Overview of the study design
The study was conducted in Marcos Paz, located in the province of Buenos Aires, Argentina19. The population is 54.181 inhabitants, mostly located in urban areas (80%) and 62% are users of the public health system, consisting in one hospital, 8 PCCs and 1 itinerant health care post.
We implemented a non-controlled feasibility study using a mixed methods approach to evaluate fidelity, adoption and acceptability of a multi-component intervention implemented in six PCCs. The eligibility criteria for study participants were uninsured individuals, age ≥ 40 years, residence in the PCC´s catchment area and 10-year CVD risk ≥ 10%. All participants had free access to antihypertensive, hypoglycemic, and lipid-lowering drugs. The intervention and follow-up lasted six months. The summary of the flow diagram of the study participants is described in Figure 1.
Screening and recruitment process
Trained and certified CHWs carried out an active search of individuals with 10-year CVD risk ≥ 10% in the catchment area of the PCCs, following the procedures used by the national program REDES. They visited the participants’ homes and estimated the 10-year CVD risk, using the CVD charts for risk stratification without cholesterol measurements developed by the World Health Organization for the Latin America and Caribbean20. In the screening/baseline visits, CHWs confirmed patients’ eligibility, and those who met the inclusion criteria were invited to participate in the study. After patients’ acceptance and consent, CHWs proceeded to deliver the intervention in the patient´s home. Subjects who did not meet the inclusion criteria received a short advice on the importance of risk factors control and were thanked for their time.
The multicomponent intervention program addressed system, provider, and individual barriers to risk factor detection and treatment by integrating some individual strategies previously proven effective with novel components (Table 1).
The intervention addressed system barriers through task shifting21-24 among the PCC´s staff by assigning CVD risk stratification, health care counseling and greater involvement in the follow up of risk factors to community health workers. Creation of protected medical appointment slots23,25. A new form specially designed for this study (Figure S1). The novelty of this form lays on including three new sections to be completed by physicians, nurses and CHWs, aiming to stimulate the teamwork. The new CVD form was designed based on the current form used by the national program REDES with the addition of new sections: I) Risk factors status showing the results in traffic light signs to facilitate its visualization to physicians and other PCC´s staff and current drug treatment (to be completed by CHWs); II) Clinical measurements (BP values, weight, height and BMI calculation, to be completed by nurses) and III) Physician´s section (notes and prescription).
Provider barriers through educational outreach visits (EOVs)26-27. The study protocol proposed that physicians attended two EOVs during the study period. Educational outreach visits have demonstrated to change health professionals’ practice, particularly the prescribing patterns of physicians27. Trained peer physicians conducted the EVOs in site, including audit of clinical cases and feedback on prescription practices. In addition, physicians were provided with aid tools (pocket-cards summarizing the guidelines algorithms as decision-trees) to facilitate both pharmacological and non-pharmacological treatment28.
Individual barriers through a home-based intervention delivered by CHWs29-31. They conducted home visits, in which they provided printed educational materials, conducted health care counselling sessions about to the importance of adhering to healthy lifestyles and taking medications to participants and their families using the traffic light signs of the new CVD form. Then CHWs scheduled a medical appointment from the protected medical appointment slots and provided participants and their families with printed educational materials and a card with the appointment record32-36. The study flow is summarized in Figure S2.
The primary outcomes were feasibility outcomes including acceptability, adoption and fidelity expressed as the proportion of 1) CHWs and physicians who completed the training sessions. 2) Eligible individuals with the new CVD form completed by CHWs, nurses and physicians 3) Counselling sessions and medical appointment provided to participants. 4) Individuals who attended the scheduled physician’s appointment at the PCC. 5) Outreach visits successfully conducted, and 6) Participants who reported to be satisfied/very satisfied with the multi-component intervention.
The secondary outcomes were constituted by intermediate and effectiveness indicators. The intermediate indicators were: 1) Proportion of individuals with hypertension who received drug treatment according to Argentinian good clinical practice guidelines (GCP)37. 2) Proportion of individuals with diabetes treated with the first-line drug proposed by Argentinian GCP; and 3) Proportion of individuals who received statins and aspirin according to their estimated cardiovascular risk. The effectiveness indicators were: 4) Proportion of participants with BP<140/90 mmHg and 5) Proportion of participants with fasting glucose <126 mg/dL.
We define moderate CVD risk as individuals with estimated 10-years risk of 10% to 19%. High CVD risk as individuals with estimated 10-years risk of 20 to 29% and very high CVD risk as ≥ 30%. Hypertension was defined as participants with BP values ≥140/90 mmHg OR under drug treatment at baseline for hypertension control. Controlled BP was defined as participants with BP values <140/90 mmHg. Diabetes was defined as participants with history of diabetes (self-reported) OR under drug treatment at baseline for diabetes control.
CHWs collected baseline data related to socioeconomic variables, medical history and use of drug treatment for risk factors control during the baseline visit in the participant´s household, using standardized questionnaires specially developed for this study. Independent personnel collected the study outcomes from administrative and medical records after six months of the enrollment visit. Qualitative data was collected with a convenient sampling to assess acceptability among providers (4 physicians and 6 CHWs) and satisfaction level in participants through a telephone interview. (Supplementary material: Study forms).
For the quantitative analysis we used descriptive statistics such as means, median and proportions to describe the general characteristics of the study population, study outcomes, process indicators and satisfaction level in participants using the RE-AIM framework38. Statistical analyses were performed using STATA version 12.0 (Stata Corp., College Station, TX, USA). For the qualitative analysis, written transcripts of the interviews were classified and then codified according to the study objectives. The written transcripts were entered into ATLAS.ti version 7 software (ATLAS.ti Scientific Software Development GmbH) combined with the manual technique of information coding. Analytical dimensions were identified as constructs for the description of findings. Finally, data were abstracted and interpreted through content analysis39,40.