Study design and setting

This study was a cluster-randomized controlled trial designed to evaluate the effectiveness of a multicomponent intervention on blood pressure control and HRQoL among adults with uncontrolled hypertension. The detailed study design has been published elsewhere [11–13]. The study was conducted in 16 communities in northern Hung Yen province, Vietnam, with 8 communities assigned to the intervention group and 8 communities to the control group. Patients were enrolled at their local community health center and followed up at 12 months after randomization. The protocol for this trial has been previously published [11], and the supporting CONSORT checklist is included in Supporting information (S1 File). The trial was registered at ClinicalTrials.gov (Registration number: NCT03590691, registration date July 17, 2018)

Patient eligibility

Potentially eligible participants were adults aged 25 years and older living in selected rural communities, without plans to relocate within 12 months, and who met the criteria for uncontrolled HTN as defined by the 8^th Joint National Commission of High Blood Pressure guidelines (JNC 8) [14]. These patients had to have no cognitive impairment, not be pregnant, and not be involved as storytellers in the parent trial intervention. Participant screening was conducted at local community health centers by trained health personnel. Screening for the parent study required two blood pressure (BP) measurements taken at least one week apart, following a standardized protocol. Individuals with persistently elevated BP—defined as >140/90 mmHg at both screening visits—were considered eligible for enrollment. Both incident (newly diagnosed) and prevalent cases of uncontrolled HTN were enrolled in the parent trial. Newly diagnosed patients were referred for antihypertensive initiation, while those with severe BP (BP greater than 180/110 mmHg) were referred for treatment adjustment. Written informed consent was obtained from all trial participants prior to study enrollment.

Study recruitment, randomization, and masking

This cluster-randomized controlled trial was conducted in 16 communities in four districts in Hung Yen Province, Vietnam. Randomization was performed at the community level, stratified by district, using a computer-generated random number sequence. In each district, two communities were allocated to the intervention arm and two to the comparison arm. Eligible consenting adults with uncontrolled HTN were assigned to either of these groups based on their residence. Both study patients and those assessing the primary and secondary trial outcomes were blinded to treatment allocation to minimize the potential for measurement bias. Recruitment occurred from October 2019 to April 2021.

Intervention content and delivery

Follow-up activities

After obtaining informed consent and trial enrollment, study participants received reminder letters before each of three follow-up visits (3-, 6- and 12 months after trial enrollment) with phone calls one week prior. If participants missed any in-person clinic visits, this led to follow-up phone calls or home visits for measurement of their BP. This study was paused twice during the COVID-19 pandemic (February – April 2021 and July-October 2021) which led to substantial missing data at the three- and six-month follow-ups. Therefore, only data from the study baseline period and the 12- month follow-up were included in the present study.

Sample size

Detailed sample size calculations for this trial have been previously described [11].

Data collection

Trained healthcare professionals conducted interviews to gather data on patient’s socio-demographic characteristics, medical history, lifestyle factors using the validated STEP survey [18]. Age (years) was treated as a continuous variable. Other variables were categorized using a similar approach with the previous publication [12] including sex (male vs. female), education level (less than primary, primary, secondary, high school, and college/university), marital status (currently married vs. other), and occupation (employed, self-employed, farmer, retired, homemaker, or other), smoking status (self-reported current smoker vs. non-smoker). Anthropometric measurements, including height, weight, waist circumference, and hip circumference, were taken using standardized equipment and protocols. Body Mass Index (BMI) was calculated and treated as a continuous variable.

BP was measured using a validated automated device. Three BP readings were taken by trained health personnel, with participants seated and rested, following the World Health Organization standard protocol [19]. The average of the last two measurements was used in subsequent analyses. HRQoL scores were measured using SF-12 survey instrument [20].

Measurement of quality of life

HRQoL was defined as the patient’s overall well-being, which was dependent on three independent factors including objective descriptors, subjective evaluations, and personal values and aspirations. These parameters were based on physical, mental, emotional, social, and material wellbeing [21]. HRQoL was assessed using the 12-Item Short Form Health Survey (SF-12), a validated and reliable instrument for measuring health-related HRQoL across diverse populations, including Vietnam. The SF-12 survey was constructed using questions drawn from the Medical Outcomes Study 36-item Short Form Survey (SF-36), which was designed to have similar performance to the SF-36, while taking less time to complete [20,22,23].

The SF-12 focuses on two main components: the Physical Component Summary (PCS-12) and the Mental Component Summary (MCS-12). It covers eight health domains: physical functioning, role physical, bodily pain, general health, vitality, social functioning, role emotional, and mental health. These domains are weighted and combined to calculate the PCS-12 and MCS-12 scores, which represent physical and mental health, respectively.

The calculation of PCS-12 and MCS-12 scores followed the standard published algorithm [20]. Each of the 12 items in the SF-12 was first recoded and normalized based on the scoring guidelines. Responses to each item were assigned specific weights, which were derived from population norms. These weights were used to calculate raw scores for the physical and mental health components.

Recoding Responses: The questions in the SF-12 use various Likert scales, such as three levels (yes, very limited; yes, limited; not limited), five levels (no influence; little influence; moderate influence; lot of influence; very much influence), six levels (always; most of the time; quite a lot; occasionally; sometimes; none of the time), two levels (yes, no), and health rating questions (excellent, very good, good, fair, poor). The scoring method produces scores for each component (PCS and MCS) ranging from 0 to 100 (50 is considered average). The total HRQoL score was calculated as the sum of the physical health and mental health component scores. Higher scores indicate better HRQoL; items related to limitations in physical functioning or emotional problems were reverse coded so that a higher score indicated fewer limitations.

Applying Weights: The recoded responses were multiplied by their respective weights for PCS-12 and MCS-12. These weights were derived from a U.S. general population reference and were specific to each item. For example, items related to physical functioning contributed more to the PCS-12 score, while items related to mental health contributed more to the MCS-12 score.

Study outcome

The primary outcome of interest was the change in HRQoL (the PCS-12 and the MCS-12) over a 12-month period (continuous outcomes), assessed at the individual patient level. The main analysis focused on comparing the differential change in HRQoL between participants in the intervention and comparison groups from baseline to the 12-month follow-up.

Data analysis

Patient’s baseline characteristics were compared between the intervention and control groups. Categorical variables were presented as percentages and compared using chi-square tests, while continuous variables were presented as means with standard deviations and compared using t-tests. Each patient contributed data at two time points (baseline and 12-month follow-up).

We examined differences in over-time changes in HRQoL between the two primary study groups using linear mixed models, including study group (intervention vs. comparison), time (baseline vs. 12 months), and the interaction term between study group and time. This was done to estimate the differential intervention effect over time, accounting for clustering of patients within the community and repeated within-person measurements by random effects. We also used the mixed models to adjust for potential covariate imbalances between the two groups at baseline, including age, sex, education level, marital status, BMI, smoking status, and baseline anti-hypertensive medication use. Additionally, we performed subgroup analyses according to sex, age group, and education level to explore the impact of the study intervention across different subgroups. All analyses were performed using STATA 17.0 statistical software (USA).

Ethical considerations

This randomized trial was approved by the Institutional Review Board at the Health Strategy and Policy Institute (HSPI) in Hanoi, Vietnam (Decision 171/QD-CLCSYT dated September 10, 2019). Written informed consent was obtained from all patients.

Baseline patient characteristics by study group

The intervention and comparison groups were generally comparable in terms of their demographic and clinical characteristics, but some differences were observed (Table 1). Patients in the intervention group were slightly younger, had a lower body mass index, and had a higher proportion of males and participants who were currently married compared with those randomly assigned to the comparison group. The intervention group had a higher proportion of participants with a high school or higher education at baseline compared to the comparison group.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Baseline characteristics of study participants by study groups.

https://doi.org/10.1371/journal.pone.0337817.t001

Baseline HRQoL according to study groups

At the baseline, there was no significant difference in total HRQoL scores between the intervention and comparison groups (Table 2). When examining each component of the HRQoL, the mean PCS-12 score was comparable between the intervention and comparison groups (41.3 vs. 42.3) (Table 2 and Fig 2a). Similarly, no significant difference in the mean MCS-12 scores was observed between the two groups at the baseline (52.6 vs. 51.1) (Table 2 and Fig 2b).

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Differential over-time change for intervention vs. comparison groups (95% CIs) in health related quality of life.

https://doi.org/10.1371/journal.pone.0337817.t002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Changes in PCS-12, MCS-12, and total HRQoL scores (95% CI) according to study groups.

https://doi.org/10.1371/journal.pone.0337817.g002

Changes in HRQoL over time according to study groups

Both study groups demonstrated improvements in their total HRQoL scores over the 12-month follow-up period; however, a greater increase was observed for patients in the intervention group compared with those in the comparison group (mean change: 6.9 points vs. 3.1 points), with a multivariable adjusted between-group difference of 5.6 points (95% CI: 2.6–8.6; p-value for the interaction term between study group and time <0.001) (Table 2 and Fig 2c).

In examining each component of HRQoL over the 12-month follow-up, the intervention group showed a significant increase in the PCS-12 compared to the comparison group (mean change: 4.7 points vs. 1.8 points), with a multivariable-adjusted difference of 4.2 points (95% CI: 2.0–6.4, Table 2 and Fig 2a). Improvements in the MCS-12 were more modest in both groups (mean change 2.1 points in the intervention group vs 1.3 points in the comparison group) with the adjusted between-group difference being 1.4 points (95% CI: −0.6; 3.5, Table 2 and Fig 2b).

Subgroup analyses

In examining the possible heterogeneity of the intervention effects, we conducted several pre-specified subgroup analyses to assess whether the impact on total HRQoL, PCS-12, and MCS-12 scores differed across various patient characteristics at the time of 12-month follow-up. The Forest plots suggest that there is no evidence of heterogeneity of the multi-component trial intervention (Figs 3a, 3b, 3c).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Changes in PCS-12, MCS-12, and total HRQoL from Baseline to 12 months by study groups.

https://doi.org/10.1371/journal.pone.0337817.g003

Serious Adverse Events (SAEs)

During the course of the trial, a total of three participants (0.9%) in the intervention group and six participants (1.8%) in the comparison group died. In addition, 27 participants (7.9%) in the intervention group and 9 participants (2.7%) in the comparison group were hospitalized for various conditions, including stroke, pleural effusion, diabetes, lung cancer, and renal failure.

Impact of the multi-component study intervention on HRQoL

Our community-based intervention showed significantly improved overall HRQoL, particularly in the PCS, in our study sample of adults with uncontrolled HTN. These results are consistent with earlier work in low- and middle-income countries [24,25]. For example, a cluster-randomized controlled trial conducted among more than 2,600 adults with HTN in rural communities across Bangladesh, Pakistan, and Sri Lanka (2016−2018) evaluated the impact of a five-component community-based intervention (BP monitoring, home health education, physician training, HTN care coordinators, compensation for additional health services and targeted subsidies on HRQoL) found improvements in overall health status measured by the EuroQol 5-Dimension 5-Level (EQ-5D-5L) visual analog scale and by the EQ-5D-5L utility index compared with participants receiving usual care [24]. Similarly, a quasi-experimental trial in rural China among more than 1,400 patients with HTN, testing an integrated care model, coordinating services across village clinics, township, and county hospitals, demonstrated enhanced SF-36 scores—particularly in the physical functioning domain, with no significant improvements observed in patient’s mental health status [25].

In contrast, our intervention resulted in modest improvement in the mental MCS, which was different from previous studies in other in low- and middle-income country settings that have shown significant benefits across both physical and mental domains. For example, a study in Egypt among 1,100 patients with HTN (2017−2019) reported significant improvements in both PCS and MCS scores in the intervention group who received a group-based educational intervention compared to a control group who received usual care [26]. A trial in rural India in 2022 among more than 300 patients with HTN found that an exercise-focused intervention had broad improvements across both physical and psychological domains [27]. A trial in rural China assessing the effectiveness of a mobile health intervention on HRQoL among more than 200 patients with uncontrolled HTN showed that the intervention significantly improved the PCS-12 and MCS-12 scores [28]. These differences may reflect differences in the MCS-12 scores at baseline, intervention content and focus, intensity, delivery mode, or contextual factors such as cultural attitudes toward mental health.

Although our delivered intervention had multiple-components, individual components likely contributed synergistically to the observed improvements in physical HRQoL. While no study has investigated the impact of a storytelling component on HRQoL in adults with uncontrolled HTN, several studies have demonstrated that this intervention can potentially improve HRQoL in patients with other non-communicable chronic diseases. For example, a randomized trial among 60 adults with type 1 diabetes in Iran found that a four-week digital storytelling intervention using peer-created videos and guided reflections significantly reduced diabetes-related distress, burnout, and depressive symptoms, together with improvements in psychological well-being as compared with routine care [29]. Evidence on the impact of home BP monitoring on HRQoL is also limited but promising. For example, a trial among nearly 200 patients with HTN in Honduras and Mexico in 2011 found that automated telephone care management plus home BP monitoring reduced depressive symptoms, medication problems, and improved patient’s general health status and satisfaction with care compared with usual care [30]. While the role and impact of CHWs on patients’ HRQoL has been studied in other chronic conditions such as diabetes [31], little is known about their effect on HRQoL in adults with uncontrolled HTN.

Our multi-component intervention improved HRQoL through strategies that included promoting a healthy lifestyle, increasing medication adherence, and disease self-management strategies. The more substantial improvements in PCS-12 scores suggest that the intervention effectively addressed physical limitations, pain, and general health perceptions in our study population. Components likely contributing to these improvements include enhanced antihypertensive medication adherence, lifestyle modifications such as increased physical activity and healthy dietary changes, improved HTN control, and access to healthcare services.

The relatively modest impact on MCS-12 scores could be explained by several factors. First, our intervention focused on heart health, and not mental health, such as stress management, social support enhancement, or psychological well-being. Second, our study population had a baseline MCS-12 score of 51.9, which was slightly above average, making further improvements in this outcome more difficult to achieve. Third, longer follow-up periods may be necessary to observe meaningful changes in mental health outcomes following physical health improvements. Finally, since the study was conducted during the COVID-19 pandemic, a period marked by significant disruptions to daily life and psychological well-being, this may have impacted the study results, particularly about the mental health outcomes.

Study strengths and limitations

This study had several strengths, including its well-designed, full-scale cluster-randomized design in multiple communities in a northern rural setting in Vietnam, testing a multi-component intervention, and use of validated measures for assessing HRQoL. Moreover, the intervention was grounded in the existing health system infrastructure and delivered in collaboration with local health authorities, enhancing its feasibility and scalability.

There are several limitations, however, that need to be considered in the context of the present study. These included the study setting in northern Vietnam, which may have resulted in limiting potential generalizability. HRQoL was self-reported, which may have introduced social desirability bias. The follow-up period was relatively short (12 months), potentially limiting the detection of long-term HRQoL changes, particularly in mental health. In addition, we were unable to evaluate each intervention enhancement separately since we examined the holistic impact of a multi-component intervention. Future studies may need to include multiple geographic and urban and rural settings with a longer follow-up period and testing a comprehensive multi-component intervention targeting both physical and mental health for patients with uncontrolled HTN.

Future intervention directions

Recent literature suggests that integrated approaches that address both physical and psychological aspects of HTN management yield meaningful improvements in HRQoL [25,32]. Given the growing accessibility of digital technologies in Vietnam, future research should explore the implementation of mHealth-based interventions as a feasible and scalable approach to enhance HTN control and improve HRQoL, especially in underserved rural populations.