Skip to main content

Table 3 Multivariate linear regressions of factors associated with grief

From: Is dying in hospital better than home in incurable cancer and what factors influence this? A population-based study

  Unadjusted Adjusted
  n M (SD) B (95 % CI) β P value B (95 % CI) β P value sr2
Grief intensity around time of death (TRIG I)          
 Place of death          
  Hospital 158 20.78 (8.25) Ref.    Ref.    
  Home 162 20.52 (7.91) −0.25 (–2.03 to 1.52) −0.02 0.779 −2.44 (–4.50 to –0.39) −0.15 0.020 0.02
 Relative’s relationship to the patient          
  Spouse/partner 130 21.85 (8.16) Ref.    Ref.    
  Son/daughter 134 20.25 (7.58) −1.61 (–3.55 to 0.33) −0.10 0.104 −1.89 (–3.86 to 0.08) −0.12 0.060 0.01
  Brother/sister 15 17.87 (6.32) −3.99 (–8.29 to 0.31) −0.11 0.069 −4.83 (–9.09 to –0.56) −0.13 0.027 0.02
  Other 40 19.45 (9.40) −2.40 (–5.25 to 0.45) −0.10 0.098 −3.16 (–6.14 to –0.19) −0.13 0.037 0.01
 Relative’s presence at time of death          
  No 94 19.04 (8.51) Ref.    Ref.    
  Yes 225 21.24 (7.72) 2.19 (0.27 to 4.12) 0.13 0.026 2.41 (0.28 to 4.54) 0.14 0.027 0.02
 Patient’s preference for place of death discussed with family          
  No 121 19.00 (8.01) Ref.    Ref.    
  Yes 188 21.57 (8.06) 2.57 (0.73 to 4.42) 0.16 0.006 3.01 (1.11 to 5.03) 0.19 0.002 0.03
Grief intensity at time of questionnaire completion (TRIG II)          
 Place of death          
  Hospital 158 43.11 (14.10) Ref.    Ref.    
  Home 158 45.02 (11.26) 1.91 (–0.92 to 4.73) 0.08 0.185 −3.63 (–6.76 to –0.50) −0.14 0.023 0.01
 Relative’s relationship to the patient          
  Spouse/partner 126 49.21 (10.00) Ref.    Ref.    
  Son/daughter 133 41.92 (11.71) −7.55 (–9.85 to –5.26) −0.29 <0.001 −6.69 (–9.98 to –3.40) −0.26 <0.001 0.04
  Brother/sister 15 40.67 (16.17) −10.93 (–15.75 to –6.21) −0.19 <0.001 −6.78 (–13.45 to –0.11) −0.11 0.046 0.02
  Other 41 36.98 (16.18) −11.92 (–15.22 to –8.63) −0.31 <0.001 −12.54 (–17.09 to –7.99) −0.32 <0.001 0.08
 Relative’s presence at time of death          
  No 90 38.77 (15.29) Ref.    Ref.    
  Yes 225 46.11 (10.94) 6.58 (4.36 to 8.81) 0.25 <0.001 4.77 (1.51 to 8.04) 0.17 0.004 0.01
 Relative’s work arrangements in last three months of patient’s life          
  0 to 3 days off work 58 37.50 (13.52) Ref.    Ref.    
  4 to 14 days off work 44 45.14 (11.92) 7.64 (2.88 to 12.40) 0.22 0.002 7.07 (2.41 to 11.73) 0.20 0.003 0.02
  15+ days off work 47 50.32 (9.19) 12.82 (8.15 to 17.49) 0.37 <0.001 8.44 (3.64 to 13.23) 0.24 0.001 0.03
  Not working 144 44.97 (12.37) 7.47 (3.77 to 11.18) 0.30 <0.001 3.87 (3.64 to 13.23) 0.15 0.048 0.01
 Patient’s preference for place of death discussed with family          
  No 123 41.83 (13.86) Ref.    Ref.    
  Yes 182 45.34 (11.96) 2.80 (0.52 to 5.09) 0.11 0.016 2.88 (0.07 to 5.82) 0.11 0.056 0.01
  1. The TRIG was used to measure grief intensity for the respondent; the first scale asks respondents about their experience at the time of death (TRIG I 8–40 scale, 8 items, measured retrospectively) and the second scale asks about their experience at the time of questionnaire completion (TRIG II 13–65 scale, 13 items; higher scores meaning greater grief intensity; [30]). TRIG I model included 307 bereaved relatives (87.2 %), with more missing amongst hospital deaths than amongst home deaths (16.4 % vs. 9.1 %, P = 0.042). Model statistics: R for regression significantly different than zero, F(9,297) = 4.492, P <0.001; R2 = 0.120, adjusted R2 = 0.093. TRIG II model included 281 bereaved relatives (79.8 %), with no significant differences in the proportion missing amongst hospital deaths and home deaths (22.6 % vs. 17.7 %, P = 0.254). Model statistics: R for regression significantly different than zero, F(12,268) = 8.181, P <0.001; R2 = 0.268, adjusted R2 = 0.235
  2. CI, Confidence interval; B, Unstandardised regression coefficient; β, Standardised regression coefficient; M, mean; sr2, Semipartial correlation; SD, Standard deviation; TRIG, Texas Revised Inventory of Grief
Back to article page

BMC Medicine

ISSN: 1741-7015

Contact us