Table 2

Adjusted Odds Ratios (95% CI) for healthcare utilization in a foreign country in 2007 by country of origin
Country of Origin Doctor
Model I Model II Model III Model IV Modul V
OR (95% CI) P OR (95% CI) P OR (95% CI) P OR (95% CI) P OR (95% CI) P
General practitioner
Denmark 1.00 1.00 1.00 1.00 1.00
Turkey 3.91 (2.61-5.86) <.000 3.86 (2.54-5.87) <.000 4.56 (2.91-7.13) <.000 4.25 (2.38-7.60) <.000 4.49 (1.92-6.32) <.000
Turkey Descendants 2.37 (1.57-3.58) <.000 1.72 (1.02-2.90) .041 1.91 (1.13-3.26) .017 1.30 (0.69-2.42) .416 1.19 (0.64-2.23) .579
Specialist doctor
Denmark 1.00 1.00 1.00 1.00 1.00
Turkey 8.73 (4.99-15.26) <.000 10.17 (5.65-18.30) <.000 12.29 (6.53-23.14) <.000 9.31 (4.02-21.58) <.000 6.74 (2.84-16.01) <.000
Turkey Descendants 2.73 (1.44-5.18) .002 6.10 (2.51-14.19) <.000 7.31 (2.91-18.36) <.000 5.92 (2.06-16.96) .001 4.97 (1.72-14.41) .003
Hospital
Denmark 1.00 1.00 1.00 1.00 1.00
Turkey 7.24 (3.96-13.24) <.000 8.28 (4.39-15.62) <.000 9.22 (4.71-18.01) <.000 7.95 (3.33-18.99) <.000 5.70 (2.33-13.93) <.000
Turkey Descendants 4.33 (2.33-8.04) <.000 4.07 (1.84-9.03) .001 4.41 (1.95-9.94) <.000 3.12 (1.19-8.18) .021 2.48 (0.95-6.48) .063
Dentist
Denmark 1.00 1.00 1.00 1.00 1.00
Turkey 3,15 (1.94-5.13) <.000 3.13 (1.89-5.18) <.000 3.34 (1.96-5.68) <.000 5.72 (2.85-11.45) <.000 5.20 (2.57-10.54) <.000
Turkey Descendants 1.95 (1.18-3.22) .010 1.58 (0.82-3.02) .170 1.72 (0.89-3.33) .109 2.28 (1.04-4.97) .039 2.17 (0.996-4.74) .051

95% CI indicates 95% confidence interval.

Statistically significant variables (P <.05) are depicted in bold type.

Model I: Using binary logistic regression to adjust for country of origin.

Model II: Using binary logistic regression to adjust for country of origin, sex, age.

Model III: Using binary logistic regression to adjust for country of origin, sex, age, and marital status.

Model IV: Using binary logistic regression to adjust for country of origin, sex, age, marital status, education, employment status, and household income.

Model V: Using binary logistic regression to adjust for country of origin, sex, age, marital status, education, employment status, household income, physical health symptoms, and mental health symptoms.

All estimates are adjusted for all covariates, even though some of the covariates are statistically non-significant.

Nielsen et al.

Nielsen et al. BMC Health Services Research 2012 12:390   doi:10.1186/1472-6963-12-390

Open Data