Table 1.
Results of logistic regression models fitted to the outcomes: used the patient portal during Harvey (Model 1) and messaged during Harvey (Model 2).
| Independent variables | Model 1: used patient portal during Harvey | Model 2: messaged during Harvey | |||
|
|
ORa (95% CI) | P value | OR (95% CI) | P value | |
| Used portal before Harvey in 2017 (reference: no portal use in before Harvey in 2017) | |||||
|
|
Yes | 13.688 (12.929-14.491) | <.001 | 14.172 (11.879-16.907) | <.001 |
| Age (years; reference: 18-44 years) | |||||
|
|
45-64 | 0.949 (0.921-0.978) | .001 | 1.168 (1.081-1.262) | <.001 |
|
|
≥65 | 1.128 (1.091-1.166) | <.001 | 1.298 (1.192-1.413) | <.001 |
| Sex (reference: female) | |||||
|
|
Male | 0.963 (0.938-0.989) | .005 | 0.899 (0.841-0.962) | .002 |
| Race (reference: White) | |||||
|
|
Black | 0.977 (0.940-1.017) | .26 | 0.859 (0.775-0.952) | .004 |
|
|
Asian | 0.987 (0.939-1.038) | .61 | 0.793 (0.690-0.912) | .001 |
|
|
Other | 1.097 (1.031-1.168) | .004 | 1.081 (0.919-1.272) | .34 |
|
|
Unknown | 0.990 (0.929-1.056) | .76 | 1.015 (0.859-1.198) | .86 |
| Ethnicity (reference: non-Hispanic/Latinx) | |||||
|
|
Hispanic/Latinx | 0.860 (0.823-0.898) | <.001 | 0.854 (0.763-0.955) | .006 |
|
|
Unknown | 0.978 (0.900-1.061) | .59 | 1.008 (0.814-1.248) | .94 |
| Outpatient (reference: no encounter in 2017) | |||||
|
|
Encounter in 2017 outside Harvey periodb | 0.898 (0.872-0.925) | <.001 | 1.037 (0.961-1.120) | .35 |
|
|
Encounter peri-Harveyb | 2.052 (1.973-2.134) | <.001 | 2.465 (2.260-2.689) | <.001 |
|
|
Encounter during Harveyb | 5.897 (5.094-6.826) | <.001 | 5.009 (3.944-6.362) | <.001 |
| Surgery (reference: no encounter in 2017) | |||||
|
|
Encounter in 2017 outside Harvey periodb | 0.953 (0.912-0.996) | .03 | 0.989 (0.890-1.100) | .85 |
|
|
Encounter peri-Harveyb | 1.810 (1.681-1.949) | <.001 | 1.762 (1.514-2.050) | <.001 |
|
|
Encounter during Harveyb | 4.337 (3.106-6.057) | <.001 | 0.583 (0.184-1.843) | .36 |
| Observations (reference: no encounter in 2017) | |||||
|
|
Encounter in 2017 outside Harvey periodb | 0.861 (0.810-0.915) | <.001 | 0.931 (0.807-1.075) | .33 |
|
|
Encounter peri-Harveyb | 1.608 (1.444-1.790) | <.001 | 1.444 (1.156-1.802) | <.001 |
|
|
Encounter during Harveyb | 4.697 (3.366-6.554) | <.001 | 1.201 (0.549-2.628) | .65 |
| Emergency department (reference: no encounter in 2017) | |||||
|
|
Encounter in 2017 outside Harvey periodb | 0.722 (0.690-0.755) | <.001 | 0.900 (0.809-1.002) | .06 |
|
|
Encounter peri-Harveyb | 1.085 (1.001-1.175) | .046 | 1.407 (1.188-1.667) | <.001 |
|
|
Encounter during Harveyb | 2.917 (2.340-3.635) | <.001 | 2.317 (1.505-3.565) | <.001 |
| Inpatient (reference: no encounter in 2017) | |||||
|
|
Encounter in 2017 outside Harvey periodb | 0.903 (0.861-0.947) | <.001 | 1.039 (0.930-1.161) | .50 |
|
|
Encounter peri-Harveyb | 1.673 (1.551-1.804) | <.001 | 1.882 (1.619-2.186) | <.001 |
|
|
Encounter during Harveyb | 2.743 (2.307-3.261) | <.001 | 2.088 (1.498-2.909) | <.001 |
aOR: odds ratio.
bOutside Harvey period defined as the periods January 1-July 27, 2017, and September 29-December 31, 2017; peri-Harvey period defined as the 4-week periods both immediately before Harvey (July 28-August 24, 2017) and immediately after Harvey (September 1-September 28, 2017); and during Harvey period defined as August 25-August 31, 2017.