who accompany their children to class.

Discussion: Advocating for Evidence-Based Group Instruction in China

This article has addressed the question of what is the most ethical way to distribute scarce resources. Specifically, in a context where there are not enough practitioners nor sufficient funding to make early intensive one-to-one intervention feasible, how might intervention be effectively provided? While we hope for a dramatic increase in the number of well-qualified practitioners in China, given the enormous number of children with ASD, any imminent government investment will not be enough to support the majority of children in the foreseeable future. Therefore, the increase in qualified practitioners will continue at a modest pace, unable to meet demands for the recommended dose or intensity of evidence-based treatment, especially using a 1:1 format.

Focusing on ways to provide evidence-based, ABA-based intervention in a group, whether in group discrete trial format (Taubman et al., 2001; Leaf et al., 2013) or using other contextually-relevant (Rehfeldt, 2011) educational methods that combine common practices and expectations in China with evidence-based methods, is essential. It is important to be cognizant of how to implement evidence-based practice in a setting that is different from what has typically been done in the research, including considering socio-cultural factors and barriers (Pellecchia et al., 2019).

Outcomes improve when students have more contact hours with an ABA interventionist (e.g., Linstead et al., 2017; Lovaas, 1987; Reed et al., 2007), and more frequent opportunities to respond (Warren et al., 2007). Without a dramatic increase in the number and affordability of qualified practitioners, an optimal way to increase the average contact hours and intensity is to have a single trained teacher provide instruction to multiple students simultaneously, using ABA-based practices that are effective in group settings. It is important that teachers of children with ASD are trained in the fundamental principles and related practices of ABA that are meaningful within their own existing educational program model.

Returning to the analogy, group instruction stretches out a relatively fixed supply of people qualified to deliver ABA-based interventions just as the hypothetical transition to an aerosol drug delivery would stretch out the fixed supply of that drug. Once we rule out distributing equal but ineffective doses, there remains some potential ethical friction from the fact that a person might be better off with a relatively high dose of 1:1 instruction than with a similarly high dose of group instruction. If there is a shift to more group instruction, some families (e.g., high socio-economic status families) whose children might have received many hours of 1:1 intervention may find that kind of intervention less available. In this way, a very small group may be worse off. Looked at differently, some practitioners or advocates may remain committed to offering only what one might call the gold standard of care, which involves 1:1 intervention, challenging our recommended move to group instruction.

Is it ethical to make a change that negatively impacts some people, in order to have a more positive impact on the general population of children requiring services? We argue that a public service organization (e.g., a hospital, center, or a school serving children with special needs) that is currently inefficiently using resources does not violate ethical principles when they are unaware of a more efficient method, and they are not culpable for this lack of awareness. However, if there is a better way to help people and they know about it, then there is a strong case to be made that they have an ethical obligation to use the method that is more broadly effective for the people they are serving (assuming that this method does not violate anyone’s rights). Therefore, even if the benefits of group instruction are less for an individual than the benefits of a full dose delivered in the traditional way (i.e., intravenous delivery of the drug, or 1:1 DTT), such a transition to this innovative delivery method is certainly morally permissible, and may even be morally required because more children will be served and receive evidence-based intervention. Indeed, teachers would surely welcome a shift that enabled more children to make greater progress.

The BACB Professional and Ethical Code for behavior analysts (BACB, 2014) notes that individuals have the “right to effective treatment,” yet in a context where resources do not allow for high intensity individualized 1:1 ABA-based intervention, how can we meet the goal of effective treatment? An educational approach that limits its use of evidence-based practices to only during very limited 1:1 sessions with instructors leads to significant ethical concerns regarding the distribution of scarce resources. The more fundamental right here is the right to effective treatment, not the right to a particular mode of receiving that treatment.

Because young children in China who attend intervention programs spend most of their day in group settings, it is urgent that teachers begin to use evidence-based practices in all classes, not just during 1:1 time. Dawson and Bernier (2013) argued that parents should “play a central role in providing intervention… [because they] spend more time with their children than any therapist...In this way, every interaction becomes a learning opportunity” (p. 1466). Similarly, if all teachers, including those running group classes, mastered ABA-based methods, every interaction and activity would be a learning opportunity and not risk leaving children passively sitting in a group. Research and implementation that focus on how practices, while remaining evidence-based, fit into various contexts in the real-world are needed (Rivard et al., 2019).

Thus, in this article we have proposed re-thinking what it means to “do ABA” in educational programs in China, or anywhere that resources do not allow for sufficient 1:1 intervention. Research findings should be effectively applied in real-life classrooms in contextually relevant ways (Rehfeldt, 2011). If students are in a whole group setting for all or most of their day, it is essential to increase OTRs in that group setting so the teacher can provide numerous antecedents for active responding, provide feedback and reinforcement, and take data on skill acquisition, while reducing wait time and off-task behavior (Haydon et al., 2012; Pennington & Courtade, 2015). In a class in China with eight children, eight parents, and one teacher, it might be feasible to provide parent training to assist the teacher in providing feedback and taking data. The unique role of parents at autism intervention programs in China has been previously examined (Feng & Wei, 2018; McCabe, 2008b) and the potential for them to assist in data collection is worth examining in more detail, but that is beyond the scope of this article.

When teachers are taught to use evidence-based practices effectively in a group format, such as for children with ASD at programs in China, this will result in significantly more effective education [“treatment”] being provided to many more students. The more limited number of families who can now afford to hire private ABA therapists to provide intensive 1:1 intervention may still be able to do so. Despite any inequality created, our argument does not claim that it is morally impermissible to create such high-quality, intensive programs for those who can afford them. Instead, we are concerned that for the vast majority of children, spending only 30 minutes receiving rigorous evidence-based intervention out of a 6-hour instructional day in an educational environment is not a good use of their time.

To help ensure more time actively engaged, there needs to be a reform of group and whole-class instruction, and a renewed understanding of what it means to practice evidence-based practices based on the science of Applied Behavior Analysis. If the students who attend these programs could go from receiving 3–5 hours of 1:1 ABA instruction per week to 18–30 hours of group ABA-based instruction, even if each of these hours is somewhat less intensive, if OTRs are kept high and data is collected, analyzed, and utilized on the children’s performance, the overall effect of this dramatic increase in ABA dosage will be highly beneficial. It is important to develop practices that are relevant to the local context and practices (Pellecchia et al., 2019; Rehfeldt, 2011), while remembering that high intensity relates to dense learning opportunities and not just hours per day (Warren et al., 2007). This goal has social validity as group instruction is the natural context for students in schools and intervention programs throughout China. These recommendations are based on the context of China but can be extended to other settings with similar resource limitations.

Given the current number of children and teachers and the costs of services in China, a new model of instruction is needed that will provide an adequate level of intervention to a much larger number of children. Given that instruction in China is pervasively in whole-group settings, training teachers in methods and strategies where they can provide intervention based on research-based practices to a group is a socially valid and urgent priority. This can both control costs and improve effectiveness of intervention. As a socially appropriate solution relevant to the local context, we are convinced that overall results will be very beneficial, and many more children will benefit.

Data Availability

n/a.

Declarations

Conflicts of Interest/Competing Interests

None.