Effectiveness of letters to patients with or without Cochrane blogshots on 10-year cardiovascular risk change among women in menopausal transition: 6-month three-arm randomized controlled trial

This parallel, three-arm randomized controlled trial (RCT) tested the effect of a 6-month educational intervention in a form of letters in decreasing the 10-year CVD risk in women aged 45 to 65 with one or more known CVD risk factors. The letters reminded the participants of their own 10-year CVD risk, with or without included Cochrane blogshots.

The inclusion criteria were (1) female sex; (2) age 45 to 65 years, which is the age of menopause transition, including postmenopause [15]; and (3) one or more CVD risk factors: overweight or obesity (body mass index (BMI) ≥ 25 kg/m², and/or central obesity, i.e., waist circumference ≥ 88 cm), high blood pressure (systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg), high blood cholesterol (≥ 5.2 mmol/L), and active smoking. The participants on antihypertensive therapy were eligible for the study.

Current CVD (ischemic heart disease, peripheral artery disease, and stroke), malignant diseases, serious systemic diseases, and/or mental diseases were the exclusion criteria.

The trial took place in family medicine offices in Croatia from February 1, 2018, to September 1, 2020. Family medicine offices are primary health care offices, i.e., the first step for patients to seek help for their health problems in the Croatian health care system, which ensures full national health coverage. Physicians working in family medicine offices are specialists in family medicine. One office was in the city of Split, the capital of the Split-Dalmatia County; one in the city of Osijek, the capital of the Osijek-Baranja County; and one in the city of Rijeka, the capital of the Primorje-Gorski Kotar County. An office from each of the following towns was the recruitment sites: Supetar (the island of Brač, Split-Dalmatia County), Bijelo Brdo (Vukovar-Srijem County), Kotoriba (Međimurje County), and Kutina (Sisak-Moslavina County). In each family medicine office, 30 participants were recruited. All participants involved in this trial had access to their family medicine office according to their place of residence. The first participant was recruited in February 2018, and the last participant was recruited in February 2020. The recruitment was performed by family medicine doctors in their medical offices. They also performed clinical measurements at the baseline and 6 months after the intervention, after sending the notification to the participants about their check-up appointment. The data from the two time points were sent to the primary investigator.

The Ethics Committee of the University of Split, School of Medicine, approved the study, No. 2181-198-03-94/10-11-0038 and No. 2181-198-03-04/19-0044. Written informed consent was obtained from all participants after they received the information about the study. The data were processed as pseudonymized (coded) data, following the General Data Protection Regulation (GDPR).

The participants were randomized into three parallel groups:

The examples of all types of letters sent to the participants in each group (in Croatian) are available in Additional file 1: Letters 1–5, including the Cochrane blogshots (in Croatian) used in the letters.

At recruitment and at 6 months after the intervention, we measured the following clinical parameters: height, weight, BMI, waist and hips circumference, systolic and diastolic blood pressure, total serum cholesterol, high-density lipoprotein (HDL), low-density lipoprotein (LDL) cholesterol, triglycerides, and glucose.

The primary outcome measure was the change in the score between the 10-year risk of CVD at the beginning of the study and the 10-year risk of CVD after 6 months. The 10-year risk of fatal CVD was estimated using an online tool (https://​www.​cvriskcalculator​.​com/​) based on the American College of Cardiology/American Heart Association (ACC/AHA) guidance [19]. The calculation is based on the following data collected from the study participants: age, gender, race, total and HDL cholesterol, systolic blood pressure, data about antihypertensive therapy, diabetes mellitus, and smoking status. The 10-year risk of fatal CVD was expressed as a percentage and was calculated at the beginning of the trial and at 6 months, after the participants received the last letter.

The secondary outcome measures were the changes in weight (in kg), body mass index (BMI), waist circumference (in cm), hip circumference (in cm), and smoking status (related to continued or excessive smoking) at 6 months post-intervention. The number of participants who changed their CVD risk category—low, moderate, and high [19]—was also calculated.

After recruitment, the participants first completed the pre-study questionnaire (in Croatian), which included (a) demographic data (available in Additional file 2), (b) Decisional Conflict Scale (DCS) [20‐22], and (c) future time perspective (FTP) [23].

The DCS consists of 16 items rated on a 5-point Likert scale and measures an individual’s uncertainty toward a course of action. The score is calculated as a sum of items, divided by the number of items and multiplied by 25, allowing for a score range from 0 (no decisional conflict) to 100 points (extreme decisional conflict) [20, 21]. The Croatian version of the scale was previously validated [22].

The FTP is a 13-item scale from the Time Perspective Inventory [23]. It assesses how respondents focus on planning and achievement of future goals. The respondent answers using a 5-point Likert scale from 1 (does not refer to me) to 5 (refers to me completely), maximum score of 65. Higher scores indicate a greater future time perspective.

As there were no previous studies on this topic, we hypothesized that in the intervention group at the end of the trial, the mean CVD risk would be 6.0 and 8.0 in the control group, with a standard deviation of 3. With a study strength of 0.8 and alpha level of 0.01 (to take multiple comparisons into account), we used an online sample size calculator (https://​epitools.​ausvet.​com.​au/​twomeansone), to estimate that we would need at least 53 participants per group (159 in total). To compensate for potential dropouts, we aimed to recruit 70 participants per trial arm.

Trial groups were formed by a random assignment, so that each respondent had one probability of falling into one of the three research groups (1:1:1) and each group of participants was exposed to only one of the study interventions. The generation of the random sequence was performed by a statistician who was not involved in the conduct of the trial, using https://​www.​randomizer.​org/​. Only the main investigator was aware of the allocation of participants into the study groups, and the participants, the family medicine doctors, or the statistician were not aware of the allocation of participants to trial arms. The family medicine doctors made physical exams and took all measurements of each eligible participant. The main investigator prepared all letters, sealed them in opaque envelopes, and sent them to the addresses collected from the medical records of the participants.

Categorical data are presented as frequencies and percentages. Numerical variables did not follow a normal distribution and were presented as medians with IQR or with 95% confidence intervals (CIs). Due to the asymmetrical distribution of the variables, a parametric 2 × 2 factorial analysis could not be performed. As there is no non-parametric analog, the post-intervention differences between the three groups were compared by using the Kruskal-Wallis test with post hoc Conover-Iman test. To address multiple comparisons, we applied the Bonferroni correction to avoid type I error. The differences in CVD risk were made by subtracting the scores at the end of the trial with the ones from the baseline assessment, which sometimes resulted in negative scores. The results were expressed as a median difference in the score (with 95% CI). The changes in the number of participants who changed the category of their 10-year CVD risk after the interventions were tested using the McNemar χ² test. The risk ratios and numbers needed to treat (NNTs) were calculated based on the number of participants who had a decrease in CVD risk after the interventions. We used linear regression to determine the variables that predicted the pre- and post-assessment changes. The results were expressed as unstandardized coefficients and R². All analyses were performed by the R programming software (R Core Team, 2021).

The trial was retrospectively registered on ClinicalTrials.​gov on October 19, 2020 (trial registration number NCT04601558). The trial was retrospectively registered because of the lack of timely official access to the registry by the first author from the University Hospital of Split.

The strengths of our study include a simple, inexpensive, and non-invasive assessment of CVD risk factors, a high rate of participant adherence, and the parallel, three-arm randomized controlled study design which reduces the chance of confounding. Furthermore, the study provides a strong piece of evidence of the usefulness of providing CVD risk information and ways to decrease it as a simple and cost-effective method in shared-decision making. The strength of the study is also the use of the ACC/AHA guidelines to calculate CVD risk, based on a full set of data collected from the study participants (age, gender, race, total and high-density lipoprotein (HDL) cholesterol, systolic blood pressure, data about antihypertensive therapy, diabetes mellitus, and smoking status) [19]. Previous studies have indicated that more individuals are recommended for the treatment according to the ACC/AHA guidelines than to other guidelines [24].

The main limitation of the study is its short-term nature. We cannot make conclusions whether the intervention would have long-term effects. We also tested only Cochrane blogshots as the simplest way of presenting the evidence synthesis about health interventions. Our study was not specifically designed to differentiate the effects of Cochrane blogshots over CVD risk reminders, and further studies are needed to investigate the separate effects of blogshots over risk reminders. We included women of specific age, from 45 to 65 years, in order to cover the full menopause transition from premenopause, menopause, and early postmenopause, which is the period for increased CVD risk for women [15]. This means that other reproductive outcomes associated with an increased CVD risk could not be evaluated, such as pre-eclampsia, hypertensive disorders of pregnancy, and gestational diabetes mellitus [15]. Finally, the study was performed in a single country in a health system that is publicly funded, provides universal health care coverage, and has family medicine as the basic point of primary health care. We balanced the geographical, cultural, and economic differences among family medicine settings by including family medicine offices from 6 different country regions, including southern coastal and island trials sites and northern sites, as well as sites in major cities and smaller towns.

The results from our trial suggest that a simple and inexpensive intervention method in a form of letters reminding patients about their CVD risk, especially when accompanied by health information in the form of Cochrane blogshots about interventions for important CVD risk factors, can play an important role in CVD management. Cardiovascular risk decreased significantly in the intervention groups compared to the control group. The intervention in the form of Cochrane blogshots—short textual information about a systematic review on a simple graphic template—has been shown to have a positive effect on the comprehension of health information among different users [1]. It is possible that other forms of presenting health information about interventions for CVD risk factor would be more effective than blogshots, especially those containing more detailed information about the results of Cochrane systematic reviews. There are several arguments against this hypothesis. First, we showed that the readability of plain language summaries of Cochrane systematic reviews is low, almost twice that of the recommended reading level for health information [25]. Secondly, in randomized trials, Cochrane infographic summaries did not produce a better understanding compared to lay summaries among different users—health consumers, doctors, and medical students, although users had greater preference for visual information [26]. Finally, we also showed in a randomized trial that blogshots were better than plain language summaries in increasing understanding of health information and had higher preference among all users—health consumers, doctors, and medical students [1]. The findings from this study suggest that a simple information such as a reminder of CVD risk sent at regular intervals is a useful intervention for decreasing CVD risk among women in menopausal transition in the setting of family medicine offices, at least early after diagnosing the risk and starting its management. Also, Cochrane blogshots may be a useful tool to structure health information so that it can be easily accessed and understood by patients.

Providing reminders about CVD risk and information about high-quality evidence about post-health risks in this study can be considered as a part of shared decision-making (SDM) process [27, 28], in which decision aids can help patients with some chronic diseases in lifestyle changes. In the linear regression model for a greater postintervention change in the 10-year CVD risk, the only significant variable was the higher score on the decisional conflict subscale Effective decision, which explained only 2.7% of variance and was a significant predictor only in the group that received only the reminder letters. It is difficult to interpret these results, which would indicate that higher scores on the Effective decision subscale, which means lower belief in one’s own ability to complete tasks or one’s own belief in the ability to make a good decision regarding the course of management, were more likely to reduce their CVD risk. Those participants may have come to the trial with lower effective decision skills, and the letters with reminders helped them increase those skills. Further studies are needed to explore the significance of this finding. A recent Cochrane systematic review showed that people exposed to decision aids are better informed and have a more active role in decision making, as well as accurate risk perception [29, 30]. The interventions such as sharing electronic health records (EHRs) may improve quality of care by providing patients with their personal health information, also involving them as key stakeholders in the self-management of their health and disease [31, 32]. In our study, doctors were the source of information for the patient. In other studies, health care providers have delegated preventive activities to practice nurses who work independently and have their own consultation. In the study by Laurant et al., experiences of chronic illness care in patients with established CVD or at high CVD risk did not change after the implementation of a tailored program aiming at nurses’ counseling skills [33]. On the other hand, a Cochrane systematic review by Huntink et al. suggested that appropriately trained nurses can provide care of the same quality as primary care doctors and achieve similar health outcomes for patients [34]. While patient health outcomes were similar for nurses and doctors, patient satisfaction was higher with nurse-led care because nurses tended to provide longer consultations, give more information to patients, and recall patients more frequently than doctors. In the public health care system in Croatia and other similar systems in central and Eastern Europe, where there are not enough nurses with relevant training [35], and where doctors still have the traditional central place in the healthcare system, family medicine doctors are still the best point of delivery of SDM information to the patient.

The prevention of CVD remains high on the agenda in health care systems, especially prevention based on shared-decision making, and the main goal of contemporary preventive medicine is therefore to encourage behavior change. However, while behavior change often seems easy in the short run, it can be difficult to sustain. Our study evaluated the short-term effects of an intervention to decrease CVD risks, but a recent systematic review [36] showed that the effects of lifestyle changes on the reduction in CVD risk factors reached their highest point at 12 months of follow-up and then gradually decreased over time. This may reflect the fact that the longer-term intervention may be more effective in reducing CVD risks but only if patients remain highly adherent to the interventions. As we showed in our study, the patients with higher risk and thus with better awareness about their status may benefit the most from being reminded about their CVD risk and potential interventions to alleviate it. Their health behavior can be influenced by a simple intervention, which is easily applicable in the primary care setting, as well as in low-resource economic settings.