Psychosocial and Behavioral Impact of COVID-19 in Autism Spectrum Disorder: An Online Parent Survey

The 2019 coronavirus disease (COVID-19) outbreak could result in higher levels of psychological distress, especially among people suffering from pre-existing mental health conditions. Young individuals with autism spectrum disorders (ASD) are particularly at risk due to their vulnerability to unpredictable and complex changes. This study aimed to investigate the impact of the COVID-19 pandemic on ASD individuals, whether any pre-pandemic sociodemographic or clinical characteristics would predict a negative outcome, and to narratively characterize their needs. Parents and guardians of ASD individuals filled out an online survey consisting of 40 questions investigating socio-demographic and clinical characteristics of their children, impact of the COVID-19 outbreak on their wellbeing and needs to deal with the emergency. Data were available on 527 survey participants. The COVID-19 emergency resulted in a challenging period for 93.9% of families, increased difficulties in managing daily activities, especially free time (78.1%) and structured activities (75.7%), and, respectively, 35.5% and 41.5% of children presenting with more intense and more frequent behavior problems. Behavior problems predating the COVID-19 outbreak predicted a higher risk of more intense (odds ratio (OR) = 2.16, 95% confidence interval (CI) 1.42–3.29) and more frequent (OR = 1.67, 95% CI 1.13–2.48) disruptive behavior. Even though ASD children were receiving different types of support, also requiring specialist (19.1%) or emergency (1.5%) interventions in a relatively low proportion of cases, a number of needs emerged, including receiving more healthcare support (47.4%), especially in-home support (29.9%), as well as interventions to tackle a potentially disruptive quarantine (16.8%). The COVID-19 outbreak has undoubtedly resulted in increased difficulties among ASD individuals.


Introduction
After the Severe Acute Respiratory Syndrome (SARS) Coronavirus outbreak of 2002-2003, the International Health Regulations (IHR) of the World Health Organization (WHO), which had been first adopted in 1969, were revised in 2005 to extend their scope to any public health risk that might affect human health, irrespective of the source. Emphasis was put on the risk that the The survey was part of a larger study which was approved by the research ethics committee at the Integrated University Hospital of Verona (CESC 2242 and CESC 2243).

Participants
Parents and guardians of individuals with an ASD diagnosis were asked by healthcare professionals affiliated with the Veneto Autism Spectrum Disorder Regional Centre at the Integrated University Hospital of Verona to fill out the online survey. Autism advocacy and family support networks were additionally used to distribute and directly encourage survey participation. Children' ASD diagnosis was self-reported.

Instrument
The parent survey was developed by a focus group of physicians, psychologists, and child life specialists, also taking the advice from parents of children with ASD. The survey consisted of 40 questions (18 multiple choice questions, 20 yes/no questions, and 2 open-response questions) in 3 categories: (i) ASD individuals' socio-demographic and clinical characteristics, (ii) impact of the COVID-19 outbreak on their wellbeing, and (iii) needs to deal with the emergency. Participants were allowed to select only 1 item for each multiple-choice question.

Analyses
The final raw data were downloaded from Google Forms into a Microsoft Excel file for analysis using SPSS software (Version 26.0; IBM Corp, Armonk, NY, USA). Descriptive statistics were used to provide baseline information concerning survey participants' ASD children. Then, multiple logistic regressions were performed to investigate whether any ASD individuals' socio-demographic or clinical characteristics would predict a greater frequency and intensity of behavior problems following the COVID-19 outbreak.
The open-response questions did not have a scoring system. The needs identified by the survey responses were gathered with the intent of informing healthcare professionals in their assessment and management of individuals with an ASD diagnosis during the ongoing COVID-19 emergency. Two authors independently evaluated such answers and pooled them into categories (e.g., healthcare, social, financial needs, etc.). In the rare instances of discrepant category attribution, consensus was reached through discussion with a third senior clinical researcher.
The survey was not intended to formally assess severity of ASD among participants' children.

Results
A total of 529 respondents participated in the survey. As 2 participants did not answer any questions at all, a total of 527 participants were included in the study. Across the 38 closed questions, survey participants provided 18,738 answers, while 373 answers were missing (<2%). Further, 34 answers and 1 missing item were deemed inconsistent (<0.2%) and excluded from the final analyses.

Socio-Demographic and Clinical Characteristics
The mean age of participants' children was 13 years (SD = 8.1). Almost all of them were from the Veneto region (99.4%) and the large majority were living in married or cohabiting couple families (88.2%). Most children had at least 1 sibling (71.6%), and one out every 10 siblings had a neurodevelopmental disorder (NDD) diagnosis (10.2%). Most children were receiving private therapy (66.2%) and most parents were members of autism advocacy and family support networks (67.1%) ( Table 1). Only one every three children had a fluent language (33.1%). About half of the entire sample was presenting with behavior problems from before the outbreak of COVID-19 (51.5%); among those, 42.2% were receiving pharmacological treatment. At least 1 comorbid medical condition was reported in 27.8% of ASD individuals, with neuromotor and gastrointestinal conditions and allergies and food sensitivity being the conditions most frequently reported (Table 1 and Supplementary Table S1).

Psychosocial and Behavioral Impact of the Emergency Outbreak
COVID-19 positivity was reported among 1.3% of nuclear family members and 4.4% of extended family members, with bereavement occurring in 2.3% of cases. Approximately one out of four parents stopped working due the emergency outbreak (26.1% of mothers and 27.5% of fathers). The large majority of them evaluated the current period of change and restrictions as challenging or very challenging (93.9%) and more challenging than before the emergency outbreak (77%) ( Table 2).
Following the emergency outbreak, a proportion of parents reported support from the Local Healthcare Services (27.7%), with the large majority of them reporting both direct (70.1%; e.g., calls, videocalls) and indirect (84%; e.g., text messages, homework assignments) school support as well as support from the private therapist (73.3%). For each type of support, an only slightly lower proportion of parents considered it from sufficiently useful to very useful during the ongoing emergency (Table 2).
A proportion of parents reported difficulties in managing their child's meals (23%), autonomies (31%), free time (78.1%), and structured activities (75.7%). For each activity, an almost overlapping proportion of parents reported such activity as more difficult than before the emergency outbreak.
Overall, compared to before the emergency outbreak, behavior problems were reported being more intense (35.5%) and more frequent (41.5%) in a substantial proportion of ASD individuals. Due to the onset of behavior problems, an emergency contact with the child's Neuropsychiatrist was required in 19.1% of cases, while an access to the Accident and Emergency (A&E) happened in 1.5% of cases (Table 2).

Predictors of Emergency Outbreak Negative Impact on Wellbeing
A multiple logistic regression tested for an effect of (i) behavior problems predating the emergency (yes/no), (ii) age, (iii) language (fluent/non-fluent), (iv) being an only child (yes/no) as a proxy of greater social isolation in quarantine, (v) medical comorbidity (yes/no), (vi) parenting couple situation (married or cohabiting/separated or single parent), (vii) support by Local Healthcare Services since COVID-19 (yes/no), (viii) direct school support since COVID-19 (yes/no), (ix) indirect school support since COVID-19 (yes/no), (x) private therapist support since COVID-19 (yes/no or no private therapist from before COVID-19) on the intensity of the behavior problems following the emergency outbreak (more intense/equally or less intense). The logistic regression model was statistically significant, χ 2 (10, N = 440) = 32.338, p < 0.001. ASD individuals with preexisting behavior problems were 2.16 times more likely to exhibit more intense behavior problems that those without preexisting behavior problems. Increasing age and living with a separated or single parent were associated with a reduction in the likelihood of exhibiting more intense behavior problems, while not receiving indirect school support during the emergency tended to be associated with an increased likelihood of exhibiting more intense behavior problems (Table 3).
A further multiple logistic regression tested for the effect of (i) behavior problems predating the emergency (yes/no), (ii) age, (iii) language (fluent/non-fluent), (iv) being an only child (yes/no), (v) medical comorbidity (yes/no), (vi) parenting couple situation (married or cohabiting/separated or single parent), (vii) support by Local Healthcare Services since COVID-19 (yes/no), (viii) direct school support since COVID-19 (yes/no), (ix) indirect school support since COVID-19 (yes/no), (x) private therapist support since COVID-19 (yes/no or no private therapist from before COVID-19) on the frequency of the behavior problems following the emergency outbreak (more frequent/equally or less frequent). The logistic regression model was statistically significant, χ 2 (10, N = 444) = 18.502, p = 0.047. ASD individuals with preexisting behavior problems were 1.67 times more likely to exhibit more frequent behavior problems that those without preexisting behavior problems (Table 4).

Needs to Deal with the Emergency: A Narrative Perspective
Out of 527 survey participants, 406 parents (77%) reported at least one need to the open-response question about what could be of help do deal with the ongoing emergency. Ten of them reported more than one need (4 participants reported 2 needs, 6 participants reported 3 needs), for a total of 422 responses. The most commonly reported need was for in-home healthcare support (29.9%), followed by center-based healthcare support (10.4%), loosening quarantine restrictions (9.7%), ending lockdown (7.1%), and in-hospital healthcare support (7.1%; Table 5).

Discussion
To our knowledge, this is the first study which systematically explored the impact of the COVID-19 outbreak in a population of individuals suffering from an autism spectrum disorder (ASD). Results from this parent survey indicate that the large majority of parents of ASD individuals consider the period of change and restrictions that has followed the onset of the emergency as challenging and requiring more commitment than before. Most support was delivered by school services, followed by private therapists and local healthcare services. Consistent with previous reports of executive functioning deficits making ASD individuals more vulnerable to routine disruption [18], an elevated number of parents reported difficulties in managing their children's daily activities, especially in terms of free time and structured activities. Despite requiring specialist intervention in a relatively small proportion of cases, and almost never ending in hospital emergency assessments, behavior problems worsened in more than one third of ASD individuals.
In line with previous evidence that pre-existing psychological difficulties seem to be the only clear predictor of poorer mental health outcomes following respiratory syndromes [14] and quarantine [13], the most relevant finding of this survey is that ASD individuals with behavior problems predating the COVID-19 outbreak are twice as likely to experience more intense and more frequent behavior problems since the beginning of the emergency. As independent evidence supports decreasing symptom levels from childhood to young adulthood in ASD, especially in verbally fluent individuals [19], we also tested whether the age and language level of survey participants' children with ASD would predict a worsening of the behavior problems. Results suggest that older age may play a protective role with regards to the emergency-induced intensification of behavior problems, while the effect of language failed to reach statistical significance. Furthermore, despite social isolation and mental health problems may co-occur in childhood [20], as well as medical conditions and ASD [21], being an only child, as a proxy of greater social isolation, and comorbid medical conditions did not predict a poorer outcome in terms of more intense or more frequent behavior problems following the COVID-19 outbreak and the implementation of restrictive measures and quarantine. Interestingly, living with a separated or single parent was associated with a better outcome in terms of intensity of behavior problems. Despite being counterintuitive, such a finding may reflect a more simplified parent-child interaction which could be effective in preventing the deterioration of the child's wellbeing during quarantine and restrictions. Future studies would need to clarify this issue. Furthermore, it is important to highlight that such evidence does not necessarily apply to other social contexts. Finally, ASD individuals not receiving school support since the COVID-19 outbreak tended to express more intense behavior problems, suggesting the importance of maintaining contact with the school during the emergency.
The last scope of this study was to narratively collect the parents' perceived needs through the emergency by offering an open-response question at the end of the survey. About half of the participants reported needing support from healthcare services, especially in-home services. Interestingly, almost one every five parents reported that loosening restrictions or ending the lockdown would be of help.
Outbreaks of emerging infections such as COVID-19 can elicit strong fear reactions and preoccupations with downstream effects on physical and mental health, especially in vulnerable individuals [22]. Moreover, such negative impacts on health could be worsened by the experience of being quarantined [13]. Autism is no exception. Even though it is a complex genetic disorder, the effect of the environment in shaping the behavioral phenotype should not be underestimated. In fact, a high emotional climate, such as that resulting from the COVID-19 emergency outbreak, has been associated with increased levels of maladaptive behavior in ASD over time [23]. Furthermore, families of individuals with ASD have been reported to experience greater stress than families whose children suffer from other disabilities [24,25], making a compelling case for the implementation of youth-oriented mental health prevention and early intervention strategies [26,27]. Evidence supports the effectiveness of such interventions in mitigating disabilities and even improving skills among young individuals with ASD [28].
Surveys performed during previous Public Health Emergencies of International Concern (PHEIC) have provided timely data to inform best practices in responding to the emergency [29]. Similarly, the online survey presented here has proved to be a powerful data collection tool, benefiting from the strengths that have been associated to this type of instrument such as having a large sample size, fast response times, timely data processing, and low costs [30]. Moreover, thanks to the solid infrastructure of autism advocacy and family support networks, we were able to mitigate the risk of poor response rates and improve sample representativeness, collecting data from over five hundred ASD individuals and their families in a catchment area (the Veneto Region) of about 5 million people. However, the findings of this study have to be seen in light of some limitations. In particular, in spite of the aforementioned advantages, the survey suffered from the lack of a standardized assessment of clinical features such as language and behavior problems. Such aspects may limit the comparability of outcomes among different studies as well as internal comparisons in any follow-up assessments. Furthermore, despite collecting information on children's language and behavior problems predating the emergency as a proxy of their baseline cognitive and adaptive functioning, due to its nature, the online survey did not allow investigating such aspects through a standard method such as psychometric tests. Furthermore, information on the gender of ASD children was not available through the online survey. Even though the main aim of the study was not to investigate gender differences in response to the COVID-19 outbreak in ASD, we acknowledge the limitation.
In conclusion, the present survey indicates that the ongoing COVID-19 emergency has resulted in a challenging period for most ASD individuals and their families, with increased difficulties in managing daily activities and at least one in every three children presenting with more frequent or more intense behavior problems. Children with behavior problems predating the COVID-19 outbreak were found to be particularly at risk to present with more intense and more frequent disruptive behavior. Even though ASD children were receiving different types of support, also requiring specialist or emergency interventions in a relatively low proportion of cases, a number of needs emerged, including receiving more healthcare support, especially from in-home services, as well as interventions to tackle a potentially disruptive quarantine.