*2.3.4 Falling action: The characters work towards resolving the main challenge(s)*

One of the aims of the SILC project was to provide services with regular analyses to inform delivery and operation. This section reports on some of the headline findings along with extract quotes from two of the SILC TSOs. Systems-level modelling demonstrates the importance of considering the interrelationships between individual practice elements as opposed to solely focusing on each in isolation [11, 21]. Although the challenges during each quarter were distinct, the areas of overlap were noteworthy. Not only was session non-attendance linked with unplanned endings, but those TSOs with the longest standing commitment to high-quality data also reported the highest rates of clinical improvement.

#### *2.3.4.1 Data quality*

One major shift during the first quarter was to adopt sessional ROM, moving from traditional pre and post-therapy measurement approaches. This process was supported by a dedicated project member auditing and feeding back information to services. By the end of the first quarter, pre-and-post outcome completion rates increased from an average of 65% at baseline to 98%, while by the end of the year, this was 97%, with all TSOs achieving above 90% and half achieving 100% completion rates (**Figure 3**). These values were almost identical to the IAPT programme's recent achievement of 98%, a decade after its first site implementation [7].

### *2.3.4.2 Session non-attendance*

At the start of the second quarter, members began to record session nonattendance, including when an appointment was cancelled (by client) or the client DNA (no advanced warning given). One of the primary areas of interest was understood when sessions being missed were most likely to occur. Aggregating each service's datasets, the total number of appointments per sequential session number was tallied to assess what proportion was recorded as either cancelled or DNA. Including only session numbers with over 10 appointments each, it was possible to chart this data (**Figure 4**). It was identified that cancellations as a

#### **Figure 3.**

*Improvement of pre-and-post outcome measures completion rates for all SILC TSOs, 1 year before-and-after the project.*

proportion tended to increase the longer therapy progressed; although this might be due to a lower number of appointments at these stages. DNAs as a proportion did not exceed 10% for any session number although they did tend to occur earlier in therapy, with sessions 2–5 reporting the highest rates of 7–8%. The occurrence of DNAs declined somewhat as therapy progressed, possibly due to contracting which discharged clients after missed appointments without prior notice. Focusing on session non-attendance helped determine the scale of the challenge and how the pattern of cancellations and DNAs differed, prompting two participating services to a revise their policy in the interests of equitable access and service efficiency.
