Skip to main content
. 2005 Jun 7;92(12):2122–2128. doi: 10.1038/sj.bjc.6602649

Table 3. Exploratory analysis of effects of prognostic factors on clinical outcome.

End point Factor RR (95% CI)
Response Peritoneal involvement: 0.43
  Yes vs no (0.13–1.47)
  Surgery of primary tumour: 1.78
  Yes vs no (0.76–4.15)
  Baseline Karnofsky status: 0.41*
  ⩽80 vs >80 (0.17–0.95)
  Treatment effect: 0.42*
  ELF vs ILF (0.18–0.97)
     
Progression Peritoneal involvement: 1.15
  Yes vs no (0.67–1.96)
  Surgery of primary tumour: 0.86
  Yes vs no (0.58–1.29)
  ELF vs ILF 1.7
  (in patients with Karnofsky status ⩽80) (0.96–3.0)
  ELF vs ILF 0.8
  (in patients with Karnofsky status >80) (0.47–1.39)
  Number of organs: 1.99**
  >1 vs 1 (1.21–3.28)
  Site of primary tumour: 1.91**
  oesophagogastric vs stomach (1.2–3.04)
     
Death Peritoneal involvement: 0.85
  Yes vs no (0.42–1.75)
  Surgery of primary tumour: 0.69
  Yes vs no (0.46–1.06)
  ELF vs ILF 2.41
  (in patients with peritoneal involvement) (0.99–5.82)
  ELF vs ILF 1.01
  (in patients without peritoneal involvement) (0.64–1.59)
  Number of organs: 2.56**
  >1 vs 1 (1.48–4.42)
  Baseline Karnofsky status: 1.84**
  ⩽80 vs >80 (1.21–2.8)

RR=risk ratio; CI=confidence interval.

A risk value >1 shows an increased likelihood of the clinical outcome in favour of the first of the two compared terms. ELF, etoposide+LV (leucovorin)+5-FU(5-fluorouracil); ILF, irinotecan+LV+5-FU; EJ, oesophagogastric junction.

*P<0.05.

**<0.01 by Wald χ2 test.