Implementation of a comprehensive intervention for patients at high risk of cardiovascular disease in rural China: A pragmatic cluster randomized controlled trial

Abstract

Objective: This study aims to assess whether a standard intervention package of cardiovascular disease (CVD) care was being delivered effectively, and if it was associated with improved lifestyle and biomedical indicators.

Methods: In rural China, we implemented a pragmatic cluster randomized controlled trial for 12 months, randomized at the township hospital level, and compared with usual care. Intervention case management guideline, training and performance monitoring meeting and patient support activities were designed to fit within the job description of family doctors in the township hospitals and comprised: 1) prescription of a standardised package of medicines targeted at those with hypertension or diabetes; 2) advice about specific lifestyle interventions; and 3) advice about medication adherence. Participants were 50-74 years old, had hypertension and CVD risk scores >20% or diabetes, but were excluded if a history of severe CVD events. We also randomly selected 100 participants from six selected clusters per arm as a panel to collect intermediate biomedical indicators over time.

Results: A total of 28,130 participants, in 33 intervention and 34 control township hospitals, were recruited. Compared with the control arm, participants in the intervention arm had substantially improved prescribing rates of anti-hypertensives, statins and aspirin (P<0.001), and had higher medication taking rates of aspirin and statins (P<0.001). Mean systolic and diastolic blood pressures were similar across both arms (0.15 mmHg, P = 0.79, and 0.52 mmHg, P = 0.05, respectively). In the panel, (950) rates of smoking (OR = 0.23, P = 0.02) and salt intake (OR = 2.85, P = 0.03) were significantly reduced in the intervention versus control arms, but there were no statistically significant improvement over the 12 month follow-up period in biomedical indicators (P>0.05).

Conclusion: Implementation of the package by family doctors was feasible and improved prescribing and some lifestyle changes. Additional measures such as reducing medication costs and patient education are required.

Trial registration: Current Controlled Trials ISRCTN58988083.

Fig 1. Trial profile.

a. Eligible subjects included those 1) aged 50–74; 2) with a calculated 10-year CVD risk≥20% and having hypertension; or 3) diagnosed as diabetic. Subjects excluded were those who had 1) mental diseases; 2) cancers; 3) acute coronary heart disease or had previously suffered stroke; 4) diastolic blood pressure <60 mmHg; or those who were 5) out of town for more than 3 months or non-locals; 6) urban residents; or those who 7) couldn’t take recommended medicines due to other severe diseases such as cancer. b. Subjects who did not sign consent forms were not recruited. c. Subjects did not participant in the survey and data collection despite being reminded 3 times within the last 2 weeks. d. Intervention package includes: 1) healthy lifestyle education; 2) drug therapies; 3) adherence support. e. Usual care refers to health care following routine procedures at the discretion of individual family doctors. f. Subjects died from reasons other than coronary heart disease or stroke during the 12 months were excluded. g. Missing cases: those who had been absent in a consecutive of 6 months or two follow-up appointments at quarterly basis.

Fig 2. Proportions of prescribing and taking of medicines among recruited hypertensive patients*.

(A) Proportion of patients having two anti-hypertensive drugs prescribed and taken in the intervention group; (B) Proportion of patients having two anti-hypertensive drugs prescribed and taken in the control group; (C) Proportion of patients having aspirin and statin prescribed and taken in the intervention group; (D) Proportion of patients having aspirin and statin prescribed and taken in the control group. * At 12 months follow-up (quarter 4) compared to the control arm in the intervention arm a significantly higher proportion of patients had been prescribed (OR = 3.55, 95% CI: 3.31 to 3.80, P<0.001), but had not taken (OR = 1.06, 95% CI: 0.98 to 1.14, P = 0.15), two antihypertensives; a significantly higher proportion of patients had been prescribed (OR = 565.77, 95% CI: 469.12 to 682.33, P<0.001) and had taken (OR = 8.85, 95% CI: 7.37 to 10.63, P<0.001), aspirin; and a significantly higher proportion of patients had been prescribed (OR = 656.24, 95% CI: 523.36 to 822.85, P<0.001) and had taken (OR = 11.79, 95% CI: 8.76 to 15.87, P<0.001) statins. Comparisons based on multivariate logistic mixed-effects models, adjusted for cluster effects, baseline measurement and demographic/socioeconomic factors.

Fig 3. Proportions of prescribing and taking of medicines among recruited patients with diabetes only*.

(A) Proportion of patients having one anti-hypertensive drug prescribed and taken in the intervention group; (B) Proportion of patients having one anti-hypertensive drug prescribed and taken in the control group; (C) Proportion of patients having aspirin and statin prescribed and taken in the intervention group; (D) Proportion of patients having aspirin and statin prescribed and taken in the control group. * At 12 months follow-up (quarter 4) compared to the control arm in the intervention arm a significantly higher proportion of patients had been prescribed (OR = 11.41,95% CI:10.24 to 12.71, P<0.001), but had not taken (OR = 0.98, 95% CI: 0.89 to 1.07, P = 0.60), one anti-hypertensive; a significantly higher proportion of patients had been prescribed (OR = 806.98, 95% CI: 597.55 to 1089.79, P<0.001) and had taken (OR = 9.53, 95% CI: 6.84 to 13.28, P<0.001) aspirin; and a significantly higher proportion of patients had been prescribed (OR = 360.86, 95% CI: 282.59 to 460.81, P<0.001) and had taken (OR = 18.93, 95% CI: 10.55 to 33.97, P<0.001) statins. Comparisons based on multivariate logistic mixed-effects models, adjusted for cluster effects, baseline measurement and demographic/socioeconomic factors.

Fig 4. Blood pressure changes (mmHg) by quarter for patients recruited in the trial*.

(A) Quarterly changes in mean systolic blood pressure in the intervention and the control group; (B) Quarterly changes in mean diastolic blood pressure in the intervention and the control group. *No significant differences between intervention and control arms by 12 months follow-up (quarter 4) in terms of either systolic blood pressure (0.15, 95% CI: -0.96 to 1.26, P = 0.79, ICC = 0.07) or diastolic blood pressure (0.52, 95% CI: -0.02 to 1.01, P = 0.05, ICC = 0.04). Comparisons based on multivariate linear mixed-effects models adjusted for cluster effects, baseline measurements and demographic/socioeconomic factors.