Study Notes: Concurrent and Discriminant Validity of the Farsi SRS-2 and SCQ

Study Overview and Objectives

• Research Title: Concurrent and Discriminant Validity of the Farsi Translation of the Social Responsiveness Scale-Second Edition (SRS-2) and Social Communication Questionnaire (SCQ).
• Primary Objective: The study aimed to determine the clinical and differential validity of the Farsi versions of the SRS-2 and SCQ within a group of children and adolescents diagnosed with Autism Spectrum Disorder (ASD) compared to a typically developing (TD) group.
• Core Rationale: Early detection of ASD facilitates earlier intervention, which is linked to substantial improvements in quality of life and functional performance. Screening tools are essential for streamlining the diagnostic process.
• Setting: The study was conducted at Roozbeh Hospital (Tehran University of Medical Sciences) in Tehran, Iran.

Background and Epidemiology of Autism Spectrum Disorder (ASD)

• Historical Context: Infantile autism was first identified by Leo Kanner in the 1940s.
• Diagnostic Criteria: ASD is characterized by qualitative deficiencies in three primary areas:     * Social interaction.     * Social communication.     * Restricted or repetitive behaviors and interests.
• Prevalence Trends:     * 1966–1998 Data: A review of 4 million subjects identified 1,533 individuals with autism, showing a median prevalence of $5.2$ cases per $10,000$.     * 2012–2021 Data: A 2022 systematic review of 71 studies showed a wide global range, with a median of $100$ cases per $10,000$ (ranging from $1.09$ to $436.0$ per $10,000$).
• Clinical Implications: The rising prevalence increases demand for assessment and treatment. Diagnosis often faces significant delays relative to the timing of initial parental concerns. Timely identification and early intervention are critical for enhancing developmental outcomes.

Methodology and Participant Selection

• Sampling Approach: Simple and convenient sampling.
• Clinical Group (ASD):     * Total participants: $52$     * Age range: $4\text{–}12$ years.     * Setting: Referrals to the child and adolescent psychiatry clinic at Roozbeh Hospital.     * Diagnosis: Established by certified child and adolescent psychiatrists using DSM-5 criteria.     * Condition: No history of significant other psychiatric disorders.
• Control Group (Typically Developing - TD):     * Total participants: $53$     * Setting: Selected from two elementary schools and one kindergarten in Tehran.     * Matching: Matched with the ASD group for age and gender.     * Condition: No prior history of psychiatric disorders or mental health service engagement.
• Screening for Comorbidity: Both groups were assessed for psychiatric disorders using the Kiddie Schedule for Affective Disorders and Schizophrenia-Present and Lifetime version-DSM-5 (KSADS-PL-5).
• Ethics: Approval granted by the Ethics Committee of Tehran University of Medical Sciences (ID: IR.TUMS.MEDICINE.REC.1398.134). Written consent was obtained from parents.

Detailed Assessment Instruments and Rating Scales

• Social Responsiveness Scale-Second Edition (SRS-2):     * Structure: 65 items, completed by caregivers in $15\text{–}20$ minutes.     * Subscales: Restricted Interests & Behavior; Social Communication & Interaction (which includes social awareness, social cognition, social communication, and social motivation).     * Psychometrics: Previous studies showed Cronbach’s alpha of $0.76$, sensitivity of $0.85$, and specificity of $0.75$ (at a cut-off of 75).
• Social Communication Questionnaire (SCQ):     * Structure: 40 questions (Yes/No), completed by primary caregivers in less than 10 minutes.     * Purpose: Evaluates communication, social interaction, and repetitive behavior.     * Versions: Lifetime and Current.     * International Stats: Recommended cut-off of 12 for general populations and 15 for clinical differentiation; sensitivity of $0.65$ and specificity of $0.73$.
• Childhood Autism Rating Scale, 2nd Edition (CARS-2):     * Structure: 15-item behavioral assessment.     * Purpose: Distinguishes ASD from intellectual disabilities and assesses severity (mild-moderate or moderate-severe).     * Farsi Reliability: Cronbach’s alpha of $0.81$.
• Asperger Syndrome Diagnostic Scale (ASDS):     * Structure: 50 Yes/No items.     * Output: Asperger Syndrome Quotient (AS Quotient).     * Farsi Reliability: Cronbach’s alpha of $0.806$.
• K-SADS-PL-5:     * Structure: Semi-structured interview based on DSM-5 criteria.     * Purpose: Reliability for current and lifetime psychiatric diagnoses in children/adolescents.

Statistical Results: Group Comparisons

• Age and Demographics:     * ASD Group Mean Age: $7.5 \pm 2.7$ years.     * TD Group Mean Age: $7.7 \pm 2.3$ years ($P = 0.656$).
• Score Disparities (Table 1):     * SRS-2 Total Score: ASD group ($122.2 \pm 33.3$) vs. TD group ($0.20 \pm 5.40$), $P < 0.001$.     * SRS-2 Social Communication & Interaction Index: ASD group ($9.26 \pm 5.99$) vs. TD group ($3.16 \pm 2.34$), $P < 0.001$.     * SCQ Total Score: ASD group ($3.25 \pm 0.7$) vs. TD group ($3.5 \pm 0.8$) - Note: Transcript contains typographical formatting error for SCQ means in Table 1, but emphasizes significant difference at $P < 0.001$.     * CARS-2 Mean (ASD group only): $42 \pm 11.3$.     * ASDS Raw Score (ASD group only): $7.6 \pm 9.32$.

Concurrent Validity and Correlations (Table 2)

• CARS-2 and SRS-2: Significant positive correlation ($r = 0.805$, $P < 0.001$).
• CARS-2 and SCQ: Significant positive correlation ($r = 0.697$, $P < 0.001$).
• SRS-2 and SCQ: Very high positive correlation ($r = 0.893$, $P < 0.001$).
• ASDS (AS Quotient) and SCQ: Significant correlation ($r = 0.691$, $P < 0.001$).
• ASDS and SRS-2: Significant correlation ($r = 0.651$, $P < 0.001$).

Diagnostic Accuracy and Cut-off Performances

Receiver Operating Characteristic (ROC) Curve Analysis

• SRS-2 Area Under the Curve (AUC): $0.976$ ($P < 0.001$).
• SCQ Area Under the Curve (AUC): $0.953$ ($P < 0.001$).
• The difference in AUC between the two questionnaires was not statistically significant ($P = 0.252$).

SRS-2 Performance Metrics (Table 3)

• Optimal Cut-off Point: $62.5$ resulted in:     * Sensitivity: $0.942$     * Specificity: $0.811$     * Youden Index: $0.753$
• Alternative Cut-offs:     * Screening focus: $61.5$ (Sensitivity $0.962$, Specificity $0.792$).     * Specific sub-group ($Age < 7$) cut-off: $65.5$ (Sensitivity $0.731$, Specificity $0.857$).     * Specific sub-group ($Age > 7$) cut-off: $61.0$ (Sensitivity $0.889$, Specificity $0.958$).

SCQ Performance Metrics (Table 4)

• Total Sample Optimal Cut-off Range: $15.5$ to $17.5$.     * At $15.5$: Sensitivity $0.865$, Specificity $0.925$, Youden Index $0.790$.     * At $17.5$: Sensitivity $0.827$, Specificity $0.981$, Youden Index $0.808$.
• Age sub-group differences:     * $Age < 7$ (Cut-off $15.5$): Sensitivity $0.889$, Specificity $0.917$.     * $Age > 7$ (Cut-off $15.5$): Sensitivity $0.962$, Specificity $0.821$.

Discussion and Comparative Analysis

• Performance Comparison: Both instruments differentiate ASD from TD effectively. SRS-2 demonstrates higher sensitivity but lower specificity compared to the SCQ.     * SRS-2 Benefit: Generates fewer false negatives; superior for initial suspicion of ASD.     * SCQ Benefit: Higher specificity suggests fewer false positives; more appropriate for true positive identification in clinical settings.
• Comparison with Global Studies:     * Charman et al. (English children): Found SCQ performed better than SRS (AUC: $0.90$ vs $0.77$).     * Fombone et al. (Mexican children): Cut-off of 60 showed sensitivity/specificity of $92.5\%/92.6\%$     * Nguyen et al. (Vietnamese children): Cut-off of 62 showed sensitivity/specificity of $93\%/98\%$     * Bölte et al. (German children): Sensitivities of $0.74–0.80$ and specificities of $0.69–1.00$.     * Aldridge et al. (Australian children): Reported extremely low specificities ($0.08$ and $0.41$) despite high sensitivity.
• Explanation of Variance: Inconsistencies across studies may stem from variations in statistical methods, cross-cultural adaptations, different target population goals, and the age ranges of participants.

Study Conclusions and Limitations

• Conclusions:     * The Farsi versions of SRS-2 and SCQ are validated tools for screening and diagnosing ASD in Iranian children.     * Both scales show high correlation with existing gold standards (CARS-2 and ASDS).
• Limitations:     * Sample size was relatively small ($N=105$).     * Age range was limited to $4–12$ years.     * Parental education levels varied, which may impact understanding of questionnaire items.     * The TD group was recruited from specific schools, which may not represent the general population perfectly.
• Recommendations: Future research should involve larger samples, broader age ranges (adolescents/adults), and exploration of cut-off points based on ASD severity levels.