Yes (disease management program available at the management time of the VBID implementation for targeted program

Transcription

1 Choudhry NK, Ficher MA, Smith BF, Brill G, Girdih C, Matlin OS, et. Five feature of vue-baed inurance deign plan were aociated with higher rate of medication adherence. Heth Aff (Millwood). 2014;33(3). Publihed online February 12, Appendix A: Claification of VBID CHARACTERISTIC FEATURE HYPOTHESIZED TO BE ASSOCIATED WITH GREATER ADHERENCE Eligibility Rik-baed (only patient with high-rik criteria a condition received copayment reduction) Benefit Generou copayment (generic copayment = $0; quity b br-name copayment $10 or coinurance 15%) Tier No tier (i.e. copayment identic for generic l br-name medication) Dieae Ye (dieae management available at the management time of the VBID implementation for targeted dieae) offered c,d Wellne offered c,e Availability Ye (wellne available at the time of the VBID implementation) Reduced copayment available at mail only (v. mail + retail) VBID = vue-baed inurance deign a High-rik patient were thoe at an increaed rik of morbidity mortity baed on dieae pecific definition. For example, high-rik patient with hypertenion were patient with hypertenion who o had chronic ren failure or at leat one other vacular condition except primary hyperlipidemia. High-rik patient with hypercholeterolemia were thoe who had experienced a cardiovacular event. b We defined the threhold for plan generoity baed upon a quitative review of the copayment/coinurance tructure of VBID plan implemented by CVS Caremark between In enitivity anye, we ued an empiricly derived definition of benefit quity baed upon the oberved changed in copayment before after VBID implementation (ee text Appendix F for detail). c The impact of dieae management wellne wa aeed baed upon whether thee were offered, a information about the extent to which patient actuly engaged in them wa not obtainable. d Dieae management are intended to promote the ue of heth care intervention preventive care for chronic dieae through uch mechanim a targeted education mailing, nure coaching, periodic monitoring. e Wellne were defined a thoe which offered any of the following: workplace education preentation/lecture education mailing, onite employee heth clinic, heth fair, lifetyle improvement campaign (e.g. weight lo, moking ceation) /or heth club memberhip or onite exercie facilitie. 1

2 Appendix B: Study ample creation flow diagram Origin Sample n plan = 221 n plan ponor = 59 n patient = 812,769 Exclude plan with fewer than 12 month pre- or pot- implementation data n plan = 64 n patient = 241,024 Retrict anyi to diabete, hypertenion, high choleterol plan n plan = 54 n patient = 160,720 Exclude plan with average copay change different from plan decription n plan = 27 n patient = 136,471 Fin Sample n plan = 76 n plan ponor = 33 n patient = 274,554 2

3 Appendix C: Characteritic of plan patient included in the tudy ample tratified by the number of addition characteritic each plan had Diabete High choleterol Hypertenion Plan, n Patient, n Patient demographic Age, mean year Feme, % Income, median $ Black race, % Comorbidity, mean number of unique medication Copayment before VBID implementation Plan ponor characteritic Number of eligible beneficiarie Mot frequent indutry grouping Govern ment School ytem Gener Utiliti e Energy ervice Conume r product Gener ervice Conume r product Hopita l ervice Media Hopita l Gener School ytem TPA Conume r product Gener Governm ent ervice Nonblue Retail/ wholea le ervice Electri c applian ce Hopita l School ytem Gener Conume r product Gener Governm ent ervice Nonblue Retail/ wholea le ervice Pharmac eutic medic device

4 Appendix D: Unadjuted change in monthly adherence by plan characteritic for patient in the diabete cohort The olid black quare open circle repreent monthly adherence rate for plan with without pecific feature (a indicated in the legend). The gray diamond repreent the difference in the monthly rate 4

5 Appendix E: Relationhip between plan characteritic change in adherence following VBID plan implementation after diaggregating the categorization of how patient were targeted Plan Diabete High choleterol Hypertenion Characteritic Univariate Multivariate Univariate Multivariate Univariate Multivariate Generou 1.4 (0.9 to 1.8) 5.0 (4.3 to 5.8) 0.6 (0.3 to 1.0) 2.2 (1.5 to 2.8) 1.0 (0.8 to 1.2) 3.5 (2.9 to 4.0) No tier -3.3 (-3.8 to -2.8) -0.7 (-1.4 to -0.1) -0.2 (-0.5 to 0.2) -0.7 (-1.2 to -0.1) -3.0 (-3.3 to -2.7) 0.0 (-0.5 to 0.5) Target highrik patient Target patient who engaged in a deired behavior Wellne Dieae management Mail only benefit 1.7 (1.1 to 2.2) -0.8 (-1.3 to -0.3) 1.5 (1.0 to 2.1) -1.8 (-2.8 to -0.8) 1.4 (0.6 to 2.2) 2.0 (1.4 to 2.6) -2.2 (-3.0 to -1.4) 4.4 (3.6 to 5.3) -6.5 (-7.7 to -5.2) 1.2 (0.0 to 2.4) 1.4 (1.0 to 1.9) -0.3 (-0.7 to -0.0) 1.2 (0.8 to 1.6) -1.6 (-2.2 to -1.0) 1.5 (0.7 to 2.4) 1.1 (0.6 to 1.6) -0.6 (-1.2 to -0.0) 3.1 (2.5 to 3.7) -4.9 (-5.7 to -4.1) 2.8 (1.6 to 4.1) 1.5 (1.1 to 1.9) 0.3 (0.1 to 0.6) 2.0 (1.7 to 2.3) -1.2 (-1.6 to -0.7) 1.0 (0.5 to 1.4) 1.3 (0.8 to 1.7) -1.2 (-1.8 to -0.7) 5.3 (4.8 to 5.9) -5.0 (-5.6 to -4.4) 3.5 (2.5 to 4.5) Plan targeted patient in one of 3 way: high-rik patient only, patient to engage in ome behavior in order to quify for reduced copayment (for example, completing a heth rik aement)( compliance plan ), thoe that lowered copayment for l patient who were precribed a drug, regardle of indication, rik level or other behavior. In our primary anyi, we compared plan targeting high-rik patient to plan that did not. In thi enitivity anyi we compared high-rik plan compliance plan eparately to thoe that offered benefit to l patient. The number repreent the percentage point ( 95% confidence interv) change in adherence after VBID plan implementation for plan with without the given characteritic. The model are preented without adjutment for the other plan characteritic (univariate) after controlling for l of the other plan characteritic (multivariate). All model control for difference in patient demographic characteritic comorbidity. 5

6 Appendix F: Relationhip between plan generoity aeed baed on the oberved difference in copayment before after VBID implementation on change in adherence Condition Univariate Multivariable Diabete 3.3 (2.6 to 4.1) 5.5 (4.5 to 6.6) High Choleterol 2.4 (1.7 to 3.0) 2.4 (1.4 to 3.4) Hypertenion 5.3 (4.3 to 6.4) 5.3 (4.1 to 6.4) The number repreent the percentage point ( 95% confidence interv) change in adherence after VBID plan implementation for plan with without the given characteritic. The univariate model conider only the effect of generoity, adjuted for difference in patient demographic characteritic comorbidity. Heart failure could not be evuated becaue there were no plan that were claified a being generou. The multivariable model evuate the impact of generoity adjuting for the other plan characteritic, difference in patient demographic characteritic comorbidity. For thee model we excluded athma/copd a there wa no variation in ome of the other characteritic evuated (ee Table 1). Oberved difference in copayment before after VBID implementation were aeed by: (a) cculating patient cotharing a the um of actu patient paid amount divided by tot lowed amount for the drug targeted by the before after VBID implementation, (b) rank ordered plan by the ratio of pre-to-pot cot-haring (c) defining thoe plan in the top decile of change (i.e. the larget change) a being the mot generou. 6

7 Appendix G: Relationhip between plan characteritic change in adherence following VBID plan implementation for diabete plan in which dieae management /or wellne exited prior to the tart of the VBID Plan Characteritic Adjuted change in adherence after VBID implementation Generou 2.7 (1.8 to 3.6) No tier -2.7 (-3.6 to -1.7) Target high-rik patient 1.3 (0.3 to 2.3) Wellne 4.7 (3.4 to 6.0) Dieae management -6.2 (-7.7 to -4.7) Mail only benefit 2.8 (1.5 to 4.1) The number repreent the percentage point ( 95% confidence interv) change in adherence after VBID plan implementation for plan with without the given characteritic. The model are preented with without adjuting for the other plan characteritic. All model adjut for difference in patient demographic characteritic comorbidity. 7

8 Appendix H: Relationhip between plan feature change in adherence following VBID implementation uing a difference-indifference deign evuating each characteritic individuly Plan Characteritic Diabete High Choleterol Hypertenion Generou 1.4 (0.9 to 1.8) 0.7 (0.4 to 1.0) 1.0 (0.8 to 1.3) No tier -2.7 (-3.0 to -2.5) 0.1 (-0.3 to 0.4) -2.7 (-3.0 to -2.4) Target high-rik 1.7 (1.2 to 2.2) 1.5 (1.0 to 1.9) 1.5 (1.1 to 2.0) Wellne 1.2 (0.7 to 1.8) 0.9 (0.5 to 1.3) 1.7 (1.4 to 2.0) Dieae management -1.9 (-2.9 to -0.9) -1.7 (-2.4 to -1.1) -1.3 (-1.7 to -0.8) Mail only benefit 1.4 (0.6 to 2.2) 1.6 (0.7 to 2.4) 1.0 (0.6 to 1.5) The number repreent the percentage point ( 95% confidence interv) change in adherence after VBID plan implementation for plan with without the given characteritic, adjuting for difference in patient demographic characteritic comorbidity. Each plan characteritic wa evuated individuly for a given condition. 8

9 Appendix I: Relationhip between plan characteritic relative change in adherence following VBID plan implementation Plan Diabete High choleterol Hypertenion Characteritic Univariate Multivariate Univariate Multivariate Univariate Multivariate Generou 1.04 (1.03 to 1.05) 1.07 (1.05 to 1.09) 1.02 (1.01 to 1.04) 1.05 (1.03 to 1.07) 1.04 (1.03 to 1.05) 1.07 (1.05 to 1.09) No tier 0.96 (0.94 to 0.97) 1.04 (1.02 to 1.07) 1.04 (1.02 to 1.05) 1.01 (0.99 to 1.04) 0.96 (0.94 to 0.97) 1.04 (1.02 to 1.07) Target highrik patient 1.04 (1.02 to 1.06) 1.06 (1.03 to 1.08) 1.07 (1.05 to 1.10) 1.07 (1.05 to 1.10) 1.04 (1.02 to 1.06) 1.06 (1.03 to 1.08) Wellne 1.01 (1.00 to 1.03) 1.18 (1.15 to 1.22) 1.00 (0.98 to 1.02) 1.12 (1.09 to 1.15) 1.01 (1.00 to 1.03) 1.18 (1.15 to 1.22) Dieae management 0.91 (0.89 to 0.94) 0.83 (0.80 to 0.86) 0.88 (0.86 to 0.91) 0.8 (0.77 to 0.84) 0.91 (0.89 to 0.94) 0.83 (0.80 to 0.86) Mail only benefit 1.07 (1.05 to 1.10) 1.25 (1.20 to 1.31) 1.10 (1.06 to 1.14) 1.21 (1.15 to 1.27) 1.07 (1.05 to 1.10) 1.25 (1.20 to 1.31) The number are odd ratio ( 95% confidence interv) repreent relative change in the likelihood that patient are fully adherent after VBID plan implementation for plan with the given characteritic a compared to thoe without it. The model are preented without adjutment for the other plan characteritic (univariate) after controlling for l of the other plan characteritic (multivariate). All model control for difference in patient demographic characteritic comorbidity. An odd ratio > 1 mean that the particular plan characteritic i aociated with an increaed likelihood of patient being fully adherent after VBID implementation, a compared to plan without thi characteritic. Patient were conidered to be fully adherent if their PDC wa greater than or equ to 80%. For thee anye we fit generized etimating equation with binary ditributed error of a log link function. 9

10 Appendix J: Margin impact of plan characteritic on adherence following VBID implementation Plan characteritic evuated Generou Generou Target highrik No tier No tier Wellne Characteritic in control plan Dieae management Wellne No tier Dieae management Wellne Dieae management Wellne Dieae management Wellne Generou Dieae management Wellne Generou No tier Dieae management Diabete High Choleterol Hypertenion 1.7 (1.0 to 2.3) 2.3 (1.9 to 2.7) 2.6 (2.3 to 2.9) 8.1 (6.5 to 9.6) 5.0 (3.4 to 6.5) 4.0 (2.8 to 5.3) 2.1 (1.5 to 2.7) 2.4 (1.8 to 2.9) 1.5 (1.0 to 1.9) -0.1 (-1.2 to 1.0) 2.0 (1.0 to 2.9) 1.2 (0.4 to 2.0) 6.4 (5.1 to 7.7) 4.8 (3.5 to 6.0) 2.8 (1.7 to 3.8) 10.3 (8.8 to 11.7) 9.7 (8.4 to 11.0) 9.0 (7.9 to 10.1) Thi anyi evuated the margin impact of each plan characteritic by comparing plan with without the characteritic of interet. For example, the firt row compare plan that were generou had concomitant dieae management wellne to thoe that only had concomitant dieae management wellne. The number repreent the percentage point ( 95% confidence interv) change in adherence after VBID plan implementation for plan with compared to thoe without the given characteritic, adjuting for difference in patient demographic characteritic comorbidity. 10

11 Appendix K: Relationhip between plan characteritic change in adherence following VBID plan implementation for patient who had initiated treatment prior to the tart of the VBID Plan Diabete High choleterol Hypertenion Characteritic Univariate Multivariate Univariate Multivariate Univariate Multivariate Generou 1.3 (0.8 to 1.7) 3.6 (3.0 to 4.1) 0.6 (0.2 to 0.9) 1.6 (1.3 to 2.0) 0.9 (0.7 to 1.2) 2.4 (2.1 to 2.7) No tier -3.6 (-4.1 to -3.0) -1.5 (-2.2 to -0.8) -0.4 (-0.7 to 0.0) -0.7 (-1.3 to -0.1) -3.3 (-3.6 to -3.0) -0.2 (-0.7 to 0.3) Target highrik patient 1.7 (1.2 to 2.3) 2.2 (1.6 to 2.8) 1.4 (0.9 to 1.8) 1.3 (0.8 to 1.8) 1.5 (1.1 to 2.0) 1.3 (0.9 to 1.8) Wellne 1.7 (1.2 to 2.3) 4.6 (3.7 to 5.4) 1.3 (0.9 to 1.7) 3.3 (2.7 to 4.0) 2.2 (1.9 to 2.5) 5.5 (4.9 to 6.0) Dieae management -1.7 (-2.7 to -0.7) -6.7 (-8.0 to -5.4) -1.5 (-2.2 to -0.9) -5.1 (-5.9 to -4.3) -1.1 (-1.6 to -0.6) -5.2 (-5.8 to -4.6) Mail only benefit 1.4 (0.6 to 2.2) 2.6 (1.5 to 3.6) 1.4 (0.6 to 2.3) 3.5 (2.4 to 4.5) 1.0 (0.5 to 1.4) 4.7 (3.9 to 5.6) The number repreent the percentage point ( 95% confidence interv) change in adherence after VBID plan implementation for plan with without the given characteritic. The model are preented without adjutment for the other plan characteritic (univariate) after controlling for l of the other plan characteritic (multivariate). All model control for difference in patient demographic characteritic comorbidity. 11

12 Appendix L: Relationhip between plan characteritic change in adherence following VBID plan implementation for patient who were continuouly enrolled for the 12 month period before after the tart of the VBID Plan Diabete High choleterol Hypertenion Characteritic Univariate Multivariate Univariate Multivariate Univariate Multivariate Generou 1.2 (0.7 to 1.6) 1.8 (1.2 to 2.4) 0.8 (0.4 to 1.1) 1.2 (0.8 to 1.6) 0.8 (0.5 to 1.1) 1.6 (1.3 to 1.9) No tier -0.2 (-0.8 to 0.4) 0.3 (-0.4 to 1.0) 0.3 (-0.1 to 0.7) 0.2 (-0.4 to 0.8) -1.3 (-1.7 to -1.0) -0.1 (-0.6 to 0.5) Target highrik patient 0.3 (-0.3 to 0.9) 1.3 (0.6 to 1.9) 1.2 (0.7 to 1.7) 1.1 (0.6 to 1.6) 0.5 (0.0 to 0.9) 0.6 (0.2 to 1.1) Wellne -0.6 (-1.3 to 0.1) 0.6 (-0.4 to 1.7) 0.6 (0.1 to 1.0) 2.4 (1.7 to 3.1) 1.1 (0.7 to 1.4) 3.0 (2.4 to 3.7) Dieae management -0.7 (-1.7 to 0.4) -1.9 (-3.4 to -0.5) -0.8 (-1.5 to -0.2) -2.9 (-3.7 to -2.0) -0.4 (-0.9 to 0.1) -2.9 (-3.6 to -2.2) Mail only benefit 0.8 (-0.0 to 1.6) 0.1 (-1.1 to 1.3) 0.8 (-0.1 to 1.7) 2.4 (1.3 to 3.6) 0.2 (-0.3 to 0.7) 2.1 (1.1 to 3.0) The number repreent the percentage point ( 95% confidence interv) change in adherence after VBID plan implementation for plan with without the given characteritic. The model are preented without adjutment for the other plan characteritic (univariate) after controlling for l of the other plan characteritic (multivariate). All model control for difference in patient demographic characteritic comorbidity. 12