The four ASD assessments
Background
The 3di has a bank of nearly 900 questions, of which around 200 are involved in ASD diagnoses. You can see every single question in the Interview Explorer.
Within ASD, there are four different assessments, developed separately and with different properties. The assessments were developed over time, meaning not all of them are present in older versions of the software. Only the latest versions of the 3di from 2023 onwards contain all four of the ASD assessments.
The four ASD assessments are:
|
|
|
|
Each ASD assessment has a corresponding set of questions (a route), a scoring algorithm, and a set of tables in the generated Word reports. In the latest 3di versions, there are therefore up to four sets of report tables, corresponding to each of the four assessments.
Recall from Key concepts that a report can be generated at any stage of completion of the 3di's questions, even before answering a single question. If you generate a report having answered no questions at all, you get a very boring but structurally complete report, full of notes telling you why scores could not be computed. The structure of the report is not determined by the questions answered, but instead by which of the two report buttons you clicked, and in later 3di versions the values you set in the Preferences form.
Picking the appropriate ASD route in the Route Explorer will allow you to complete just the necessary questions for that assessment, thus populating the corresponding report components.
You can also fill in sufficient questions to complete more than one assessment: the 3di report will fill in whatever report components it can. It was common in earlier versions of the 3di to complete questions on a combined route which covered both the ICD-10 brief and DSM-5 extended assessments. If you set out to fill in the full route for an algorithm but have to miss some questions for whatever reason, that's fine too: the report will tell you about the missing data and attempt to make a sensible adjustment.
There are two flavours of report corresponding to two buttons in the Case Manager:
Report (with ASD_extended). For patient number 1, this generates a Word document named Report_000001.docx.
Report ASD_brief. For patient number 1, this generates a Word document named ASD_000001.docx.
History of the 3di
The World Health Organisation's International Statistical Classification of Diseases and Related Health Problems, 10th revision (ICD-10) came into effect on 1 Jan 1993. ICD-11 came into effect on 1 Jan 2022.
Meanwhile the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) was published in 1994, followed by DSM-5 in 2013.
The 3di was first developed as a research tool starting around the year 1999 at the Great Ormond Street Institute of Child Health. Its first ASD assessment was built around an amalgamation of ICD-10 and DSM-IV criteria, the two systems seen as broadly equivalent. The interview was fairly long, based on 134 questions, and designed for research use where there was sufficient time for such a thorough assessment.
Over the next 25 years there were several developments. Among young people in the UK, the prevalence of autism grew from a fraction of a percent to several percent. There was increasing demand for a quicker autism informant interview for use in CAMHS clinics where there was much less time available for the extended assessment. This led to the development of the ICD-10 brief assessment.
With the worldwide adoption of DSM-5 and more recently ICD-11, the 3di was updated with a new DSM-5 compatible extended ASD assessment in 2013. Finally in late 2022 the brief DSM-5 assessment was released.
Comparing DSM-5 and ICD-11, we consider that by design the two sets of criteria are so similar that ASD outcomes (and many others) can be reported as DSM-5 or as ICD-11. The 3di always references DSM-5 because the APA got there about 7 years before the WHO; you can substitute one term for the other in your reports or perhaps reference neither.
The ASD assessments
The ASD routes and report sections were renamed in 2023.
1. ICD-10 ASD extended
This was the first ASD algorithm implemented in the 3di. Its scoring is based on 134 questions.
Route
The corresponding route is ASD_extended (ICD-10).
Before 2023 the route was called simply ASD_extended, but it still contained the exact same questions.
Report
In recent 3di versions, if you still want to see ICD-10 report components you need to visit the Preferences form.
This output of this algorithm appears only in the extended Word report, and not in the brief Word report. In the extended Word report the corresponding tables appear under the heading ICD-10 ASD.
Before 2023, these results appeared under the heading ASD Outcomes.
2. ICD-10 ASD brief
This was the second ASD algorithm implemented in the 3di. Its scoring is based on 61 questions.
Route
The corresponding route is ASD_brief (ICD-10).
Before 2023 the route was called simply ASD_brief, but it still contained the exact same questions.
Report
In recent 3di versions, if you still want to see ICD-10 report components you need to visit the Preferences form.
This output of this algorithm appears only in the brief Word report, and not in the extended Word report. In the brief Word report the corresponding sections appear under heading ICD-10 ASD.
Before 2023, these sections were at the top level of the report.
3. DSM-5 extended
This was the third ASD algorithm implemented in the 3di. Its scoring is based on 177 questions.
Route
The corresponding route is ASD_extended (DSM-5 Crit A + B).
Before 2023, there was no route corresponding to only the questions of this algorithm. Instead there was a choice of two routes named ASD_brief+DSM-5 and ASD_extended+DSM-5, which contained the DSM-5 extended questions plus those from one of the ICD-10 routes already described. In other words, a single route covered all the questions from two algorithms.
Report
This output of this algorithm appears only in the extended Word report, and not in the brief Word report, under the heading DSM-5: ASD extended assessment.
Before 2023, in the extended Word report the two corresponding tables appear under the heading DSM-5 scales for ASD, Criterion A and DSM-5 scales for ASD, Criterion B. Note that the immediately following section ASD Outcomes relates to the ICD-10 ASD extended algorithm as described above.
Before 2023, despite the existence of the ASD_brief+DSM-5 route, you cannot get the output from both of those algorithms in the same report document: the ICD-10 ASD brief output appears only in the brief report, while the DSM-5 extended algorithm output appears only in the extended report.
4. DSM-5 brief
This was the fourth ASD algorithm implemented in the 3di. Its scoring is based on 66 questions.
Before 2023 this assessment did not exist, so there is no corresponding route and no corresponding report components.
Route
Introduced in 2023, the corresponding route for this algorithm is named ASD_brief (DSM-5 Crit A+B).
Report
This output of this algorithm, despite having 'brief' in the name, appears in both the brief and extended Word reports from 2023 onwards, under the heading DSM-5: ASD brief assessment.
How the algorithms work
The details of the ICD-10 algorithms 1 and 2 are explained in three files found in the 3di installation folder, under the Reference subfolder. ASD_three-domain_scales_defined and ASD_diagnosis_flowcharts_3_domain describe the algorithm, while ASD_scales_Ref details which individual questions contribute to each subscale.
The details of the DSM-5 algorithms 3 and 4 are explained in two files found in the 3di installation folder, under the Reference subfolder. DSM-5_ASD_scoring describes the algorithm, while again ASD_scales_Ref details which individual questions contribute to each subscale.
3di-adult
You may have heard about this assessment. This was part of a research project and you can read the paper here. It was a standalone, pen-and-paper tool which drew on the 3di but was never implemented in the software, and presently there are no plans to do so.
Currently, the 3di software is not suitable for assessments of adults.
Last updated