Table 2. Health System Engagement by Study Group.
| Time | Participants with any, % | Group differencea | ||
|---|---|---|---|---|
| LINKAGE group | Usual care group | RR (95% CI) | P value | |
| Substance use discussion with primary care practitionerb | ||||
| Year 1 | 89/186 (47.8) | 72/194 (37.1) | 1.30 (1.03-1.65) | .03 |
| Year 2 | 62/164 (37.8) | 56/166 (33.7) | 1.14 (0.85-1.53) | .39 |
| Year 3 | NA | NA | NA | NA |
| Year 4 | NA | NA | NA | NA |
| Year 5 | 24/122 (19.7) | 17/116 (14.7) | 1.33 (0.76-2.33) | .33 |
| All yearsc | 124/209 (59.3) | 109/217 (50.2) | 1.20 (0.99-1.45) | .07 |
| Annual electronic patient portal login | ||||
| Year 1 | 181/206 (87.9) | 144/221 (65.2) | 1.34 (1.20-1.50) | <.001 |
| Year 2 | 139/175 (79.4) | 123/182 (67.6) | 1.17 (1.03-1.33) | .01 |
| Year 3 | 127/164 (77.4) | 119/164 (72.6) | 1.06 (0.93-1.20) | .38 |
| Year 4 | 117/153 (76.5) | 105/144 (72.9) | 1.05 (0.92-1.19) | .51 |
| Year 5 | 120/152 (78.9) | 105/141 (74.5) | 1.07 (0.94-1.21) | .32 |
| All yearsc | 181/206 (87.9) | 144/221 (65.2) | 1.18 (1.08-1.29) | <.001 |
| Online medication refills | ||||
| Year 1 | 164/206 (79.6) | 100/221 (45.2) | 1.75 (1.49-2.06) | <.001 |
| Year 2 | 119/175 (68.0) | 97/182 (53.3) | 1.27 (1.07-1.51) | .01 |
| Year 3 | 105/164 (64.0) | 93/164 (56.7) | 1.13 (0.94-1.34) | .19 |
| Year 4 | 98/153 (64.1) | 83/144 (57.6) | 1.12 (0.93-1.34) | .24 |
| Year 5 | 98/152 (64.5) | 75/141 (53.2) | 1.22 (1.01-1.49) | .04 |
| All yearsc | 192/225 (85.3) | 158/230 (68.7) | 1.38 (1.22-1.56) | <.001 |
| Secure electronic messages sent to health care professional | ||||
| Year 1 | 144/206 (69.9) | 115/221 (52.0) | 1.34 (1.15-1.57) | <.001 |
| Year 2 | 116/175 (66.3) | 98/182 (53.8) | 1.24 (1.04-1.47) | .02 |
| Year 3 | 106/164 (64.6) | 92/164 (56.1) | 1.16 (0.97-1.38) | .10 |
| Year 4 | 98/153 (64.1) | 88/144 (61.1) | 1.05 (0.88-1.25) | .59 |
| Year 5 | 100/152 (65.8) | 85/141 (60.3) | 1.09 (0.91-1.30) | .35 |
| All yearsc | 186/225 (82.7) | 165/230 (71.7) | 1.18 (1.05-1.33) | .01 |
Abbreviations: ED, emergency department; NA, not applicable; RR, risk ratio.
a
Group differences were estimated from a modified Poisson regression model per follow-up time point, adjusting for baseline age, Charlson Comorbidity Index, and emergent-type ED use.
b
Telephone interviews were conducted at 1, 2, and 5 years after study enrollment.
c
Group differences were estimated from a modified Poisson generalized estimating equations model, adjusting for (linear) time and baseline age, Charlson Comorbidity Index, and emergent-type ED use.