Skip to main content
. 2017 Nov 9;20(1):1–16. doi: 10.1093/biostatistics/kxx058

Table 2.

RR of hospital admission: PMInline graphic higher than 25 Inline graphic (chosen cut point) in Greater London. The main analysis with Inline graphic5 subjects per wards is presented in bold, while the others are used as sensitivity analyses

Areas Data used Inline graphic1 subject (12.9% missing) Inline graphic5 subjects (24.6% missing) Inline graphic10 subjects (37.4% missing) Inline graphic20 subjects (68.0% missing)
    RR (95% CI) 95% CI width RR (95% CI) 95% CI width RR (95% CI) 95% CI width RR (95% CI) 95% CI width
Inline graphic X,Inline graphic 0.91   0.94   0.96   0.96  
    (0.86–0.96) 0.10 (0.89–1.00) 0.11 (0.90–1.03) 0.13 (0.88–1.04) 0.16
  X,Inline graphic,EPS 1.04   1.06   1.07   1.05  
    (0.94–1.14) 0.10 (1.01–1.12) 0.12 (1.01–1.15) 0.14 (0.95–1.16) 0.21
Inline graphic X,Inline graphic 0.89   0.89   0.89   0.89  
    (0.84–0.94) 0.10 (0.84–0.94) 0.10 (0.84–0.94) 0.10 (0.84–0.94) 0.10
  X,Inline graphic,EPS 1.05   1.02   1.03   1.06  
    (0.99–1.12) 0.13 (0.96–1.09) 0.13 (0.97–1.10) 0.13 (0.99–1.13) 0.14
  X,Inline graphic,Inline graphic 0.98   0.93   0.92   0.89  
  (MICE) (0.93–1.04) 0.11 (0.88–0.99) 0.11 (0.87–0.98) 0.11 (0.84–0.94) 0.10