Fig. 4

A comparative application of the session 'dosing' model to inform treatment planning for two patients with different sensitivities to lapses in session cadences, based on their historical data. The "q**d" nomenclature indicates schedules at a particular frequency, where 'q7d' translates to every seven days. The observed and model fit plot (far left) is the same as that in Fig. 3. The dashed horizontal line across the image represents a hypothetical 'response' threshold/reference and maps to sub-moderate symptom severity on the DSM L1 survey. All simulations were generated with 180 days of treatment followed by an additional 180 days of no treatment. For the top profile, consistent care appears to be paramount. In all the simulations for this profile, (despite the response magnitude) historical data suggests that this patient will return to a severe level of symptoms within 6 months. Historical data for the bottom profile suggests that this patient would be amenable to less frequent sessions. For these profiles, and all other applications of this framework, as new data is collected the relationships may shift, highlighting that these models should not be applied without review and regular collection of new data