Table 2.

Overview of European screening studies

Setting and users	Screening procedure	Study sample and resultsa	Comments
United Kingdom—South East Thames region  Primary health care practitioner to parents	CHAT (high + medium risk) + CHAT (high + medium risk)	N = 16.235, M age = 18.7 (1.1) PPV = 0.59; NPV = 1.00; Se. = 0.21; Sp. = 1.00	Extremely low false-positive rate High false-negative rate Specifically, combination of joint attention items + pretend play indicates ASD risk Discriminating protodeclarative acts may be difficult for parents (Baron-Cohen et al. [17], Baird et al. [62])
The Netherlands—Province of Utrecht Well-baby clinics + home  Physicians to parents + psychologist to parents	4-item + 14-item ESAT	N = 31.724, M age = 14.91 (1.37) PPV = 0.25; NPV = *; Se. = *; Sp. = *	High false-positive rate but no TD children At young age, hard to discriminate between ASD and TD/DD At young age, failure to detect higher functioning children/milder ASD variants/children who regress or develop autism later Drop-out because parents not yet willing to cooperate Physicians cautious in referring for ASD Screen-negative cases not followed up (Dietz et al. [43])
The Netherlands—Nijmegen Primary care setting + child psychiatry  Primary care worker  Primary care worker + parents’ self-administered test  Primary care worker + parents’ self-administered test  Primary care worker + parents’ self-administered test	Procedure 1: Clinical concern + 14-item ESAT Procedure 2/3: 14-item ESAT + SCQ 11 14-item ESAT + SCQ 15 Procedure 4: 14-item ESAT + CSBS-DP Procedure 5/6: 14-item ESAT + CHAT high risk 14-item ESAT + CHAT high + medium risk	N = *, M age = PPV = 0.68; NPV = 0.37; Se. = 0.88; Sp. = 0.14 PPV = 0.71; NPV = 0.47; Se. = 0.84; Sp. = 0.28 PPV = 0.79; NPV = 0.48; Se. = 66; Sp. = 0.64 PPV = 0.78; NPV = 0.50; Se. = 0.71, Sp. = 0.59 PPV = 0.97; NPV = 0.37; Se. = 0.18; Sp. = 0.99 PPV = 0.88; NPV = 0.45; Se. = 48; Sp. = 0.87	No screening instrument clearly better than any other in differentiating ASD from non-ASD Trade-off between sensitivity and specificity (F.1) High false-positive rate Explore different cut-offs/item-selection within screening instruments. CHAT not administered in original form, constructed from SCQ and CSBS-DP items Screen-negative cases not followed up: where true sensitivity and specificity could not be calculated, they were calculated with the percentage of children about whom there was already some concern (Oosterling et al. [63])
Belgium—Flanders Child day-care setting + home  Child care worker + parents’ self-administered test  Child care worker + parents’ self-administered test  Child care worker + parents’ self-administered test  Child care worker + parents’ self-administered test	Procedure 1: CESDD + 14-item ESAT Procedure 2/3: CESDD + SCQ 11 CESDD + SCQ 15 Procedure 4: CESDD + M-CHAT Procedure 5: CESDD + FYI	N = 7.092, M age = 16.70 (8.19) PPV = 0.55; NPV = 0.95; Se. = 0.40; Sp. = 0.97 PPV = 0.44; NPV = 0.94; Se. = 0.70; Sp. = 0.84 PPV = 0.83; NPV = 0.91; Se. = 0.43; Sp. = 0.98 PPV = 0.29; NPV = 0.98; Se. = 0.71; Sp. = 0.87 PPV = 1.00; NPV = 0.93; Se. = 0.33; Sp. = 1.00	First screening to include report by child care workers High false-positive rate but many developmental disorders/delays among false positives Low parent compliance rate Adaptation of original screening protocol: no telephone interview included in M-CHAT, ESAT completed by parents alone. (Dereu et al. [24])
Spain—Salamanca and Zamora; Madrid Well-baby clinic + home  Parents’ self-administered test + researcher to parents  +paediatrician  Parents’ self-administered test + paediatrician/nurse to parents through web interface	Procedure 1: M-CHAT + M-CHAT telephone interview(by researchers at Univ. when needed) Procedure 2: M-CHAT + M-CHAT web-based interview	Salamanca and Zamora N = 8,122, M age = 20.58 (3.2) PPV = 0.38; NPV = 0.99; Se. = 0.83; Sp. = 0.99 Madrid N = 2,910, M age = 23.14 (4.0) PPV = 0.26; NPV = 0.99; Se. = 0.90; Sp. = 0.99  N = 1,402, M age = 20.21 (3.0) PPV = 0.50; NPV = 0.99; Se. = 0.67; Sp. = 0.99	Translated and adapted; M-CHAT results similar to original M-CHAT study Explore adaptation with screening instrument, such as web-based interview instead of telephone interview Need for coordination of health services and ASD intervention units in Spain Screen-positive children followed up for 2 years Locating and contacting families for telephone interview proved very time-consuming (García-Primo et al. [64])
Sweden—Gothenburg (Home +) child health centre  Nurse  Parents’ self-administered test  Parents’ self-administered test + nurse	Procedure 1: JA-OBS Procedure 2: M-CHAT (including interview) Procedure 3: M-CHAT (including interview) + JA-OBS	N = 3.999, M age = 36.00 (no SD reported) PPV = 0.92.5; NPV = .*; Se. = 0.86; Sp. = * PPV = 0.92; NPV = .*; Se. = 0.76; Sp. = * PPV = 0.89.6; NPV = .*; Se. = 0.95.6; Sp. = *	Interview M-CHAT was necessary; many parents had difficulties understanding questions JA-OBS raised nurse awareness about ASD Combining different instruments for professionals and parents is effective. Screen-negative cases not followed up Screening procedure implemented in developmental programme (Nygren et al. [27])
France—Toulouse Well-baby clinic  Parents’ self-administered test + professional	M-CHAT + CHAT	N = 1,227, M age = 24 Preliminary data: TP = 17; TN = 1,192; FN = 1; FP = 17	Difficulty in obtaining participation of professionals Follow-up at 30 and 36 months in order to check the diagnosis status
Italy  Paediatrician to parents	M-CHAT + M-CHAT interview by paediatrician directly	N = 1,000, M age = 24.4 (3.2) Preliminary data: TP = 4; TN = *; FN = *; FP = 8 PPV 0.28	Difficulties in re-screening children with “pass result” in order to find false-negative cases
Finland  Nurse + Nurse to parents	Procedure 1(first study attempt): At 18 m.o.:CHAT + ICQ and CBCL +BITSEA	N = 200	CBCL (Children’s Behavioural Checklist) No longer ongoing
	Procedure 2(started later): At 12 m.o.: nurse checklist + BITSEA + ICQ + ESAT	N = 677	Small sample, no cases with ASD yet Planning modifications in short future

PPV positive predictive value, NPV negative predictive value, Se. sensitivity, Sp. specificity, ASD autism spectrum disorder, DD developmental disorder/delay, TD typical development. Mage in months

^aNote that the results presented here need to be taken with caution since some of the tools have been used in unusual or adapted conditions and for that reason cannot be considered as the unique psychometric properties of the too

* Number is unknown and could neither be extracted from the literature nor calculated from the data