Behavioral factors of medical center patients regarding treatment compliance: confirmatory factor analysis model

Introduction

According to the World Health Organization (WHO) (1), adherence to medical treatment can be understood as the degree to which an individual’s behavior (for example, taking medications) aligns 36 with the recommendations of a health care provider.

Therapeutic recommendations arise from outpatient or hospital care to solve health problems. The medical staff elaborates and delivers treatment and detailed in-formation on the therapy (2), with the intention that the patient will follow the indications precisely. However, this presents an important challenge for health professionals (3) because the levels of therapeutic adherence are usually low as indicated by Ortega Cerda et al. (4).

A lack of adherence to treatment means that the patient’s medication-taking be-haviour does not correspond to that recommended by the doctor. This means that patients do not obtain the benefits of the prescribed treatment, achieving effects contrary to the desired outcomes, such as worsening of symptoms, early appearance of complications and an increase in the frequency of hospital admissions (5).

Nonadherence negatively affects the efficacy and safety of therapies. This is a complex problem derived from multiple factors, so solutions from different approaches are required (6,7). Otherwise, if the expected health benefits are not obtained, both patients and society in general must face the economic burden derived from health complications (8).

Preventing non-adherence to medication is an important aspect for the effective-ness of health systems worldwide (1). The high costs associated with medication non-compliance contrasts with the inherent difficulty of improving therapeutic adherence. This phenomenon negatively impacts both clinical outcomes and healthcare system efficiency, representing a persistent challenge for healthcare professionals and medical institutions (5).

Thus, achieving improved adherence to medication is an important challenge for health care providers (8). Such adherence can be influenced by various factors, such as perceptions of the disease, health literacy, self-efficacy, cognitive skills such as memory, coping skills and problem solving, as well as psychosocial factors such as personal beliefs and cultural issues related to taking medications (3), among others. Therefore, it is necessary to promote the study of factors that may have a negative influence on adherence, since this would improve health care, reduce comorbidities and mortality and promote patient safety (1). Additionally, it would facilitate the identification of patients at high risk of noncompliance, barriers to adherence that could be eliminated, and individualized interventions for the improvement of adherence (6).

Thus, adherence to medical treatment is a critical problem in global health care, with significant implications for both individual health and health systems in general. Non-adherence to medical recommendations not only compromises the effectiveness of therapies, but also generates patient complications, increases hospitalization rates and raises associated economic costs. Therefore, this study is relevant in seeking to understand the individual, social and educational barriers that impact adherence, which can lead to more personalized and effective interventions.

The aim of this study is to analyze the factors that influence adherence in patients with chronic diseases and to determine how elements such as patient motivation, social environment, access to health services, and socioeconomic factors affect their ability to adequately comply with prescribed treatment regimens. The foregoing is relevant since therapeutic adherence continues to be blamed for the most part on patients, but they need the support of health professionals to successfully adhere to medical treatment (1).

In this context, adherence is considered to be the result or consequence of multiple influencing factors, rather than an isolated factor. Therefore, the study focuses on understanding the determinants that affect adherence, with adherence being seen as a manifestation of the interaction of these factors.

This research explores six factors from the literature reviewed using confirmatory factor analysis (CFA). These factors are perceived disease severity, adaptation of medical regimen, adherence to treatment, likelihood of taking preventive measures, provider preference, and educational strategies.

Although there is literature examining factors such as educational level, emotional impact of the disease, and family support on treatment adherence (6), gaps persist in understanding how these factors interact with technological tools and educational strategies to promote long-term adherence. In addition, evidence related to specific populations or complex chronic conditions remains limited (3). This study provides a model that not only identifies key determinants of adherence but also suggests contributions to expand understanding of therapeutic adherence, laying the groundwork for future research aimed at integrating digital technologies and educational programs in the management of adherence. Also, the application of methodologies such as CFA opens up new possibilities for validating models that integrate individual and contextual factors more precisely.

Methods

Research hypotheses

This research explores six factors from the literature examined by means of a CFA. These factors are perceived severity of the disease, adaptation of the medical regimen, adherence to treatment, probability of taking preventive action, provider preference, and educational strategies.

Perceived severity of the disease

Among the relevant factors to consider in adherence to medical treatments is the perception of the severity of a disease. This refers to the judgement that a person makes about the seriousness of a disease and its consequences, which is mediated by emotional and cultural issues to which the person is exposed (9). Therefore, the following hypothesis is proposed:

H1: perceived severity positively influences the probability of taking actions to adhere to medical prescriptions.

Adaptation of the medical regimen

This construct refers to the ability of the patient to incorporate the recommended treatment into his or her daily routine, taking advantage of existing habits; the medication should be presented in the most pleasant way for the user (10). The hypotheses are as follows:

H2: the adaptation of the medical regimen positively influences the probability of adherence to medical prescriptions.

H3: the adaptation of the medical regimen negatively influences the probability of adherence to medical prescriptions.

Adherence to treatment

This is a key behaviour in the treatment of patients since it increases the effectiveness of the recommended treatments and interventions, which in the future means a positive improvement in their health (4). Likewise, the WHO indicates that this refers to the “the extent to which a person’s behaviour taking medication, following a diet, and/or executing lifestyle changes—corresponds with agreed recommendations from a health care provider” (1,11). From this, the following hypotheses are formulated:

H4: adherence to treatment has a positive influence on the probability of adherence to medical prescriptions.

H5: adherence to treatment negatively influences the probability of adherence to medical prescriptions.

Probability of taking preventive action

This construct can be considered a consequence of the individual’s attitude towards adherence to medical treatment due to the possible adaptation of the medical regimen that has been offered; these actions are also mediated by access to preventive campaigns that motivate the willingness to comply with the recommended medical treatment (12). Due to the above, the following hypothesis is proposed:

H6: preventive actions have positive effects on the fulfilment of medical prescriptions ordered by health professionals.

Provider preference

The patient-provider interaction is essential for the fulfilment of a treatment. It is a point of intervention in the attempts to improve levels of adherence, which depends on the accuracy with which the patient remembers the provider’s instructions (in non-technical language) and the tone or emotional impact of the interaction (where a socioeconomic characterization is made in nonnegative terms) (10). This prompted the following hypothesis:

H7: the health professional’s preference determines the possibility of fulfilling the medical prescriptions by patients.

Educational strategies

This construct refers to the pedagogical actions deployed in order for the patient to adopt a positive attitude towards the treatment of his or her disease from diagnosis to treatment and the importance of adherence (11). In this sense, health personnel must evaluate the capacity of the patient to understand the provider’s instructions and to adhere to them (10). In accordance with the above, the following hypotheses are pro-posed:

H8: the adoption of educational strategies positively impacts the fulfilment of medical prescriptions ordered by health professionals.

H9: the educational strategies used constitute a positive stimulus for taking preventive actions.

Study design

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This research was approved by the Ethics Committee of the Institución Universitaria Escolme (under the act No. 10042023) and informed consent was obtained from all individual participants.

A nonexperimental investigation was carried out under a quantitative approach of descriptive and explanatory scope to establish and verify the relationships of factors that intervene in the fulfilment of medical prescriptions ordered by health professionals.

Sample and data collection

The sample size of 137 participants was determined based on accessibility and proximity criteria, using non-probabilistic convenience sampling. This sample size was deemed sufficient for the present research, as it meets the generally accepted minimum statistical power requirements (typically 0.80) to detect significant effects with a 95% confidence level and a moderate effect size. The selection of this sample size was informed by the principle of ensuring adequate variability in responses and the robustness of the statistical analyses performed.

For the present investigation the inclusion criteria for participants were patients over 18 years of age with a formal medical diagnosis, in active treatment (in which they consult regularly), and affiliated with the social security health system. As for the exclusion criteria, patients with mental disabilities that limited their understanding of the questionnaire or who did not agree to participate voluntarily in the study were excluded from the study. We were recruited using nonprobability and convenience sampling according to the possibility of access and proximity (13). This method was chosen due to limitations in access to participants. The study was conducted during the first semester of 2023 in two medical centers in the city of Medellín, Colombia. Participants were recruited in the waiting rooms and primary care offices of these institutions.

Regarding the sociodemographic characteristics of the participants, 67% were male and 33% were female. Regarding their age, most of the participants were ages 21–30 years (60%), 31–40 years (14%), and older than 40 years (26%). Additionally, 94% were direct contributors to their social security, 4% were beneficiaries, and 2% were subsidized by the state. Regarding the use of health services, more than 50% of the participants indicated that they had consulted a physician at least twice in the past year. Regarding the responses on the health problems mentioned by the participants, the most common ailment was headache, which affected 34.3% of the population. This was followed by colds or flu (23.8%) and digestive disorders (12.8%). Other important health problems were joint (16.9%), muscle (16.3%) and throat ailments (12.2%). The least frequent complaints were gastric (11%), bone (8.1%) and diarrhea (6.4%). Also, the participants reported that they were receiving treatment for other types of diseases such as respiratory diseases (12%), hypertension (13%), diabetes (5%), heart problems (1%). This sociodemographic information on the participants can be seen in Table 1.

For data collection, a self-administered questionnaire was used, distributed in person and in an online form, and anonymity and confidentiality were guaranteed. A total of 147 surveys were conducted, which were reviewed upon receipt. Of these, 10 (6.8%) were not approved because the respondent did not answer all the necessary questions for the analysis.

For the latter, a unique code was assigned to each participant, eliminating any personally identifiable information. The online questionnaires were stored in an encrypted, password-protected database, and only the principal investigators had access to the data. This survey consisted of questions asking for basic data for the characterization of the population, followed by reasons for consultation (i.e., headaches, bone, joint, cold or flu, digestive, throat, diarrhea, gastric or muscular ailments), diagnoses, use of health services and behaviour regarding the administration of medications (with questions such as: did I receive a prescription after my consultation, Number of medications prescribed, when I feel well do I stop taking my medications, and what percentage do I estimate I stop taking my medications?). Finally, Likert-type questions were incorporated, which helped to measure the 6 factors of interest in this study. These questions were measured on a scale of 1 to 5, where 1 corresponded to strongly disagree, up to 5, which meant strongly agree, as presented in Appendix 1. The six selected dimensions were defined based on a review of the existing literature on adherence to treatment. During this process, relevant constructs that reflect both theoretical frameworks and empirical evidence were identified and categorized. To this end, priority was given to those factors that in previous studies showed consistent associations with adherence behaviors. The questionnaire was designed from scratch, ensuring that each dimension was grounded in the reviewed literature, and then empirically validated using CFA statistical techniques.

Statistical analysis

A CFA was carried out. In this type of analysis, the factors or constructs are previously identified from the theory, and what is expected is that the CFA will verify if the previous theoretical structure (model) fits the data through hypothesis tests (14,15). Thus, Cronbach’s alpha was calculated to measure the internal consistency of the instrument, the Kaiser-Meyer-Olkin (KMO) test was used to evaluate the viability of the model, and the Bartlett specificity test was used to confirm the applicability of the factor analysis. In this way, the reliability or statistical validation of the proposed measures was ensured. Cramer’s V coefficient was used to estimate the degree of association of the variables. The data obtained were analyzed with Statistical Product and Service Solutions (SPSS) in its 12th version, manufactured by International Business Machines Corporation (IBM).

Results

Proposed model

Given that studies carried out in developed countries indicate that 50% of patients do not comply with medical treatment (4), in this research, a model was proposed for the purpose of understanding the factors that intervene in compliance with medical prescriptions ordered by health professionals. The model underpinning the research is depicted in Figure 1.

Figure 1 Proposed model, Cramer’s V. Source: representation of the Anderson and Gerbing model (16).

Internal consistency of the questionnaire: validity of measurement scales

The use of the evaluation scales is based on psychophysics and psychometrics. Psychophysics refers to the process of quantification of perception. Thus, to transfer intangible phenomena to a numerical system, analogies must be established that trace the path to arrive at numerical valuation, and psychometry allows studying the adequacy of the scale to the phenomenon being measured and the quality of the measurement. Accordingly, it is observed that the process of construction and validation of a questionnaire is relatively complex and requires clear theoretical knowledge of the aspect we want to measure, as well as having advanced statistical knowledge and knowing how to handle computer programs to perform statistical tests (17).

For the present study, the verification of the validity of the measurement scales used, of each of the constructs and of the instrument in general was developed by means of a CFA with the statistical software SPSS. This analysis represents a set of various technical procedures for the study of the interdependent relationship between a set of variables in order to group them based on “shared variability”; discover the underlying structures (factors), dimensions or latent concepts, fulfilling the purpose of summarizing and reducing the data; thus, this is a very useful statistical technique if we evaluate the multidimensionality of a construct, since it allows an empirical exploration, considering that the objective is to select those items that correlate the most with the set of items that are measuring the construct (18).

Given the above, it should be considered that the reliability of the model is the degree to which an instrument measures accurately, without error. It indicates the condition of the instrument to be reliable, that is, to be able to offer truthful and constant results in its repeated use under similar measurement conditions (17). Further, it is valued at two levels: the reliability of the observable items and the reliability of the constructs (19). A reliability greater than 0.6 is considered evidence that the model is reliable (20). Likewise, the reliability of the constructs refers to the degree to which an observable variable reflects a factor, considering a value greater than 0.7 (21). Thus, convergent validity evaluates the degree to which the measurement of the items that include the same concept is correlated (19).

In this research, to achieve the union or convergence of the model, it was necessary to eliminate the indicators AR1 and VI1, given that their standardized factorial loads were less than 0.6, following the recommendations of Bagozzi and Yi (20). However, the average obtained from the loads of the indicators on each factor was greater than 0.7 for all the constructs (22), indicating the presence of convergent validity, as shown in Tables 2,3.

Subsequently, the Bartlett test of sphericity and the KMO measure of adequacy of the sample were calculated, and the level of conditioning of the model was determined to carry out a factor analysis. The first value to calculate is a statistical test that detects the presence of correlation between variables, offering the probability that the correlation matrix collects significant values, its p must be less than the critical levels 0.05 or 0.01; further, this test is very sensitive to increases in the sample size, and when samples are increased it is easier to find significant correlations (23). Bartlett’s test of sphericity tests the null hypothesis that the observed correlation matrix is actually an identity matrix. If the critical level (significance level) is greater than 0.05, it is not possible to reject the null hypothesis of sphericity, and consequently, it cannot be ensured that the factorial model is adequate to explain the data. Given that, in the model proposed in this research, the Bartlett values are less than 0.05, so it can be said that there are significant correlations between the variables.

Using a similar approach, the value of the KMO sampling adequacy measure is defined as an index that compares the magnitudes of the observed correlation coefficients with the magnitudes of the partial correlation coefficients, and its value is between 0 and 1. Kaiser (24) characterizes these values on a scale that considers KMO measurements close to 0.90 as wonderful, 0.80 as meritorious, 0.70 as moderate, 0.60 as mediocre and below 0.50 as unacceptable (25).

It is observed that in Table 4, the coefficients produced by the SPSS software for each of the constructs meet the criteria mentioned above, which indicates that it is feasible to perform the data reduction technique, that is, to obtain the minimum number of explanatory elements (factors) that clarify the reality of the factors involved in the processes of adherence and compliance with medical prescriptions.

Discriminant validity

Advancing in the analysis of the model, the stage of discriminant validity is reached, which refers to the fact that each factor must represent a different dimension, and for this to occur, each observable variable must be loaded to a single factor. In this sense, the further the Phi value is from 1, the greater the discriminant validity (25,26).

In accordance with the aforementioned, the analysis of discriminant validity was carried out by verifying that the confidence interval in the estimation of the correlation between each pair of factors did not contain the value 1 (16). Table 5 shows that all cases meet discriminant validity. It is necessary to mention that, in Table 5, the ellipsis (…) indicates that the correlation of a factor with itself is not calculated, since it would be redundant.

Subsequently, the reliability of the measurement scale was identified, and the explanatory capacity of the proposed model was verified, for which Cronbach’s alpha was calculated for the respective scales of each construct. This procedure is necessary because Cronbach’s alpha is an index used to measure the reliability of the internal consistency of a scale, that is, to evaluate the extent to which the items of an instrument are correlated (27). The test reaches positive values between 0 and 1, where 0 indicates total absence of internal consistency, and 1 indicates total redundancy between the items. Nunnally (28) proposed values between 0.75 and 0.9; Cea-D’Ancona (29), not less than 0.8. Peterson (30) considered a minimum of 0.7 for preliminary investigations and 0.8 for basic investigations. Morales et al. (31) recommended a value of 0.5 for basic investigations and more than 0.85 for diagnostic investigations and interventions, and (23,32) recommended a higher value of 0.70.

As seen in Table 6, the measurement instrument seems to have adequate reliability of internal consistency of the measurement scale, since all the values for Cronbach’s alpha are within the range of values recommended by the aforementioned authors.

The results of the confirmatory analysis show the existence of a sustainable factorial model for the analysis of the factors that affect compliance with medical orders by patients. The presence of convergent validity and discriminant validity within the instrument, together with an acceptable reliability, confirms that the instrument evaluates fundamental variables that directly or indirectly affect the understanding of the behavior of patients in response to their doctors’ prescribed treatment.

Hypothesis testing: analysis of results

We proceeded to estimate the proposed structural model for the process of adaptation of medical treatment by patients, where the various hypotheses raised are collected and their degree of association is measured by means of Cramer’s V statistic. This coefficient corresponds to a measure of association between two ordinal variables that takes a value between 0 and 1, where values close to 1 in absolute value indicate a strong relationship between the two variables and values close to 0 indicate that there is little or no relationship between the two variables (33).

Table 7 shows the values obtained from the SPSS software for the evaluated statistic and the model used. We can conclude that only hypotheses 1, 6 and 9 present a mean degree of association with values of 0.363, 0.374 and 0.304, respectively, for Cramer’s V statistic.

Cramer’s V statistic was extracted from the SPSS software and placed in a table of crossed factors, which allowed us to observe the degree of association between the variables that were part of the hypotheses and those that were not. This enabled us not only to verify the degree of association for the hypothetical relationships but also to corroborate that a high level of association was not presented among the other constructs. Table 8 shows all the relationships established between the variables of the proposed model. Next, the respective proposed model is presented with the association value between the variables.

According to the model (Figure 1), patients who receive an awareness workshop in their medical centers that inform them about the complexity of the medicine, side effects and even about the degree of behavioral change that such treatment can cause are more likely to carry out preventive actions (R=0.304). However, their behavior is not modified, and it is unlikely that they will adhere to the treatment (R=0.176) since the behavioral change required to follow a medical prescription implies the development of new behaviors and the termination of old habits.

Although the patient-provider interaction is very important, factors such as the cost and duration of the treatment can cause an individual not to adhere to the treatment completely (R=0.179). In this scenario, the individual variables take on a dominant role and are decisive in regard to adapting to the medical regimen and complying with the treatment (R=0.234).

Therefore, although the severity of the disease is a relevant factor in continuing medical treatment, it is even more significant to understand the individual lifestyle and preferences of patients (R=0.393).

Finally, because Cramer’s V is a measure of directional association, the results presented in Figure 1 demonstrate the relationships established between the constructs, and it is concluded that the variables corresponding to hypotheses 1, 6 and 9 showed significant degrees of association with the constructs used to establish the hypothetical relationship. In short, it is verified that the main variables of the model are concentrated in the constructs: individual variables, which show the psychosocial variables of the patient, such as their education, emotional impact of the disease and their environment; perceived severity of the disease, which can influence how likely a patient is to adopt or not adopt a medical treatment given the level of complexity of their disease; adaptation to the medical regimen, which reveals when patients feel that the treatment generates well-being and offers a cure; probability of taking preventive action, in which happiness increases and patients take the initiative to implement a better lifestyle; and educational strategies, which demonstrate the beliefs, behaviors and environment needed to adhere to a medical regimen.

Discussion

The results indicate that individual variables are the most important factors influencing patient adherence to medical treatment. Specifically, the data show that individual variables, such as the patient’s educational level, the emotional impact of the disease, and environmental factors, have the highest degree of association with adherence (Cramer’s V=0.363). This suggests that understanding the unique psychosocial characteristics of each patient is critical to predict and improve their likelihood of adapting to and adhering to prescribed medical regimens. In contrast, factors such as perceived severity of illness or cost/duration of treatment, although relevant, have a weaker association with adherence. Therefore, a patient-centered approach that addresses individual patient needs, beliefs, and behaviors appears to be key to promoting better adherence to medical treatment.

The development of promotion and prevention programs are proposed as a way to ensure that patients with chronic diseases adhere to the suggested medical treatment. Studies show that constant participation in such programs leads to patient compliance with their medical prescriptions, which is essential to improve self-management behaviors in patients (34). This finding is in agreement with those findings of this study that patients are predisposed to follow medical recommendations after participating in awareness workshops and that this adherence depends on repetitive actions that lead to sustainable behavioral change over the long term. That is, a patient suffering from hypertension who participates in regular educational programs on the management of their condition (repetitive activities such as daily blood pressure monitoring, reduction of salt intake and regular physical exercise), show a significant improvement in adherence to their medical treatments and reinforces learning and facilitates the incorporation of these habits into their daily life and long term. Regarding the latter, the article by Keyser et al. (35) found that for greater adherence to the medical regimen through educational programs, clear criteria for the inclusion of participants must be in place to lead to more positive results.

About perceived severity that is relevant to determining adherence to medical treatments, highlights that patients who perceive their illness as serious tend to prioritize medical recommendations to avoid major complications, so it is important to understand the individual schemes of each person. A study on the intention to be vaccinated against coronavirus disease 2019 (COVID-19) (9) indicates that perceived susceptibility as a consequence of interaction with friends and family helps individuals take the recommendations of experts and epidemiologists to combat this disease with vaccination, although the authors of that study indicate that the media also plays a role in the perception of the severity of a disease.

In the study carried out by Ruiz-Aranjuelo et al. (36), adherence to medical prescriptions is also associated with the levels of family support, education level and the preexistence of mental health diseases, which is in line with the results of the current study.

Practical and theorical implications

In terms of practical implications, the relevance of effective educational and communication strategies to promote adherence to treatment is highlighted. The implementation of awareness workshops that highlight both the perceived severity of the diseases and their impact on the patient’s quality of life is recommended. These workshops should include interactive sessions where patients can express their personal barriers and receive personalized guidance, thus strengthening their commitment to treatment. Also, the use of information and communication technologies (ICTs), such as mobile applications and automated reminders, is presented as a key tool to facilitate follow-up and improve adherence. These tools can be integrated into treatment plans to remind medication schedules or preventive activities, helping patients to incorporate these actions into their daily routines and reducing perceived barriers. In addition, healthcare professionals can use the proposed structural model to identify patients with low perceived severity and implement targeted educational interventions, improving not only health outcomes, but also empowering the patient by providing them with a greater understanding of their condition. Similarly, for hospital management, these findings suggest establishing indicators to measure and monitor therapeutic adherence as part of institutional performance evaluation. In terms of health policy, the findings can support the creation of integrated programs that combine health education and information technologies, optimizing resources and strengthening coordination between health care levels.

In terms of theoretical implications, the structural model presented validates the importance of individual and contextual factors in adherence to treatment, reinforcing previous theories on patient behavior in the face of chronic diseases. It also highlights the need to explore complementary constructs, such as the impact of ICTs on the modification of health-related behaviors. These explorations open new areas of research, allowing us to expand our knowledge of the relationship between critical variables, such as perceived severity, educational strategies, and the likelihood of taking preventive actions. The associations identified in this structural model suggest possible lines of future research to delve deeper into the factors that influence adherence to treatment, especially in populations with historically low adherence.

Strength and limitation of the study

Strengths

The main strength of this study is the comprehensive and empirical perspective to examine the various factors that influence patients’ adherence to medical treatment. Using CFA and hypothesis testing, the study identifies key determinants, such as the strong association between individual patient variables, perceived severity of illness, and educational strategies. This provides valuable guidance for developing more effective and targeted interventions to improve adherence, focusing on improving patient education, addressing disease perceptions, and promoting preventive actions.

It should be noted that although the sample size does not allow statistical generalization to the entire population, it does provide a detailed and useful insight into the factors influencing adherence in this specific group of patients. In addition, as detailed in the Methods section, the sociodemographic characteristics and distribution of the sample (gender, age, employment status, and type of health insurance) are described, allowing us to contextualize the results within a specific patient profile. This characterization of the sample contributes to a better understanding of the factors affecting adherence in this particular group, which may be relevant to similar studies in similar contexts and populations.

It is recognized that a study with a larger and more representative sample might allow for greater generalization, and therefore we suggest that future research consider a design with random or stratified sampling to improve the representativeness of the results and allow for broader inferences. However, it is believed that this study provides valuable information about the patterns and determinants of adherence in the group studied, laying the groundwork for broader and more diverse research in the future.

Limitations

The study used non-probability convenience sampling, selected based on participants’ accessibility and willingness to collaborate. While this approach allowed for efficient data collection, it introduces limitations related to the representativeness of the sample. For example, the predominance of participants from a single geographic region could influence cultural or social variability, limiting the generalizability of the findings to other populations. Furthermore, the sample size of 137 individuals, although sufficient for the statistical analysis performed, might not capture the heterogeneity needed to explore specific subgroups, such as those with complex chronic conditions or low levels of health literacy.

Also, this study has methodological limitations in that it focuses on a variety of pathologies without specializing in any one disease. The conditions covered include diseases such as diabetes, hypertension, rheumatoid arthritis, cardiovascular diseases, among others. This lack of specificity in diagnoses could affect the results related to adherence, as the characteristics and treatments of each disease may differ significantly. In addition, the sample size, although adequate and accessible for the purposes of the study, is limited. Therefore, it would be useful for future research to include a larger number of participants and use more diverse sampling methods, such as statistical and random selection, which would allow greater generalization of the results obtained. As well as developing comparative studies according to group of diseases.

Another limitation of this study is related to the statistical technique selected for hypothesis testing, because although the Cramer’s V statistic has theoretical support, more advanced techniques such as path analysis, P value or regressions are currently used. Therefore, it is recommended that future studies on the subject be analyzed using techniques such as structural equation modeling (SEM) to improve the quality of the model.

Finally, future research could focus on the development of qualitative studies to explain the intrinsic motivation of patients to adhere (or not) to medical treatments and to identify the most common practices among patients. In addition, their differences according to geographic location (urban/rural) and cultural, gender or ethnic characteristics should be considered. New constructs to help explain phenomena such as the relationship between ICTs and patient compliance behavior. These findings could identify the best methods of patient care and follow-up to achieve greater compliance. Perceived severity of illness was found to be a variable that affects adherence to prescribed medical treatment, so it is necessary to continue research that analyzes this phenomenon from the perspective of patients with different diagnoses, such as chronic diseases, orphan diseases, and high-cost diseases. It is also recommended to fund experimental studies that seek to examine the study variables and systematize the best interventions. Such studies should favor therapeutic adherence, taking into account the individual characteristics of the participating populations.

Conclusions

Therapeutic adherence remains a critical challenge in healthcare. This study provides a structural model of the variables that influence adherence. The adaptation to the medical regimen and the likelihood of taking preventive actions may indicate that when patients manage to integrate the treatment into their daily lives, they are more likely to adopt preventive behaviors. The perception of disease severity shows a significant relationship with individual variables, indicating that the patient’s personal characteristics influence how they perceive the seriousness of their condition. These findings suggest the need for a personalized approach in treatment design, considering not only education about the disease, but also the integration of the regimen into the patient’s daily life. Future interventions should focus on strategies that address individual barriers and facilitate sustainable behavioral changes to improve therapeutic adherence.

The results of the CFA and hypothesis testing showed that the proposed model is statistically valid for examining the factors that influence adherence to medical treatment. In addition, it demonstrates that medical regimen adaptation, individual variables and provider preference act as barriers to medical regimen adherence. Favorable relationships were found for individual variables versus perceived severity of illness, as well as for medical regimen adaptation and educational strategies in relation to the likelihood of taking preventive measures.

Acknowledgments

We thank the Institución Universitaria Escolme for the allocation of in-kind contributions for the development of the research and the Universidad Señor de Sipán for the payment of article processing fees.

Footnote

Data Sharing Statement: Available at https://jhmhp.amegroups.com/article/view/10.21037/jhmhp-24-123/dss

Peer Review File: Available at https://jhmhp.amegroups.com/article/view/10.21037/jhmhp-24-123/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jhmhp.amegroups.com/article/view/10.21037/jhmhp-24-123/coif). All authors report that Universidad Señor de Sipán covered the article processing fees, while Institución Universitaria Escolme provided in-kind contributions for the development of the research. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This research was approved by the Ethics Committee of the Institución Universitaria Escolme (under the act No. 10042023) and informed consent was obtained from all individual participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Organización Mundial de la Salud. Adherencia a Terapias a Largo Plazo: Evidencia Para la Acción. 2004. Available online: https://iris.paho.org/handle/10665.2/41182
2. Campos-Caicedo C, Velasquez-Uceda R. Calidad de la receta médica en dos hospitales de Lambayeque y su influencia en la comprensión de la información brindada. Acta Méd Peru 2018;35:100-7. [Crossref]
3. Chia LR, Schlenk EA, Dunbar-Jacob J. Effect of personal and cultural beliefs on medication adherence in the elderly Drugs Aging 2006;23:191-202. [Crossref] [PubMed]
4. Ortega Cerda JJ, Sánchez Herrera D, Rodríguez Miranda ÓA, et al. Adherencia terapéutica: Un problema de atención médica. Acta Méd. Grupo Ángeles 2018;16:226-32.
5. Basu S, Garg S, Sharma N, et al. Improving the assessment of medication adherence: Challenges and considerations with a focus on low-resource settings. Tzu Chi Med J 2019;31:73-80. [Crossref] [PubMed]
6. Gast A, Mathes T. Medication adherence influencing factors-an (updated) overview of systematic reviews. Syst Rev 2019;8:112. [Crossref] [PubMed]
7. Muench U, Whaley C, Coffman J, et al. Scope-of-Practice for Nurse Practitioners and Adherence to Medications for Chronic Illness in Primary Care. J Gen Intern Med 2021;36:478-86. [Crossref] [PubMed]
8. Kvarnström K, Westerholm A, Airaksinen M, et al. Factors Contributing to Medication Adherence in Patients with a Chronic Condition: A Scoping Review of Qualitative Research. Pharmaceutics 2021;13:1100. [Crossref] [PubMed]
9. Cavazos-Arroyo J, Pérez de Celis-Herrero C. Severidad, susceptibilidad y normas sociales percibidas como antecedentes de la intención de vacunarse contra COVID-19. Rev Salud Pública 2020;22:178-84. [Crossref] [PubMed]
10. Masur FT, Anderson K. Adhesión del paciente al tratamiento: Un reto para la psicología de la salud. Rev Latinoam Psicol 1988;20:103-26.
11. Dilla T, Valladares A, Lizán L, et al. Adherencia y persistencia terapéutica: Causas, consecuencias y estrategias de mejora. Aten Primaria 2009;41:342. [Crossref] [PubMed]
12. Juárez-García DM, Téllez A, García-Cadena CH. Escala del modelo de creencias de salud para la autoexploración de mama en estudiantes universitarias. Acta Investig Psicol 2019;9:7-14. [Crossref]
13. Otzen T, Manterola C. Técnicas de muestreo sobre una población a estudio. Int J Morphol 2017;35:227-32. [Crossref]
14. López-Aguado M, Gutiérrez-Provecho L. Cómo realizar e interpretar un análisis factorial exploratorio utilizando SPSS. REIRE Rev Innov Recer Educ 2019;12:1-14.
15. Arango-Botero D, Valencia-Arias A, Bermúdez-Hernández J, et al. Factors that promote social media marketing in retail companies. Contad Adm 2021;66:1-22.
16. Anderson JC, Gerbing DW. Structural equation modeling in practice: A review and recommended two-step approach. Psychol Bull 1988;103:411. [Crossref]
17. Arribas M. Diseño y validación de cuestionarios. Matronas Prof 2004;5:23-9.
18. Pérez ER, Medrano LA. Análisis factorial exploratorio: Bases conceptuales y metodológicas. Rev Argent Cienc Comport 2010;2:58-66. (RACC).
19. Calvo-Porral C, Martínez-Fernández VA, Juanatey-Boga O. Análisis de dos modelos de ecuaciones estructurales alternativos para medir la intención de compra. Investig Oper 2013;34:230-43.
20. Bagozzi RP, Yi Y. On the evaluation of structural equation models. J Acad Mark Sci 1988;16:74-94. [Crossref]
21. Hair J, Andreson R, Tatham R, et al. Multivariate Data Analysis, 5th ed. Prentice-Hall Inc.: New Jersey, USA; 1998.
22. Hair JF, Anderson RE, Tatham RL, et al. Análisis Multivariante, 5th ed. Prentice Hall Iberia: New Jersey, USA; 2001.
23. Aldás Manzano J, Currás Pérez R, Ruiz Mafé C, et al. Factores determinantes de la lealtad en el comercio electrónico B2C Aplicación a la compra de billetes de avión. Rev Esp Investig Mark ESIC 2010;14:113-44.
24. Kaiser HF. An index of factorial simplicity. Psychometrika 1974;39:31-6. [Crossref]
25. Lévy JP, Martín MT, Román MV. Optimization according to covariance structures. In: Lévy JP, Varela J, editors. Modeling with covariance structures in the social sciences. Coruña: Netbiblo; 2006:21-2.
26. Ratchford B. New insights about the FCB grid. Journal of Advertising Research 1987;27:24-38. [Crossref]
27. Oviedo HC, Campo-Arias A. Aproximación al uso del coeficiente alfa de cronbach. Rev Colomb Psiquiatr 2005;34:572-80.
28. Nunnally JC. Psychometric Theory, 3rd ed. McGraw-Hill: New York, USA; 1994.
29. Cea-d’Ancona M. Metodología Cuantitativa. Estrategias y Técnicas de Investigación Social; Editorial Síntesis: España; 1998.
30. Peterson RA. A meta-analysis of cronbach's coefficient alpha. J Consum Res 1994;21:381-91. [Crossref]
31. Morales VP, Urosa SB, Blanco BA. Construcción de escalas de actitudes tipo Likert. Una guía práctica; Editorial la Muralla: Madrid; 2003.
32. Churchill GA. A Paradigm for Developing Better Measures of Marketing Constructs. Journal of Marketing Research 1979;16:64-73. [Crossref]
33. López-Roldán P, Fachelli S. Análisis de tablas de contingencia. Metodol Investig Soc Cuantitativa 2015; cap. III. 6.
34. Eshete A, Mohammed S, Shine S, et al. Effect of physical activity promotion program on adherence to physical exercise among patients with type II diabetes in North Shoa Zone Amhara region: a quasi-experimental study. BMC Public Health 2023;23:709. [Crossref] [PubMed]
35. Keyser LE, McKinney JL, Pulliam SJ, et al. A digital health program for treatment of urinary incontinence: retrospective review of real-world user data. Int Urogynecol J 2023;34:1083-9. [Crossref] [PubMed]
36. Ruiz-Aranjuelo A, Chabbar-Boudet MC, Gambó-Ruberte EP, et al. Analysis of medium-term adherence after the implementation of a cardiac rehabilitation program. Arch Cardiol Mex 2023;93:131-8. [PubMed]