178 6 Chapter 6 signi cantly between subgroups (Chi² = 10.93, df = 3, P = 0.01). Pooling of studies conducted in North America showed a signi cant reduction in respiratory-related hospital admissions (OR 0.69, 95% CI 0.50 to 0.94; I² = 44), as did pooling of studies conducted in Southern Europe (OR 0.35, 95% CI 0.18 to 0.68; I² = 25%). Pooling of studies from Northwestern Europe and Oceania was not possible due to the small numbers (Analysis 1.21) (Lenferink 2019; Smith 1999). In addition to regional di erences in e ects of IDM on respiratory-related hospital admissions, there was a marked di erence in the mean rate of respiratory-related hospital admissions per patient. Among IDM groups, the mean rate per patient was 0.19 admissions per patient in studies from North America, 0.21 per patient from Northwestern Europe, 0.59 per patient for Southern Europe, and 0.70 per patient from Oceania. Similarly, for controls, the rate from North America was 0.26 per patient, from Northwestern Europe 0.26 per patient, from Southern Europe 0.64 per patient, and from Oceania 0.56 per patient. 3.4. Hospital admissions, all causes We were able to pool ten studies that reported on patients experiencing at least one hospital admission for all causes and included a total of 9030 participants. Pooling showed an overall statistically signi cant e ect in favour of IDM (OR 0.75, 95% CI 0.57 to 0.98). This means that compared with usual care, there were 72 fewer (range 138 fewer to 5 fewer) hospitalisations per 1000 with IDM. Pooling based on follow-up period indicated slight di erences in short-, medium-, and long-term e ects (Analysis 1.22). 3.5. Hospital admissions, all causes - short-term Only one study including 112 participants reported on the number of hospital admissions for all causes after 6 months’ follow-up and therefore could not be pooled (Bernocchi 2017). Study authors reported a signi cant reduction in the number of patients having at least one hospital admission, in favour of the intervention group (OR 0.31, 95% CI 0.14 to 0.67). 3.6. Hospital admissions, all causes - medium-term Five studies with a total of 1212 participants provided data on the number of participants admitted at least one time for all causes at 6 to 15 months’ follow-up (medium-term) (Fan 2012; Kessler 2018; Lenferink 2019; Littlejohns 1991; Rea 2004). Kessler 2018 did not directly report the number of participants, so the number was approximated based on the percentage of people with 0 hospitalisation days. Pooling showed that results were homogeneous and there was no signi cant di erence between groups (OR 0.93, 95% CI 0.71 to 1.21; I² = 14%). A sensitivity analysis of only high-quality studies showed a similar result (OR 0.91, 95% CI 0.66 to 1.26; I² = 0%). 3.7. Hospital admissions, all causes - long-term Four studies including a total of 7706 participants assessed the number of participants admitted after 15 months’ follow-up (Kalter-Leibovici 2018; Lou 2015; Sridhar 2008; van
RkJQdWJsaXNoZXIy MTk4NDMw