The Development and Validation of DSM 5-Based AIIMS-Modified INDT ADHD Tool for Diagnosis of ADHD: A Diagnostic Test Evaluation Study
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.280638
Source of Support: None, Conflict of Interest: None
Keywords: Attention deficit hyperactivity disorder, All India Institute of Medical Sciences, Conners, Diagnostic and Statistical Manual-5, INDTKey Messages: The AIIMS modified INDT-ADHD tool is a standardised, well structured and culturally acceptable diagnostic tool based on DSM-5 for the diagnosis of ADHD with an added advantage of severity scoring of symptoms of ADHD.
Current Affiliation for Dr. Lokesh Saini: Division of Pediatric Neurology, Department of Pediatrics, PGIMER, Chandigarh, Chandigarh - 160 012, India.
Attention deficit hyperactivity disorder (ADHD) is one of the commonest neurodevelopmental disorders of childhood. The global prevalence in school-aged children is 3–5 percent. In India, however, it is reported to be as 1.7 per cent. Globally, the International Classification of Disease-10 (ICD-10) and the Diagnostic and Statistical manual of Mental Health Disorders criteria are used to diagnose ADHD., Currently, the latest version of Diagnostic and Statistical Manual (DSM) being used is DSM-5. Its construct is based on core symptom clusters of inattention (I) and hyperactivity/impulsiveness (H/I). The DSM-5 version consists of two symptom domains, hyperactivity/impulsivity and inattention associated with impaired social, academic, adaptive and occupational functioning.,
In India, diagnosis of ADHD is challenging due to multiple factors. Caregivers as well as health care professionals lack awareness and keep perceiving deviant behaviour to be normal and eventually outgrown in adolescence. Discrepant outcomes are also seen across different racial or cultural groups due to differences in clinical judgement and parental perception. This implies that the criteria are not standardised according to relevant cultural and socioeconomic contexts for universal applicability. Some tools additionally require teacher's perspectives of the children's behaviors, which may not be as reliable in scenarios with high student–teacher ratios.,
The present study has developed and validated a reasonably well-structured, physician-administered, culturally acceptable tool for diagnosis of ADHD based on DSM-5 criteria. Recently, there are studies on regional adaptation and validation of ADHD diagnostic criteria based on DSM-5 in different regions of world to suit their cultural context., Most of the available diagnostic tools for ADHD are based on older versions of DSM (i.e., DSM-IV). There are some key differences in DSM-IV and DSM-5 [Table 1], and DSM-5 holds some merit points over DSM-IV [Box 1], so there was a need to come up with this new diagnostic tool. Furthermore, there is a lack of standardised Indian tools for diagnosis of ADHD based on DSM-5.
In this direction, the INDT-ADHD (International Clinical Epidemiology Network [INCLEN] Diagnostic Tool for Attention Deficit Hyperactivity Disorder) tool was developed and validated in the age group of 6–9 years based on the DSM-IV criteria as the gold standard. The current tool is a modification of the same and the first DSM-5-based tool validated in an Indian population.
The All India Institute of Medical Sciences (AIIMS)-modified INDT-ADHD tool is easy to administer and requires minimal training. The tool has retained core 18 symptoms of inattention and hyperactivity–impulsivity symptom domains as per DSM-5, with modifications according to Indian cultural and socioeconomic contexts.
This study was conducted at a tertiary care teaching hospital in northern India between January and December 2015. Children aged 6–18 years with 'suspected ADHD' were enrolled in the study. Suspected ADHD was screened according to Practice parameter recommendations by the American academy of Pediatrics. Children with IQ <70 (calculated by Malin's Intelligence Scale for Indian Children (MISIC) were excluded from the present study.
The primary objectives of the study was to develop and validate AIIMS-modified INDT-ADHD tool for diagnosis of ADHD in children aged 6–18 years. The secondary objective was to develop a clinical score for describing symptom severity in ADHD.
Development of AIIMS-modified INDT ADHD tool for ADHD
A team of pediatric neurologists, clinical psychologists and psychiatrists reviewed the diagnostic criteria for diagnosis of ADHD by DSM-5. In addition, hyperactivity, impulsivity and inattentiveness symptoms were also reviewed from Conners rating scale, and the child behaviour check list (CBCL)., A pool of items was selected by the investigators using the modified Delphi technique. Two focused group discussions were held for selecting the items for questionnaire. Appropriateness criteria comprising of 18 symptoms based on parental interview and direct observation, were finalized based on clarity, importance and frequency of endorsement. Finally, questions from the DSM-IV-based INDT-ADHD tool were modified as per DSM-5 criteria and cultural context.
The modified tool has two sections: Sections A and B. Section A has A1 that represents nine questions addressing inattention symptoms and A2 consisting of nine questions addressing hyperactivity and impulsivity symptoms. The response is marked as 'yes' and 'no'. Questions were designed in such a way that each favourable response for presence of symptoms of ADHD corresponds to a 'yes' response. Section B has analysis of items in Section A as well as other questions that are history based. As per section B, at least six out of nine items should have a response of 'yes' in A1 and A2 or both for a diagnosis of ADHD. There are three additional prerequisites for making a diagnosis of ADHD: symptom onset before 12 years of age, presence of symptoms for at least six months duration and at least in two settings. It takes 20–25 minutes to administer the tool. Investigator assessed each child by face-to-face interview of parents or primary caregivers.
Pilot testing was conducted among 10 subjects who were already diagnosed with ADHD based on DSM-IV criteria and were on regular follow up. The ability to diagnose ADHD and the degree of agreement with DSM-5 criteria were checked. The internal consistency consistency of items as a whole construct was found good with Cronbach's alpha 0.91 and individual subscale Cronbach's alpha ranged 0.84–0.88 across each of the symptom domain, that is, inattentiveness and hyperactivity-impulsivity.
Consecutive eligible children (aged 6–18 years) with 'suspected ADHD' were enrolled and subjected to IQ assessment followed by DSM-IV-based INDT-ADHD tool (Investigator 1). Subsequently, patients were subjected to Conners parent rating scale (Conners 3–P[S]) and AIIMS-modified INDT ADHD in two groups in reverse sequences (Investigators 2 and 3). These study participants were finally evaluated independently by the gold standard team on the same day. The gold standard assessment was performed by a team including a pediatric neurologist, clinical psychologist and child psychiatrist with more than 10 years of experience in the field. The assessments of each team member were sealed in separate opaque envelopes, and Investigators 1, 2 and 3 were blinded to each other. The patient's treatment and management plan were decided by the experts. The flow of the study has been depicted in [Figure 1].
The sample size of 50 children in each group (ADHD present and ADHD absent as per gold standard) was computed for diagnostic accuracy of modified INDT-ADHD assuming sensitivity of 85 per cent of INDT-ADHD for diagnosing ADHD in children with suspected ADHD. As it was a time bound study, for logistic reasons, a feasibility sample size of minimum 30 children was calculated in each group.
Data was entered in Microsoft Excel and statistical analysis was done using SPSS 15.0 version. Sensitivity, specificity, positive predictive value and negative predictive value were calculated for modified INDT-ADHD against gold standard diagnosis of ADHD. The AIIMS-modified INDT-ADHD tool was used in conjunction with Conners severity score for drawing Receiver operative characteristics (ROC) curves to find the severity cut-off scores for the newly developed tool.
Overall, 97 children (aged 6–18 years) with suspected ADHD were enrolled in the study. None refused consent to participate. Thirty one children were excluded as their IQ was found to be <70. The remaining 66 children were evaluated and finally analysed. Among 66 children, 36 (55 percent) were diagnosed as ADHD and 30 (45 percent) were diagnosed with no ADHD as per the gold standard diagnosis. The clinico-demographic variables have been depicted in [Table 2]. The developed tool was administered to 66 participants suspected for ADHD based on inclusion criteria. It was observed that sensitivity and specificity of the tool were 100 per cent and 90 per cent, respectively. Diagnostic accuracy was estimated at 95.45 per cent for diagnosing ADHD in suspected ADHD population [Table 3].
Development of clinical monitoring score
AIIMS-modified INDT ADHD tool has 18 items, 9 items each in two domains (inattention and hyperactivity-impulsivity). The number of items marked as favourable response for attention deficit and hyperactivity-impulsivity symptoms were counted, and a total score was generated. Each 'Yes' response was awarded a score of 1 and 'No' as 0. Similarly, all 18 questions were marked and a raw score was generated at the end that varied from 0 to 18.
As all 18 tool questions were based on DSM-5 criteria, each individual questions were analysed against their corresponding DSM-5 question and specificity and likelihood ratio of positive test (LR+) of each individual tool questions were calculated. Sum of specificity and LR+ for all 'Yes' responses for each individual child was determined. Since AIIMS-modified INDT ADHD tool is a diagnostic tool, only specificity and LR+ were taken into account.
An ROC curve was used to determine the cut-off for total score of the tool (raw score [Figure 2]a, specifity score [Figure 2]b, LR+ score [Figure 2]c) with clinically significant or severe ADHD symptoms as per Conners score (t-score >63). The ROC cut-offs have been determined for all the three categories. The area under ROC curve for all three categories were almost same and statistically comparable, that is, 0.87, 0.86 and 0.87 [Figure 2]a, [Figure 2]b, [Figure 2]c, respectively, for raw scoring, LR + scoring and specifity scoring. The cut-off for raw scoring using ROC curve is ≥12, which corresponds to Conners score of ≥63 for clinically significant symptoms.
The current study revealed acceptable psychometric properties of AIIMS-modified INDT-ADHD tool with sensitivity and specificity of 100 per cent and 90 per cent, respectively, in comparison to gold standard diagnosis of ADHD. However, one needs to interpret this with caution, in view of wider dispersion of confidence interval. This needs confirmation in future studies with increased sample size.
In the present study, the tool been has validated among children with suspected ADHD presenting to a tertiary based on a definition recommended by American Academy of Pediatrics (AAP). Thus, it is recommended that this tool be tested in community settings for external validity.
To the best of our knowledge, this is the first diagnostic tool for ADHD in which a diagnosis can be established and at the same time, the disease severity can be determined with the help of a score. The evaluation of severity using the current scale is in concordance with the established Conners severity rating scale for ADHD.
The presence of severity scale within the tool makes it a very promising indicator of disease course, response to therapy and it can be used for future intervention trials in ADHD. In addition to this, in a scenario where symptoms are present only in one setting, and the child cannot be labelled as ADHD, the score can guide appropriate use of behavioural and pharmacotherapy.
The present tool was modified from the INDT-ADHD tool for diagnosis of ADHD, which was based on DSM-IV. In the current study, both AIIMS-modified INDT-ADHD tool and the initial INDT-ADHD tool were applied and their diagnostic performances were compared. The latter had a sensitivity and specificity of 87.7 per cent and 97.2 per cent, respectively. This is because it was based on DSM-IV, and in the present study, DSM-5 was used as the gold standard diagnostic criteria. Moreover, the criteria for age at onset for ADHD has now been increased to less than 12 years in DSM-5 from less than seven years of age in DSM-IV. This justifies the need for modification of the existing tool according to DSM-5 for optimal clinico-epidemiological utilization of the diagnostic tool.
Recently, there have been studies on regional adaptation and validation of ADHD diagnostic criteria based on DSM-5 in different regions of world to suit their cultural context., One such tool has been validated in Japanese populations for the age group of ≥17 years and they have concluded that the tool has acceptable psychometric properties. Another tool has been developed in Egypt among school-aged children (6–14 years) to find out the prevalence of ADHD. Although the psychometric properties of these tools are comparable with the current tool, the advantages of the current tool are its applicability in the pediatric population and a scoring system to grade the severity of ADHD. The psychometric properties of the current tool are also in consonance with existing tools such as Conners, SNAP-IV and ADHD rating scale-IV  [Table 4].
The strengths of the present study is a robust design for development of a new diagnostic tool for ADHD suitable for Indian cultural context based on current gold standard DSM-5 covering a wider age group with acceptable psychometric properties (6–18 years) and development of severity score, which can aid in prognostication, the formulation of treatment plans and in the design of intervention trials.
Limitations of the study include a limited sample size and a possible referral bias, the current center being a tertiary care referral teaching hospital. This opens up avenues for future research in the form of testing the current tool in the community settings.
To conclude, ADHD is associated with impaired academics, dysfunctional peer relationships and school dropouts.,,, The positive aspect is that symptoms in ADHD are significantly treatable and modifiable with behavioural and pharmacotherapy. Thus, the need of the hour is a diagnostic tool based on latest global criteria that can go a long way in diagnosing and initiating appropriate treatment in ADHD across various sociocultural settings.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4]