Comparison of adherence and persistence among multiple sclerosis patients treated with disease-modifying therapies: a retrospective administrative claims analysis

Figures & data

Figure 1 Subject selection and attrition.

Table 1 Demographic and baseline clinical characteristics by index disease-modifying therapy

	IM-IFNβ-1a N = 2,305		IFNβ-1b N = 894		GA N = 2,270		SC-IFNβ-1a N = 1,211		P-valuesa
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Overall comparison	IM-IFNβ-1a vs IFNβ-1b	IM-IFNβ-1a vs GA	IM-IFNβ-1a vs SC-IFNβ-1a
Age (years)	42.2	10.1	42.4	10.5	42.6	9.9	41.1	10.2	<0.001	0.739	0.193	0.001
	N	%	N	%	N	%	N	%
Female	1,774	77.0	678	75.8	1,817	80.0	900	74.3	0.001	0.500	0.011	0.081
Charlson scoreb
0	1,853	80.4	688	77.0	1,748	77.0	913	75.4	0.029	0.081	0.041	0.005
1	255	11.1	106	11.9	304	13.4	170	14.0
2	142	6.2	76	8.5	160	7.1	99	8.2
≥3	55	2.4	24	2.7	58	2.6	29	2.4
MS-related utilizationc
≥1 ER visit	34	1.5	31	3.5	41	1.8	24	2.0	0.003	<0.001	0.378	0.262
≥1 inpatient stay	122	5.3	68	7.6	123	5.4	95	7.8	0.002	0.013	0.850	0.003
≥1 corticosteroid	799	34.7	311	34.8	738	32.5	507	41.9	<0.001	0.947	0.123	<0.001
Days observed for the index DMTd	743	322	741	312	738	307	692	319	<0.001	0.831	0.555	<0.001

Notes:

a overall comparison provides P-value associated with comparison over all cohorts; for paired comparisons, Bonferonni estimation-adjusted significance level is P < 0.017;

b measured in the 6-month baseline period;

c proportions of patients with at least one instance of designated utilization category measured in the 6-month baseline period;

d days observed for the index DMT = days from index date to earlier of end of follow-up period or switch to another DMT.

Abbreviations: IM-IFNβ-1a, intramuscular interferon beta-1a; IFNβ-1a, interferon beta-1b; GA, glatimer acetate; SC-IFNβ-1a, subcutaneous interferon beta-1a; SD, standard deviation; MS, multiple sclerosis; ER, emergency room; DMT, disease-modifying therapy.

Table 2 Unadjusted disease-modifying therapy measures

	IM-IFNβ-1a N = 2,305		IFNβ-1b N = 894		GA N = 2,270		SC-IFNβ-1a N = 1,211		P-valuesa
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Overall comparison	IM-IFNβ-1a vs IFNβ-1b	IM-IFNβ-1a vs GA	IM-IFNβ-1a vs SC-IFNβ-1a
MPR	0.77	0.30	0.70	0.32	0.72	0.31	0.74	0.31	<0.001	<0.001	<0.001	0.010
Persistence (days)b	508	374	482	380	471	373	471	358	0.003	0.075	0.001	0.005
	N	%	N	%	N	%	N	%
Adherent patients (MPR ≥ 0.80)	1,437	62.3	467	52.2	1,257	55.4	708	58.5	<0.001	<0.001	<0.001	0.025
Patients with 60-day therapy gap	1,240	53.8	491	54.9	1,229	54.1	662	54.7	0.930	0.566	0.815	0.623

Notes:

a Overall comparison provides P-value associated with comparison over all cohorts; for paired comparisons, Bonferonni estimation-adjusted significance level is P < 0.017;

b patients were considered persistent throughout follow-up if they had no 60-day gaps in therapy and the last observed first-line DMT fill was within 60 days of the end of follow-up.

Abbreviations: IM-IFNβ-1a, intramuscular interferon beta-1a; IFNβ-1a, interferon beta-1b; GA, glatimer acetate; SC-IFNβ-1a, subcutaneous interferon beta-1a; SD, standard deviation; MPR, medication persistence ratio.

Figure 2 Regression-adjusted odds ratios of adherence compared with IM-IFNβ-1a: all patients.

Abbreviation: IM-IFNβ-1a, intramuscular interferon beta-1a; IFNβ-1a, interferon beta-1b; GA, glatimer acetate; SC-IFNβ-1a, subcutaneous interferon beta-1a.

Table 3 Results of logistic regression modeling adherence: all patients

	Adherence (MPR ≥ 0.80)
	Odds ratio	95% confidence interval	P-value
IFNβ-1b	0.656	(0.561–0.768)	<0.001
GA	0.749	(0.665–0.844)	<0.001
SC-IFNβ-1a	0.829	(0.719–0.957)	0.011
Age	1.011	(1.006–1.016)	<0.001
Male	1.266	(1.125–1.426)	<0.001
Charlson comorbidity index score = 1	1.093	(0.939–1.273)	0.250
Charlson comorbidity index score = 2	1.081	(0.888–1.316)	0.436
Charlson comorbidity index score >2	0.790	(0.575–1.086)	0.147
Baseline MS-related inpatient stay	1.061	(0.853–1.319)	0.595
Baseline MS-related ER visit	0.757	(0.529–1.083)	0.128
Baseline corticosteroid	1.138	(1.024–1.265)	0.016
Time observed for the index DMT (months)	0.990	(0.986–0.995)	<0.001
Observations	6,680

Notes: Reference disease-modifying therapy = IM-IFNβ-1a; reference Charlson comorbidity index score = 0.

Abbreviations: IM-IFNβ-1a, intramuscular interferon beta-1a; IFNβ-1a, interferon beta-1b; GA, glatimer acetate; SC-IFNβ-1a, subcutaneous interferon beta-1a; MS, multiple sclerosis; ER, emergency room; DMT, disease-modifying therapy.

Figure 3 Kaplan–Meier failure curve of non-persistence.

Abbreviations: IM-IFNβ-1a, intramuscular interferon beta-1a; SC-IFNβ-1a, subcutaneous interferon beta-1a.

Table 4 Results of Cox proportional hazards regression modeling persistence: all patients

	Persistence defined with 60-day gap
	Hazard ratio	95% Confidence interval	P-value
IFNβ-1b	1.020	(0.912–1.141)	0.728
GA	1.022	(0.939–1.112)	0.612
SC-IFNβ-1a	1.117	(1.013–1.233)	0.027
Age	0.985	(0.981–0.988)	<0.001
Male	0.826	(0.759–0.899)	<0.001
Charlson comorbidity index score = 1	1.068	(0.961–1.188)	0.221
Charlson comorbidity index score = 2	1.119	(0.979–1.279)	0.098
Charlson comorbidity index score >2	1.489	(1.208–1.835)	<0.001
Baseline MS-related inpatient stay	0.920	(0.791–1.071)	0.282
Baseline MS-related ER visit	1.149	(0.896–1.473)	0.275
Baseline corticosteroid	1.005	(0.934–1.082)	0.884
Observations	5,880

Notes: Reference disease-modifying therapy = IM-IFNβ-1a; reference Charlson comorbidity index score = 0.

Abbreviations: IM-IFNβ-1a, intramuscular interferon beta-1a; IFNβ-1a, interferon beta-1b; GA, glatiramer acetate; SC-IFNβ-1a, subcutaneous interferon beta-1a; MS, multiple sclerosis; ER, emergency room.