Measures

Depressive symptom severity over the past 7 days was assessed using the QIDS-SR,⁴⁷ a 16-item questionnaire covering the nine DSM IV-TR⁴⁸ major depressive disorder symptoms. Ratings are made on a four-point scale (0–3) anchored at all points by a description. For example, Question 11, ‘view of myself' is anchored at 0=‘I see myself as equally worthwhile and deserving as other people', 1=‘I am more self-blaming than usual', 2=‘I largely believe that I cause problems for others' and 3=‘I think almost constantly about major and minor defects in myself'. Following local ethics committee advice, item 12 ‘thoughts of death or suicide' was removed from the online mood monitoring (weekly and daily). Note: the complete scale including suicide item was used for face-to-face assessments. The QIDS-SR correlates highly with established clinician rating scales such as the Hamilton Rating Scale of Depression (r=0.86) and has high internal consistency (α=0.87).⁴⁷

Anxiety symptom severity over the past 7 days was measured using the BAI,⁴⁶ comprising 21 items. Ratings are made on a four-point scale (0–3) from ‘not at all' to ‘severely'. Example items include ‘scared', ‘fear of losing control' and ‘heart pounding or racing'. The BAI has high internal consistency (α=0.92) and test–retest reliability over 1 week, r=0.75.⁴⁹

Depressive symptoms were assessed using a modified version of the QIDS-SR⁴⁷ anchored to the last 24 h. The suicide item was again removed, as were two items referring to an unfeasible time frame for a daily rating—‘decreased/increased weight within the last 14 days', yielding 13 items. The structure of the daily QIDS-SR scoring is appropriate for time-series analysis with Gamma errors (see Supplementary Materials and also ref. 28). For further individual patient analysis of the time-series (using Markov chains — see the methods below), the cut-off score for moderate depression on the original QIDS-SR (⩾11)⁴⁷ was adapted to ⩾9 to account for the omitted items.

Daily anxiety symptoms were assessed via two bespoke ratings on a similar scale: the extent to which the participant felt anxious and fearful, and the severity of physical anxiety symptoms.

The computerized monitoring system was implemented via two secure servers: a web server hosting the website for completing questionnaires, and a database server to generate automatic prompts. Automated prompts were sent via email to all the participants (N=14), and by a mobile phone short message service where requested (N=12). If a participant did not complete a weekly or daily measure online within the scheduled time frame, researchers would attempt to contact them to complete the measure at the earliest opportunity, either via the online system, phone or at a therapy session. At pre- and post-treatment combined, 27 outstanding primary outcome weekly assessments (of 504, that is, 5.36%) and 34 outstanding daily assessments (of 784, that is, 4.34%) were collected following this procedure.

Measures taken at the face-to-face pre- and post-treatment assessments and weekly scores of mania symptoms are reported in the Supplementary Materials.