Factors Associated With Pharmacological and Non-pharmacological Treatments Adherence in Patients With Borderline Personality Disorder in Iran

Borderline personality disorder is a major mental illness characterized by a sustained relationship instability, impulsive behavior and intense affects. Adherence is a complex behavior, from minor refusals of treatment to inappropriate use of health services or even abandonment of treatment, which can be affected by various factors. Therefore, the present study aimed to investigate the factors affecting pharmacological and non-pharmacological adherence in patients with borderline personality disorder referred to an outpatient referral clinic in Tehran, Iran. The study was a retrospective cohort. The les of patients with borderline personality disorder referred to the outpatient clinic of the Tehran Psychiatric Institute were reviewed as the rst step. In the next step, we contacted the patients and asked them to ll out the questionnaires. Data were collected using the Drug Attitude Inventory (DAI) questionnaire and a researcher made questionnaire to determine the attitude of patients toward pharmacological and non-pharmacological treatment and therapeutic adherence. After collecting data, patients’ therapeutic acceptance was divided into three groups: poor, partial, good compliance. The data were analyzed by SPSS software version-22. job (P=0.303), marital status (P=0.759), history of self-injury (P=0.364), suicide history (P=0.649), history of substance use (P=0.577), comorbidity with psychiatric disorders (P=0.230), family history of mental illness (P=0.524) and type of medication used (P=0.063) and psychotherapy adherence. There was a statistical signicant difference between education level and psychotherapy adherence (P=0.029). The best adherence to psychotherapy was obtained in postgraduate education (with 66% good adherence among this subgroup) and the worst medication adherence in undergraduate education (with 22% good adherence among this subgroup) (P=0.04). In terms of history of psychiatric hospitalization, there was a signicant difference between people with different psychotherapy adherence (P=0.01); those who had a history of hospitalization had better adherence to psychotherapy than non-hospitalized patients (66% vs. 36.9% of the relevant subgroup).

Borderline personality disorder (BPD) is characterized by severe instability in impulsive feelings, identities, relationships, and behaviors (1). Patients with BPD are hypersensitive to emotional stimuli due to abnormal activity in the brain's prefrontal areas during stimulus processes (2). BPD is a major mental illness that has a prevalence of approximately 1-3% among general population. With the constant change in the instability of relationships, the creation and regulation of impulses and the sense of life were determined (3). BPD is a severe disorder that accounts for 20 to 40% of psychiatric admissions, and it is estimated that 84% of patients show suicidal behaviors, and 8% of them die due to suicide (4,5). BPD causes many problems for the patient and the community. The repeated and severe damages caused by this disorder affect the entire life of an individual, including occupational problems, being a dropout, and disruption of relationship with friends and colleagues, family members, failure in marriage and high-risk and unstable sexual relations. Almost 11-69% of patients are substance users, 22% of these patients suffer from alcoholism and 25 to 50% of female prisoners are patients with BPD (6-11).
An accurate prevalence rate is not available in Iran. Due to the cultural and civil changes in recent years in Iran, and the existence of religious thoughts in the Iranian community and families, in most cases, the behaviour of patients with borderline personality disorder are contrary to social norms and severely complicating among Iranian society (12,13).
In 2017 Little et al. showed that behavioral therapy alters supportive relationships, improves self-e cacy, and alters individual attitudes (14). In a study by Bellino et al., patients with more severe BPD and higher levels of symptoms such as fear of abandonment, emotional instability, and identity showed a better chance of combining psychotherapy with uoxetine (15). Although psychotherapy is still the rst line of treatment for patients with BPD, medications are commonly prescribed for them. The use of antidepressants has decreased, and mood stabilizers and second-generation antipsychotics have increased (16).
Adherence is a complex behavior, from minor refusals of treatment to inappropriate use of health services or even abandonment of therapy, which can be affected by various factors. These factors are often patient-dependent (17). Rejection is often associated with irrational fears and misjudgments, and the most common cause is fear of medication dependence. Various studies have been performed on patients with personality disorders, schizophrenia and bipolar showed that one of the critical issues in the treatment is the acceptance of treatment by patients. Due to the prevalence of BPD and the need for longterm pharmacological and non-pharmacological treatment, the high cost of treatment imposed on the health care system and the fact that no study was conducted in Iran, it seems necessary to evaluate the medication and non-medication adherence in these patients. Hoping that by working on changeable factors, patients' adherence and, consequently, their function improves and the recurrence of symptoms and re-admission will be reduced.
In the present study, we investigate the factors of rejection and reasons for discontinuation of treatment to provide a way to improve adherence in these patients and to introduce practical methods into the clinical practice of therapists in future studies.

Study design & Participants
The study was conducted as a retrospective cohort. In this study, the les of patients with BPD who were referred to the outpatient adult psychiatric clinic of Tehran Psychiatric Institute, Tehran, Iran in April of 2014 to April 2019 were reviewed as the rst step. Patient information including name, gender, age, occupation, education, marital status, family history of psychiatric illness, type of treatment, history of substance use, history of hospitalization, history of suicide and self-harm, comorbidity and medication side effects based on the les, was recorded. In the next step, patients were contacted to collect the information, including receiving or not receiving medical or non-pharmacological treatment, last referral date, con rmation of demographic data, current life status (alone, with family), suicide and self-harm, nancial status, legal issues, medication side effects, and location (far or close to the treatment center).
In addition, they were asked to ll out the Drug Attitude Inventory (DAI) questionnaire and a researcher made questionnaire to determine the attitude of patients toward pharmacological and nonpharmacological treatment and therapeutic adherence. Patients who did not respond were called three times within a week, and if they did not respond, they were given a message and excluded from the study if they did not respond again.

Instruments
In this study, in addition to demographic items, two questionnaires were used. The rst questionnaire, was DAI-10 based that evaluate attitude of patients toward medication with 10 yes-no questions and higher total score between +10 and -10 shows more positive attitude. Shariati et al. con rmed psychometric properties of the Persian version of the scale (19). They reported Test-retest reliability was 0.805 and Cronbach's α 0.787. Concurrent validity between DAI-10 score and Medication possession rate (MPR) at the study time was 0.676, and positive predictive value of DAI-10 score for medication compliance at assessment time was about 88.9%. The spearman correlation coe cient between the two tests was 0.822 (95% con dence interval (CI): (0.652 -0.901). They also reported cut off point +1, with the speci city of 81.5% and sensitivity of 89.1%. (19).
The second questionnaire was a researcher made questionnaire to determine attitude of patients toward pharmacological and non-pharmacological treatment and therapeutic adherence. This questionnaire consisted of 17 questions which was validated by eight psychiatric faculty members of Iran University of Medical Sciences. To standardize and validate the questionnaire, qualitative assessment was done using an expert panel and quantitative assessment was done using content validity ratio (CVR) and content validity index (CVI) (20). According to the overall CVR (0.75<) and the CVI (0.79<), the questionnaire had content validity. This questionnaire was divided into two parts. The rst part, which included questions (1-6), was lled by all 94 patients, and the second part, (questions 7-17) was lled by 64 patients with moderate or poor drug compliance.
Finally, based on the data, the patients' current medical acceptance was evaluated by the researcher, and according to the Medication possession rate (MPR), patients were divided into three groups: poor, good, and partial pharmacological adherence. Based on MPR more than 80% compliance equal good adherence, between 50 to 80% equal partial adherence and less than 50% equal poor adherence. The MPR represents the proportion of days of medication supply in a speci ed time period, in our study last month, and is given by the sum of days divided by the number of days within the time interval (21 There was a signi cant relationship between attitudes toward medication and the belief that "lack of treatment disrupts relationships with others". 87.5% of patients with a positive attitude stated that lack of treatment could disrupt their relationships with others. 81.5% of patients who disagreed with this view had a negative medication attitude (P=0.03). There was a signi cant relationship between attitudes toward medication and the belief that "their disorder can be treated". 95% of patients with a positive medication attitude believed that their disorder could be treated (P=0.001). 72.2% of patients who had a negative attitude toward medication, believed that the medication side effects had interrupted their treatment (P=0.022).

Discussion
The aim of this study was to determine the factors contributing to the pharmacological and nonpharmacological adherence in 94 patients with BPD, who were referred to the outpatient clinic of Tehran Psychiatric Institute. The results of this study showed that in gender, occupation, education, marriage, history of psychiatric hospitalization, history of self-harm and suicide attempt and history of substance use were not signi cantly different in people who had good, bad and relative adherence to medication. However, education and history of hospitalization were signi cantly different in patients with good, bad and relative adherence to psychotherapy. Also the analysis showed that there was no signi cant difference between gender, occupation, education, marriage, history of psychiatric hospitalization, history of self-harm and suicide and history of drug use and adherence to medication.
In 2020, Zarei et al. examined the improvement of BPD at different time intervals of transference-focused treatment. Their ndings showed that gender, occupation, education, marriage, history of psychiatric hospitalization and history of self-harm were not signi cantly different in people with different adherence to treatment, which was consistent with our results (22). In another study, there was no signi cant difference between history of self-harm, substance use, psychiatric comorbidity, history of psychiatric illness and type of medication and medication adherence that was consistent with our results (23).
Soltaninejad et.al showed that self-injury, substance use, psychiatric comorbidity, history of psychiatric illness and type of pharmacotherapy were not signi cantly different in groups with different adherence to medication that was consistent with our results (24).
The ndings of our study showed better attitude toward psychotherapy, negative attitude toward medication, fear of medication dependence were the most common reason for poor adherence. In 2019, Timäus et al. conducted a study to evaluate the pharmacological treatment, medication and psychotherapy adherence in patients with BPD. In this study, discontinuation of treatment due to dissatisfaction with the physician, lesser education, history of psychiatric hospitalization and discontinuation of treatment due to the high cost of psychotherapy were seen signi cantly different among individuals with different psychotherapy compliance. But in other variables such as gender, occupation, education, marital state, history of psychiatric hospitalization, history of self-injury and history of substance use were not signi cantly difference among people with different medication and psychotherapy adherence; all of these results were similar to our study (25). In general, demographic factors, history of self-harm and suicide, history of substance use and comorbidity did not affect patient's adherence in our study.

Limitations
There were several limitations; the rst one was the small sample size. Also, some inaccurate recording of documents and the lack of facilities to follow up patients and failure to assess the severity of the disorder was other limitation. It is suggested considering these limitations for future studies.

Conclusion
In conclusion, working on changeable factors such as patients fear of dependence to medications, dissatisfaction with their doctor, medication side effects and patients' attitude toward medication, belief in the curability of the disorder, and belief in the negative impact of the treatment on the person's life increase the patients' adherence to treatment, therefore their function and interpersonal relationships improve. This research received no speci c grant from any funding agency in the public, commercial, or not-forpro t sectors.

Competing interests
The authors have no con icts of interest to report.
Availability of data and materials The dataset used and analyzed during the current study can be shared by the corresponding author upon reasonable request.

Ethics approval
In this study, no intervention was performed on patients. If a person refused to participate in the study, he/she wouldn't be deprived of any medical services. All personal information remained con dential.