Stress (PSS-10) (n=2422)* | Fatigue (Pichot) (n=2431)* | |||||
Estimates | CI | P value | Estimates | CI | P value | |
Fixed effects† | ||||||
Time | −0.00 | −0.16 to 0.16 | 0.957 | 0.48 | 0.31 to 0.66 | <0.001 |
Interviewer (ref=1 (for hospitals A and C)) | ||||||
2 (for hospitals B and D) | 1.62 | 0.85 to 2.40 | <0.001 | −0.36 | −1.47 to 0.76 | 0.531 |
HCW-level variables | ||||||
Gender (ref=female) | ||||||
Male | −0.83 | −1.73 to 0.06 | 0.068 | −1.52 | −2.63 to −0.40 | 0.008 |
Age (ref=<25) | ||||||
26–35 | 1.26 | 0.09 to 2.44 | 0.035 | |||
36–45 | 0.50 | −0.79 to 1.78 | 0.450 | |||
46–55 | −1.01 | −2.37 to 0.35 | 0.147 | |||
>55 | −2.83 | −4.49 to −1.17 | 0.001 | |||
Commuting time to work (ref=>1 hour) | ||||||
>2 hour | 1.82 | 0.80 to 2.83 | <0.001 | |||
1 hour–2 hours | 0.30 | −0.37 to 0.96 | 0.382 | |||
Personal life event (ref=no) | ||||||
Yes, negative | 2.27 | 1.81 to 2.74 | <0.001 | |||
Yes, positive | 0.46 | −0.19 to 1.12 | 0.164 | |||
Professional life event (ref=no) | ||||||
Yes, negative | 1.62 | 1.11 to 2.12 | <0.001 | |||
Yes, positive | 0.40 | −0.34 to 1.14 | 0.289 | |||
Break cancelled due to work overload (ref=never) | ||||||
Almost never | 1.04 | 0.27 to 1.82 | 0.008 | |||
Quite often | 1.54 | 0.77 to 2.32 | <0.001 | |||
Very often | 1.78 | 0.92 to 2.65 | <0.001 | |||
Professional status (ref=temporary CDD) | ||||||
Contract (contractuel CDI) | 2.23 | −0.53 to 4.99 | 0.114 | |||
Permanent (titulaire) | 0.45 | −1.04 to 1.93 | 0.554 | |||
Knowing work schedule in advance (ref=never) | ||||||
Sometimes | −2.30 | −4.81 to 0.22 | 0.074 | |||
Always | −1.87 | −4.23 to 0.49 | 0.121 | |||
Support from colleagues—Karasek subscale | −0.12 | −0.24 to −0.01 | 0.035 | |||
Support from supervisor—Karasek subscale | −0.23 | −0.33 to −0.13 | <0.001 | −0.34 | −0.45 to −0.24 | <0.001 |
HSOPSC item-1‡ (ref=agree) | ||||||
Neutral | 0.00 | −1.01 to 1.02 | 0.995 | 0.70 | −0.52 to 1.91 | 0.260 |
Not agree | 0.81 | −0.10 to 1.71 | 0.082 | 1.76 | 0.81 to 2.70 | <0.001 |
HSOPSC item-8§ (ref=agree) | ||||||
Neutral | 0.46 | −0.58 to 1.51 | 0.385 | |||
Not agree | 0.96 | 0.07 to 1.84 | 0.034 | |||
Work overcommitment—Siegrist subscale | 0.71 | 0.63 to 0.78 | <0.001 | 0.65 | 0.56 to 0.73 | <0.001 |
Sickness presenteeism (ref=never) | ||||||
Almost never | 0.04 | −0.48 to 0.57 | 0.871 | 0.63 | 0.04 to 1.22 | 0.037 |
Quite often | 1.04 | 0.45 to 1.63 | 0.001 | 2.48 | 1.82 to 3.15 | <0.001 |
Very often | 1.61 | 0.55 to 2.66 | 0.003 | 4.03 | 2.84 to 5.22 | <0.001 |
Ward-level variables | ||||||
Specialty (ref=surgery/obstetrics) | ||||||
Other medical specialties¶ | −0.36 | −1.30 to 0.59 | 0.459 | 1.59 | 0.24 to 2.94 | 0.021 |
ICU/reanimation | −1.08 | −2.01 to −0.15 | 0.023 | −0.24 | −1.60 to 1.13 | 0.733 |
Number of beds per ward | −0.03 | −0.05 to −0.01 | 0.009 | −0.04 | −0.07 to −0.00 | 0.024 |
Going outside the ward (ref=never) | ||||||
Sometimes | 1.28 | −0.87 to 3.43 | 0.243 | 0.98 | −2.09 to 4.04 | 0.533 |
Often | 1.53 | −0.70 to 3.76 | 0.178 | 0.86 | −2.30 to 4.03 | 0.592 |
Always | 1.87 | −0.44 to 4.18 | 0.112 | 2.59 | −0.70 to 5.89 | 0.123 |
Use of interim nurses (ref=never) | ||||||
Sometimes | 0.05 | −0.81 to 0.91 | 0.915 | −0.36 | −1.59 to 0.88 | 0.571 |
Often | 0.96 | −0.14 to 2.06 | 0.088 | 1.96 | 0.35 to 3.56 | 0.017 |
Random effects * | ||||||
Level 1—time | 17.80 (4.23) | 22.04 (4.69) | ||||
Level 2—healthcare worker | 12.30 (3.51) | 19.69 (4.44) | ||||
Level 3—ward | 0.00 (0.00) | 0.50 (0.70) | ||||
Marginal R²/Conditional R²** | 0.507/NA | 0.299/0.634 | ||||
AIC | 14 743.16 | 15 433.57 |
*The slightly different samples for each outcome refer to the difference between the factors included in the final model for each outcome, following the variable selection. Indeed, the number of missing values per variable has an impact on the number of observations that will be considered in the final mode.
†In multilevel models, fixed effects are usually equivalent to the regression coefficients, while random effects usually account for the underlying structure of the data and characterised by estimates of variability σ2(σ). Fixed effects can be interpreted as slopes in the traditional sense.
‡HSOPSC item-1: item-1 from the Hospital Survey on Patient Safety Culture. ‘Hospital management provides a work climate that promotes patient safety’.
§HSOPSC item-8: item-8 from the Hospital Survey on Patient Safety Culture. ‘The actions of hospital management show that patient safety is a top priority’.
¶The marginal R² considers only the variance of the fixed effects, while the conditional R² takes both the fixed and random effects into account.
AIC, Akaike information criterion; CDD, Fixed-term contract; CDI, Permanent employment contract; HCW, healthcare worker; ICC, intra-class correlation coefficient; ICU, intensive care unit; PSS-10, Perceived Stress Scale 10-item scale.