Medication Adherence in Schizophrenia Study
Medication Adherence in Schizophrenia Study
2-2024
Abstract
Background and Aims: Schizophrenia is a chronic and severe mental disorder. Treatment by antipsychotic
medication is one of the main therapies to control the symptoms. Medication non/poor adherence is a one
of the key factors leading to relapse and declined social and occupational function in patients. However,
prevalence of schizophrenic patients with non/poor adherence to medication was quite high in previous
studies. Therefore, this study aimed to identify the prevalence of medication non/poor adherence and
its related psychosocial and clinical factors among schizophrenic patients. Subjects and Method: a cross -
sectional analytical study of 126 schizophrenic patients at National Psychiatric Hospital 2 were conducted to
evaluate the rate of medication non-adherence by Morisky Medication Adherence Scale 8 item (MMAS - 8).
Data were collected on patients’ sociodemographic factors such as sex, age, marital and economic status,
level of education and clinical factors including duration of schizophrenia, clinical symptoms evaluated by
Positive and Negative syndrome scale (PANSS), type, and side effects of medication. Logistic regression
was used to analyze the factors associated with medication non-adherence among schizophrenic patients.
Results: Rate of poor medication adherence among participants was 78.57% by using MMAS – 8. The
associated factors with poor medication adherence among schizophrenic patients were level of education
under high school; unstable job/unemployment; poor insight about schizophrenia, poor family care/support,
high scores of negative PANSS and general PANSS scale, duration of schizophrenia above 5 years, treated by
typical antipsychotics and having more than two side effects of antipsychotics. Conclusion: The rate of poor
medication adherence/ non - adherence among schizophrenic patients is high. Mental health staff should be
aware of this risk and screening individuals for relevant risk factors is highly recommended.
Keywords: medication adherence, schizophrenia, social - psycho - clinical factors.
and occupational impairment. Furthermore, non- including descriptive and statistical analyses.
compliance with treatment also aggravates the Data measurement
consequences of the disease such as violence, The Vietnamese version of MMAS-8 was used
homelessness, delinquency, and suicide, adversely to assess medication adherence in the study
affecting social security. Investigating factors related participants. This instrument, a self-administered
to treatment adherence can reduce the rate of questionnaire, consisted of 8 items assessing the
non-adherence, significantly reducing relapse medication-taking behavior and has been widely
and cost of care. This will improve the quality of used in various cultures [6]. The scale consists of
treatment for schizophrenic patients and reduce 8 items in which the first 7 items are answered
financial burden on families, healthcare resources with “yes” or “no”. If the answer is “yes”, the item
and society. This study investigated the prevalence is scored as 1, and the answer is “no”, the item is
of non – adherence to medications especially to scored as 0, except for item 5, in which each “yes”
antipsychotics among patients with schizophrenia answer is scored as 1 and each “no” answer is scored
by using Morisky Medication Adherence Scale 8 as 0. Item 8 is rated on a 5-point scale ranging with A
item (MMAS - 8) in National Psychiatric Hospital 2 (never/rarely), B (once in a while), C (Sometimes), D
and examined the psycho–social and clinical risk (usually) and E (all the time/ always). If the answer
factors associated with medication non - adherence is A, the item is scored as 0, the answer is from B-E,
in these patients. the item is scored as 1. Response options were used
to calculate a continuous total score ranging from 0
2. SUBJECTS AND METHODOLOGY to 8; scores of 3 or more indicating low adherence,
2.1. Subjects 1 or 2: medium adherence, 0: high adherence [7].
A total of 126 patients over 18 years old When analyzing the statistical correlations, we
with schizophrenia, using ICD 10 (International grouped high adherence and moderate adherence
Classification of Diseases, 10th Revision) into one group and the low adherence group in
Schizophrenia diagnostic criteria of WHO were a separate group. To collect data on risk factors
selected from their medical records in National associated with medication adherence, participants
Psychiatric Hospital 2, Vietnam from April, 2018 to were evaluated by a structured questionnaire
May, 2019. This psychiatric hospital is one of the two on socio-demographic information such as age,
biggest psychiatric hospitals in Vietnam and is the gender, occupation, economic, marital status,
biggest psychiatric hospital in Southern Vietnam. and level of education, support from family and
Patients were invited to enroll in the study if they met society; and clinical information including duration
the following inclusion criteria: (1) the patients had of schizophrenia, clinical symptoms evaluated by
been previously diagnosed with schizophrenia and Positive and Negative syndrome scale (PANSS),
were experiencing a return of symptoms at the time type of medication, and side effects of medication.
of the study after a full or partial recovery, (2) signed The Positive and Negative Syndrome Scale (PANSS)
the written informed consent form and (3) had the is a widely used instrument for measuring the
ability to answer the questionnaire or had relatives severity level of adult patients with schizophrenia.
or caregivers who know patients’ medication use The PANSS consists of 30 items divided into three
well. The exclusion criteria were as follows: (1) dimensions including negative, positive and
patients with the first stage of schizophrenia, (2) general psychopathological dimensions. Positive
schizophrenic patients who had not been previously and negative dimension has 7 items each, while
treated, (3) outpatients, (3) inability to understand general psychopathological dimension consists
Vietnamese, (4) having acute and severe physical of 16 items. Each item is rated on a seven-point
diseases that did not allow the participants to severity scale (1=absent, 2=minimal, 3=mild,
answer the interview questions correctly. 4=moderate, 5=moderate severe, 6=severe, and
Eligible participants were selected by convenient 7=extreme). The lowest possible total score scales
sampling technique and during the time of the are 7 and the highest score are 49 on both negative
study we selected a total of 126 patients with and positive dimension of PANSS. For the general
schizophrenia. psychopathological scale, the minimum score is 16
2.2. Methods and maximum score is 112. A higher score indicates
Research design the a more severe level of schizophrenia [8,9].
This was a cross-sectional analytical study, For data quality control, research team
members were psychiatrists with at least 9 years of minutes. If patients were calm down, they filled out
experience and were carefully trained in a group in a social - demographic and clinical questionnaire
administering all measures. and MMAS - 8 with or without their relatives’ help.
Statistical analysis If patients were in agitated state, they would fill out
All statistical analyses were performed using this questionnaire right after the agitation is under
SPSS version 20.0. Frequency analyses were used control.
to describe the sample. Chi-squared tests were Ethics
applied to measure the difference of characteristics The research protocol was approved by the
between those with high/medium and low professional council and the medical ethics
adherence group. The odds ratio (OR) and its committee of Hue University of Medicine and
95% confident intervals (Cis) were calculated. Pharmacy, Hue University. This study was conducted
Multivariate logistic regression was used to analyze in accordance with the Declaration of Helsinki.
the factors associated with medication non-
adherence among schizophrenic patients. 3. RESULTS
Procedure A total of 126 patients with schizophrenia,
126 schizophrenic patients admitted at National with the mean age of 39.60 (SD 11.48), 56 women
Psychiatric Hospital 2, Vietnam from April 2018 (44.44%) and 70 men (55.56%), participated in the
to May 2019 were invited to participate in the study and completed the questionnaire. Among
study. Participants and their relatives were fully them, 7 participants (5.56%) were classified as high
explained of the study. After that they signed adherence, 20 participants (15.87%) were classified
the written consent form. Psychiatrists who had as medium adherence and 99 participants (78.57%)
received PANSS training accessed patients at the were classified as low adherence according to the
time of admission. This examination took about 50 MMAS - 8 (Table 1).
Table 1. Prevalence of medication adherence according to MMAS-8
Prevalence of medication adherence
Number (n) Percentage (%)
by using MMAS-8
Low 99 78.57
Medium 20 15.87
High 7 5.56
Total 126 100
Table 2. Socio-demographic characteristics of the participants and differences
between low adherence and high/medium adherence groups
Adherence
Characteristics Low Medium/high Total P
n % n %
Age (year) ± SD 40.18 ± 11.73 37.48 ± 10.44 0.28
Gender Men 41 73.2 15 26.8 56
0.19
Women 58 82.9 12 17.1 70
Level of Secondary or less 69 87.3 10 12.7 79
education High school or 0.002
30 63.8 17 36.2 47
higher
Marital status Married 43 76.8 13 23.2 56
Single/divorced/ 0.661
widowed 56 80.0 14 20.0 70
Table 2 shows that the prevalence of unemployment or less) in low adherence group were significantly
in low adherence group was significantly higher higher than in medium or high adherence group
than in medium or high adherence group (94,6% vs (p < 0.05). There were not significant differences
71.9%, p < 0.01). Similarly, the rates of patients with between the high adherence, medium adherence
poor support from family and poor economic status and low adherence groups in gender, marital status
as well as low education level (secondary school (p > 0.05).
Table 4. Multivariate logistic regression model predicting low adherence (MMAS-8 score > 2)
by socio-demographic and clinical factors
Characteristics OR 95% CI P
Level of education Secondary or less 1.65 0.27 - 9.95 0.586
High school or higher 1
Job status Unemployment 0.10 0.00 - 2.53 0.163
Employment 1
Support from families/ Poor 30.33 2.00 - 459.08 0.014
communities Good 1
Duration of ≥ 5 years 0.32 0.04 - 2.73 0.300
schizophrenia < 5 years 1
Number of ≤ 5 times 1
hospitalizations > 5 times 40.68 3.87 - 427.60 0.002
Type of antipsychotics SGA 1
FGA 3.05 0.14 - 64.78 0.475
FGA + SGA 0.67 0.11 - 4.10 0.665
Number of side effects <2 1
≥2 0.31 0.02 - 4.44 0.391
PANSS score Negative 1.05 0.84 - 1.30 0.667
General 1.31 1.03 - 1.66 0.027
Multivariate logistic regression analysis found that with schizophrenia were reported as non-adherence
several factors associated with low adherence to (PDC < 80%) to treatment by using the proportion
medication among schizophrenic patients including of days covered (PDC) adherence measure [12].
poor support from family/community (OR = 30.33, Valenstein et al. studied on approximately 34,000
p < 0.05), number of admissions over 5 times (OR Veterans Affairs patients with schizophrenia
= 40.68, p < 0.05), and higher score of general using Medication possession ratios (MPRs) to assess
psychopathological scale of PANSS (OR = 1.31, p < adherence to medication in patients, with good
0.05) (Table 4). adherence defined as an MPR ≥ 0.8 during a year
and poor adherence with MPR ≥ 0.8 in all years, for
4. DISCUSSION 4 consecutive years showed that the prevalence
Previous studies showed that the prevalence of of poor adherence was 36% of study population in
low/poor adherence to treatment was high among each year [13]. The criteria used to determine non/
patients with schizophrenia. In this study, we found low/poor adherence in the studies varied according
that 78.57% of the study population was classified to the usage of different instruments such as MMAS
as low adherence to medication according to MMAS - 8, PDC, MPR...Thus, the difference in the rate of
- 8. The prevalence of non/poor compliance among non - adherence in schizophrenic patients among
schizophrenic patients varied widely in different the studies could be explained by using different
studies. Non-adherence is estimated to range from assessment tools of non-adherence.
40% to 90% of schizophrenic patients [10]. Chaudhari Our study results showed that the prevalence of
B et al. (2017) studied 50 patients with schizophrenia patients with low education level (secondary or less)
and found that among the studied patients, 52% of in low medication adherence group was significantly
patients were low adherers according to MMAS – 8 higher compared to those in the high or medium
[11]. Desai R and Nayak R (2019), in a retrospective group in bivariate analysis (Table 2) but this factor
cross-sectional study with data from the Medical was not a risk factor in multivariate logistic regression
Expenditure Panel Surveys (MEPS) for the years analysis (Table 4). Our study findings were consistent
2010 - 2014, found that 71% of 1.2 million people with those reported in literature. Research by Desai
R. and Nayak R. (2019) showed that people with less 1,3%, p < 0.05, Table 2). This difference was also
than 15 years of education often had poor adherence found in multivariate logistic regression (OR = 30.33,
to treatment compared to those with higher 95% CI: 2.00 - 459.08, p < 0.05). In Vietnam, there is a
education and longer study time [12]. Patients with close relationship between members in each family.
high level of education often had better insight and Family plays an important role in caring patients
awareness of importance of treatment adherence, with chronic diseases generally and schizophrenia
thereby making them adhere to treatment better. particularly. Families remind and control patients’
The relationship between treatment adherence medication intake. Cohen A.N and Pedersen E.R et
and education level is reciprocal. High adherence al. studied on 801 patients with schizophrenia, and
to treatment would help control symptoms in the the results found that patients with good support
patients, help prolong the stable time of symptoms, from their families were often associated with high
thereby giving the patients the opportunity to study adherence [15]. However, in another study, Glick
to achieve a higher level of education. Patients with I.D and Davis J.M et al. (2017) reported that there
low education level often combined with other risk was no association between support of families and
factors such as poor insight, unemployment, and medication adherence [16]. Contrasting results on
poor economic status. All these factors lead patients the association between family/community support
to poor adherence to medication treatment. This and treatment adherence may suggest that this
view was supported by the data in table 2 in our factor is not an independent factor but treatment
study and other previous studies. Poor or lack of adherence may be influenced by other factors
insight include denial of illness, rejection to taking such as severity of symptoms, duration of illness,
medication because of unawareness of benefit combined substance use… Our study also showed
of medication and negative medication belief; that there were no significant differences between
and unawareness of social consequences due to low adherence versus high/medium adherence
schizophrenia. Chang J.G et al (2019) studied on 81 groups in gender, age and marital status (p > 0.05).
outpatients with schizophrenia realized that there Among clinical characteristics, our study’s results
was a correlation between insight and medication indicated some factors associated with medication
adherence [14]. Higashi K et al. in their systemic adherence in the participants including type of
review with thirty-seven papers indicated that schizophrenia, duration of schizophrenia, number of
one of the key drivers of non - adherence was lack hospitalizations, severity of schizophrenia assessed
of insight [3]. Chaudhari B et al. also found that by PANSS, type of antipsychotics and number
patients from low-income households often had of side effects in bivariate analysis. Evidence
poor adherence to treatment [11]. The correlation from other studies points to severity of positive
between unemployment, poor economic status symptoms, negative symptoms and other symptoms
and adherence to treatment was a bidirectional of schizophrenia could impact on medication
relationship. Patients with poor adherence to adherence in patients. Bajaj V et al. (2009) found
treatment often have uncontrolled symptoms, that when the score of PANSS was higher, treatment
short stabilization time, increased number of adherence was lower [17]. Chaudhari B et al. also
hospitalizations, so their possibility of finding stable shared this view [11]. Negative symptoms such as
jobs are very low, leading to an increased risk of memory and cognitive impairment, impaired self-
unemployment. On the other hand, non-adherence care could predispose patients to low adherence
to treatment increases the cost of treatment due to to treatment. Our results in table 3 showed that
its consequences, thereby making economic status the mean score of negative dimension of PANSS
more difficult. In contrast, patients with difficult in the low adherence was significantly higher than
economic status and unemployment have more in medium/high adherence group (29.15 ± 6.09 vs
difficulty accessing health services and using good 23.11 ± 3.31, p < 0.01). This result was similar with
medications, thus leading them to low adherence. general psychopathological dimension of PANSS
Lack of support from families/communities was a (54.20 ± 5.46 vs 46.48 ± 4.53, p < 0.01) (Table 3). But in
predictor of low adherence among patients with multivariate logistic regression analysis, only higher
schizophrenia. Our result showed that the rate of score of general psychopathological dimension of
patients with lack of support from their families in PANSS were still risk factors (OR = 1.31, 95%, CI 1.03
low adherence group was significantly higher than – 1.66, p < 0.05) (Table 4). Prevalence of patients
those in high/medium adherence group (98.7% vs with duration of schizophrenia more than 5 years
was significantly higher in low adherence group with schizophrenia or bipolar found that patients’
compared to those in high/medium adherence dissatisfaction attitude towards medication and
group (82.8% vs 12.7%, p < 0.05) (Table 3). Besides their physical symptoms was associated with low
that, the percentage of patients with number of post-hospitalization drug adherence in severe
admission more than 5 times was also significantly psychiatric patients including schizophrenia [20].
higher in the low adherence group than in the Our data in table 3 indicated that prevalence of
medium/high adherence group (95.1% vs 4.9%, p participants having 2 or more side effects was
< 0.05) (Table 3). Number of hospitalizations was significantly higher in low adherence group than in
also a risk factor in multivariate logistic regression high/medium adherence.
(OR = 40.68, 95% CI 3.87 - 427.60) (Table 4). An There were limitations to our study. First, the
increased number of admissions usually means that sample in our study was small. Second, this was a
the treatment was less effective, the symptoms descriptive cross - sectional study not a longitudinal
were less improved and negative symptoms were follow up study so it is difficult to investigate causal
progressively worse. All of these factors contributed relationship. Third, some potential variables that
to low adherence in patients with schizophrenia. could be associated with low adherence may involve
Treatment by typical antipsychotics or first route of medication administration, substance
generation of antipsychotics was suggested as a abuse/dependence, therapeutic alliance...
significant predictor of poor adherence in previous
studies. El Abdellati K et al. reviewed 26 studies on 5. CONCLUSION
factors related to antipsychotics adherence revealed The prevalence of low adherence to medication
that patients using first-generation antipsychotics is high in patients with schizophrenia. Bivariate
(FGA) monotherapy had poor adherence at analysis indicated several factors associated with
discharge than patients using second-generation low adherence in participants including level of
antipsychotics (SGA) either as monotherapy or in education below high school; unstable job/un-
combination in some studies but this result was employment; poor insight about schizophrenia, poor
not consistent in other studies [18]. In bivariate caring/support from family, high scores of negative
analysis in our study, low medication adherence was PANSS and general PANSS, duration of schizophrenia
significantly associated with using FGA (p < 0.001) above 5 years, treated by typical antipsychotics and
(Table 3). In multivariate logistic regression analysis, having more than two side effects of antipsychotics.
this risk factor was not significant (p > 0.05) (Table Therefore, to improve adherence to medication
4). FGA only impact on positive symptoms but not among this population, following tasks should be
on negative symptoms and cognitive symptoms and done: (1) patients’ families should support patients
even make these symptoms worsen. As mentioned after discharge, (2) psychiatrists should choose
above, patients with negative or cognitive symptoms and adjust dose of medication to enhance effects
with poor self - care, memory and cognitive and reduce side effects to improve medication
impairment was a predictor of low adherence. In adherence in patients. (3) Mental health care staff
addition, FGA often associated with side effects should educate patients and their relatives about
including parkinsonism, acute dystonia, tardive the important role of medication in treatment and
dyskinesia, akathisia, anticholinergic effects such as schizophrenia.
constipation, blur vision, mouth dry and side effects
on cardio - vascular system such as palpitation, Acknowledgments
orthostatic hypotension. These factors lead patients The authors wish to thank the patients and the
to low adherence to FGA. Wubeshet Y.S et al. (2019) hospital staff who participated and facilitated us in
studied on 356 patients with schizophrenia found the study.
that 293 patients (97.7%) had side effects caused
by FGA in which cardiovascular side effects occurred Funding
in 169 patients (56.3%), sedation and side effects This research was not supported by any funds.
on the central nervous system occurred in 149
patients (49.6%) and extrapyramidal side effects Disclosure
occurred in 114 patients (38.0%) [19]. Lee Y, Lee The authors have no conflict of interest to
MS, Jeong HG, et al. (2019) studied on 81 patients declare.
REFERENCES