To summarize baseline data and lessons learned from the Autism Learning Health Network, designed to improve care and outcomes for children with autism spectrum disorder (ASD). We describe challenging behaviors, co-occurring medical conditions, quality of life (QoL), receipt of recommended health services, and next steps.
A cross-sectional study of children 3 to 12 years old with ASD receiving care at 13 sites. Parent-reported characteristics of children with ASD were collected as outcome measures aligned with our network’s aims of reducing rates of challenging behaviors, improving QoL, and ensuring receipt of recommended health services. Parents completed a survey about behavioral challenges, co-occurring conditions, health services, and the Patient-Reported Outcomes Measurement Information System Global Health Measure and the Aberrant Behavior Checklist to assess QoL and behavior symptoms, respectively.
Analysis included 530 children. Challenging behaviors were reported by the majority of parents (93%), frequently noting attention-deficit/hyperactivity disorder symptoms, irritability, and anxiety. Mean (SD) scores on the Aberrant Behavior Checklist hyperactivity and irritability subscales were 17.9 (10.5) and 13.5 (9.2), respectively. The Patient-Reported Outcomes Measurement Information System Global Health Measure total score of 23.6 (3.7) was lower than scores reported in a general pediatric population. Most children had received recommended well-child (94%) and dental (85%) care in the past 12 months.
This baseline data (1) affirmed the focus on addressing challenging behaviors; (2) prioritized 3 behavior domains, that of attention-deficit/hyperactivity disorder, irritability, and anxiety; and (3) identified targets for reducing severity of behaviors and strategies to improve data collection.
Challenging behaviors are common in children with autism and have the potential to negatively impact health, well-being, and quality of life. Clinicians and parents are united in a goal to improve outcomes for children with autism.
Children enrolled in the Autism Learning Health Network demonstrate high rates of challenging behavior (93%) and have significantly lower quality of life compared with the general population. These data inform baseline rates and specific improvement targets for the network.
The established multisite clinical research network, Autism Speaks Autism Treatment Network (ATN)–Autism Intervention and Research Network on Physical Health (AIR-P), was established in 2008 and includes academic medical centers1 in the United States and Canada, which conduct research and develop clinical practice standards for the identification and treatment of co-occurring medical conditions in autism spectrum disorder (ASD).2–4 In December 2014, the ATN–AIR-P began to transition to an Autism Learning Health Network (ALHN) with the goal to integrate quality improvement (QI) to improve care and outcomes for children with ASD. A learning health network, based on the Institute of Medicine’s Learning Healthcare System,5,6 provides a platform for QI, research, innovation, and sharing of ideas and data to drive advances in care and translation of research into practice.5,7
The ALHN design process focused on outcomes identified as clear priorities by both parents and clinicians during an in-person design session of key stakeholders, which is described elsewhere6 : optimizing the physical health of children with ASD and treating challenging and interfering behaviors, which are common in children with ASD. Although it is well known that children with ASD can often have significant behavioral difficulties, we considered the following in developing the key driver diagram (KDD) framework for the network: (1) the heterogeneity of the group of children with autism across the spectrum with respect to core symptom severity, cognitive ability, communication levels, and severity of challenging behaviors; (2) difficulty of defining a “good outcome” in ASD because this is likely subjective; and (3) a lack of standardized measures for assessing outcomes in children with ASD. The KDD provides the framework for the network and highlights 3 areas of focus8 (Fig 1): (1) ensuring that children with ASD receive recommended routine health services to optimize physical health, (2) reducing rates of challenging behaviors, and (3) improving quality of life (QoL). Our theory was that optimizing both the physical and behavioral health of children with ASD would result in improved QoL. We decided to collect data directly from families about behavioral concerns and QoL because family members are best positioned to report on the impact of ASD and associated behaviors on the child’s day-to-day functioning as well as receipt of routine health care services. We recognized that using patient- or parent-reported outcomes (PROs) as a primary outcome could be challenging but was critical for collecting the information we needed.
KDD for the ALHN. AAP, American Academy of Pediatrics; ECHO, Extension for Community Healthcare Outcomes; SMART, specific, measurable, achievable, realistic, and timely.
KDD for the ALHN. AAP, American Academy of Pediatrics; ECHO, Extension for Community Healthcare Outcomes; SMART, specific, measurable, achievable, realistic, and timely.
The purpose of this article is to describe the results of data collected on the first 530 children enrolled in the ALHN, how these data would help to set targets for network outcomes, and the feasibility of using PROs. Our hypotheses were that children with ASD enrolled in ALHN would (1) commonly demonstrate challenging behaviors (≥65% of children on the basis of studies of co-occurring psychiatric conditions9–11 ) (2) have a lower parent-reported QoL compared with a general pediatric population, and (3) may not consistently receive routine recommended health services (eg, well-child care, dental visits) per pediatric guidelines12 because of the impact of challenging and interfering behaviors.
Methods
The KDD
The first draft of the KDD was developed once the aims of improving receipt of routine health care, QoL, and the impact of challenging behaviors were prioritized at the in-person design day meeting and potential drivers for these aims were hypothesized.6 Whereas the aims for ALHN remain the same over time, the KDD is an iterative document, in which the key drivers and interventions can be modified as new information about the system becomes available and the aims may be modified to become more specific. Descriptive behavioral, QoL, and health care data from children enrolled in ALHN were analyzed to set more precise targets for improvement.
Study Population
Participants were recruited from a population of children with ASD receiving care at 13 sites participating in the ATN–AIR-P network between September 2016 and November 2018 (Supplemental Table 4). Participants included children 6 to 12 years old initially (September 2016–September 2017), with the age range for inclusion expanding to 3 to 12 years old in October 2017. To be eligible for participation, children needed to have an established diagnosis of ASD on the basis of Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, or Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, criteria and receive ongoing care at one of the ALHN clinical sites. Families were invited to participate in ALHN by mail, e-mail, phone calls, or face-to-face conversations at clinic visits. Written informed consent to join ALHN and allow data to be used for research was obtained from parents. This study was approved by the Massachusetts General Hospital (coordinating center) Institutional Review Board and by the institutional review boards or research ethics boards at all 13 sites.
Measures
Data collection tools, processes, and measures were developed on the basis of the aims and drivers outlined in the KDD. Families consenting to participate in the ALHN were asked to complete surveys either through a mobile app on a smart phone or tablet, using an online Web portal, or on paper. Parents were asked to complete 3 surveys: a parent survey developed specifically for the ALHN and 2 previously published and validated tools, the Aberrant Behavior Checklist (ABC)13 and the Patient-Reported Outcomes Measurement Information System Global Health Measure (PGH-7).14 The ALHN parent survey included demographics; parent perceptions about challenging behaviors and co-occurring medical conditions (gastrointestinal [GI] and sleep problems), identifying which behaviors and/or symptoms were problematic and describing severity, frequency, and level of concern; and receipt of routine, recommended health care services (eg, well-child visits, dental care). The ABC is a 58-item caregiver report form developed to assess maladaptive behaviors in individuals with developmental disabilities and is commonly used in studies of children with ASD.15 The ABC contains the following subscales: irritability, lethargy and/or social withdrawal, stereotypic behavior, hyperactivity and/or noncompliance, and inappropriate speech. Items are rated on a 4-point Likert-type scale, with higher scores on each subscale indicating greater levels of problematic behavior. Parents were asked to complete the PGH-7 parent-proxy report, which is a brief, valid, and reliable (internal consistency α = .84) tool for assessing overall health, well-being, and QoL in pediatric populations.14 Items are scored on a 5-point Likert-type scale, with higher scores indicating better health and well-being, and the total score can range from 7 to 35.
Statistical Analysis
We conducted cross-sectional analyses of parent-reported survey data. Descriptive statistics were tabulated for categorical variables as frequencies and percentages and for continuous variables as means with SD, medians, and ranges. The ABC and PGH-7 scales were scored according to published instructions16 and were reported as described above for continuous variables, with the ABC reported by subscale and the PGH-7 reported as a total score. Because standardized percentile or clinical cutoff scores are not available for the ABC and the PGH-7, to provide a general context of how the ALHN population’s scores compare to other published study populations, mean (SD) scores were compared between our sample and published samples by using a 2-sided Student’s t test with a Bonferroni correction for multiple comparisons (significance cutoff of P < .005 with correction). Of note, because of age differences between the ALHN and comparison groups, these statistics provide only a crude estimate of similarities and differences between the groups.
Results
Demographics
Data were collected on 619 children across 13 ALHN sites out of a total of 1319 who consented to participate in the ALHN (47% overall survey completion rate; range: 25%–89% by site). Detailed information about each ALHN site, labeled as sites A through M, can be found in Supplemental Table 4. For this analysis, demographic data were available for 617, behavior data from the parent survey for 530, ABC data for 510, and PGH-7 data for 526 subjects. Missing data were due to incomplete parent survey responses. Participating children had a mean age of 8.3 (2.6) years, and 80% were male. In terms of ethnicity and race, 86% were non-Hispanic, 87% white, and 9% African American (Table 1). There was wide variation in the number of children enrolled in the registry by center, with the number of participants per site ranging from 2 to 200, with a median of 31 children enrolled per site. The site with the fewest enrollees joined ALHN most recently in March 2018, and another lower enrolling site left ALHN in 2017.
Demographics
| Characteristic | n (%) |
|---|---|
| Age, mean: 8.3 y ± 2.6 | |
| Age, y (n = 617) | |
| 0–3 | 35 (5.7) |
| 4–6 | 150 (24.3) |
| 7–9 | 223 (36.1) |
| 10–12 | 209 (33.9) |
| Sex (n = 614) | |
| Male | 491 (80.0) |
| Female | 123 (20.0) |
| Ethnicity (n = 618) | |
| Hispanic or Latino | 66 (10.7) |
| Non-Hispanic or non-Latino | 532 (86.1) |
| Unknown | 20 (3.2) |
| Race (n = 595) | |
| Asian American | 21 (3.5) |
| African American | 51 (8.6) |
| White | 519 (87.2) |
| Other | 41 (6.9) |
| Characteristic | n (%) |
|---|---|
| Age, mean: 8.3 y ± 2.6 | |
| Age, y (n = 617) | |
| 0–3 | 35 (5.7) |
| 4–6 | 150 (24.3) |
| 7–9 | 223 (36.1) |
| 10–12 | 209 (33.9) |
| Sex (n = 614) | |
| Male | 491 (80.0) |
| Female | 123 (20.0) |
| Ethnicity (n = 618) | |
| Hispanic or Latino | 66 (10.7) |
| Non-Hispanic or non-Latino | 532 (86.1) |
| Unknown | 20 (3.2) |
| Race (n = 595) | |
| Asian American | 21 (3.5) |
| African American | 51 (8.6) |
| White | 519 (87.2) |
| Other | 41 (6.9) |
Co-occurring Behavioral and Medical Conditions
Challenging behaviors were the most commonly noted co-occurring concern, with 93% of parents noting problematic behaviors in the previous month and, of those, 85% indicated that their child’s biggest behavior problem was of moderate or worse severity (moderate [47%], severe [28%], extremely severe [10%]) and 61% reported high frequency (occurring every day or many times per day). The types of behavior challenges most commonly reported related to 3 areas: attention-deficit/hyperactivity disorder (ADHD), irritability and/or aggression, and anxiety (Fig 2). Similarly, when parents were asked to indicate the single most challenging behavior, these 3 behavioral complexes were most commonly endorsed (Fig 3). More than half (53%) of children were taking medication to address behavior challenges.
Types of behavior challenges reported by parents who indicated their child had challenging or difficult behaviors in the past month. Note that parents can give >1 answer and that categories are not mutually exclusive.
Types of behavior challenges reported by parents who indicated their child had challenging or difficult behaviors in the past month. Note that parents can give >1 answer and that categories are not mutually exclusive.
The mean (SD) scores for the ABC subscales reported for the ALHN population are listed in Table 2, which also reveals ABC scores reported in other published study populations.17,18
ABC Scores by Subscale, Comparing ALHN Sample Scores to Other Published Study Populations
| ALHN Scores | Scores Reported in Other Published Study Populations | ||
|---|---|---|---|
| Children With ASD Mean (SD), N = 511 | ASD Sample Mean (SD), N = 604a | General Special Education Sample Mean (SD), N = 195b | |
| ABC subscale | |||
| Hyperactivity and/or noncompliance | 17.9 (10.5)c | 17.3 (10.9) | 12.92 (11.70)c |
| Irritability | 13.5 (9.2)d | 11.8 (9.8)d | 8.50 (8.11)c |
| Lethargy and/or social withdrawal | 8.9 (7.6)c | 9.4 (7.1) | 5.47 (6.05)c |
| Stereotypic behavior | 4.0 (4.1)c | 4.6 (4.2) | 2.71 (3.99)c |
| Inappropriate speech | 3.5 (3.2)c | 3.9 (2.9) | 2.01 (2.60)c |
| ALHN Scores | Scores Reported in Other Published Study Populations | ||
|---|---|---|---|
| Children With ASD Mean (SD), N = 511 | ASD Sample Mean (SD), N = 604a | General Special Education Sample Mean (SD), N = 195b | |
| ABC subscale | |||
| Hyperactivity and/or noncompliance | 17.9 (10.5)c | 17.3 (10.9) | 12.92 (11.70)c |
| Irritability | 13.5 (9.2)d | 11.8 (9.8)d | 8.50 (8.11)c |
| Lethargy and/or social withdrawal | 8.9 (7.6)c | 9.4 (7.1) | 5.47 (6.05)c |
| Stereotypic behavior | 4.0 (4.1)c | 4.6 (4.2) | 2.71 (3.99)c |
| Inappropriate speech | 3.5 (3.2)c | 3.9 (2.9) | 2.01 (2.60)c |
Because full distribution of data for published populations was not available for review, and there are some differences in age groups between samples, it is possible that data may violate the assumptions of a t test and that the statistical significance of the comparisons may be overstated.
ASD sample: scores from Kaat et al18 reflect children ages 6 to 12 y old.
General special education sample: scores from Brown et al17 include combined sample mean scores (ages 6–10 y), as was used by Schmidt et al24 .
Statistically significant differences (P < .005 with Bonferroni correction) between ALHN and Brown et al17 sample; on all ABC subscales, P < .0002.
Statistically significant difference (P < .005 with Bonferroni correction) between ALHN and Kaat et al18 sample; irritability scale only, P = .003.
Both GI problems (44%) and sleep problems (55%) were commonly reported by parents. Of those reporting GI problems, 31% reported being moderately or extremely concerned, and constipation was the most commonly endorsed GI problem (65%). Of those reporting sleep problems, 41% indicated a moderate or extreme level of concern.
QoL, Well-being, and Recommended Health Services
On the PGH-7, the mean (SD) total score of ALHN participants was 23.6 (3.7) (Table 3). The majority of children enrolled in ALHN had received well-child care from a primary care clinician in the past 12 months (94%) and had received a routine dental checkup (85%) per parent report.
Patient-Reported Outcomes Measurement Information System PGH-7 Parent-Proxy Scores: Total Score and Item-Level Scores Comparing ALHN Sample to Published Normative General Pediatric Sample
| ALHN (N = 526), Mean (SD) | Normative General Population Sample (N = 1807a), Mean (SD) | |
|---|---|---|
| PGH-7 domain | ||
| Total score | 23.6 (3.7) | 28.9 (4.2) |
| Overall health | 3.8 (0.9) | 4.3 (0.8) |
| QoL | 3.9 (0.9) | 4.3 (0.79) |
| Physical health | 3.8 (1.0) | 4.3 (0.81) |
| Mental health | 2.9 (1.0) | 4.1 (0.96) |
| Feel sad | 2.6 (0.8) | 3.8 (0.91) |
| Fun with friends | 3.0 (1.0) | 4.1 (0.84) |
| Parents listen to ideas | 3.6 (1.0) | 4.1 (0.83) |
| ALHN (N = 526), Mean (SD) | Normative General Population Sample (N = 1807a), Mean (SD) | |
|---|---|---|
| PGH-7 domain | ||
| Total score | 23.6 (3.7) | 28.9 (4.2) |
| Overall health | 3.8 (0.9) | 4.3 (0.8) |
| QoL | 3.9 (0.9) | 4.3 (0.79) |
| Physical health | 3.8 (1.0) | 4.3 (0.81) |
| Mental health | 2.9 (1.0) | 4.1 (0.96) |
| Feel sad | 2.6 (0.8) | 3.8 (0.91) |
| Fun with friends | 3.0 (1.0) | 4.1 (0.84) |
| Parents listen to ideas | 3.6 (1.0) | 4.1 (0.83) |
The total score range is 7 to 35, and each subscale score range is 1 to 5.
Modifications to the KDD
A specific goal of decreasing the proportion of children with moderate-to-extremely severe behavior challenges from 85% to 70% by September 2020 was set on the basis of current performance of the system and consensus opinion of the ALHN leadership team of a feasible improvement goal. A plan was developed to achieve this goal through a more targeted focus on improved identification and treatment of the most common behavior challenges and priorities reported by parents: ADHD, irritability, and anxiety. Similarly, the aim for improving “very good” or “excellent” QoL scores was specifically set to increase from 68% to 75%. The goal for receipt of routine health services was to maintain current performance of the system at >90%.
Discussion
Parents of children with ASD participating in the ALHN reported high levels of concern about challenging behaviors and co-occurring conditions of sleep and GI problems, relatively poorer QoL compared with a general US population of children, and high levels of receipt of well-child and dental care. ALHN, designed with a goal of improving care and outcomes for children with ASD, has used these data to prioritize an initial focus on challenging behaviors and to define specific, measurable targets for improvement.
Co-occurring Behavioral and Medical Conditions
The high rates of reported behavioral concerns and co-occurring sleep and GI problems in children with ASD enrolled in the ALHN is not surprising. Children with ASD are at high risk for challenging behaviors such as aggression19 and commonly have co-occurring medical conditions such as GI20 and sleep problems.21–23 Co-occurring psychiatric conditions are extremely common in children with ASD (70%), especially ADHD, anxiety, and oppositional defiant disorder, and 41% to 66% are reported to have 2 or more psychiatric conditions.10,11
The ALHN sample’s ABC scores were compared to published mean scores for 2 similar-aged study populations, including a study of children in general special education programs by Brown et al17 and a study of children with ASD by Kaat et al18 (Table 2). ABC subscale scores for children in the ALHN were similar to the sample of children with ASD,18 with the only significant difference being a higher irritability mean score for the ALHN sample (13.5 vs 11.8; P = .003). ABC scores on all subscales were significantly higher (P < .001) in the ALHN sample than those reported for children from a general special education population, which included children with a range of disabilities (deafness, blindness, epilepsy, cerebral palsy, medication use, and “other” conditions).17,24
Scores on the hyperactivity and irritability subscales of the ABC in particular were elevated in the ALHN population compared with a normed special education population.17 These 2 ABC scales represent behavioral symptoms similar to the prevalent psychiatric diagnoses described10,11 and the types of behavior challenges most commonly reported by parents in the ALHN: ADHD symptoms, irritability, tantrums, repetitive thoughts and/or behaviors, not following directions, and anxiety.
QoL, Well-being, and Recommended Health Services
PGH-7 QoL Data
The ALHN participants’ mean total score of 23.6 was significantly lower than the published normative pediatric population mean of 28.9 reported by Forrest et al14 (P < .001), with a comparison between these populations shown in Table 3. Scores on each of the 7 scale items were also significantly lower than published means (P < .001).14 Our findings of significantly lower QoL in children with ASD relative to the general pediatric population were consistent with our hypothesis and with a validation study of the PGH-7 indicating children with ASD scored 1.03 SD units lower than those without ASD.25 Studies have consistently found that children with ASD are reported to have lower QoL than typically developing peers,26–28 with similar results reported from studies using the Pediatric Quality of Life Inventory29 and a range of other QoL assessment tools.26,27 Behavior problems in individuals with ASD have been found to be significantly related to QoL, with greater levels of behavioral difficulty associated with poorer QoL in both children and adults.26,30
Recommended Health Services
The majority of children enrolled in the ALHN had received routine well-child care and dental care in the past 12 months. This contrasts with other data suggesting that children with ASD are more likely than children with other special health care needs to have unmet needs for health care services.31 In national studies, it is reported that 15% of children with ASD have unmet dental needs, and although 91.5% of children with ASD have a usual source of sick- and well-child care, only 18.9% have a medical home.32,33 Because children enrolled in ALHN are receiving treatment at specialized autism centers, they may have better access to health care services than the general population of children with ASD.
Use of Data To Inform QI Priorities
Although high rates of challenging behaviors and lower QoL are expected findings in children with ASD, an understanding of the rates of challenging behaviors and population mean scores on the ABC and PGH-7 provides an important baseline to focus improvement work as described in the KDD (Fig 1). Our data have reinforced that our aims are addressing an important aspect of care and have allowed us to set specific targets for improvement as noted in the aims of the KDD (eg, decreasing the percent of children with moderate, severe, or extremely severe behavior, as reported by parents, from 85% to 70%) and supported our efforts to focus on the identification and treatment of 3 common drivers of behavioral difficulty: ADHD symptoms (>60%), anxiety (47%), and irritable behavior (55%). The ATN–AIR-P has developed practice pathways, tool kits, and resources for addressing these areas, but there is variation in their use across centers. We will use QI methods to support clinical teams to standardize the use of evidence-based approaches for the identification and treatment of these co-occurring conditions, reduce practice variability, and, thereby, improve care and outcomes for individuals with ASD and their families. The ALHN will use statistical process control methods to monitor changes over time in aggregate and by site, from which we will be able to learn from variation and identify best practices. These methods can then be applied to improving other behavioral challenges.
We anticipated many challenges inherent in designing a learning health network for ASD, including lack of standardized outcome measures; heterogeneity in the population with respect to ASD severity, communication skills, and cognitive level; subjectivity and variation in how families rate behaviors; and inconsistency between patients and sites in the frequency of subspecialty follow-up visits for ASD. Given the nature of the outcomes we were targeting, we designed the ALHN data collection system to rely heavily on PROs. This has been an exciting and novel aspect to our network design and also an area of challenge. We learned that many families are eager to join the ALHN but may not activate the mobile app designed for survey completion and network participation, leading to low completion rates (only 47% of consented families completed initial surveys). Completion rates varied widely across sites, ranging from 25% to 89%. Sites with the highest completion rates attribute their success to providing families the option to complete paper surveys during the clinic visits, as an alternative to the app, and through response tracking and reminders. On the basis of this feedback, 2 significant changes are being made in the data collection: enhancing the app to allow families to track their own data, a feature that families report would motivate their participation, and enrolling families during clinic visits. The ALHN is also considering future options for PRO integration directly into the electronic health record. We will continue to obtain feedback to learn from families about how to optimize the collection and clinical integration of PRO data.34
There are some limitations to our data. We did not collect information about completion rates, parent preferences, or completeness of data using different survey collection modalities (paper, app, Web portal). A better understanding of this information would be helpful for improving family participation. These changes are being implemented in a new registry design. Variation across sites in both participation rates and modality of questionnaire completion may have led to response bias. It is possible that families enrolled in the ALHN are not representative of the broader population of children with ASD, in that families attending specialty clinics may have better access to health care and Internet access or familiarity with online surveys. Compared with a nationally representative ASD population, our ALHN sample had a similar sex distribution but less racial diversity.35 However, a goal of the ALHN is to enroll 75% of all children with ASD who receive care at ALHN sites, reflecting the broader population seen at those sites. With more diversity in enrollment, we will be better able to identify and address barriers to care.
Despite the challenges in developing and launching a learning health network for autism, there are many benefits to this approach, which allows for coproduction between families and clinician-scientists to better identify and treat the problems that negatively impact the ability of individuals with ASD to reach their full potential.
Conclusions
After analyzing PRO data from >500 children with ASD, the ALHN has reaffirmed its focus on reducing the severity of challenging behaviors and identified the areas of ADHD, anxiety, and irritability as priority targets. Specific, measurable targets for improvement were determined on the basis of baseline data, and outcomes will continue to be collected longitudinally. The application of a learning health system model is an innovative and important advance in developmental-behavioral pediatrics. This initial ALHN effort provided valuable information about data collection, concerns of parents, and engagement of families and clinicians. Lessons learned will improve data collection methods and integration of these data into clinical care, support the ongoing partnership with families, and will ultimately improve care and outcomes.
Acknowledgments
We thank Logan Herbers and Matthew Fenchel for data management and conducting statistical analyses, Sarah McGovern and RaeAnne Davis for support with preparation of this article, and Maurizio Macaluso for analytic consultation.
Dr Anixt conceptualized and designed the study, designed data collections instruments, conducted some analyses, drafted the initial manuscript, and reviewed and revised the manuscript; Drs Murray, Coury, and Kuhlthau, Ms Eskra, and Dr Lannon conceptualized and designed the study, designed data collection instruments, interpreted results, and reviewed and revised the manuscript; Ms Seide, Ms Kelly, and Ms Hess as family coproducers conceptualized and designed the study, contributed to the design of data collection instruments, and critically reviewed the manuscript for important intellectual content; Drs Lipkin and Law conceptualized the study design and critically reviewed the manuscript for important intellectual content; Ms Fedele coordinated and supervised data collection and critically reviewed the manuscript for important intellectual content; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
FUNDING: Supported by Autism Speaks and the Health Resources and Services Administration of the US Department of Health and Human Services under cooperative agreement UA3 MC11054 (Autism Intervention Research Network on Physical Health). This information or content and conclusions are those of the authors and should not be construed as the official position or policy of nor should any endorsements be inferred by the Health Resources and Services Administration, US Department of Health and Human Services, or US government. This work was conducted through the Autism Speaks Autism Treatment Network serving as the Autism Intervention Research Network on Physical Health.
- ABC
Aberrant Behavior Checklist
- ADHD
attention-deficit/hyperactivity disorder
- AIR-P
Autism Intervention and Research Network on Physical Health
- ALHN
Autism Learning Health Network
- ASD
autism spectrum disorder
- ATN
Autism Treatment Network
- GI
gastrointestinal
- KDD
key driver diagram
- PGH-7
Patient-Reported Outcomes Measurement Information System Global Health Measure
- PRO
patient- or parent-reported outcome
- QI
quality improvement
- QoL
quality of life
Comments