Review of: Hollis C et al, J Child Psychol Psychiatry 2018;59(12):1298–1308

With busy clinic schedules and the ever-burgeoning load of documentation, computerized diagnostic aids are in more demand than ever. For ADHD, the gold standard is still a clinical assessment with information from parents and teachers, but those reports are difficult to obtain and time-consuming to go through. In these situations, computerized testing may help boost clinical decision-making.

One common testing procedure is continuous performance testing (CPT), which involves a subject’s ability to quickly respond to a given stimulus while not responding to distracting stimuli. QbTest is a specific testing method that combines computerized CPT and an infrared camera measuring how much the patient moves around during the 20-minute test. In 2014, the FDA cleared QbTest as a tool to supplement a clinical assessment for ADHD, meaning that it reached the diagnostic sensitivity and specificity thresholds required by the FDA. However, like all such tests, it is not meant to be a stand-alone diagnostic test. This study attempted to see how useful QbTest is for clinicians.

The randomized, controlled trial analyzed data from 250 youth between ages 6 and 17 years referred for an ADHD assessment. Funding came from the National Institute for Health Research in the UK, but equipment and training were provided directly from QbTech Ltd (the makers of QbTest). The device’s website (www.qbtech.com/qbtest) has descriptions of the testing equipment: an infrared camera, a reflector that fits on the patient’s forehead, and the computer software. The sample, drawn from UK outpatient clinics, was nearly 80% male and 90% white. All participants took the QbTest at the beginning of the study period, then were divided into two groups. The QbOpen group had the results revealed to the clinician immediately, while the QbBlind group withheld the results. The primary outcome was the number of appointments it took to rule in or out an ADHD diagnosis, with secondary outcomes including appointment duration and clinician’s confidence in the diagnosis.

At the end of 6 months, the youth in the QbOpen group were 44% more likely (hazard ratio = 1.44, p = 0.029) to have reached a diagnostic decision than those in the QbBlind group. However, over 30% of the entire sample had still not reached a diagnostic decision at 6 months. Interestingly, ADHD was excluded at double the rate when clinicians had access to the QbTest report (p = 0.049), and they were more confident in their decision overall (p = 0.022). The appointment duration for the QbOpen group was reduced by about 15% (p = 0.001). The authors also did a cost analysis concluding that QbTest was largely cost-neutral to the healthcare system.

CCPR’s Take
As clinicians, we need to maintain diagnostic pre-eminence over supplemental tests for ADHD. While QbTest may increase the expediency of diagnosis and boost diagnostic confidence for clinicians, we need to be careful that it is neither masking other reasons for symptoms nor ruling them out when, for instance, the child being tested is inattentive but not overactive. It would also be interesting to see more comparison studies with more established measures such as the TOVA, GDS, IVA, or Connors CPT.