Table 3.
Estimate of the association between demographic, practice, and HIT characteristics and 1 or more symptoms of burnout among physician respondents with EHRs (N = 1630)
| Characteristic | Unadjusted OR (95%CI) | P value | Adjusted ORa (95%CI) | P value |
|---|---|---|---|---|
| Age | ||||
| 30–50 | ref | ref | ||
| 51–64 | 1.01 (0.80–1.29) | 0.912 | 0.89 (0.63–1.26) | 0.508 |
| 65–90 | 1.07 (0.79–1.45) | 0.678 | 1.04 (0.69–1.58) | 0.839 |
| Gender | ||||
| Male | ref | ref | ||
| Female | 1.59 (1.27–1.98) | <.001 | 1.41 (1.02–1.94) | 0.037 |
| Practice setting | ||||
| Hospital/inpatient | ref | ref | ||
| Office/outpatient | 1.33 (1.06–1.68) | 0.015 | 1.07 (0.44–2.62) | 0.884 |
| Practice size | ||||
| 1–3 clinicians | ref | ref | ||
| 4–9 clinicians | 1.00 (0.74–1.35) | 0.987 | 0.70 (0.47–1.02) | 0.066 |
| 10–15 clinicians | 1.24 (0.85–1.81) | 0.263 | 1.03 (0.62–1.73) | 0.905 |
| 16 or more clinicians | 0.92 (0.68–1.24) | 0.577 | 0.90 (0.57–1.41) | 0.651 |
| Degree type | ||||
| MD | ref | ref | ||
| DO | 1.29 (0.84–1.97) | 0.241 | 1.30 (0.73–2.31) | 0.372 |
| Primary care physicianb | ||||
| No | ref | ref | ||
| Yes | 1.34 (1.03–1.75) | 0.031 | 1.04 (0.75–1.44) | 0.805 |
| Uses medical scribe | ||||
| No | ref | ref | ||
| Yes | 0.69 (0.47–1.00) | 0.050 | 0.91 (0.52–1.58) | 0.728 |
| Remote EHR use | ||||
| No | ref | ref | ||
| Yes | 1.32 (0.97–1.80) | 0.079 | 1.03 (0.61–1.75) | 0.905 |
| EHR adds to the frustration of my dayc | ||||
| Strongly disagree/Disagree | ref | ref | ||
| Strongly agree/Agree | 3.15 (2.41–4.12) | <.001 | 2.44 (1.60–3.74) | <.001 |
| Time spent on the EHR at homec | ||||
| Minimal/none/modest/satisfactory | ref | ref | ||
| Moderately high/excessive | 2.59 (2.07–3.24) | <.001 | 1.93 (1.36–2.75) | <.001 |
| Sufficiency of time for documentationc | ||||
| Sufficient | ref | ref | ||
| Insufficient | 3.83 (2.98–4.92) | <.001 | 2.81 (1.95–4.06) | <.001 |
EHR – electronic health record; OR—odds ratio; CI—confidence interval; ref – reference group.
aVariables in the adjusted model included age, gender, practice setting, practice size, PCP status, degree type, use of a medical scribe, remote EHR use, and the 3 health information technology-stress measures (EHR adding to daily frustration, time spent on the EHR at home, and sufficiency of time for documentation).
bSurvey respondents who replied “yes” to the question: Do you provide primary care?.
cWe performed a sensitivity analysis that included each HIT-related stress variable with its ordinal response categories (vs. the dichotomized response categories shown in the table above). For the variable EHR adds to the frustration of my day, with “strongly disagree” as the reference category, the OR (95%CI) for “disagree” was 1.57 (0.43–5.72), for “agree” was 2.58 (0.72–9.22), and for “strongly agree” was 5.38 (1.50–19.31). For the variable Time spent on the EHR at home, with “minimal/none” as the reference category, the OR (95%CI) for “modest” was 0.93 (0.48–1.80), for “satisfactory” was 0.95 (0.47–1.90), for “moderately high” was 1.38 (0.76–2.51), and for “excessive” was 1.95 (1.04–3.66). For the variable Sufficiency of time for documentation, with “optimal/good” as the reference category, the OR (95%CI) for “satisfactory” was 1.30 (0.68–2.49), for “marginal” was 2.27 (1.21–4.24), and for poor was 3.80 (1.95–7.40). The “optimal” and “good” response categories were combined due to a low number of responses in the “optimal” category. A table with complete results from this sensitivity analysis is available in the Appendix (Supplementary Appendix Table S3).