Therapeutic compliance has been defined as the  degree to which the behaviour of a person corresponds  with the recommendations of the health professional [1].  Several terms associated with the concept of adherence  to treatment are used: therapeutic alliance, cooperation,  compliance, mutuality, collaboration; among others. Each  term has a particular nuance but all focus on highlighting  the ability of the patient to engage, participate and  respect the agreement on treatment recommendations  co-established by the health professional [2].  In reality, there has been a process of transition from  the term "compliance" to that of "adherence", and successively to "therapeutic alliance". This is the step from an "objective" vision focused on the doctor, to an "objective-subjective" one focused on the patient and the doctor-patient relationship. Or in other words, from a hierarchical relationship (active doctor, passive patient), to another "hierarchical-paternalistic" (paternalistic doctor, passive patient), and another based on "collaboration" (inter-subjective dialogue), and finally to a "Shared decision model" (active participation) [3]. A large body of research finds that in various settings, from 30%-60% of patients with chronic illness (such as hypertension, bronchial asthma, depression, etc.) are not adherent to regimens, and this figure can be considerably higher in developing countries [3,4-11]. Medication nonadherence is a major impediment to the management of diseases and risk factors. Pharmaceutical treatment is essential to the management of most chronic diseases, but patients' failure to take medications as prescribed often results in failure to meet treatment goals. Medication nonadherence has been associated with a worse evolution, a greater number of relapses and a higher economic cost. Even in the case of diseases where the treatment saves the patient's life, many of them do not comply adequately.

Suboptimal adherence can result not only in  progression of disease, but in drug resistance, often  to multiple classes of drugs. Improving adherence is  thus of vital clinical and public health importance, and  leads to preventable costs and hospitalizations [12-14].  Medication adherence is thought to be the principal  clinical predictor of positive outcomes, not only  for serious mental illnesses such as schizophrenia,  bipolar disorder, or depression, but also for physical  conditions such as diabetes. Consequently, research on  medication often looks not only at medication condition  (e.g., placebo, standard medication, investigative  medication), but also at adherence in taking those  medications within each medication condition [15]. As  average life expectancy increases, so do the incidence of  chronic diseases and the number of persons receiving  long-term drug therapy. Thus elderly patient's noncompliance  with medication regimens has the potential  for sweeping medical and economic consequences and is  likely to become increasingly important in the design of disease-management programs for this population [16].

Actually, two "therapeutic continuities" can be  differentiated, referring to chronic or long-term  pathological processes with slow variations of their  clinical picture, which makes the patient live with  his disease such as diabetes, hypertension, stroke,  schizophrenia, dysthymia, and a pharmacological  discontinuity in acute processes or episodes that resolve  in a short time or not very long. A through-stages-oflife  compliance can also be differentiated (with the  passage from child to adolescent and adult, and the  rest of generational turnings points) [17], for example,  during the transition to adolescence of children with  type 1 diabetes, a reduction is observed of the frequency  of glycaemic monitoring, with marked repercussion  on metabolic control, even after statistical adjustment  for important covariates [18], and a longitudinal adherence, during each longitudinal stages of life [3].

Despite the potential impact on the health and  expenditure of the population, it has been found that  interventions that promote adherence to medication are  only moderately effective. Understanding the relative  influence of the factors that affect adherence decisions  to the patient's medication and the characteristics of the  individuals associated with the variation in adherence  will lead to a better understanding of how future  interventions should be designed and directed [19].  While much research has focused on determining the factors responsible for poor compliance, studies have tended to focus on the easy-to-measure characteristics of patients, such as poly-pharmacy, sex, age, or illness, which, unfortunately, generally they are neither predictive nor alterable [20]. On the other hand, accurate measurement of adherence behaviours remains a difficult problem without direct resolution; one measure used is the number of patients who have left the clinical trials, but extrapolation of this finding to the usual situations in real life or to all healthcare systems is not be easy [10,15,21].

Fundamentally, the construct of adherence is not one  behavior, but a complex of behaviours, and therefore  is multiply determined. Behavior in one domain  (e.g., diet) may not correlate well with behaviours in  another domain (e.g., medication taking). Furthermore,  research on various approaches to assessment and  measurement has failed to produce a widely agreed  upon 'gold-standard' for assessment. Even more  objective assessment methods (e.g., pill counts vs.  electronically-monitored dosing) are associated with  measurement errors and bias [22]. Specific issues of  measurement and assessment may differ across disease,  time course, and patient, among other factors [23,24].  Examining the relationships between evaluation  methods can also provide useful information about the robustness and consistency of observed effects [10].

Research should seek to understand the best way  to capture the dynamic and heterogeneous nature  of adherence over time, whether by developing and  validating new methods, refining existing ones or, perhaps  more importantly, addressing broader conceptual  issues that may present problems for empirical study. If  adherence is really a dynamic process that changes over  time, it is unlikely that single assessments of adherence  at any given time, whether evaluated by subjective  or objective measures, completely capture the selfcontrol  behaviours of the disease and they probably  account for variation in the results of the disease [13].  In this scenario, from general medicine, how can  we understand the lack of adherence to medical  advice or pharmacological prescription? Why do not patients comply? What are their reasons?

Different models or explanatory dimensions

The World Health Organization described 5  "interacting dimensions" that affect adherence,  including social and economic factors, conditionspecific  factors, therapy-related factors, patient  factors, and factors related to the health care  team, including provider-patient interactions [25].  From a broader perspective, adherence to treatment  can be analyzed from different perspectives considering different explanatory models: psychological, biological, or sociological. The biomedical model (numerical, mathematical), try to identify factors that can cause adherence and seeking to use this information to lead the patient. The behavioural/social model builds adherence as a behavior that is learned. Another explanatory model is the communicational model that associates poor adherence to rupture or asynchronous communication between the doctor and the patient; consequently the improvement of the communication improves the adherence. Another model is the rational model of beliefs, applied to beliefs in the area of health; this model assumes that adherence can be modified based on rational thinking, and It argues that if a patient works and uses logical thinking and balanced considers the risks versus the benefits of the treatment, it will lead to greater adherence. The self-regulatory system model that takes cognitive factors and the patient's planning capacity into consideration; patients are seen as a subject capable of solving their problems and therefore develop a consistent behavior. Finally, the community model, which assesses the factors that contribute to the lack of adherence to medication among adults with chronic disadvantages (acceptance of the disease, healthy eating habits, preventive behaviours, positive mental attitude, health practice, sex, levels of education and the short duration of the disease, etc.) [2,4,12,26].

However, no single model incorporates all  the variables that explain a d herence to treatment,  and cannot be applied to all patients to a specific  medical condition or a specific context [20,27].  In any case, there are a number of factors associated with  adherence to treatment that are the most researched and  in which it is admitted that they are the most important  for predicting adherence, such as the beliefs and attitudes  of the subjects about their disease and treatment,  which explain up to 20% of the variance of adherence  in patients with chronic diseases. There are other  factors that influence the adherence to treatment and  that are not directly related to cognitive variables. Table  1 shows these factors investigated more frequently [28].  However, the findings are inconsistent with respect to  the effects of patient's age, sex, socio-economic status,  housing, co-morbidities, number of visits to the doctor,  knowledge, attitudes, beliefs about health, intervention by  mobile phone, etc. And many studies present numerous  errors and statistical biases in their measurement [15].  Until the results, of additional comprehensive studies,  become available, current knowledge should be taken  into account with great caution when designing and  implementing disease management programs [16,29-31].  So, really, why do our patients not take medications  properly? Does non-compliance occur simply by chance?  The usual studies show statistical-mathematicalrelationship  between many treatments and non-adherence  factors, but they lack practical interest, because they do not really know the reason for non-compliance. These studies, generally of quantitative design, cannot investigate the truth of why patients do not comply. We would have to say with the philosopher Blaise Pascal that "The heart has reasons, which the reason cannot understand" [32].

Table 1: Factors more investigated as associated to the adherence to the treatment. 

Heart drives a variety of sometimes contradictory  emotions. That heart that makes us say things that  we would never have revealed or do what we never  thought. According to Blaise Pascal, the man endowed  with reason is susceptible to open his heart; it is this  ability that he calls "thought" that makes his humanity.  However, the heart is often stronger than the reason.  Because the heart is not reasonable, it does not  listen to reason, it has its own reasons. The heart has  reasons that the reason ignores! Thus, those reasons  are mostly personal, and there is nothing reasonable about them, they are driven by emotions and desires.

Consequently, there are reasons (of the patients) that  the reason (of the professional) does not understand.  In this way, the authors show some reasons that their  patients do not comply with treatment, but they are still  very superficial reasons, since they start from reason and  not from emotion. Behind those reasons there would still  be more reasons. The more we delve into the reasons,  the more we can understand not only why patients "do  not comply with our treatments", but why "patients do  not want, they do not know, or they cannot fulfil it". In  short, the design of studies or individual reflection to  bring light to the subject, should not be quantitative, or  focus-exclusively on a biomedical model, but qualitative,  "emotional", or if you want, qualitative-quantitative,  to be able to inquire into the "true reasons". So far, the  data and results about non-compliance are predictable  and do not add anything to the known ("they do not  shed light on the problem"). Biomedical research fails to  clarify the causes of noncompliance, insists on analyzing  "superficial" factors of noncompliance, and thus has  little practical interest, especially in general medicine.  The reality is that beyond the various models of adherence  and evaluation techniques, the patient often convincingly  affirms that he remains adherent. The explanation of these  phenomena is not easy because of everything said before,  but a first approach may be that, when motivational,  neuropsychological and behavioural factors have been  ruled out, it is important to discover and analyze the  presence of sociocultural factors with the patient [33]. And  to deepen the true reasons, how is it done? If the patient  feels that their voice is listened to carefully, they will begin  to identify more clearly the meaning of each symptom  and event related to their illness. You can also name, little  by little, the knots that "slow you down in your journey"  due to the disease. Sometimes we hear the words as  hollow words of personal connotation and classify them  as symptoms, but we do not hear the key elements of the  tone of the voice, the linguistic expressions and their  psychological world. The general practitioner (GP) has to find/know some reasons why his patients do not comply with the advice, but without it happening to us as it happens many times: we stay with the superficial; we are left with what are still superficial reasons. Behind those reasons there would still be more reasons. The more we delve into the reasons, the more we can understand it [34,35]. To increase the capacity to understand why the patient does not comply, it is necessary to return to the basics: "understand" the patients and ourselves in the respective contexts. The patient has his reasons that the doctor's reason does not understand, but the doctor has reasons that the patient's reason does not understand. Only by tuning in both directions can a "click" of understanding arise based on both the psychological and physical aspects of both actors and significantly increase adherence. This is a dialogue or conversation between experiences, patient and doctor: a "double tuner" bio-psychosocial [36], where we recognize ourselves and recognize the relationship that exists between the disease and the patient's personal and social life [37]. The adherence or lack of it depends mainly on the doctor-patient relationship. Diagnosis ("evaluation") means discovering what brings the patient and how he sees the problem; but the patient also makes a diagnosis (evaluation) of the therapist and of his ability to understand and treat his problem [36,38,39]. Improving doctor's communication skills improves outcomes for both patients and doctors [40,41]. We could summarize, that the main one-perhaps the only reason why many patients comply with the treatment is the good relationship they have with their doctor. And many doctors are horrible in the human relationship [42,43].

Although there has been a process of moving from  the concept of "compliance" (centered on the doctor)  to that of "adherence", and successively to "therapeutic  alliance" (centered on the patient and the doctor-patient  relationship), the basic concept It remains complex  and has multiple facets and models of approach. In  any case, more than half of the population abandons  the pharmacological treatment, and medication nonadherence  is a major impediment to the management  of diseases and risk factors, with great public health  importance, and leads to preventable costs and  hospitalizations. Many investigations have been directed  to determine the factors responsible for low compliance,  and there are a number of factors associated with adherence  to treatment that are the most investigated. However, the  findings are inconsistent with respect to the effects of  numerous variables, and lack practical interest, because  they do not really know the reason for non-compliance.  Non-adherence does not occur at random, but it cannot  be correctly predicted by medication beliefs, chronic  disease, and socio-demographics characteristics.  Health care providers, especially GPs, may consider  routinely inquiring about non-adherence in order to proactively address patient's suboptimal medication beliefs, but you have to remember that usually you are left with superficial reasons of little practical utility. "The heart has reasons, which the reason cannot understand"; there are reasons (of the patients) that the reason (of the professional) does not understand. The more we delve into the reasons, the more we can understand not only why patients "do not comply", but why "patients do not want, do not know, or cannot" comply with our advice and pharmacological prescriptions. Medication-taking decisions are not the subject of rational choices by patients, influenced only by the attributes of treatments and psychosocial, socio-demographic, and disease-related factors, etc. The GP should try to know "the reasons for the reasons" for the lack of adherence of the patient, and this can only be done by "understanding" the patients and ourselves in the respective contexts: a kind of double tuner to listen to the other and to listen to ourselves about our reasons for how we understand the other. If the patient feels that their voice is listened to carefully, they will begin to identify more clearly the meaning of each symptom and event related to their illness. Once this is understood, each professional has to decide what to do next. Understanding individual treatment preferences is an important step to improving adherence support provision in practice. Satisfaction with the doctor is the only reason why many patients meet the treatment. But some doctors are horrible in the human relationship.