Decomposing the effect of drug benefit program on antihypertensive medication adherence among the elderly in urban China

Figures & data

Table 1 Characteristics of study population^a

Variables	Total (n=1,969)	Enrollees (n=586)	Non-enrollees (n=1,383)	P
Medication adherence				0.001
No	291 (14.8%)	62 (10.6%)	229 (16.6%)
Predisposing factors
Gender				0.001
Female	1181 (60.0%)	318 (54.3%)	863 (62.4%)
Age^b	72.26 (7.69)	72.80 (7.05)	72.04 (7.94)	0.021
Education				0.001
Illiterate	757 (38.4%)	189 (32.3%)	568 (41.1%)
Primary school	644 (32.7%)	206 (35.2%)	438 (31.7%)
Middle school or above	568 (28.8%)	191 (32.6%)	377 (27.3%)
Enabling factors
Annual household income (yuan)				<0.001
<48,000	618 (31.4%)	131 (22.4%)	487 (35.2%)
48,000–68,400	682 (34.6%)	200 (34.1%)	482 (34.9%)
≥68,400	669 (34.0%)	255 (43.5%)	414 (29.9%)
Urbanity				0.853
Urban	599 (30.4%)	180 (30.7%)	419 (30.3%)
Frequency of clinic check-ups (times over the past 12 months)				<0.001
0	393 (20.0%)	82 (14.0%)	311 (22.5%)
1–2	557 (28.3%)	158 (27.0%)	399 (28.9%)
≥3	1019 (51.8%)	346 (59.0%)	673 (48.7%)
Need
Diabetic status				0.011
Yes	615 (31.2%)	207 (35.3%)	408 (29.5%)
Duration of hypertension (years)				<0.001
1–5	305 (15.5%)	44 (7.5%)	261 (18.9%)
5–10	461 (23.4%)	111 (18.9%)	350 (25.3%)
≥10	1203 (61.1%)	431 (73.5%)	772 (55.8%)
Blood pressure control				0.106
Controlled	786 (39.9%)	250 (42.7%)	536 (38.8%)

Notes: ^aUnless indicated, numbers presented are n, %. ^bNumbers presented are mean, SD.

Figure 1 OR for medication nonadherence by drug benefit program enrollment status.

Notes: The x-axis refers to magnitude of the OR. Black horizontal lines reflect 95% CIs.

Figure 2 Contribution of drug benefit program effects to antihypertensive medication nonadherence.

Notes: Other observed characteristics adjusted for in the decomposition analysis were gender, age, education, annual household income, urbanity, other frequency of clinic check-ups, and other duration of hypertension.

Table 2 Blinder-Oaxaca nonlinear decomposition of the differences in medication nonadherence between drug benefit program enrollees and non-enrollees

	Non-enrollees	Enrollees	Difference (non-enrollees –- enrollees)	Difference attributed to differences in program effect (unexplained component)	Difference attributed to differences in observed characteristics (explained component)
Probability of nonadherence	0.17	0.11	0.06	0.037	0.023
P-value	—	—	<0.001	0.024	<0.001
% of contribution	—	—	—	60.90%	39.10%

Figure 3 Contribution of observable characteristics to antihypertensive medication nonadherence.

Notes: Other observed characteristics adjusted for in the decomposition analysis were gender, age, education, annual household income, urbanity, other frequency of clinic check-ups, and other duration of hypertension.

Table S1 Logistic regression analysis between drug benefit program enrollees and non-enrollees

	Enrollees		Non-enrollees
	OR value	95% CI	OR value	95% CI
Predisposing factors
Gender (reference: male)
Female	0.937	0.523,1.678	0.917	0.665,1.266
Age	0.978	0.940,1.018	0.999	0.981,1.019
Education (reference: illiterate)
Primary school	1.110	0.566,2.178	1.088	0.759,1.560
Middle school or above	0.972	0.456,2.071	1.272	0.846,1.914
Enabling factors
Annual household income (reference: <48,000)
48,000–68,400	0.782	0.383,1.595	0.697	0.482,1.006
≥68,400	0.887	0.442,1.781	1.078	0.751,1.547
Urbanity (reference: rural)
Urban	0.981	0.507,1.896	0.786	0.552,1.118
Frequency of clinic check-ups (reference: 0)
1–2	0.497	0.233,1.060	1.013	0.690,1.487
≥3	0.285	0.141,0.576	0.608	0.421,0.877
Need
Diabetic status (reference: no)	0.783	0.430,1.425	0.585	0.411,0.832
Yes
Duration of hypertension (reference: 1–5)
5–10	0.965	0.332,2.803	0.610	0.409,0.911
≥10	0.714	0.270,1.885	0.454	0.315,0.655
Blood pressure control (reference: uncontrolled)
Controlled	0.317	0.167,0.600	0.714	0.524,0.972

Table S2 Contribution of drug benefit program effect to the differences in medication nonadherence between drug benefit program enrollees and non-enrollees

	Coefficient	P	95% CI	% of contribution
Predisposing factors
Gender (reference: male)
Female	−0.001	0.952	−0.030,0.028	−1.51%
Age	0.128	0.334	−0.131,0.387	213.92%
Education (reference: illiterate)
Primary school	−0.001	0.960	−0.022,0.021	−0.94%
Middle school or above	0.007	0.545	−0.015,0.030	11.89%
Enabling factors
Annual household income (reference: <48,000)
48,000–68,400	−0.003	0.779	−0.026,0.019	−5.35%
≥68,400	0.007	0.629	−0.021,0.035	11.49%
Urbanity (reference: rural)
Urban	−0.006	0.566	−0.024,0.013	−9.23%
Frequency of clinic check-ups (reference: 0)
1–2	0.016	0.104	−0.003,0.034	25.98%
≥3	0.036	0.038	0.002,0.070	60.55%
Need
Diabetic status (reference: no)
Yes	−0.008	0.432	0.029,0.012	−13.98%
Duration of hypertension (reference: 1–5)
5–10	−0.007	0.436	−0.025,0.011	−11.76%
≥10	−0.026	0.406	−0.090,0.037	−45.01%
Blood pressure control (reference: uncontrolled)
Controlled	0.028	0.005	0.008, 0.048	46.93%

Table S3 Contribution of observed characteristics to the differences in medication nonadherence between drug benefit program enrollees and non-enrollees

	Coefficient	P	95% CI	% of contribution
Predisposing factors
Gender (reference: male)
Female	−0.001	0.603	−0.005,0.003	−1.61%
Age	0.000	0.992	−0.002,0.002	0.17%
Education (reference: illiterate)
Primary school	−0.000	0.647	−0.002,0.001	−0.68%
Middle school or above	−0.002	0.252	−0.005,0.001	−2.95%
Enabling factors
Annual household income (reference: <48,000)
48,000–68,400	−0.000	0.057	−0.001,0.000	−0.60%
≥68,400	−0.001	0.684	−0.008,0.005	−2.35%
Urbanity (reference: rural)
Urban	0.000	0.184	−0.000,0.000	0.23%
Frequency of clinic check-ups (reference: 0)
1–2	0.000	0.948	−0.001,0.001	0.06%
≥3	0.007	0.009	0.002,0.012	11.86%
Need
Diabetic status (reference: no)
Yes	0.004	0.001	0.002,0.007	7.18%
Duration of hypertension (reference: 1–5)
5–10	−0.004	0.022	−0.008,-0.001	−7.22%
≥10	0.019	<0.001	0.010,0.029	32.13%
Blood pressure control (reference: uncontrolled)
Controlled	0.002	0.034	0.000, 0.003	3.02%