163 6 Cochrane review on integrated disease management for COPD 6. Depression (10 studies). 7. Dyspnoea (13 studies). 8. Process-related outcomes (14 studies). Details of the included studies and outcomes are provided in Characteristics of included studies, Table 3, Table 4, Table 5, and Table 6. We requested additional data from 21 study authors; 14 (67%) responded. Nine studies provided additional data that we used in the analysis (Bernocchi 2017; Kalter-Leibovici 2018; Kennedy 2013; Kessler 2018; Khan 2019; Lenferink 2019; Titova 2017; Vasilopoulou 2017; Wang 2017) . Seven studies provided su cient data for calculation of correlation coe cients used to impute missing data (Aboumatar 2019; Engstrom 1999; Fan 2012; Kalter-Leibovici 2018; Lilholt 2017; Sridhar 2008; Vasilopoulou 2017) (see Dealing with missing data). Excluded studies We excluded 118 full-text articles from the current update during the full-text screening process. The Characteristics of excluded studies table provides full details on reasons for exclusion. Risk of bias in included studies Results of the risk of bias assessment are presented in Figure 2. All but one of the included studies were judged to be at high risk of bias for blinding of participants, which is a result of the nature of the intervention. With regard to the other domains, the likelihood that bias was present (high risk of bias) varied across studies, from 4% for random sequence generation (selection bias) to 27% for blinding of outcome assessment (detection bias). Allocation We judged 43 included studies as having low risk of bias in sequence generation (Aboumatar 2019; Aiken 2006; Bernocchi 2017; Bourbeau 2003; Boxall 2005; Cambach 1997; Fan 2012; Fernandez 2009; Freund 2016; Gottlieb 2011; Haesum 2012; Jimenez-Reguera 2020; Kalter-Leibovici 2018; Kennedy 2013; Kessler 2018; Khan 2019; Ko 2016; Ko 2009; Kruis 2014; Lenferink 2019; Lilholt 2017; Littlejohns 1991; Mendes 2010; Öztürk 2020; Rea 2004; Rice 2010; Rose 2017; Sanchez-Nieto 2016; Silver 2017; Smith 1999; Sridhar 2008; Tabak 2014; Theander 2009; Trappenburg 2011; van Wetering 2010; Vasilopoulou 2017; Vianello 2016; Wakabayashi 2011; Wang 2017; Wijkstra 1994; Wood-Baker 2006; Zhang 2020; Zwar 2016). Information from eight studies was insu cient to permit a decision (Bendstrup 1997; Dheda 2004; Engstrom 1999; Farrero 2001; Güell 2000; Güell 2006; Lou 2015; Strijbos 1996). One study was judged to have high risk of bias, as participants were randomised based on district (Titova 2017). With regard to allocation bias, we judged 27 studies as having low risk of bias and ve studies as having high risk of bias. For the remaining 20 studies, provided information was insu cient to permit a rm conclusion (unclear risk of bias).
RkJQdWJsaXNoZXIy MTk4NDMw