Mechanical heart valve replacement (MHVR) is still one of the most common surgeries to reduce the mortality of patients with advanced valvular disease (van Geemen et al., 2016). Every year, more than 200,000 heart valve surgeries are performed in China (Chen et al., 2015). However, patients after mechanical heart valve replacement (MHVR) require life-long anticoagulant treatment. The use of anticoagulants is complicated, and the therapeutic range is narrow. Incorrect medication use can lead to thromboembolism or bleeding and even death (da Silva Praxedes et al., 2019). Medication adherence is considered as the extent to which patients take medications as recommendations by healthcare providers (Huang et al., 2020). Stringent adherence to anticoagulants can reduce the duration of hospitalizations and mortality (Smet et al., 2018). However, according to the World Health Organization, the average adherence rate for long-term medications is only 50% and even lower in low-and middle-income countries. (Thomson Mangnall et al., 2016).

Family is regarded as a critical factor in promoting the health of patients with cardiovascular disease (Vedanthan et al., 2016). Family functioning is the main concept exploring the effect of family in illness and is defined as how family members communicate with each other, fulfill family roles, cope with and adjust family stress, and relate to other members (Zhang 2018). Family functioning played an integral role in long-term self-management and adherence, especially in Chinese families (Luo et al., 2019). Patients with adequate family functioning, that is, receiving considerable support and care, were proved to have better medication-taking behaviors (Psihogios et al., 2019; Sun et al., 2019). However, the research exploring the potential mechanisms underlying the relationship between family functioning and medication adherence is still limited.

Family functioning is closely related to the emotions of individuals living in their families (Sullivan et al., 2012). There is evidence that patients with poor family functioning suffer from more emotional problems (Zhang et al., 2019). Illness perceptions are identified as the emotions and thoughts individuals formed about their illnesses (Alyami et al., 2021). Maintaining or developing negative illness perceptions can lead to worse health status and poorer quality of life after valve replacement (Kohlmann et al., 2012). According to the Common Sense Model, illness perceptions could activate individuals to change their behaviors (Gibbons et al., 2013). Prior research has also found that illness perception was a predictor of adherence behavior in patients with cardiovascular disease (Mosleh and Almalik 2016). Patients with negative illness perceptions were more likely to be non-adherent to receive anticoagulation as prescribed (Miyazaki et al., 2018). Therefore, based on the findings of previous studies, illness perceptions may play the mediating role in the association between family functioning and medication adherence such that sufficient family functioning may associate with less threatening illness perceptions, which in turn led to better medication adherence.

Medication literacy is health literacy in the context of medication use, which refers to the ability of individuals to obtain, understand, calculate, and process information to the appropriate use of medication (Lee et al., 2017; Cordina et al., 2018). Ostini et al. found that there was not always a linear but U-shaped relationship between health literacy and medication adherence, affected by factors like education, social support, or cognitive functioning (Ostini and Kairuz 2014). Although, a study carried out in Chinese patients initially revealed the connection between medication literacy and medication adherence (Shi et al., 2019). It is still uncertain whether medication literacy has interactions with other influencing factors on adherence. Furthermore, a previous study revealed that negative illness perceptions can lead to poor medication adherence, however, not all patients were equally affected by perceptions because of different levels of literacy (Shiyanbola et al., 2018). In other word, literacy played a moderating role in the relationship between illness perceptions and medication adherence. On the other hand, for patients with limited literacy, family functioning promote skills to access, share and evaluate information to enhance medication adherence (Kraenbring et al., 2019; Wong et al., 2020). However, patients with adequate literacy have the ability to process information independently to improve medication adherence (Shen et al., 2020). Thus, literacy may also be a potential moderator in the association between family functioning and medication adherence. However, no research has examined the moderating effect of medication literacy.

Family plays an important role in medication adherence of patients receiving life-long anticoagulant treatment, especially in domestic life after discharge. However, the underlying mechanisms of the association remain unknown. To fill the gaps in the literature, this study 1) examined the mediating effect of illness perceptions between family functioning and medication adherence, and 2) tested the moderating effect of medication literacy on the relationship between family functioning and medication adherence, and the relationship between illness perceptions and adherence. Knowledge about the mediating and moderating mechanisms between family functioning and medication adherence could elucidate the mean of improving medication adherence to develop more effective strategies.

The medication adherence of MHVR patients is closely related to adverse events (Smet et al., 2018). Therefore, paying attention to medication adherence and its influence mechanisms is of great significance to improving patients’ long-term survival. In this study, we examined the mediating effect of illness perceptions on the relationship between family function and medication adherence, and the moderating role that medication literacy played in the relationship. To the best of our knowledge, this is the first study to reveal the underlying mechanisms between family functioning and medication adherence, which provides a basis for further targeted intervention.

In a previous study, family functioning was found to be closely correlated with medication adherence in Chinese patients (Sun et al., 2019). Consistent with the prior study, our results also showed that good family functioning was a positive predictor of medication adherence in patients with MHVR. Family is the main source of support in the management of disease, for paying attention to members’ needs and having a profound impact on the behavior of the individual nested in it (Pereira et al., 2014; Barnes et al., 2020; Yeom and Lee 2020). Good family functioning provides adequate family cohesion and sufficient family coping ability to give practical support, including reading medication labels, picking up medications, and supervising medication-adverse reactions timely to improve medication adherence (Pereira et al., 2014; Zhang 2018). Early detection of risks caused by medications may also reduce the patients’ worries to improve adherence to oral anticoagulation therapy (Hwong et al., 2017).

In addition, the effect of family functioning on medication adherence was partially mediated by illness perceptions in this study. In other words, good family functioning helps to improve patients’ illness perceptions, which in turn impacts medication adherence. Family members provide emotional and appraisal support, bringing social interaction and relieving negative emotions including fear of illness (Mayberry and Osborn 2012; Ao et al., 2020). It was also mentioned in McMaster’s Family Functioning Model that sufficient family functioning helps improve the individual emotional response (Pires et al., 2016). Patients whose families act openly, express emotions directly, solve problems effectively, and communicate information directly, have lower negative illness perceptions (Postolica et al., 2017). In Common Sense Model, individuals’ emotional representations of illness are also precursors of behavioral change like adhering to medication (Shiyanbola et al., 2018). Thus, healthcare professionals should value the family status of patients with MHVR, and could integrate family involvement into the medication guidance and management.

In recent years, medication literacy has received increasing attention. A cross-sectional study conducted on 420 Chinese hypertension patients reported that medication literacy was an important predictor of medication adherence (Shi et al., 2019). Same as the previous study, a high level of medication literacy was also closely associated with medication adherence in patients with MHVR. In addition, medication literacy also played a moderating role. Specifically, the relationship between family functioning and medication adherence of patients with limited medication literacy was significant, but this relationship does not hold for patients with adequate medication literacy. Medication literacy contains elements including knowledge, attitude, skill, and behavior (Shen et al., 2020). Patients with limited literacy have difficulty in obtaining drug information actively, understanding drug label information correctly, and calculating drug dosage exactly (Soones et al., 2017). Moreover, patients may have skeptical perceptions about oral anticoagulants, resulting in prolonged duration of medication, adjustment of drug dose, or even unauthorized discontinuation (Lee et al., 2017). Limited literacy also increases doctor-patient communication barriers, resulting in patients being less involved in medical decision-making (Cordina et al., 2018). Therefore, patients with poor literacy are more likely to seek assistance and supervision from their families about how to use the medication correctly. Conversely, medication intervention with family involvement may not be effective for patients with adequate literacy. Therefore, medication literacy may be regarded as a promising indicator to distinguish whether people require family involvement in promoting adherence. For healthcare providers, they should also evaluate the level of medication literacy before guiding self-management of medication and pay more attention to patients with marginal literacy.

Literacy was reported as a moderator in a study conducted on 174 patients with diabetes in the United States (Shiyanbola et al., 2018). Similarly, our results also showed that medication literacy buffered the negative impact of illness perceptions on medication adherence. In other words, an increase in threatening illness perceptions may not always be associated with a sharp decrease in medication adherence for individuals with adequate literacy. Negative illness perceptions are independent of the surgery or medical changes and result in negative attitudes or actions for a long time after valve replacement (Kohlmann et al., 2012). Patients with adequate medication literacy increase their self-efficacy and sense of mastery during the process of actively acquiring and correctly utilizing information (Shen et al., 2020). This allows them to participate in their treatment more actively. On the other hand, in obtaining information, patients may also have a deeper understanding of the necessity and importance of receiving anticoagulants after MHVR. Thus, patients with high medication literacy can show better medication adherence even if they perceive a severe illness threat. However, for patients with limited literacy, improving negative illness perceptions may also contribute to medication adherence. Overall, our findings suggest that in focusing on medication use, it is necessary not only to understand the severity of patients’ illness perceptions but also to consider differences in individual literacy levels.

There are still several limitations. First, we cannot infer any causal relationships from the cross-sectional design, nor can we observe the dynamic relationships between variables over time. Future research may apply longitudinal or experimental designs to validate relationships assumed among these variables. Second, a convenience sampling method was applied in this study, resulting in selection bias. Third, the data of this study were collected from patients’ self-reports. Multiple sources of information (such as family reports) can provide more rigorous tests for research hypotheses. Finally, the model of this study was developed for patients after MHVR. The results of this study may not be generalizable to other populations. Future research may also benefit from testing this model in other samples.

This study indicated illness perceptions can be one possible mediating mechanism underlying the relationship between family functioning and medication adherence. Moreover, medication literacy moderated the direct and indirect relationship between family functioning and medication adherence, with the relation being more significant for patients with low level of medication literacy. Efforts to improve medication adherence by targeting at improving family functioning may be more effective when considering illness perceptions, especially for patients with limited medication literacy.