Table 3.
Results of Controlled Interrupted Time Series Analyses of the Impact of a Visiting Ban on Loneliness Among Swedish Nursing Home Residents
| Effect | Model 0a | Model 1b | Model 2c |
|---|---|---|---|
| Year (2020 vs 2019) | 1.266 (1.078; 1.486) | 1.234 (1.053; 1.446) | 1.070 (0.919; 1.245) |
| Period (post vs pretest) | 1.096 (0.908; 1.323) | 1.048 (0.869; 1.265) | 0.968 (0.809; 1.158) |
| Time (date, 2-day intervals) | 0.994 (0.978; 1.012) | 0.995 (0.978; 1.012) | 0.998 (0.982; 1.014) |
| Year × Period | 1.236 (0.886; 1.725) | 1.223 (0.878; 1.702) | 1.251 (0.913; 1.715) |
| Year × Time | 1.029 (1.007; 1.052) | 1.027 (1.006; 1.049) | 1.016 (0.996; 1.037) |
| Period × Time | 1.009 (0.989; 1.029) | 1.010 (0.99; 1.030) | 1.010 (0.991; 1.029) |
| Year × Period × Time | 0.974 (0.949; 0.998) | 0.975 (0.951; 0.999) | 0.984 (0.961; 1.008) |
Notes: N = 23,564 observations. Numbers are risk ratios with 95% confidence intervals.
aModel 0: one unadjusted model only including the design effects.
bModel 1: adjusted for living alone, area of residence, provider, questionnaire form, questionnaire assistance, age, and sex.
cModel 2: additionally adjusted for self-rated general health, mental health, and mobility limitations.