Evaluating Patient-Reported Adherence and Outcomes in Specialty Disease States: A Dual Site Initiative
1 Demographics
| Characteristic | N | Site 1, N = 33461 | Site 2, N = 3311 | Overall, N = 36771 |
|---|---|---|---|---|
| Age (at the time of first assessment in study period) | 3677 | |||
| Mean (SD) | 47.3 (18.2) | 50.0 (14.0) | 47.5 (17.9) | |
| Median (IQR) | 50.0 (37.0 - 61.0) | 50.0 (40.0 - 61.0) | 50.0 (37.0 - 61.0) | |
| Range | 0.0 - 90.0 | 16.0 - 81.0 | 0.0 - 90.0 | |
| Gender | 3677 | |||
| Female | 2374 (71.0%) | 264 (79.8%) | 2638 (71.7%) | |
| Male | 972 (29.0%) | 67 (20.2%) | 1039 (28.3%) | |
| Race | 3677 | |||
| White | 2719 (81.3%) | 61 (18.4%) | 2780 (75.6%) | |
| Black or African American | 304 (9.09%) | 167 (50.5%) | 471 (12.8%) | |
| American Indian or Alaska Native | 12 (0.36%) | 2 (0.60%) | 14 (0.38%) | |
| Other Asian | 27 (0.81%) | 7 (2.11%) | 34 (0.92%) | |
| Other | 27 (0.81%) | 89 (26.9%) | 116 (3.15%) | |
| Decline to Answer | 6 (0.18%) | 1 (0.30%) | 7 (0.19%) | |
| Unknown | 77 (2.30%) | 4 (1.21%) | 81 (2.20%) | |
| Chinese | 2 (0.06%) | 0 (0%) | 2 (0.05%) | |
| Filipino | 1 (0.03%) | 0 (0%) | 1 (0.03%) | |
| Japanese | 1 (0.03%) | 0 (0%) | 1 (0.03%) | |
| Korean | 2 (0.06%) | 0 (0%) | 2 (0.05%) | |
| Asian Indian | 4 (0.12%) | 0 (0%) | 4 (0.11%) | |
| Other Pacific Islander | 5 (0.15%) | 0 (0%) | 5 (0.14%) | |
| NULL | 159 (4.75%) | 0 (0%) | 159 (4.32%) | |
| Ethnicity | 3677 | |||
| Not Hispanic, Latino/a, or Spanish origin | 2967 (88.7%) | 243 (73.4%) | 3210 (87.3%) | |
| Hispanic or Latino | 1 (0.03%) | 79 (23.9%) | 80 (2.18%) | |
| Unknown | 35 (1.05%) | 3 (0.91%) | 38 (1.03%) | |
| Decline to Answer | 11 (0.33%) | 6 (1.81%) | 17 (0.46%) | |
| Mexican, Mexican American, or Chicano/a | 9 (0.27%) | 0 (0%) | 9 (0.24%) | |
| Puerto Rican | 4 (0.12%) | 0 (0%) | 4 (0.11%) | |
| Cuban | 1 (0.03%) | 0 (0%) | 1 (0.03%) | |
| Other Hispanic, Latino/a, or Spanish origin | 80 (2.39%) | 0 (0%) | 80 (2.18%) | |
| NULL | 238 (7.11%) | 0 (0%) | 238 (6.47%) | |
| 1 n (%) | ||||
| Characteristic | N | Site 1, N = 33461 | Site 2, N = 3311 | Overall, N = 36771 |
|---|---|---|---|---|
| assess_clinic | 3677 | |||
| MS | 798 (23.8%) | 82 (24.8%) | 880 (23.9%) | |
| Rheumatology | 2546 (76.1%) | 249 (75.2%) | 2795 (76.0%) | |
| Rheumatology, MS | 2 (0.06%) | 0 (0%) | 2 (0.05%) | |
| 1 n (%) | ||||
1.1 Number of Monthly Medication Assessments
There are approximately 30 months of study time.
| Characteristic | N | N = 3677 |
|---|---|---|
| n | 3677 | |
| Mean (SD) | 16.8 (10.6) | |
| Median (IQR) | 16.0 (7.0 - 27.0) | |
| Range | 1.0 - 63.0 |
1.1.1 Medications assessed by clinic type
| Characteristic | N | Rheumatology, N = 454201 | MS, N = 165061 | Overall, N = 619261 |
|---|---|---|---|---|
| Medication prescribed (radio) | 61926 | |||
| Cimzia (certolizumab) | 787 (1.73%) | 0 (0%) | 787 (1.27%) | |
| Humira (adalimumab) | 16643 (36.6%) | 17 (0.10%) | 16660 (26.9%) | |
| Orencia (abatacept) | 2915 (6.42%) | 0 (0%) | 2915 (4.71%) | |
| Rasuvo (methotrexate) | 0 (0%) | 0 (0%) | 0 (0%) | |
| Simponi (golimumab) | 641 (1.41%) | 0 (0%) | 641 (1.04%) | |
| Stelara (ustekinumab) | 151 (0.33%) | 0 (0%) | 151 (0.24%) | |
| Xeljanz/Xeljanz XR (tofacitinib) | 4408 (9.70%) | 0 (0%) | 4408 (7.12%) | |
| Actemra (tocilizumab) | 1673 (3.68%) | 0 (0%) | 1673 (2.70%) | |
| Benlysta (belimumab) | 1107 (2.44%) | 0 (0%) | 1107 (1.79%) | |
| Cosentyx (secukinumab) | 1879 (4.14%) | 0 (0%) | 1879 (3.03%) | |
| Enbrel (etanercept) | 11060 (24.4%) | 0 (0%) | 11060 (17.9%) | |
| Kevzara (sarilumab) | 303 (0.67%) | 0 (0%) | 303 (0.49%) | |
| Kineret (anakinra) | 869 (1.91%) | 7 (0.04%) | 876 (1.41%) | |
| Olumiant (baricitinib) | 120 (0.26%) | 0 (0%) | 120 (0.19%) | |
| Otezla (apremilast) | 1101 (2.42%) | 0 (0%) | 1101 (1.78%) | |
| Rinvoq (upadacitinib) | 1082 (2.38%) | 0 (0%) | 1082 (1.75%) | |
| Taltz (ixekizumab) | 681 (1.50%) | 0 (0%) | 681 (1.10%) | |
| Ampyra/XR (dalfampridine) | 0 (0%) | 1826 (11.1%) | 1826 (2.95%) | |
| Aubagio (teriflunomide) | 0 (0%) | 1331 (8.06%) | 1331 (2.15%) | |
| Avonex (interferon beta-1a) | 0 (0%) | 996 (6.03%) | 996 (1.61%) | |
| Betaseron (interferon beta-1b) | 0 (0%) | 427 (2.59%) | 427 (0.69%) | |
| Copaxone (glatiramer acetate) | 0 (0%) | 1391 (8.43%) | 1391 (2.25%) | |
| Extavia (interferon beta-1b) | 0 (0%) | 54 (0.33%) | 54 (0.09%) | |
| Gilenya (fingolimod) | 0 (0%) | 3065 (18.6%) | 3065 (4.95%) | |
| Glatopa (glatiramer acetate) | 0 (0%) | 69 (0.42%) | 69 (0.11%) | |
| Mayzent (siponimod) | 0 (0%) | 30 (0.18%) | 30 (0.05%) | |
| Plegridy (peginterferon beta-1a) | 0 (0%) | 367 (2.22%) | 367 (0.59%) | |
| Rebif (interferon beta-1a) | 0 (0%) | 1522 (9.22%) | 1522 (2.46%) | |
| Tecfidera (dimethyl fumarate) | 0 (0%) | 3373 (20.4%) | 3373 (5.45%) | |
| glatiramer acetate | 0 (0%) | 1561 (9.46%) | 1561 (2.52%) | |
| Otrexup (methotrexate) | 0 (0%) | 0 (0%) | 0 (0%) | |
| Vumerity (diroximel fumarate) | 0 (0%) | 369 (2.24%) | 369 (0.60%) | |
| Bafiertam (monomethyl fumarate) | 0 (0%) | 45 (0.27%) | 45 (0.07%) | |
| Zeposia (ozanimod) | 0 (0%) | 56 (0.34%) | 56 (0.09%) | |
| 1 n (%) | ||||
1.1.2 Number of patient monthly medication assessments by medication
1.2 Number of patient reported outcomes (PROs)
| Characteristic | N | N = 36771 |
|---|---|---|
| any.missed.doses | 3677 | |
| No | 2418 (65.8%) | |
| Yes | 1259 (34.2%) | |
| any.tolerability | 3660 | |
| No | 3188 (87.1%) | |
| Yes | 472 (12.9%) | |
| Missing | 17 | |
| any.hosp_er | 3677 | |
| No | 3156 (85.8%) | |
| Yes | 521 (14.2%) | |
| any.fair.poor | 3677 | |
| No | 2621 (71.3%) | |
| Yes | 1056 (28.7%) | |
| any.excellent | 3677 | |
| No | 955 (26.0%) | |
| Yes | 2722 (74.0%) | |
| 1 n (%) | ||
2 Monthly Medication Assessments
2.1 Site
| Characteristic | N | N = 619261 |
|---|---|---|
| Research site | 61926 | |
| Site 1 | 57162 (92.3%) | |
| Site 2 | 4764 (7.69%) | |
| 1 n (%) | ||
2.2 Clinic and medications
| Characteristic | N | Site 1, N = 571621 | Site 2, N = 47641 | Overall, N = 619261 |
|---|---|---|---|---|
| Clinic medication prescribed through | 61926 | |||
| Rheumatology | 42139 (73.7%) | 3281 (68.9%) | 45420 (73.3%) | |
| MS | 15023 (26.3%) | 1483 (31.1%) | 16506 (26.7%) | |
| Medication prescribed (radio) | 61926 | |||
| Cimzia (certolizumab) | 754 (1.32%) | 33 (0.69%) | 787 (1.27%) | |
| Humira (adalimumab) | 14912 (26.1%) | 1748 (36.7%) | 16660 (26.9%) | |
| Orencia (abatacept) | 2742 (4.80%) | 173 (3.63%) | 2915 (4.71%) | |
| Rasuvo (methotrexate) | 0 (0%) | 0 (0%) | 0 (0%) | |
| Simponi (golimumab) | 595 (1.04%) | 46 (0.97%) | 641 (1.04%) | |
| Stelara (ustekinumab) | 150 (0.26%) | 1 (0.02%) | 151 (0.24%) | |
| Xeljanz/Xeljanz XR (tofacitinib) | 4202 (7.35%) | 206 (4.32%) | 4408 (7.12%) | |
| Actemra (tocilizumab) | 1634 (2.86%) | 39 (0.82%) | 1673 (2.70%) | |
| Benlysta (belimumab) | 1086 (1.90%) | 21 (0.44%) | 1107 (1.79%) | |
| Cosentyx (secukinumab) | 1794 (3.14%) | 85 (1.78%) | 1879 (3.03%) | |
| Enbrel (etanercept) | 10306 (18.0%) | 754 (15.8%) | 11060 (17.9%) | |
| Kevzara (sarilumab) | 289 (0.51%) | 14 (0.29%) | 303 (0.49%) | |
| Kineret (anakinra) | 876 (1.53%) | 0 (0%) | 876 (1.41%) | |
| Olumiant (baricitinib) | 113 (0.20%) | 7 (0.15%) | 120 (0.19%) | |
| Otezla (apremilast) | 1005 (1.76%) | 96 (2.02%) | 1101 (1.78%) | |
| Rinvoq (upadacitinib) | 1056 (1.85%) | 26 (0.55%) | 1082 (1.75%) | |
| Taltz (ixekizumab) | 649 (1.14%) | 32 (0.67%) | 681 (1.10%) | |
| Ampyra/XR (dalfampridine) | 1826 (3.19%) | 0 (0%) | 1826 (2.95%) | |
| Aubagio (teriflunomide) | 1243 (2.17%) | 88 (1.85%) | 1331 (2.15%) | |
| Avonex (interferon beta-1a) | 936 (1.64%) | 60 (1.26%) | 996 (1.61%) | |
| Betaseron (interferon beta-1b) | 335 (0.59%) | 92 (1.93%) | 427 (0.69%) | |
| Copaxone (glatiramer acetate) | 1289 (2.25%) | 102 (2.14%) | 1391 (2.25%) | |
| Extavia (interferon beta-1b) | 54 (0.09%) | 0 (0%) | 54 (0.09%) | |
| Gilenya (fingolimod) | 2875 (5.03%) | 190 (3.99%) | 3065 (4.95%) | |
| Glatopa (glatiramer acetate) | 69 (0.12%) | 0 (0%) | 69 (0.11%) | |
| Mayzent (siponimod) | 30 (0.05%) | 0 (0%) | 30 (0.05%) | |
| Plegridy (peginterferon beta-1a) | 367 (0.64%) | 0 (0%) | 367 (0.59%) | |
| Rebif (interferon beta-1a) | 1391 (2.43%) | 131 (2.75%) | 1522 (2.46%) | |
| Tecfidera (dimethyl fumarate) | 2783 (4.87%) | 590 (12.4%) | 3373 (5.45%) | |
| glatiramer acetate | 1341 (2.35%) | 220 (4.62%) | 1561 (2.52%) | |
| Otrexup (methotrexate) | 0 (0%) | 0 (0%) | 0 (0%) | |
| Vumerity (diroximel fumarate) | 369 (0.65%) | 0 (0%) | 369 (0.60%) | |
| Bafiertam (monomethyl fumarate) | 45 (0.08%) | 0 (0%) | 45 (0.07%) | |
| Zeposia (ozanimod) | 46 (0.08%) | 10 (0.21%) | 56 (0.09%) | |
| 1 n (%) | ||||
2.3 Missed doses
2.3.1 Summary
| Characteristic | N | Site 1, N = 571621 | Site 2, N = 47641 | Overall, N = 619261 |
|---|---|---|---|---|
| Have you missed doses of your medication this month? | 61926 | |||
| Yes | 1861 (3.26%) | 621 (13.0%) | 2482 (4.01%) | |
| No | 55300 (96.7%) | 3890 (81.7%) | 59190 (95.6%) | |
| Data not available (NULL) | 1 (<0.01%) | 253 (5.31%) | 254 (0.41%) | |
| 1 n (%) | ||||
2.3.2 Reasons
2.3.3 Figures
2.3.4 Intersections
2.4 Tolerability
| Characteristic | N | Site 1, N = 571621 | Site 2, N = 47641 | Overall, N = 619261 |
|---|---|---|---|---|
| Site 1: Are you experiencing any side effects? Site 2: Have you had any problems tolerating the medication? | 61835 | |||
| Yes | 519 (0.91%) | 170 (3.57%) | 689 (1.11%) | |
| No | 56552 (99.1%) | 4340 (91.1%) | 60892 (98.5%) | |
| Data not available (NULL) | 0 (0%) | 254 (5.33%) | 254 (0.41%) | |
| Missing | 91 | 0 | 91 | |
| 1 n (%) | ||||
2.5 Effectiveness / hospitalizations
| Characteristic | N | Site 1, N = 571621 | Site 2, N = 47641 | Overall, N = 619261 |
|---|---|---|---|---|
| Since we last spoke for your refill, have you been hospitalized or had to visit the ER or ED for any reason? | 61926 | |||
| Yes | 567 (0.99%) | 191 (4.01%) | 758 (1.22%) | |
| No | 55673 (97.4%) | 4318 (90.6%) | 59991 (96.9%) | |
| Data not available (NULL) | 922 (1.61%) | 255 (5.35%) | 1177 (1.90%) | |
| How well do you think your medication is working for you? | 61926 | |||
| Excellent | 13055 (22.8%) | 1230 (25.8%) | 14285 (23.1%) | |
| Good | 41409 (72.4%) | 2422 (50.8%) | 43831 (70.8%) | |
| Fair | 1608 (2.81%) | 793 (16.6%) | 2401 (3.88%) | |
| Poor | 122 (0.21%) | 34 (0.71%) | 156 (0.25%) | |
| NULL/Data not available | 968 (1.69%) | 285 (5.98%) | 1253 (2.02%) | |
| 1 n (%) | ||||
2.6 Missed doses and tolerability
| Characteristic | N | Missed doses | p-value2 | Overall, N = 615781 | |
|---|---|---|---|---|---|
| Yes, N = 24781 | No, N = 591001 | ||||
| Site 1: Are you experiencing any side effects? Site 2: Have you had any problems tolerating the medication? | 61578 | <0.001 | |||
| Yes | 102 (4.12%) | 587 (0.99%) | 689 (1.12%) | ||
| No | 2376 (95.9%) | 58513 (99.0%) | 60889 (98.9%) | ||
| 1 n (%) | |||||
| 2 Pearson's Chi-squared test | |||||
2.7 Missed doses and effectiveness
| Characteristic | N | Missed doses | p-value2 | Overall, N = 606731 | |
|---|---|---|---|---|---|
| Yes, N = 24331 | No, N = 582401 | ||||
| How well do you think your medication is working for you? | 60673 | <0.001 | |||
| Excellent | 479 (19.7%) | 13806 (23.7%) | 14285 (23.5%) | ||
| Good | 1711 (70.3%) | 42120 (72.3%) | 43831 (72.2%) | ||
| Fair | 234 (9.62%) | 2167 (3.72%) | 2401 (3.96%) | ||
| Poor | 9 (0.37%) | 147 (0.25%) | 156 (0.26%) | ||
| 1 n (%) | |||||
| 2 Pearson's Chi-squared test | |||||
2.8 Missed doses and hosp/er
| Characteristic | N | Missed doses | p-value2 | Overall, N = 607461 | |
|---|---|---|---|---|---|
| Yes, N = 24401 | No, N = 583061 | ||||
| Since we last spoke for your refill, have you been hospitalized or had to visit the ER or ED for any reason? | 60746 | <0.001 | |||
| Yes | 200 (8.20%) | 558 (0.96%) | 758 (1.25%) | ||
| No | 2240 (91.8%) | 57748 (99.0%) | 59988 (98.8%) | ||
| 1 n (%) | |||||
| 2 Pearson's Chi-squared test | |||||
2.8.1 Missingness
| Characteristic | N | Missing missed_doses | Overall, N = 619261 | |
|---|---|---|---|---|
| No, N = 616721 | Yes, N = 2541 | |||
| missing.hosp_er | 61926 | |||
| No | 60746 (98.5%) | 3 (1.18%) | 60749 (98.1%) | |
| Yes | 926 (1.50%) | 251 (98.8%) | 1177 (1.90%) | |
| 1 n (%) | ||||
3 Analyses
3.1 Missed doses
To test for associations between the missed dose Monthly Medication Assessments (MMA) outcome, we used mixed effects logistic regression model, with patient ID as a random effect. Included in the model were age (modeled as a nonlinear effect), gender, race, clinic and site. We also included the other MMA outcomes in the model (tolerability, hospitalization and effectiveness) to assess correlation among the MMA outcomes. A small fraction of race was missing (~6.5%). As that was the only variable with missing data, we imputed the values to the most frequent category, conditioned on the study site (i.e. Site 1 missing race imputed to White, and Site 2 missing race imputed to Black).
We found strong evidence of a nonlinear association between age and reported missed doses (p < 0.001), with the highest likelihood of a missed dose occurring at roughly 40 years. Males were 28% less likely than females to report a missed dose at any MMA (OR = 0.72, 95% CI 0.61 - 0.85, p <0.001). The additional MMA outcomes were highly associated with a reported missed dose as well, with people who reported a hospitalization being 8.4 times more likely to report a missed dose than those who did not report one (OR = 8.42, 95% CI 6.82 - 10.39, p < 0.001). The odds ratio of reporting a missed dose increased to 2.46 when tolerability issues were reported (95% CI 1.87 - 3.23, p <0.001). An effectiveness rating of ‘Fair’ was associated with a 61% increase in the odds of a missed dose compared to a rating of “Good/ Excellent” (OR 1.61, 95% CI 1.33 - 1.94). We also found evidence of a very strong site effect (p < 0.001), likely due to the way the MMAs are administered across sites.
3.1.1 Results
The predicted effects in the following plots are based off of an incorrectly specified model. These are for illustrative purposes only
3.2 Tolerability
3.2.1 Results
The predicted effects in the following plots are based off of an incorrectly specified model. These are for illustrative purposes only
3.3 Hospitalization
3.3.1 Results
The predicted effects in the following plots are based off of an incorrectly specified model. These are for illustrative purposes only
3.4 Effectiveness
3.4.1 Results
Test for significance of non-linear age effect
## Likelihood ratio tests of cumulative link models
##
## Response: effect2
## Model
## 1 gender + race.imp + age + assess_clinic + site + missed_doses + tolerability + hosp_er | |
## 2 gender + race.imp + rcs(age, 5) + assess_clinic + site + missed_doses + tolerability + hosp_er | |
## Resid. df -2logLik Test Df LR stat. Pr(Chi)
## 1 60569 17630.29
## 2 60566 17604.04 1 vs 2 3 26.24889 8.458856e-06The predicted effects in the following plots are based off of an incorrectly specified model. These are for illustrative purposes only
