In the present study, we developed a novel measure of intentional non-adherence to medical recommendations. The final version of the instrument consisting of 12 behaviors, demonstrated good psychometric properties and correlated strongly with the total score from the initial item pool, indicating practically no loss of information after the abridgement.
Importantly, a single common latent dimension was found to underlie a diverse set of non-adhering behaviors, suggesting that adherence is a unitary construct, broader than medication-taking behavior. Namely, self-medication, changing or adapting the prescribed therapy without or contrary to medical advice, ignoring symptoms, and postponing to seek treatment proved to form a relatively homogeneous category of behaviors reflecting the same general tendency to disregard official medical recommendations.
Non-adherence to public health and other health recommendations, such as avoiding using sunscreen or toothpaste with fluoride, rejecting tooth filling, refusing or postponing a child’s vaccination proved not to be a part of the same construct as the rest of the iNAR behaviors. There was also virtually no correlation between iNAR and vaccination against COVID-19 and only a weak negative correlation with following guidelines to protect against COVID-19. Taken together, our results suggest that these types of non-adherence behaviors have somewhat different precursors.
iNAR behaviors were more strongly negatively related to both self-reported health status and the presence of chronic diseases. It is possible that participants with more health issues have had more opportunities to disregard health recommendations than people who are generally healthy. Likewise, it may be that intentional non-adherence behaviors contributed to negative health outcomes.
Relationship Between Inar, Irrational Beliefs And Experiences With The Healthcare System
iNAR correlated only modestly with conspiracy mentality and superstition, while there was no relationship between iNAR and COVID-19 vaccination, which has previously been strongly related to a conspiratorial worldview [32]. This suggests that the reasons for intentional non-adherence to medical recommendations do not primarily stem from core irrational beliefs, but from other individual or contextual factors.
In our study, the irrational mindset did not significantly improve the prediction of non-adherence when other predictors were accounted for. This might be because, in contrast to Oliver & Wood’s study [31], in which the authors measured medical conspiracist beliefs and found them strongly related to specific health behaviors (including using sunscreen, going for an annual physical examination and visiting the dentist), we measured general conspiracist beliefs and related them to an index drawn from a comprehensive set of non-adherence behaviors (which ultimately excluded precisely the types of behaviors Oliver & Wood tracked). It is important to note that, while our findings clearly suggest that other factors are of crucial importance for non-adherence, we did observe zero-order correlations with both conspiracy mentality and superstition. Superstition remained a marginally significant predictor when other predictors were controlled for, indicating a role, albeit small, of an irrational mindset in intentional non-adherence.
iNAR behaviors were consistently rooted in negative healthcare-related beliefs and experiences. Negative correlations with trust in the healthcare system and trust in healthcare practitioners are in line with previous findings that individuals with higher trust are more prone to follow medical recommendations (see [54] for a review). Further, a negative correlation between the normalization of patients’ passivity and iNAR indicates that the individuals who are more likely to hand over decision-making to their physicians are less prone to iNAR behaviors. When it comes to the predictive power of this set of beliefs and experiences, only negative experiences and, to a lesser extent, normalization of patients’ passivity contributed to the explanation of iNAR behaviors. This supports previous findings that patients who experienced poor communication with their physicians are less likely to adhere to therapy [55]. It also suggests that giving up personal control over the treatment contributes to higher adherence to the therapy, which was found to be the case in healthcare systems in developing countries [56]. It is possible that an unequal relationship between the patient and the medical professional benefits patients who are more submissive in a relation to their physicians, but repels patients who value autonomy and agency, who may thus refuse the medical recommendation they are given. Striving for shared decision making about the treatment would likely benefit both types of patients in improving their health outcomes. This is also in line with our results that individuals who reported more negative communication experiences with medical professionals were also less likely to adhere to recommendations.
The lack of additional predictive power of trust in the healthcare system and professionals might have occurred due to their relatively high correlations with experiences with the healthcare system. Although the literature proposes that trust is one of the key determinants of (non)adherence to the prescribed treatment, it also documents a relatively high overlap between trust and previous experiences (e.g. [57]), similarly to the pattern we observed.
Importantly, iNAR behaviors were shown to be distinct from an unhealthy habitual non-adhering behavior, i.e., smoking, as demonstrated by a weak relationship between the two. In addition, our novel intentional non-adherence measure and smoking were predicted by entirely different sets of variables. For instance, while educational level correlated negatively with smoking, consistently with previous internationally and locally relevant findings [58–59], we did not observe any socio-demographic differences in intentional non-adhering. It seems that system-related, rather than patient-related predictors are important for this type of behavior: non-adherence is mainly rooted in negative experiences and distrust of the medical system.
Limitations And Future Research
While we found meaningful relations between iNAR on one side, and health status, healthcare-related beliefs and experiences, and irrational beliefs on the other, the correlations were only weak to moderate. One reason might be that in order to obtain the clinically most relevant measure, we focused on behaviors instead of, for example, attitudes or intentions. The variability of behaviors is, on the other hand, dependent on a myriad of internal and external factors such as opportunity and available resources. To confirm the validity of the iNAR instrument, future studies should examine its relationship with other adherence scales.
Despite their brevity, almost all of our measures showed adequate reliability, with the exception of the Normalization of patient passivity scale. Since this scale was adapted from the longer Passivity normalization during childbirth scale [43], this might have affected its reliability. Future studies should thus aim to replicate the finding on its relation with iNAR using scales that are longer and developed to measure generalized normalization of patient passivity.
As for the scale format, we opted to exclude the ‘non-applicable’ option in our iNAR scale, which could have led to us capturing “accidental” adherence - participants that had no opportunity to adhere were clumped together with those who always adhered to recommendations. We excluded this option since we assumed that most people were at some point in their life in a situation where adherence was required, and that those who had no opportunity to adhere (e.g., take medication or visit a dentist) were very rare.
Since we used only smoking as a representative of habitual non-adherence behaviors, future studies could include a more comprehensive set (e.g., adding behaviors such as exercising, or healthy eating) to better disentangle the differences between different predictors of habitual non-adherence and iNAR. In addition, although patient-related factors examined here did not prove central for predicting iNAR, these behaviors might be rooted more deeply; personality traits, such as conscientiousness and agreeableness, could also be tested as predictors.
As we aimed to develop a novel measure of intentional non-adherence (i.e. we primarily focused on psychometric properties and its relationships with the irrational mindset and experiences with the healthcare system), a convenience sample was optimal for this study. To gain further insights into the prevalence of iNAR, future studies should use probability samples.
Implications
The results of the current study suggest that interventions need to focus on establishing trust in the healthcare system, the medical profession and prescribed therapies. Implementing systemic changes in healthcare to support patients and build trust is a long-term process; broken trust is difficult to repair. Although our results indicated that passive patients are less likely to engage in iNAR, we would not suggest fostering further passivization. Instead, we would focus on improving patients’ experiences with the system and their healthcare providers: non-adhering patients reported mistrust in the healthcare system and repeated negative experiences, which proved to be more important than positive ones. This may seem self-evident, but our study demonstrated that such negative experiences are by far the strongest predictor of intentional non-adherence, outperforming those proven relevant to other types of questionable health behaviors: for example, sociodemographic for habitual non-adherence such as smoking [58–59], or irrational beliefs for the use of traditional, complementary and alternative medicine [34].