Table 1 Synoptic table of study measures Sample size and justification The sample size calculation was based on an audit study data from the Department of Community Pediatrics at the Medway NHS Trust (K Selby,
2013, unpublished data). Calculations based on this audit study data showed that the mean number of visits needed selleck products to achieve an ADHD diagnosis before introduction of the QbTest (control rate) for children aged 6–14 year olds was 2.94 visits and following the introduction of QbTest a diagnosis was reached in a mean of 2.18 visits. Following consultation with stakeholders, it was agreed that this difference (2.94–2.18) represented the minimum clinically important difference, with any smaller difference in mean clinic visits being of debatable value. Therefore, 71 patients in each study group will be required to detect a mean count difference of the above magnitude with 80% power at two tailed 0.05 significance level36 37 assuming the number of visits follows a Poisson distribution. Given the evidence that the intraclass correlation coefficients of mental health measures across General Practitioner (GP) centres is extremely low,38–40 and results from the Medway audit data indicate that the number of visits needed
to achieve an ADHD diagnosis was homogeneous across centres, we will assume that centre effects will not influence the sample size calculation for this study. After taking into account a 20% attrition rate, the final total sample size will be 178. The same calculation performed with 90% power would require a total sample of 234 participants. We aim to recruit 178 participants as a minimum and 234 participants as a maximum. Software Stata V.13 was used for power analysis. Randomisation and blinding Once consent has been obtained from participants, their information will be entered onto a web-based randomisation system (set up by University
of Nottingham Clinical Trials Unit; CTU). The arm to which a participant is assigned will be determined by a computer generated pseudo-random code using random permuted blocks of varying size, created by the Nottingham CTU Carfilzomib in accordance with their standard operating procedure and held on a secure server. Participants will be allocated with equal probability to each arm (QbO and QbB) with stratification by site. All participants will undergo the same research measures, including the QbTest. It is the time at which the report is made available to the clinician and patient that is randomised (immediately vs 6 months later). Outcome assessors for all measures will be blind to which arm the participant is in. There are no anticipated events. In which participant unblinding would be necessary.