183 6 Cochrane review on integrated disease management for COPD 6. Lung function Lung function was expressed as FEV1 in litres and as FEV1% predicted. Following Kruis 2013 , we pooled data from a total of six studies for FEV1 (litre) (Bourbeau 2003; Kalter-Leibovici 2018; Öztürk 2020; Sridhar 2008; Wood- Baker 2006; Zhang 2020), and from 14 studies for FEV1% predicted (Farrero 2001; Fernandez 2009; Güell 2000; Jimenez-Reguera 2020; Kalter-Leibovici 2018; Khan 2019; Ko 2016; Lenferink 2019; Littlejohns 1991; Lou 2015; van Wetering 2010; Wakabayashi 2011; Wood-Baker 2006; Zhang 2020). Wang 2017 and Wood-Baker 2006 reported on short-term e ects on FEV1 in litres, but data from Wang 2017 could not be pooled due to reporting error. Pooling of FEV1 in litres showed no di erences between groups for medium- and long-term follow-up (Analysis 1.31). Pooled MDs in FEV1% predicted showed a shortterm e ect in favour of the IDM group (MD 2.88, 95% CI 1.35 to 4.40). This e ect was statistically signi cant but was not clinically signi cant. Medium-term e ects were less pronounced and were not statistically signi cant (MD 0.95, 95% CI -0.20 to 2.11). After 24 months, there was no di erence between groups (MD 1.18, 95% CI -0.82 to 3.18). Results were homogeneous across studies (Analysis 1.32). However, except for Lou 2015 , 95% con dence intervals for the di erent studies were consistently large, suggesting large between-patient variation. 7. Anxiety and depression Ten studies assessed depression, anxiety, or both as an outcome (Engstrom 1999; Güell 2000; Kessler 2018; Lenferink 2019; Littlejohns 1991; Öztürk 2020; Rose 2017; Titova 2017; Trappenburg 2011; Vianello 2016). Engstrom 1999 used the Mood Adjective Check List (MACL), and Güell 2006 used a Revised Symptom Checklist. Kessler 2018 used the Hospital Anxiety and Depression Scale (HADS) but reported only the combined score. The other studies reported depression and anxiety scores from the HADS, and results were pooled. Pooled data from the anxiety domain of the HADS showed no di erences between groups (MD 0.09, 95% CI -0.30 to 0.47; I² = 38%). Pooled data for the depression domain of the HADS showed a non-signi cant e ect in favour of the intervention group (MD -0.20, 95% CI -0.45 to 0.05; I² = 38%: Analysis 1.33). 8. Process-related outcomes 8.1 Compliance/Adherence Patient adherence to the programme or to intervention uptake was evaluated in ve studies by review of programme attendance rate and programme completers (Bernocchi 2017; Rose 2017; Tabak 2014; Vasilopoulou 2017; Zwar 2016). Bernocchi 2017 reported a high adherence rate, with 93% of participants performing activities at home as part of the programme. Rose 2017 reported that 29% of participants were fully compliant and 22% were non-compliant (< 50% compliant with separate components). In addition, only 7% of study participants attended respiratory rehabilitation despite this being a component of usual care. Study authors also noted that 38% of intervention group participants who met the eligibility criteria for pulmonary rehabilitation were unable to attend due to unavailability of classes. Tabak 2014 monitored use of the web
RkJQdWJsaXNoZXIy MTk4NDMw