Skip to main content
. 2021 Jun 22;37(5):1115–1121. doi: 10.1007/s11606-021-06950-y

Table 3.

Associations Between Misdiagnosis Experience and the Trust in Physician Scale

Trust in Physician Scale, points Mean difference (95%CI) P-value
Unadjusted*
Patient’s experience with a misdiagnosis −5.80 (−9.09 to −2.51) 0.001
Patient’s family experience with a misdiagnosis −5.19 (−8.23 to −2.14) 0.001
Multivariable-adjusted†
Patient’s experience with a misdiagnosis −4.30 (−8.12 to −0.49) 0.028
Patient’s family experience with a misdiagnosis −3.20 (−6.34 to −0.05) 0.047
Age, per 10 years 1.17 (0.06 to 2.29) 0.040
Sex, female 1.17 (−1.49 to 3.83) 0.379
Duration with patients’ physician
< 1 year Ref
1–< 3 years 1.13 (−2.86 to 5.11) 0.573
≥ 3 years 4.28 (0.28 to 8.27) 0.036
Education
Junior high school Ref
High school −1.34 (−6.6 to 3.88) 0.607
Junior college 1.48 (−5.6 to 8.58) 0.677
University −0.43 (−5.7 to 4.84) 0.869
Graduate school −3.67 (−12.0 to 4.61) 0.377
Not answered −1.16 (−9.2 to 6.85) 0.772
Total household income
< 1,000,000 yen Ref
1,000,000–< 3,000,000 yen 5.93 (−1.56 to 13.4) 0.118
3,000,000–< 5,000,000 yen 4.43 (−2.25 to 11.1) 0.188
5,000,000–< 10,000,000 yen 4.81 (−1.54 to 11.1) 0.134
≥ 10,000,000 yen 9.22 (3.38 to 15.1) 0.003
Reported disease
Cardiac disease, arrhythmia −0.34 (−6.2 to 5.55) 0.907
Cardiac disease, angina pectoris or myocardial infarction −1.30 (−5.2 to 2.56) 0.502
Cardiac disease, heart failure 2.41 (−5.0 to 9.78) 0.514
Diabetes −2.22 (−6.3 to 1.89) 0.282
Connective tissue disease −1.38 (−8.6 to 5.81) 0.701
Cancer 2.08 (−0.5 to 4.69) 0.114
Depression −2.49 (−6.4 to 1.44) 0.208

Analysis of 661 patients from 46 prefectures. Bold font indicates significance at P <0.05 in P-value column and corresponding mean difference and 95% confidence interval.

*General linear models with consideration of prefectural-level correlation using cluster-variance

†General linear model adjusted for age, sex, duration of the relationship with the physician, comorbidities, education, and total household income with consideration of prefectural-level correlation using cluster variance. The variable in each unadjusted model was included in the multivariable model