Table 2

Multivariate models predicting the early stage, treatment receipt and mortality in patients diagnosed with oesophageal adenocarcinoma (EAC) in the Veterans Health Administration

Early stage*
OR (95% CI)
Treatment†
OR (95% CI)
Overall mortality
HR (95% CI)
BE surveillance programme
 NoRef.Ref.Ref.
 Yes2.62 (1.67 to 4.11)1.71 (1.11 to 2.63)0.49 (0.37 to 0.66)
BE fiscal year
 2004Ref.Ref.–‡
 20051.70 (0.88 to 3.31)1.65 (0.89 to 3.09)
 20061.05 (0.49 to 2.25)2.20 (1.04 to 4.65)
 20071.15 (0.49 to 2.71)3.81 (1.53 to 9.46)
 20081.68 (0.59 to 4.78)3.39 (1.20 to 9.61)
 20092.80 (0.55 to 14.33)4.36 (0.88 to 21.49)
Age at EAC diagnosis
 <60Ref.Ref.Ref.
 60–640.82 (0.46 to 1.48)1.30 (0.71 to 2.37)0.93 (0.59 to 1.46)
 65–691.13 (0.60 to 2.13)0.75 (0.41 to 1.34)1.22 (0.77 to 1.92)
 70+0.90 (0.49 to 1.67)0.98 (0.54 to 1.77)1.70 (1.17 to 2.46)
Race
 WhiteRef.Ref.Ref.
 Black0.69 (0.20 to 2.41)1.94 (0.49 to 7.65)1.00 (0.46 to 2.22)
 Other1.68 (0.81 to 3.46)1.16 (0.60 to 2.24)0.64 (0.39 to 1.03)
Propensity score (quartiles)
 1Ref.Ref.Ref.
 21.34 (0.38 to 4.69)0.98 (0.28 to 3.39)1.77 (0.99 to 3.16)
 31.61 (0.41 to 6.35)1.47 (0.38 to 5.71)2.33 (1.25 to 4.34)
 41.72 (0.41 to 7.29)2.48 (0.59 to 10.42)1.78 (0.89 to 3.58)
PPI
 NoRef.Ref.
 Yes0.29 (0.13 to 0.67)0.45 (0.21 to 0.96)
GI visit
 NoRef.Ref.
 Yes1.52 (0.98 to 2.37)1.25 (0.93 to 1.67)
Comorbidity score at BE diagnosis
 0Ref.
 11.46 (1.04 to 2.06)
 2+2.06 (1.42 to 2.99)
Number of VA visits between BE and EAC diagnosis
 1st quartile (8–36)Ref.
 2nd quartile (37–68)0.64 (0.34 to 1.20)
 3rd quartile (69–125)0.41 (0.21 to 0.79)
 4th quartile (126–720)0.46 (0.23 to 0.92)
Year of EAC diagnosis
 2004–2006Ref.
 2007–20091.33 (0.94 to 1.88)
 2010–20111.71 (1.14 to 2.55)
  • Full models were adjusted for the variables indicated for each outcome.

  • *Early stage: 0 and 1 vs 2, 3, 4 and missing.

  • †Any treatment: surgery, endoscopic ablation/or resection, chemotherapy or radiation versus no treatment.

  • ‡All variables with ‘–’ were not included in multivariable model due to exclusion by stepwise regression.

  • BE, Barrett's oesophagus; PPI, proton pump inhibitor; VA, Veteran Affairs.