Intensity of Care

In Italy, public vs private sectors were significantly different with respect to the proportion of units delivering outpatient services (97.1% vs 81.6%, public vs private χ2 (1) = 35.50, p < 0.0001), day patient services (16.4% vs 3.5%, χ2 (1) = 12.50, p = 0.0004), semi-residential (9.9% vs 37.7%, χ2 (1) = 50.17, p < 0.0001), and residential services (5.2% vs 17.5%, χ2 (1) = 18.19, p < 0.0001). Differences among geographical macro-areas were found within public CAMHs with respect to the proportion of units delivering semi-residential (χ2 (3) = 18.79, p = 0.0003) and residential services (χ2 (3) = 19.69, p = 0.0002) (see Table 3 for post-hoc comparisons). No significant differences among geographical macro-areas were found for private units (all ps > 0.05).

Table 3 Percentage of public and private CAMHs across categories of intensity of care, split by macro-areaUse of Digital Technology and Regional Health Information Exchange

More than half (60.1%) of the surveyed CAMHs reported the use of an electronic data system for collecting, storing, and retrieving clinical data, as well as data sharing with the general Regional Health Information System (68.2%). No significant differences were found between public and private units for both these variables (χ2 (1) = 0.97, p = 0.3247; χ2 (1) = 1.93, p = 0.1648). By contrast, the use of digital technology resulted to be unevenly distributed across the Italian territory (use of an electronic data system: χ2 (3) = 69.49, p < 0.0001; data sharing: χ2 (3) = 80.83, p < 0.0001). In particular, CAMHs based in the North reported the largest use of an electronic data system (post-hoc comparisons: North 82.4% vs Center 49.3%, South 50.5% and Islands 40.7%; all ps < 0.01) and data sharing with the Regional Health Information System (post-hoc comparisons: North 88.1% vs Center 67.1%, South 37.9% and Islands 59.3%; all ps < 0.01). The smallest percentage of units reporting the adoption of electronic clinical data sharing was found in the South (post-hoc comparisons: South vs all other macro-areas, all ps < 0.05).

Availability and Qualification of Health Professionals

Table 4 shows the average number of weekly working hours per CAMH per health professional category. As expected, health professionals distribute differently in public and private units. In particular, mean weekly working hours of therapists (Neurodevelopmental Disorders Therapists, Speech Therapists and Educators) were larger in private than in public CAMHs (95.6 vs. 27.3, Welch’s t = 4.55, df = 92, p < 0.0001; 127.4 vs. 62.6, Welch’s t = 3.98, df = 101, p < 0.0001; 102.8 vs. 19.3, Welch’s t = 4.85, df = 92, p < 0.0001 respectively), while mean weekly working hours of Child and Adolescent Psychiatrists were larger in public than in private CAMHs (56.70 vs 33.40, Welch’s t = 4.13, df = 210, p < 0.0001). Within public units, territorial heterogeneity was observed in mean weekly working hours of Child Adolescent Psychiatrists, Psychologists, Neurodevelopmental Disorders Therapists, Speech Therapists, Educators and Social workers (Brown-Forsythe ANOVA, all ps < 0.0003). Specifically, the availability of human resources (Child Adolescent Psychiatrists, Psychologists, Neurodevelopmental Disorders Therapists and Speech Therapists) was higher in CAMHs based in the North and Center in comparison with those in the South and Islands (all post-hoc comparisons ps < 0.05). Moreover, CAMHs based in the Islands showed a lower availability of educators (Islands vs Center, p < 0.01) and of social workers (Islands vs Center, p < 0.01). When private unit were examined, the difference among macro-areas was significant only for neurodevelopmental disorders therapists, speech therapists, and educators (Brown-Forsythe ANOVA, p = 0.004, 0.0068, and 0.0243, respectively), with lower availability of resources in Islands especially compared to the South (p < 0.05; see Table 4 for more details).

Table 4 Average number of weekly working hours of each health professional category per CAMH unit, split by macro-areaProportion of ASD Patients and Working Time Allocated to ASD Patients

To estimate a proxy of the average medical burden allocated to ASD by the surveyed CAMHs, we asked participants to indicate the percentage of ASD patients on the total number of patients served in their units, and the percentage of weekly working hours allocated to ASD patients (independently of the service provided, i.e., diagnosis or treatment). The reported percentages were classified in three classes: low proportion, < 30%; medium proportion, 31–70%; high proportion, > 70%. The difference between public and private CAMHs was significant for both the variables (ASD patients: χ2 (2) = 22.57, p < 0.0001; weekly working hours dedicated to ASD patients: χ2 (2) = 22.96, p < 0.0001), with public units reporting low burden more frequently than private units (Table 5). Within public units, no significant differences among macro-areas were observed as for percentage of ASD patients (Fisher’s exact probability test p = 0.058). A significant difference across macro-areas was observed with respect to the percentage of weekly working hours dedicated to ASD patients (Fisher’s exact probability test p = 0.002), in particular between CAHMs located in the Center and in the South of Italy (p < 0.05 at post-hoc comparisons). In particular, about 11% of CAMHs based in the South reported a high proportion of weekly working hours vs 0% of CAMHs in the Center. Within private units, no significant differences were found for both variables (Fisher’s exact probability test p = 0.209 and 0.514, for percentage of ASD patients and percentage of time dedicated to ASD patients, respectively).

Table 5 Percentage of public and private CAMHs reporting low, medium, and high proportion of ASD patients and professionals’ weekly working hours dedicated to ASD patients, split by macro-areaAutism Training for Health Professionals

The majority of the CAMHs surveyed reported to provide autism training to personnel working in the unit (76.6%) and/or to offer financial support for the specialist training of their staff (56.7%). We did not observe significant differences between public and private CAMHs with respect to both variables (staff training: χ2 (1) = 1.73, p = 0.1888; financial support for training: χ2 (1) = 1.16, p = 0.2823). The survey revealed territorial heterogeneity with respect to direct provision of training on autism (χ2 (3) = 16.5, p = 0.0009). In particular, a significantly higher percentage of CAMHs based in the North (84.9%), in comparison with those in the Center (72.1%) and in the South (65.3%), reported to offer autism training to their staff (p < 0.05 and p < 0.01, respectively). Instead, no significant difference was observed in comparison to Islands (75.9%). Financial support for specialist training also resulted to be unevenly distributed across the territory (χ2 (3) = 81.9, p < 0.0001), with the highest percentage of CAMHs offering financial support in the North in comparison to the other macro-areas (North 81.7% vs. Center 49.3%, South 47.4%, and Islands 31.5%, all ps < 0.01).

Diagnosis and Intervention Services for Children with ASD

As for provision of diagnosis and intervention, a significantly higher percentage of public CAMHs reported to provide diagnosis than private units (χ2 (1) = 129.4, p < 0.0001). By contrast, intervention was more frequently reported by private than public units (χ2 (1) = 15.4, p < 0.0001). A significant difference among macro-areas was observed both in public and private CAMHs, either for diagnosis or interventions services (Fisher’s exact probability test: public p = 0.0003 for both diagnosis and intervention; private p = 0.0345 for diagnosis and p = 0.0002 for intervention). As for public CAMHs, a significant higher number of units located in the North reported to provide diagnosis compared to units based in Center (p < 0.01) and interventions compared to units in the South (p < 0.01). As for private units, a higher percentage of units based in the Center reported to provide intervention than units based in the Islands (p < 0.05) (see Table 6).

Table 6 Percentage of public and private CAMHs providing diagnosis and intervention services for children with ASD, split by macro-areaASD Diagnostic Assessment and Patient Characterization

The provision of ASD diagnostic assessment and characterization was analyzed exclusively in the public units, as the overall number of private CAMHs reporting to provide diagnosis was very low (n = 51).

As for diagnostic system, the majority (81.2%) of respondent public units reported using the diagnostic system ICD-10, while the use of the DSM-IV was reported by 23.2% of the surveyed CAMHs. Heterogeneity among macro-areas was found for the use of ICD-10 (χ2 (3) = 48.53, p < 0.0001) and DSM-IV (χ2 (3) = 53.64, p < 0.0001). In particular, CAMHs based in the North reported the use of ICD-10 more frequently than units based in the other macro-areas (North 96.2% vs Center 73.3%, South 82.3%, and Islands 62.1%, all ps < 0.05), while the use of DSM-IV was more frequently reported in the South (38.7%) and Islands (46.0%) than in the North (9.4%) and Center (15.8%) (all ps < 0.05). As for the presence of a formalized protocol for the diagnosis of ASD in the unit, 62.2% of the respondents reported the adoption of agreed standard operating procedures in their practice. Data highlight a territorial heterogeneity (χ2 (3) = 35.5, p < 0.0001), specifically a higher percentage of CAMHs adopting a formalized diagnostic protocol in the North than in the other macro-areas (North 78.0% vs Center 60.6%, South 45.6%, and Islands 47.4%; all ps < 0.01).

The majority of surveyed public units (96.3%) reported the use of standardized diagnostic tools. The use of the Autism Diagnostic Observation Schedule-Generic (ADOS-G; Lord et al. 2001) was reported by 62.5% of the public units, while the use of the Autism Diagnostic Interview-Revised (ADI-R; Lord et al. 1994) was reported by 48.5% of the units (Fig. 2). The other most commonly used measures were the Childhood Autism Rating Scale (CARS; Schopler et al. 1988) used by 67.0% of the units, the Gilliam Autism Rating Scale (GARS; Gilliam 1995) used by 22.3% of the units, and the Autism Behavior Checklist (ABC; Krug et al. 1993), the use of which was reported by 12.6% of the surveyed units. As regards patient characterization, the survey considered four main areas of assessment: cognitive, language, adaptive behavior/global measures of functioning, and comorbid behavioral problems. Concerning the assessment of cognitive functions, the standardized tools more frequently reported in the surveyed CAMHs were: the Wechsler Preschool and Primary Scale of Intelligence (WPPSI; Wechsler 2012), reported by 73.7% of the units; the Leiter International Performance Scale-Revised (LEITER-R; Roid and Miller 1997), used by 67.8% of the units; the Raven’s Progressive Matrices (RPM, Raven et al. 2000), used by 63.8% of the CAMHs. With regard to language assessment, less than half of the respondents (37.5%) reported using the MacArthur–Bates Communicative Development Inventories (MCDI; Fenson et al. 2007), while the use of Grammar Comprehension Test for Children (TCGB; Chilosi and Cipriani 1995) was reported by 21.7% of the surveyed CAMHs.

Fig. 2

Percentage of CAMHs reporting to use standardized tools for: a ASD assessment, b cognitive assessment; c language assessment; d assessment of behavioral problems; e adaptive behavior/global assessment of functioning. Post-hoc comparisons among macro-areas: *p < 0.05, **p < 0.01

About half (53.4%) of the surveyed CAMHs reported to assess adaptive behavior using the Vineland Adaptive Behavioural Scales (VABS; Sparrow et al. 2005), while 73.5% of units reported the use of the Psychoeducational Profile-Revised (PEP-R; Schopler et al. 1990) as a global measure of functioning and psychoeducative profiling. The Child Behaviour Checklist (CBCL; Achenbach and Edelbrock 1983) and the Schedule for Affective Disorders and Schizophrenia for School-Age Children (K-SADS; Kaufman et al. 1996) were the two most commonly used tools for assessing comorbid behavioral problems (38.6% and 24.4%, respectively). Statistical differences among macro-areas in the use of standardized tools was reported for: ADOS-G (χ2 (3) = 16.5, p = 0.0009), CARS (χ2 (3) = 31.0, p < 0.0001), GARS (χ2 (3) = 24.5, p < 0.0001), ABC (χ2 (3) = 11.0, p = 0.0117), WPPSI (χ2 (3) = 10.4, p = 0.0155), LEITER-R (χ2 (3) = 15.9, p = 0.0012), MCDI (χ2 (3) = 10.9, p = 0.0123), TCGB (χ2 (3) = 36.4, p < 0.0001), CBCL (χ2 (3) = 17.3, p = 0.0006), KSADS (χ2 (3) = 23.2, p < 0.0001), PEP-R (χ2 (3) = 77.3, p < 0.0001). Post-hoc comparisons are shown in Fig. 2. No differences among macro-areas were found in the use of ADI-R (χ2 (3) = 5.5, p = 0.1404), RPM (χ2 (3) = 5.6, p = 0.1345), and VABS (χ2 (3) = 0.6, p = 0.8895).

Intervention Services for ASD Patients

Most of the CAMHs providing outpatient treatment services reported to serve patients in the overall 0–17 years age range (88.3%). Surveyed CAMHs reported to offer various intervention programs to ASD patients, selecting one or more items among the following: Augmentative and Alternative Communication (AAC, used by the 74.8% of the units), the Training and Education of Autistic and Related Communication Handicapped Children (TEACCH, 45.2%), the Applied Behavioral Analysis (ABA, 29.2%), and the Early Start Denver Model (ESDM, 24.9%), the Thérapie d’Echange et Developpment (TED, 13.4%), the Developmental, Individual Difference, Relationship-based (DIR, 8.7%) as well as other (no autism-specific) interventions, such as Speech Therapy (ST, 84.2%) and Psychomotor Therapy (PMT, 82.4%), Expressive-motor rehabilitation (EMR, 54.1%), Family support (FS, 64.3%), and Cognitive therapy (CT, 60.9%). To evaluate if CAMHs adopted more than one intervention strategy and in which combination, we first grouped interventions in the following categories: (i) specific for ASD, including TEACCH, ABA, ESDM, TED, DIR; (ii) specific for communication, AAC; (iii) not specific for ASD, including ST and PMT; (iv) other interventions, including CT, FS, and EMR. For each category, the CAMHs were classified as YES when they offered at least one of the interventions included in the category. Finally, we estimated the frequency of combinations in the CAMHs providing outpatient intervention to patients in the whole age range 0–17 years (88%). Table 7 presents the percentage of CAMHs offering the different combinations of interventions, according to the macro-area. Heterogeneity in the pattern of treatments offered was found across macro-areas (χ2 (12) = 50.1, p < 0.0001), and in particular between North and other macro-areas (all ps < 0.05).

Table 7 Percentage of CAMHs offering autism-specific and other intervention programs, split by macro-areaCollaborative Network for Children with ASD

Collaborative treatment plans between clinicians, family pediatricians, teachers, and adult health services are reported for the public CAMHs, which are appointed by INHS to take charge of children with ASD in collaboration with other public health, social, and educational agencies.

Most of the units (72.1%) reported to collaborate with pediatricians for ASD care, but only 18.4% through a formalized agreement. The percentage of CAMHs reporting the implementation of a formalized agreement was higher in the North and Center than in the South and Islands (North 29.1% and Center 23.9% vs. South 6.3% and Islands 5.6%; post-hoc comparisons p < 0.05 for all). A specific collaboration with school for children with ASD (i.e., school-based clinical interventions for children with autism and training/tutoring on ASD management for teachers) was reported by 59.3% of the CAMHs, 12.5% through a formalized agreement. Macro-areas did not differ with respect to the percentage of public CAMHs collaborating with schools through a formalized protocol. A collaboration with adult local public services (i.e., adult mental health services and services for adults with disabilities) aimed at guaranteeing continuity of care for children and adolescent with ASD (i.e., specific interventions and tutoring to support/improve transition process) was reported by 56.1% of the surveyed CAMHs. Of them, only 27.9% referred to a formalized agreed protocol. The percentage of CAMHs reporting the implementation of a formalized agreement was higher in the North and Center than in the South and Islands (North 37.7% and Center 38.6% vs. South 2.2% and Islands 4.9%; post-hoc comparisons p < 0.01 for all).

Regional ASD Policies

Results show that 11 Italian regions out of 20 (55%) have issued and released formal recommendations for ASD’s management, and have adopted the IAAP. The majority of regions (70%) have reported at least one ASD funded action plan relative to the period 2010–2015. In half of the regions, a formal agreement defining the role of pediatricians and ASD early referral protocol have been approved, although in only four of them the agreement has been implemented across the whole region. As for CAMHs’ knowledge of—and involvement in—regional plans, about 50% of the surveyed CAMHs reported to be aware of regional ASD policies; only 70% of these reported to have been involved in regional policies.