Semahegn et al. Systematic Reviews            (2020) 9:17 
https://doi.org/10.1186/s13643-020-1274-3
RESEARCH Open Access
Psychotropic medication non-adherence
and its associated factors among patients
with major psychiatric disorders: a
systematic review and meta-analysis
Agumasie Semahegn1,2* , Kwasi Torpey1, Adom Manu1, Nega Assefa2, Gezahegn Tesfaye2 and
Augustine Ankomah3
Abstract
Background: Major psychiatric disorders are growing public health concern that attributed 14% of the global
burden of diseases. The management of major psychiatric disorders is challenging mainly due to medication non-
adherence. However, there is a paucity of summarized evidence on the prevalence of psychotropic medication
non-adherence and associated factors. Therefore, we aimed to summarize existing primary studies’ finding to
determine the pooled prevalence and factors associated with psychotropic medication non-adherence.
Methods: A total of 4504 studies written in English until December 31, 2017, were searched from the main
databases (n = 3125) (PubMed (MEDLINE), Embase, CINAHL, PsycINFO, and Web of Science) and other relevant
sources (mainly from Google Scholar, n = 1379). Study selection, screening, and data extraction were carried out
independently by two authors. Observational studies that had been conducted among adult patients (18 years and
older) with major psychiatric disorders were eligible for the selection process. Critical appraisal of the included
studies was carried out using the Newcastle Ottawa Scale. Systematic synthesis of the studies was carried out to
summarize factors associated with psychotropic medication non-adherence. Meta-analysis was carried using Stata
14. Random effects model was used to compute the pooled prevalence, and sub-group analysis at 95% confidence
interval.
Results: Forty-six studies were included in the systematic review. Of these, 35 studies (schizophrenia (n = 9),
depressive (n = 16), and bipolar (n = 10) disorders) were included in the meta-analysis. Overall, 49% of major
psychiatric disorder patients were non-adherent to their psychotropic medication. Of these, psychotropic
medication non-adherence for schizophrenia, major depressive disorders, and bipolar disorders were 56%, 50%, and
44%, respectively. Individual patient’s behaviors, lack of social support, clinical or treatment and illness-related, and
health system factors influenced psychotropic medication non-adherence.
Conclusion: Psychotropic medication non-adherence was high. It was influenced by various factors operating at
different levels. Therefore, comprehensive intervention strategies should be designed to address factors associated
with psychotropic medication non-adherence.
Systematic review registration: PROSPERO CRD42017067436
Keywords: Medication non-adherence, Psychiatric disorders, Systematic review, Meta-analysis
* Correspondence: agucell@yahoo.com
1Department of Population, Family and Reproductive Health, School of
Public Health, College of Health Sciences, University of Ghana, Accra, Ghana
2College of Health and Medical Sciences, Haramaya University, Po Box 235,
Harar, Ethiopia
Full list of author information is available at the end of the article
© The Author(s). 2020 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 2 of 18
Background Methods
Psychiatric disorders have been a global public health Protocol development and registration
challenge. Almost 450 million people are affected by This systematic review and meta-analysis has been regis-
psychiatric disorders worldwide. It contributes 14% of tered in the international Prospective Register of System-
the overall global burden of diseases, and 30% of the atic Reviews (PROSPERO 2017:ID:CRD42017067436) [16]
non-fatal diseases burden, which is worsened by medica- and written in accordance with the Preferred Reporting
tion non-adherence [1–3]. Psychiatric disorders cost ap- Items for Systematic Review and Meta-analysis (PRISMA)
proximately US$2.5 trillion in 2010 and are expected to statements guidelines [17] (see Additional file 1). The de-
rise up to US$6.0 trillion by 2030. Lost resources and tail of this systematic review and meta-analysis protocol
production, unemployment, absences from work, and has been published elsewhere [18].
premature mortality are some of the indirect economic
costs [3]. The World Health Organization (WHO) has Search methods for identification of studies
designed a comprehensive strategic action plan (2013– The Medical Subject Headings (MeSH) and keywords were
2020) to promote mental well-being, prevent psychiatric constructed based on the review question. Studies were
disorders, and provide care and support to reduce mor- searched using search engines, from the main electronic da-
bidity, disability, and mortality [4]. tabases (PubMed (Medline), EMBASE, CINAHL, Web of
Nearly one third (31.7%) of people who suffer major Science, and PsycINFO), and other sources (Google
psychiatric disorders end up with a long-term disability Scholar, reports, thesis, or dissertation). Search strings were
and dependency [5]. Psychiatric disorders are associated constructed using a combination of MeSH terms such as
with individual factors as well as community social sup- psychotropic non-adherence, non-compliance, compliance,
port, cultural, social protection, living standards, and adherence, determinants, barriers, associated factors, risks,
other environmental factors [4]. Compliance to medica- correlates, influencing factors, and major psychiatric disor-
tion is essential but challenging in the management of ders (see Additional file 2). The search strings were modi-
major psychiatric disorders [6–8]. The WHO defines fied to suit to the corresponding database interface. All of
medication non-adherence as, “a case in which a per- the identified studies were exported to the EndNote citation
son’s behavior in taking medication does not correspond manager [19], and duplicates were removed.
with agreed recommendations from health personnel”
[9]. Patients with major psychiatric disorders are most Eligibility criteria
likely to be non-adherent to medication due to poor rea- Studies were included in the systematic review and or
soning and lack of insight about their illness and treat- meta-analysis if they fulfill the following eligibility cri-
ment [8, 10, 11]. teria. The criteria were as follows:
Psychotropic medication non-adherence can lead to
exacerbation of their illness, reduce treatment effective- 
 Studies had been conducted among adult patients
ness, or leave them less responsive to subsequent treat- (18 years and older);
ment. Other consequences of non-adherence include re- 
 Studies had been conducted on one or more of the
hospitalization, poor quality of life or psycho-social out- major psychiatric disorders (major depressive
comes, relapse of symptoms, increased co-morbid med- disorders, schizophrenia, or bipolar disorder) were
ical conditions, wastage of health care resources, and eligible;
increased suicide [7, 8, 12–15]. Research evidence on the 
 Studies reported psychotropic medication non-
level of psychotropic medication non-adherence and its adherence or adherence and or factors associated
associated factors among patients with major psychiatric with medication non-adherence;
disorders is essential to design appropriate interventions 
 Studies conducted at community and/or facility-
to achieve desired treatment goals for both patients and based;
health care providers. Although several primary studies 
 Studies used observational study designs (cross-
have been conducted on this issue, there has not been sectional, case-control cohort, and or survey);
any systematic review and meta-analysis carried out to 
 Studies were written in English before December 31,
inform policy. Thus, a systematic review and meta- 2017;
analysis on the level and factors associated with psycho- 
 Documents (both published and unpublished
tropic medication non-adherence is useful to inform pol- studies, survey reports, thesis, or dissertations)
icy makers and program planners. Therefore, the main which were accessible with full text.
aim of this systematic review and meta-analysis was to
summarize available findings of primary studies to deter- Selection of studies into systematic review
mine the level of psychotropic medication non- Studies were systematically selected using predeter-
adherence and associated factors. mined eligibility criteria. Studies’ title and abstract that
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 3 of 18
clearly mentioned either patients with major psychiatric Quality assurance of the systematic review
disorder psychotropic medication non-adherence or ad- We searched both published and non-peer reviewed
herence were selected for the subsequent evaluation. studies comprehensively for the systematic review and
Then, to minimize bias during screening, two authors meta-analysis to minimize publication bias. The elec-
(AS and GT) independently screened the title and ab- tronic or computerized, manual, and email searching
stract of the studies to proceed to the next step of the methods were applied to have comprehensive search.
studies selection. Studies overview such as aim of the Eligibility criteria, selection method, quality assessment,
study, design of the study, participants, and main out- data extraction template, and regular meeting for discus-
come of the study were screened. In this stage, the sion schedule were pre-designed by authors to assure
studies potentially eligible for the full text were selected the quality. The studies’ methodological quality critical
based on the title and abstract. The full text of the appraisal was carried out using the Newcastle-Ottawa
studies selected based on the title and abstract were re- Scale [20] (see Additional file 3).
assessed independently by two of the authors (AS and
GT) for details. The body of the studies’ (aims, mainly Data abstraction, synthesis, and statistical analysis
design, participants, sampling method, findings, conclu- The two authors (AS and GT) abstracted the data from the
sions, and recommendations) were assessed. Finally, included studies and recorded in the data extraction tem-
studies have reported the medication adherence or plate. Studies’ detail descriptions such as an author, study
non-adherence among major psychiatric disorders pa- area or country, aim, design, sample size, sampling proced-
tients (schizophrenia, major depressive, or bipolar dis- ure, and response rate were presented on the table using
orders) and associated factors and fulfill the eligibility Microsoft Word (2013) (Table 1). Meanwhile, the raw data
criteria were selected for the systematic review and of medication non-adherence and total sample size were
meta-analysis. All studies that consider psychiatric dis- extracted and stored using Microsoft Excel (2013) template
orders as a factor for medication non-adherence were (see Additional file 4). All the meta-analysis were carried
excluded, because studies that consider psychiatric dis- out using Stata SE-64 version 14.2 (Stata Corporation, Col-
orders as predictor for the non-adherence to treatment lege Station, TX) [66] and based on the recommendation
for other illness may not fully assess the adherence level for the meta-analysis of observational studies [67]. Hetero-
of psychotropic medications. Overall, the studies’ selec- geneity between studies was assessed and substantial het-
tion process was adhered to the PRISMA flow diagram erogeneity was anticipated when I2 greater than 75% [68,
[17] (Fig. 1). Any difference during studies selection 69]. The pooled prevalence (proportion) was estimated
process was resolved through consensus. using the inverse variance method [66]. The 95% confi-
dence interval for pooled and sub-group proportion of pa-
tients’ medication non-adherence was computed.
Measurement of outcome and exposure Moreover, the sub-group-pooled proportion of patients’
According to the WHO, medication non-adherence is medication non-adherence was performed for schizophre-
defined as “a case in which a person’s behavior in taking nia, major depressive disorder, and bipolar disorders separ-
medication does not correspond with agreed recommen- ately. Random effects model [70] was used for the overall
dations from a health personnel”. It can be either pooled estimate and sub-group meta-analysis.
intentional or unintentional, including failing to initially
fill or refill a prescription, discontinuing a medication Publication bias
before completing the course of therapy, taking more or Potential publication bias was assessed by inspecting the
less of a medication than prescribed, and taking a dose funnel plot [71]. The funnel plots were constructed using
at the wrong time [9]. Thus, the main outcome of inter- the plot-observed studies only and plot standard error
est for this systematic review was the level of psycho- with logit event rate (see Additional file 5). In addition,
tropic medication non-adherence. Medication non- statistical tests Egger’s regression test (one-tailed test), p =
adherence was measured either as direct report from 0.683, and Begg’s rank correlation (one-tailed), p = 0.831,
studies or indirectly by subtracting adherence report were computed to make sure that there is no evidence of
from total observations (sample size). Studies’ reported publication bias on studies included in this systematic re-
non-adherence in another way such as medication non- view and meta-analysis. In addition, the tests confirmed
compliance, non-persistence, dropout, discontinuation, that there are no small-study effects in the meta-analysis.
missing, and other alternatives was considered. More-
over, exposure or explanatory variables for the medica- Results
tion non-adherence were measured using synonymous A total of 46 studies were included in this systematic re-
terms such as determinants, predictors, barriers, associ- view and meta-analysis. Each study’s key findings and
ated factors, risk factors, and influencing factors. conclusion has summarized in detail (Table 2).
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 4 of 18
Fig. 1 Diagrammatic presentation of the selection process of studies for systematic review and meta-analysis
Magnitude of psychotropic medication non-adherence was 56% (95% CI 48%, 63%). The prevalence in the sub-
Thirty-five studies were used for meta-analysis to compute group analysis was relatively consistent with the overall
the pooled proportion of the psychotropic medication non- pooled prevalence (Fig. 3).
adherence. In 35 studies with 63,957 cases from a sample
of 120,134, the pooled prevalence of medication non- Major depressive disorder medication non-adherence
adherence among major psychiatric disorders was 49% From 16 studies with 42,255 participants, medication
(95% CI 44%, 55%). In addition, the psychotropic medica- non-adherence among patients with major depressive dis-
tion non-adherence was 48%, 48%, 49%, and 57% in Africa, orders was 50% (95% CI 40%, 59%). The prevalence in the
North America, Europe, and Asia, respectively (Fig. 2). sub-group analysis was relatively consistent with the over-
all pooled prevalence, but a bit lower in Europe (Fig. 4).
Medication non-adherence among schizophrenia patients
Sub-group analyses were conducted for studies that re- Bipolar disorder patients’ medication non-adherence
ported medication non-adherence among schizophrenia From 10 studies with 73,250 study participants, medica-
patients. From nine studies with 2643 participants, the tion non-adherence among patients with bipolar disor-
medication non-adherence among schizophrenia patients ders was 44% (95% CI 43%, 45%) (Fig. 5).
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 5 of 18
Table 1 Description of studies included for systematic review and meta-analysis (n = 46)
Author, country Study aim Design Population Sampling Sample Scale used Response
methods size rate (%)
Ibrahim et al., To determine the socio-demographic and CS Schizophrenia SRS 390 MMAS 94.8
Nigeria [21] clinical predictors of sub-optimal MA & depression
patients
Alene et al., To evaluate MA and associated factors CS Schizophrenia Purposive 336 CFR 87.5
Ethiopia [22] patients
Eticha et al., To investigate factors associated with MA CS Schizophrenia Consecutive 393 MARS 97.5
Ethiopia [23] among patients with schizophrenia patients
Kenfe et al., To assess the magnitude and associated CS Psychiatric Consecutive 422 MMAS 100
Ethiopia [24] factors of MNA patients
Hibdye et al., To assess the prevalence and factors CS Bipolar Systematic 410 MMAS 97
Ethiopia [25] associated with MNA disorders
patients
Anne et al., USA To examine the barriers of antidepressant MA Longitudinal Depression Multistage 134 Brief 90
[26] patients interview
Hill et al., Ireland To examined concurrent predictors of MNA cohort Psychosis Restrictive 171 Interview NR
[27] Patients & DAI
Moritz et al., To investigate attitudes toward psychotic Cohort Schizophrenia Restrictive 113 Self-report NR
Germany [28] symptoms affect MNA patients questions
Mert et al., Turkey To evaluate factors resulting in MNA CS Schizophrenic, Patients 203 SCID-I NR
[29] depressive receiving
patients treatment
Novick et al., To explore the relationship between insight CS Schizophrenia SRS 903 MARS NR
Multi-country- and MA and bipolar
European [30] patients
Hillary, Nigeria To evaluate the level of patients’ MNA and CS Psychiatric Convenient 200 MMAS NR
[31] associated factors disorders
patients
Ibrahim et al., To assessed the prevalence and exclusively CS Schizophrenia Convenient 358 MMAS 94.2
Nigeria [32] X-rayed medication-related factors of MNA and bipolar
patients
Dibonaventura To examine the relationship between these CS Adults Convenience 876 MMAS NR
et al., USA [33] variables among community-dwelling patients schizophrenia
with schizophrenia patients
Gurmu et al., To determine the statistical significance CS Patients who Convenience 209 MARS 96.3
Ethiopia [34] of the association of variables with adherence visited
psychiatric
clinic
Magura et al., To identify predictors of MA among CS Psychiatric Patients 131 MARS NR
USA [35] psychiatric patients disorders fulfilled
patients eligibility
criteria
Kikkert et al., 4 To explore factors influencing MA of qualitative Schizophrenia Purposive 91 Qualitative NA
European schizophrenia patients study patients
countries [36].
Teferra et al., To improve understanding of the underlying Qualitative Schizophrenia Purposive 43 FGDs NR
Ethiopia [37] reasons for MA study patients &
caregivers
Sher et al., USA To evaluate the effects of caregivers’ causal longitudinal MDD patients Multistage 47 Link’s NR
[38] beliefs about depression and their perceptions study scale
of stigma on MA
Mohamed et al., To examine the strength of association of CS Chronic Purposive 1432 DAI, pill NR
USA [39] measures of both insight and attitudes toward schizophrenia count,
MA patients ITAQ
Sava, Turkey [40] To investigate the relationship between CS Comprised of Patients 147 Self-report NR
treatment adherence and the level of MA euthymic attending question
patients their follow-
up
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 6 of 18
Table 1 Description of studies included for systematic review and meta-analysis (n = 46) (Continued)
Author, country Study aim Design Population Sampling Sample Scale used Response
methods size rate (%)
Sirey, USA [41] To examine the extent to which perceived CS Psychiatric Multi-stage 92 SCS NR
stigma affected treatment discontinuation patients
Sajatovic, USA To examined MA among patients with bipolar Longitudinal Bipolar patients All 44,637 MPR NR
[42] disorder
Sajatovic, USA To examine antipsychotic MA among bipolar Longitudinal Bipolar patients All 73,964 MPR NR
[43] disorder
John, USA [44] To investigate the factors associated with CS Bipolar disorder Interactive 469 Adapted NR
non-adherence patients panel tool
Iseselo et al., To determine the psychosocial problems of Qualitative Patients Purposive 14 Interview NR
Tanzania [45] mental illness study families/care and FGDs
givers
Olivares, Spain To evaluate long term treatment outcomes in Cohort Schizophrenia Prospective 1622 GAF score NR
[46] routine clinical practice patients chart review
Charlotte, To identify predictors of MNA to CS MDD patients SRS 1031 CRF & NR
Sweden [47] antidepressant treatment TDM
Adeponle et al., To assess relationship of family engagement cohort Psychiatric Purposive 81 Case Note NR
Nigeria [48] and MA patients review
Rashid, Malaysia To determine the treatment related risk factors CC MDD patients Convenient 148 Self- 86
[49] with the default of depression treatment reported
question
Roy, India [50] To examine factors associated with poor drug CS Psychiatric Consecutive 100 Checklist NR
compliance. patients
Omran, Iran [51] To describe psychiatrists’ attributions on CS Psychiatric SRS 500 Interview NR
non-compliance patients using
checklist
Tara et al., To assess levels of MNA and determinants CS Psychiatric SRS 80 Self-report NR
Canada [52] patients
Banerjee, India To assess the correlates of MNA to unipolar CS Psychiatric with Purposive 239 MMAS 97.2
[53] patients depression
Oliver, Spain [54] To describe MA among patients with CS Psychiatric SRS 212 Medical NR
depression patients with record
Depression
Dave, UK [55] To assess the patterns, incidence and Cohort MDD patients SRS 13,927 PHQ 91.2
predictors of therapy discontinuation (2006–2008)
Mahaye, South To assess the levels of MA and its associated CS Psychiatric Convenient 95 MMAS NR
Africa [56] factors patients
Sundell, Sweden To analyze whether socio-economic factors CS Depression SRS 6536 MARS NR
[57] influence early discontinuation patients
Akincigil, [58] To describe patient and provider level factors Cohort Psychiatric Convenient 4312 pharmacy NR
associated with treatment adherence. patients claims
Fawad, Pakistan To elucidate predictors of non-adherence CS Psychiatric Convenient 128 Adapted 94.8
[11] patients question
Prukkanone et al., To quantify the adherence rate to and Cohort Depression Convenient 1058 MPR NR
Thailand [59] associated factors patients
Shigemur, Japan To identify predictors of antidepressant CS MDD patients Online survey 1151 Checklist NR
[60] adherence
Bambouer et al., To examined compliance and faxed alerts to Cohort Psychiatric Purposive 13,128 MMAS NR
USA [61]. physicians in 2003 patients
Demyttenaere To investigate of compliance in patients with CS Mdd SRS 85 MEMS NR
et al., Belgium MDD
[62]
Mascha, [63] To evaluate adherence to antidepressant Cohort Depression Purposive 131 MMAS NR
among depressed patients Patients
Baldessarini et al., To sought risk factors to guide clinical CS Bipolar patients SRS 429 Self-report NR
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 7 of 18
Table 1 Description of studies included for systematic review and meta-analysis (n = 46) (Continued)
Author, country Study aim Design Population Sampling Sample Scale used Response
methods size rate (%)
USA [64] prediction of non-adherence PRFs
Nega et al., To assess psychotropic MNA and associated CS Psychiatric SRS 613 MMAS 92.9
Ethiopia [65] factors disorder
patients
CC case control, CS cross-sectional, CRFs case report forms, FGD focus group discussion, GAF Global Assessment of Functioning score, ITAQ Insight and Treatment
Attitudes Questionnaire, MA medication adherence, MNA medication non-adherence, MMAS Morisky Medication Adherence Scale, CFR compliant fill rate, MARS
Medication Adherence Rating Scale, MEMS Medication Monitoring System, MPR Medication Possession Ratio, NRR no-response rate, PRFs Patient Record Forms,
PHQ Patient Health Questionnaire, RR response rate, SRS simple random sampling, SCS Stigma Coping Scale, SCID-I Structural Clinical Interview Diagnosis I, TDM
therapeutic drug monitoring
Determinants of psychotropic medication non-adherence cigarette smoking) was a factor associated with psycho-
Medication non-adherence is influenced by various fac- tropic medication non-adherence among major psychi-
tors. We systematically mapped the factors that affect atric patients [23, 25, 65]. Likewise, three studies
medication non-adherence among patients with major conducted in Ethiopia [23, 25, 65] have reported that
psychiatric disorder into individual patient, social sup- “Khat” chewing was a factor associated with psycho-
port, clinical or treatment and illness, and health tropic medication non-adherence among psychiatric pa-
system-related factors based on the review of 46 studies. tients. In addition, in six studies, having a history of
concurrent alcohol dependency was the main factor as-
Factors related with individual behaviors sociated with psychotropic medication non-adherence
Patient’s socio-demographic factors [23, 25, 37, 47, 64, 65].
Some psychiatric patients’ socio-demographic characteris-
tics were associated with medication non-adherence. Patient attitude toward medication
However, the association was inconsistent across studies. In four studies, patients’ attitude toward medication was
In four studies, unemployment was one of the factors as- a crucial factor affecting treatment adherence and thera-
sociated with medication non-adherence [25, 31, 56, 60]. peutic alliance. Patients having negative attitude towards
On the other hand, the nature of the job (for example, en- their medication was a factor associated with psycho-
gaging in farming activities, being busy) influenced pa- tropic medication non-adherence [23, 25, 39, 65]. More-
tients’ adherence to their medication [50]. Educational over, in two studies, patients having negative attitudes
status was one of the influencing factors of medication toward the psychotropic medication were more likely to
non-adherence. In six studies, psychiatric patients having seek alternative treatment such as traditional or religious
lower education level (lower than secondary education) treatment practices [21, 24]. Likewise, where patients
were more likely to be non-adherent to their psychotropic were suspicious about the medication, believes that the
medication compared to those patients having higher edu- medication would harm them, heard voices telling them
cational level [23, 40, 44, 52, 56, 57]. Patients’ non- not to take the medication, and taking medication is un-
adherence to their psychotropic medication was associated natural were less likely to adhere [36]. In three studies,
with some non-modifiable demographic factors such as psychiatric patients may also attribute to antipsychotic
age and gender) [28]. In three studies, patients aged 60 medication non-adherence due to the alterations in cog-
years and older were more likely to be non-adherent to nitive and attitudinal functioning and therefore be un-
their medication [23, 55, 58]. Nevertheless, one study [60] willing to use the medication [29, 37, 38].
reported that young age (less than 34 years) patients were
also more likely to have medication non-adherence. In Patients’ perceived stigma
three studies, the relationship of gender and medication In eight studies, the perception or the feeling of psychi-
non-adherence was inconsistent. Being female was a factor atric patients being stigmatized by their families, neigh-
associated with medication non-adherence [52, 53, 65], bors, health professionals, and other community
but in two studies, being male also linked with medication members was a factor associated with medication non-
non-adherence [21, 54]. adherence [24, 25, 28, 36, 37, 41, 45, 65]. In one study,
both internal and external triggering factors caused the
Patients’ substance abuse patients to feel being stigmatized. Some of these in-
In eight studies, both psychostimulant and psycho- cluded patient believe that they can get better without
depressant substances misuse were associated with psy- medicine were afraid of medication dependency and felt
chotropic medication non-adherence [23, 27, 29, 37, 42, too embarrassed to take the medicine [24]. On the other
43, 58, 65]. In three studies, psycho stimulants (e.g., hand, patients perceived the effects of the medication to
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 8 of 18
Table 2 The key findings and conclusions of studies included in the systematic review and meta-analysis (n = 46)
Author, country Key findings (prevalence and associated factors) Conclusion
Ibrahim et al., MNA was 55.7%. Seeking for traditional treatment (OR, 6.5), male Psycho-education on adherence and the active
Nigeria [21] (OR, 3.3), low levels of insight (OR, 1.8), and low social support involvement of the family has significant in the prevention
levels (OR, 1.5) were predictors of MNA.
Alene et al., The prevalence of MNA was 42.5%. MA is low and associated with pill burden, side-effect, and
Ethiopia [22] exposure to social drugs.
Eticha et al., MNA was 26.5%. Positive attitude (AOR, 1.4), awareness of illness Schizophrenia patients were highly non-adherence to their
Ethiopia [23] (AOR, 1.4), and relabel symptoms (AOR, 1.6). Khat (AOR, 0.2), medication. Intervention strategies focused on patient edu-
illiteracy (AOR, 0.13), and older age (AOR, 0.03) were the predictors cation can be helpful to improve adherence.
of MNA.
Kenfe et al., MNA was 41.2%. Forgetfulness was attributed to 78.2% of their MNA among psychiatric patients in Southwest Ethiopia is
Ethiopia [24] MNA. Irregular follow-up, poor social support, and complex drug high and revealed possible associated factors.
regimen were associated with MNA.
Hibdye et al., MNA was 51.2%. Poor social support (AOR, 5.2), stigmatized (AOR, MNA was found to be high. It has significant implications
Ethiopia [25] 2.2), negative attitude (AOR, 4.6), medication frequency (AOR, 1.7), to enhance level of adherence by tackling factors through
unemployment (AOR, 2.1), and Khat chewing (AOR, 2.1) were intervention program.
predictors.
Anne et al., USA MNA was 28%. It was associated with perceived stigma (0.05), Clinicians’ should give psychological support to improve
[26] patient-rated severity of illness (0.05), interpersonal problems (0.02), adherence
and age 60 years or older (0.04).
Hill. et al., Ireland MNA was 24%. It was associated with less insight, negative Longer treatment duration is associated with non-
[27] attitudes toward medication, substance misuse, and treatment adherence
duration.
Moritz et al., MNA was 20%. Side-effect, missing voices, feeling of power as a Approximately 1-in-5 patient had discontinued anti-
Germany [28] motive for non-compliance, stigma, mistrust against the physician, psychotic treatment due to forgetfulness and ambivalence
and rejection of medication were the most frequent reasons for toward symptoms.
drug discontinuation
Mert et al., Turkey MNA for bipolar disorder, schizophrenia, and MDD was 12.1%, MNA is a serious problem. Ensuring regular follow-up ap-
[29] 18.2%, and 24.2%, respectively. Irregular follow-up (OR, 5.7) and pointments and improving their thoughts are needed.
diagnosis (OR, 1.5).
Novick et al., MA was higher in bipolar patients than in schizophrenia, which Insight and MA were found to be closely related. Insight
Multi-country- might be schizophrenic patients had lower insight than in bipolar. impacts on the therapeutic alliance with mental health
European [30] Better insight was associated with higher MA and had stronger and associated to treatment outcomes.
therapeutic alliance, which reduce the clinical severity.
Hillary, Nigeria [31] Adherence varied from poor adherence (55.5%) through moderate More than half of the psychiatric out-patients had MNA.
(36%) to high adherence (8.5%).
Ibrahim et al., MNA was 54.2% (schizophrenia = 62.5%, bipolar = 45.8%). Multiple Encourage rational pharmacotherapy, consider routine
Nigeria [32] dosing frequency (OR, 7.8), side-effects (OR, 6.8), cost of medications lower dosing prescriptions, integrating side effects
(OR, 4.1), and poly-therapy (OR, 2.3) were factors associated with surveillance, and early intervention are recommended
MNA.
Dibonaventura MA was 42.5%. Medication side-effect and forgetfulness were Medication side-effects and resource are associated with
et al., USA [33] 86.19% and 48.4%, respectively. Agitation (OR = 0.6), sedation/cogni- MNA. Prevention, early detection, and effective manage-
tion (OR = 0.7), prolactin/endocrine (OR = 0.7), and side-effects (OR = ment of side-effects are crucial to avert it.
0.6) were significantly associated with MNA.
Gurmu, et al., MNA was 50.2%. Schizophrenia (75.7%), bipolar disorder (37.5%), The observed rate of antipsychotic MNA in this study was
Ethiopia [34] and depression (52.6%). Factors were perceived recovery (26.7%), high. Interventions to increase adherence are therefore
drug unavailability (18.1%), adverse effect (12.7%), forgetfulness crucial.
(10.6%), and being busy (8.6%).
Magura et al., USA Lower social support, alcohol use, lower satisfaction with Health care providers should encourage to address
[35] medication, side-effects, lower self-efficacy for avoidance and recov- patients’ adherence strategies via education about side-
ery, forgetfulness, unnatural to be controlled by medication, care- effects and benefits of the medication.
less at times, and felt better were the reasons for MNA.
Kikkert et al., Medication efficacy, external factors (such as patient support and Professionals, care-givers, and patients do not have a
European therapeutic alliance), insight, side-effects, and attitudes had influ- shared understanding of which factors are important.
countries [36] ence on MA.
Teferra et al., Inadequate availability of food, perceived strength of medications, Greater attention to provision of social and financial
Ethiopia [37] social support and safety net, lack of insight, failure to improve, side assistance will potentially improve MNA.
effects, substance abuse, stigma, and poor attitude of the care
provider were some of the main reasons for MNA.
Sher et al., USA Caregivers’ attribution of depression to cognitive and attitudinal Involving caregivers on the treatment plan, social support,
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 9 of 18
Table 2 The key findings and conclusions of studies included in the systematic review and meta-analysis (n = 46) (Continued)
Author, country Key findings (prevalence and associated factors) Conclusion
[38] problems, which significantly predicted patients’ MNA. Perceived and attitude may improve adherence.
stigma was also another predictor of non-adherence.
Mohamed et al., Insight and drug attitudes were associated with declining Better insight, positive attitudes toward medication, and
USA [39] schizophrenia symptoms but increasing levels of depression. educational interventions can be an important part of
Change toward more positive medication attitudes was associated psychosocial rehabilitation services.
with changes in insight, improve community functioning, and
greater medication compliance.
Sava, Turkey [40] MNA was 26.5% and associated with education, lack of insight, Lower education level, having thought of inadequate
thought they had recovered, believed that treatment had no-effect information about illness, and lack of insight about
on their disorder, thinking that had recovered, not taking medica- treatment were significantly associated with MNA.
tion, and thought of treatment not effective.
Sirey, USA [41] MNA was 82%. Elderly (24%) and younger (13%) patients Patients’ perceptions of stigma at the start of treatment
discontinued treatment completely. Patients perceived more stigma had influence their subsequent treatment behavior.
than older patients, stigma predicted treatment discontinuation.
Sajatovic. USA [42] MNA was 45.9%. Younger age, unmarried, homeless, substance Almost half of the patients had MNA that reduce the
abuse, or fewer outpatient psychiatric visits were predictors. effectiveness treatments in clinical settings.
Sajatovic. USA [43] MA was 51.9%. Factors associated were younger age, comorbid MNA is common in bipolar disorder medication.
substance abuse, and homelessness were the factors associated
with MA level.
John, USA [44] MNA was 77%. Weight gain and cognitive effects of a medication Patients’ satisfaction is seriously affect adherence. Health
most significantly affected patients’ likelihood of MA. care providers can optimize prescribing patterns.
Iseselo et al., Financial constraints, lack of social support, family disruption, A collaborative approach between the care providers,
Tanzania [45] stigma, discrimination, and disruptive behavior were some of the leader,and family is needed.
influencing factors for MNA.
J.M. Olivares, Spain Minimize patients waiting stay was significantly associated with MA. Treatment retention had greater improvement in clinical
[46] symptoms, reduce hospital stay, and increase efficacy.
Charlotte, Sweden Antidepressant MNA was 61.4%. Age (< 35 or > 64 years), having Patient and illness-related factors may imply an increased
[47] personality disorder, sensation-seeking traits, substance abuse, and risk of MNA.
unavailability of concomitant medications were predictors.
Adeponle, et al., Half (50.6%) of patients were adherent with appointments. Family support was significantly associated with
Nigeria [48] appointment, which can improve MA.
Rashid, Malaysia The type of antidepressant medication prescribed, not given a Involvement of patients, caregivers, flexible schedule, place
[49] choice to choose the treating doctor, and the preference to choice, drug, and doctor can help to prevent MNA.
traditional medicine were significant risk factors.
Roy, Ranchi (India) Poor infrastructure and lack of proper information about mental Develop community mental health care facilities and
[50] illness to patients and caregivers were some of the reasons for provide adequate information to patients and caregivers.
MNA.
Omran, Iran [51] Non-compliance was reported as a possible cause of admission in Providing a better insight about disease to patients to take
(88.2%) of the re-hospitalized cases. No insight to disease (59%) and their medications, even feeling of cure is important.
feeling of cure (27.6%) were causes for MNA.
Tara et al., Canada MA was 73%. Forgetting, change in routine, side effect, had lower Clinicians should be simple and easy to address
[52] self-efficacy, female, and had not completed post-secondary educa- medication efficacy, tolerability, and social moderator
tion were the most frequently identified reasons for MNA.
Banerjee, India MNA was 66.9%. Women (OR 2.7), consume extra pills (OR 2.8), and Interventions focusing on individuals and intersectoral
[53] had a considerably lower internal locus of control (OR 4.5) were system-oriented approach to improve MA are needed.
predictors
Oliver, Spain [54] MNA was 33.9%. Long-term treatment duration is a factor for MNA. Designing proper drug collection at pharmacies can
Women were more adherent than men. improve the MA of patients.
Dave, UK [55] Discontinuation was 80%. Lower discontinuation in the first Lack awareness was a risk for discontinuation.
6 months after initiation was associated with higher age, weight
gain, and comorbid irritable bowel syndrome.
Mahaye, South MNA was 50.8%. Age and race become predictors of MNA. Age and race were significant predictors for MNA.
Africa [56]
Sundell, Sweden MNA was 26.1%. It was less in women (OR, 0.8) and least 2 years of MNA occurred more commonly among social support
[57] higher education (OR, 0.7), and those who received social recipient
assistance (OR, 1.3).
Akincigil, [58] MNA was 49%. Care from a psychiatrist and higher general Substance abuse is one of the main risk factor for MNA
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 10 of 18
Table 2 The key findings and conclusions of studies included in the systematic review and meta-analysis (n = 46) (Continued)
Author, country Key findings (prevalence and associated factors) Conclusion
pharmacy utilization were associated with better adherence. and needs to be targeted for intervention.
Younger age, substance abuse, and comorbidity were associated
MNA.
Taj, Pakistan [11] MA among major depressive and bipolar disorders was 61.5% and MNA is a common and important issue. Treatment cost
73.9%, respectively. Reasons were sedation (30%), cost (22%), and co-morbidity are common factors
forgetting (36%), and no explanation by doctors (92%).
Prukkanone et al., MA was 41% but all patients who attended only once were non- MA to antidepressant therapy for treatment was high.
Thailand [59] adherent, adherence may be as low as 23%.
Shigemur, Japan MNA was 33.1%. It was associated with lower age, unemployed MNA was predicted by lower age and unemployment.
[60] (OR, 1.9), higher daily dosing frequency, low drug satisfaction, and
poor doctor–patient dyad, and age (> 34 years) (OR, 1.6).
Bambouer et al., MNA was 75%. Rates of antidepressant non-adherence significantly Effectiveness of electronically triggered, patient-specific,
USA [61]. increased over time were 40%. and faxed feedback should be carefully evaluated.
Demyttenaere MA was 70%, and it was decreased by 2.5% per month and more MA decreases with time is influenced by demographic and
et al., Belgium [62] than three times more rapidly in drop-outs. clinical variables.
Mascha C. Ten D MNA ranged from 39.7 to 52.7%. It did not significantly differ MNA is high on MDD. Doctors continuously have to be
[63]. between intermittent ad continuation antidepressant users (37.2% aware of this problem
versus 25%).
Baldessarini et al., MNA was 33.8%. Prescribing psychiatrists considered only 6% as Underestimation of the problem may encourage
USA [64] MNA. Alcohol, youth, comorbidity, side effects, obsessive- increasingly complex treatment regimens of untested
compulsive disorder, and recovering from mania-hypomania and value, added expense, and risk of adverse effects
drug-complexity were the predictors.
Nega et al., MNA was 61.2%. It was associated with female (AOR, 2.3), Psychotropic MNA was high. We recommend the
Ethiopia [65] combined drug (AOR, 2.7), long treatment duration (AOR, 2.3), > 24 concerned bodies to design and implement programs
months (AOR, 2.5), substance use (AOR, 2.6), perceived stigma (AOR, focused on associated factors in order to improve MA.
2.2), patient’s poor attitude (AOR, 3.0), and poor social support
(AOR, 1.8).
AOR adjusted odds ratio, CI confidence interval, MA medication adherence, MNA medication non-adherence, SCID-I Structural Clinical Interview Diagnosis I, TDM
therapeutic drug monitoring
be unnatural and reported feeling better after terminat- Medication side-effects
ing them were the factors associated with medication In several studies, psychotropic medication non-
non-adherence [35]. Similarly, in two studies, those pa- adherence was associated with medication-related side-
tients who perceived that the treatment had no effect on effects [11, 22–25, 28–30, 32–37, 39, 44, 50, 52, 55, 64,
their illness were more likely to be non-adherent to their 65]. In seven studies, patients feeling dizziness, fatigue,
medication [26, 40]. In seven studies, behavioral factors tiredness, sedation, lethargy, and sleepiness were the most
such as forgetting the right dose and right time of taking frequently reported side-effects that contributed to medi-
medication were the factors associated with medication cation non-adherence [11, 33, 36, 37, 50, 52, 65]. In two
non-adherence [11, 22, 24, 28, 34, 35, 52]. In six studies, studies, sleepiness during day time (medication dose time)
patients and caregivers reported being busy with daily and potentially life-threatening or distressing side-effects
routines, careless about the timing, forgetting to remem- seriously affected patients’ medication non-adherence [37,
ber medication time, and irregular follow-up were asso- 44]. Another two studies, feeling of powerlessness, insom-
ciated with medication non-adherence [11, 22, 24, 34, nia, difficulty thinking or concentrating, restlessness, or
35, 52]. In the worst scenario, patients’ complete rejec- feeling jittery were found to be associated with medication
tion of the medication was a main cause of discontinu- non-adherence [28, 37]. Likewise, in five studies, weight
ation and non-adherence to their medication [28]. gain was another medication-related side-effect that asso-
ciated with medication non-adherence and patients’ per-
ception toward their medication [25, 33, 36, 44, 55].
Clinical factors In two studies, side-effects such as decreased sexual
The clinical factors of medication non-adherence were interest and having a symptom of sexual dysfunction
re-categorized into medication side-effect, lack of insight were associated with patients’ medication adherence [33,
about their illness and treatment, comorbidity, medica- 36]. Moreover, patients and caregivers’ perceived medi-
tion efficacy, long treatment duration, and complexity of cation adverse drug reaction was a factor associated with
the prescribed medication. psychotropic medication non-adherence [34, 36].
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 11 of 18
Fig. 2 Pooled estimate of medication non-adherence (n = 35)
Likewise, extra pyramidal symptoms or agitation [33], and memory problem [50] were common factors associ-
other medication-related side-effects such as cognitive ated with medication non-adherence.
deterioration or impairment [44], missing voice [28],
paralysis of body parts, twisting of the neck, drooling, Lack of insight about illness and medication
weakness, appetite stimulation [37], severe depressive In seven studies, patients’ lack of insight (level of awareness
symptoms and episodes [39, 44], salivation, dry mouth, or understanding) about their illness and medication was a
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 12 of 18
Fig. 3 Pooled estimate of medication non-adherence of the schizophrenia patients (n = 9)
common factor associated with psychotropic medication perceived medication efficacy such as subjective relief of
non-adherence [21, 23, 28, 39, 40, 50, 51]. Likewise, misun- symptoms, patients’ feel drugs have no effect on the illness,
derstanding about the treatment consequences, lack of not helpful, being ineffective [36, 46, 50], and feeling of
awareness of their illness and or mental disorder in general, cured [51] were side-effect-related factors associated with
and sometimes appreciating subjective relief symptoms [30, medication non-adherence.
36, 37] were the factors associated with medication non-
adherence among major psychiatric disorder patients. Medication duration
In five studies, long treatment duration (6–12months and
Medication efficacy longer) was an associated factor for medication non-
The pharmacological management of psychiatric disorders adherence [22, 25, 37, 43, 65]. Similarly, having long-term
needs safe and efficacious medication to achieve desired medication prescriptions, long duration maintenance ther-
treatment goals. The fact that lower medication efficacy apy [29, 54], and irregular follow-up [29] were associated
and patient self-rating of efficacy were also factors associ- with psychotropic medication non-adherence.
ated with psychotropic medication non-adherence. Taking
lower potent concomitant psychotropic medications [47, Treatment complexity
65], recovery from illness [34], felt better [24, 52], and fail- In three studies, multiple dose, frequency and drug com-
ure to improve with medication [37] were the factors asso- binations, or complex drug regimen were seriously
ciated with medication efficacy related with psychotropic linked with medication non-adherence [24, 32, 64]. In
medication non-adherence. Likewise, patients’ or caregivers’ two studies, pill burden or consuming extra pills was
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 13 of 18
Fig. 4 Pooled estimate of medication adherence of the major depressive disorders patients (n = 16)
also one of treatment-related factors that negatively in- Lack of social support
fluenced patients’ adherence to their psychotropic medi- In seven studies, poor or lack of social or family support
cation [22, 53]. In another two studies, taking was associated with psychotropic medication non-
medication twice per day was a negative factor for medi- adherence [21, 24, 25, 35, 37, 45, 65]. In two studies,
cation adherence [46, 49]. In addition, the route of medi- limited or inadequate patient information, weak profes-
cation administration had a significant effect on sional or family support, therapeutic alliance, social in-
medication non-adherence. The patients on injectable volvement, and low education were some of the social
medication were more likely to be adherent than the pa- support-related factors [36, 37]. Cohesiveness, family
tient taking drugs orally [46]. reminding, and transport to hospital [37], lower family
harmony or lack of resilient family support, discrimin-
Co-morbidity ation by nearby people, disruption of family functioning
In three studies, psychotropic medication adherence was or household routine and religious practices [45], weak
compromised where there were co-morbidities of mental community functioning [39], homelessness [42, 43], had
illness and other physical illnesses. Studies [11, 43, 64] old age caregivers or lack of caregivers [50], lack of fam-
reported that medication non-adherence was associated ily compliance of follow-up [51], lack of advice about
with patients having co-morbidities with their current their medication intake from friends and relatives [28],
psychiatric disorders. Of these, affective morbidity, not receiving social assistance [57], and caregivers’ attri-
obsessive-compulsive disorders, recovering from mania- bution of depression to cognitive and attitudinal prob-
hypomania [64], personality disorders and sensation of lem [38] were the factors associated with psychotropic
seeking personality traits [47], and alcohol abuse disor- medication non-adherence.
ders [61] were negatively associated with medication ad-
herence. Irritable bowel syndrome as a co-morbidity was Health system-related factors
also significantly associated with medication non- The health system-related factor was the crucial area for
adherence [55]. getting quality mental health service. In three studies,
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 14 of 18
Fig. 5 Pooled estimate of medication non-adherence of bipolar disorder patients (n = 10)
medication non-adherence was associated with lack of In three studies, patient-physician or therapist relationship
free access to medicine due to inadequate or unavailabil- was crucial for better medication adherence. Consequently,
ity of psychotropic drug supplies in health facilities [24, unfriendly, judgmental behavior, inflexible appointment sys-
25, 34]. In one study, although psychotropic medications tems, mistrust, and having negative patient-physician rela-
were normally provided free of charge in the govern- tionships were the factors associated with patients’
ment health facilities, patients were suffering unavailable psychotropic mediation non-adherence [28, 30, 60]. In one
of medication in the government pharmacies. Thus, pa- study, health care providers’ negative attitude had influenced
tients need to buy from private pharmacies which are patients’ adherence to the medication and their follow-ups
very expensive and lead to interruption of the medica- [37]. Similarly, in two studies, health professional shortages
tion. In addition, health care provider sometimes had also affected medication adherence [49, 58]. Patient
changes the drug but it may not be found in the govern- preference for traditional/complementary medicine was an-
ment hospital pharmacy [45]. Therefore, the lack of al- other cause of medication non-adherence [49]. Medication
ternative drug or therapy affects psychotropic non-adherence was affected by the number of hospitaliza-
medication adherence [39]. On the other hand, the lack tions [28], irregular hospitalization and frequently discharge
of sufficient and quality health education to psychiatric of patients, length of stay [42, 46], lack of patients’ satisfac-
patients and or their caregivers/relatives/families about tion with health care services [44], and long distance to ac-
the medication and illness influenced patients’ adherence cess the health service/recollect medications [50]. In three
to their prescribed medication [22, 36, 39, 40, 51]. studies, health care providers would be unable to explain
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 15 of 18
and optimize prescribing pattern, timing, and dose benefit from a systematic review which revealed that adherence in
of medication. In addition, the lack of friendly deal with psychiatric patients ranged from 10.7 to 38% [76].
medication complexity, tolerability, efficacy, and health be- This systematic review of factors influencing psycho-
lief issues were critical factors influence medication adher- tropic medication non-adherence which is consistent
ence [11, 44, 52]. Furthermore, the health care system has with other systematic reviews [14, 72, 73, 77] has shown
also associated with medication non-adherence. These fac- that medication adherence is mainly affected by patients’
tors were poor service structure and cumbersome purchas- negative attitude toward their medication, lack of insight,
ing procedure (affect access), availability and timely use or negative health belief, and perceived stigma. Similarly,
collection of psychotropic medication during follow-up visit, medication non-adherence is consistently associated
and patients not covered by health insurance scheme [58]. with patients behavioral practices (e.g., substance abuse)
[14, 74] and also patients’ socio-demographic character-
istics (such as educational status, age, gender, and em-
Medication cost
ployment) [14, 72]. The present systematic review has
In seven studies, psychiatric patients and their caregivers
identified that the lack of social support is associated
having financial constraints to buy medicines were fac-
with medication non-adherence among major psychi-
tors associated with medication non-adherence. In
atric disorder patients. This is similar with other reviews,
addition, the lack of money for transportation, to pur-
which have reported that the lack of family involvement,
chase proper food, and to buy medications were the fac-
care/dyad support, and other social supports are strongly
tors associated with patients’ adherence to their
negatively associated with poorer therapeutic alliance
medication. Psychotropic medications had an appetite
[14, 72, 75, 76]. In addition, medication non-adherence
stimulation that has been increasing food demand which
is associated with clinical- or medication-related factors
incurs an additional economic burden [11, 24, 32, 37, 42,
[14, 72–74, 77]. This finding is supported by another
45, 50] and contribute for medication non-adherence.
systematic review which revealed that psychiatric dis-
order comorbidities with other physical disorders influ-
Discussion ence medication adherence and increase re-admission of
This systematic review and meta-analysis determined psychiatric patients [73, 78, 79].
the pooled proportion of psychotropic medication non- In the present systematic review, medication non-
adherence and synthesized the associated factors with adherence is associated with poor functioning of the
medication non-adherence among major psychiatric dis- health system such as lack of psychotherapy, lack of in-
order patients. Almost half (49%) of patients with major formation, long treatment duration with little health
psychiatric disorders did not adhere to their psycho- personnel follow-up, inadequate discharge planning, in-
tropic medication. Medication non-adherence among creased hospitalizations, poor support and care environ-
patients with schizophrenia, major depressive disorder, ment, experiencing access barriers to high-quality care
and bipolar disorder were 56%, 50%, and 44%, respect- and health care providers unable to provide elicit infor-
ively. Medication non-adherence is influenced by various mation on adherence, inadequate medication coverage,
factors such as patients’ individual behavior, social or and poorer therapeutic alliance [14, 72, 74–76, 79]. Fi-
family support, clinical or illness and treatment-related, nancial factors seriously affected medication adherence.
and overall health care system-related factors. These included unaffordability of medication, increased
Previous systematic reviews have indicated that medica- health care cost [73, 74], lack of health insurance [75],
tion non-adherence is a common challenge in the treat- patients’ poor capacity, and limited resources [73].
ment of psychiatric disorders [72, 73]. This meta-analysis A large amount of heterogeneity in the definition and
finding is consistent with a systematic review revealed an measurement methods used to assess medication ad-
overall medication adherence level of 58% (ranged from 24 herence have been reported in some reviews. The het-
to 90%), and medication adherence to antidepressants was erogeneity of factors related to non-adherence calls for
65% [6]. Another earlier systematic review has shown that individually tailored approaches to promote adherence
the level of medication non-adherence was 60% [74]. The [80, 81]. Non-adherence contributes enormously to
present systematic review and meta-analysis finding is con- poor health outcome and needs substantial work to im-
sistent with a finding from a comprehensive systematic re- prove treatment outcomes [80]. Evidence showed that
view on schizophrenia which reported that a mean rate of improving adherence to psychotropic medications
non-adherence was 41.2%. The sub-group analysis indi- could have a positive impact on patients and society.
cated that a mean non-adherence rate was 49.5% [14], and Non-adherence issues need to be looked at from many
another systematic review has shown that psychotropic me- angles and taking a multifaceted approaches with pa-
diation non-adherence was 44% [75]. Nevertheless, the tients and healthcare providers to address identified
present meta-analysis finding is a bit lower than a finding challenges [81].
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 16 of 18
Conclusions Availability of data and materials
Almost half of patients with major psychiatric disorder did The data that support the review findings are available upon submitting a
reasonable request to the corresponding author.
not adhere to their psychotropic medication. Patients’ indi-
vidual behavior, lack or poor social/family support, treat- Ethics approval and consent to participate
ment and illness-related clinical conditions, and the health Not applicable.
system barriers are influencing factors of psychotropic
Consent for publication
medication non-adherence among patients with major psy- Not applicable.
chiatric disorders. Therefore, multifaceted intervention is
needed to create supportive environment for patients and Competing interests
caregivers to minimize psychotropic medication non- The authors declare that they have no competing interests.
adherence. Additionally, supportive social and health care Author details
system programs should be designed to alleviate major psy- 1Department of Population, Family and Reproductive Health, School of
chiatric disorder patients’ medication non-adherence. Com- Public Health, College of Health Sciences, University of Ghana, Accra, Ghana.2College of Health and Medical Sciences, Haramaya University, Po Box 235,
prehensive approaches targeting the factors that affect Harar, Ethiopia. 3Population Council/Ghana, Yiyiwa Drive, Accra, Ghana.
medication non-adherence can bring tremendous positive
outcomes. This systematic review and meta-analysis finding Received: 9 April 2018 Accepted: 6 January 2020
can be helpful to inform policy-makers, clinicians, and
other caregivers to undertake necessary decisions to estab- References
lish an integrated approach to boost therapeutic alliance 1. WHO. The bare facts, Geneva. Geneva: WHO; 2008.
2. WHO. Mental disorders fact sheet. Reviewed April. Geneva; 2016. Available
and improve medication adherence. from: http://www.who.int/mediacentre/factsheets/fs396/en/. Accessed 16
Mar 2017
3. World Bank Group and, WHO. Out of the shadows: making mental health a
Supplementary information global development priority organized by World Bank Group&WHO, April 13–
Supplementary information accompanies this paper at https://doi.org/10. 14. Washington D.C; 2016. Available from: http://www.worldbank.org/en/
1186/s13643-020-1274-3. events/2016/03/09/out-of-the-shadows-making-mental-health-a-global-priority
4. WHO. Comprehensive Mental Health Action Plan 2013-2020. Geneva: World
Additional file 1. PRISMA checklist Helath Organziation; 2013. https://www.who.int/mental_health/action_
plan_2013/en/.
Additional file 2. Sample searching strategies 5. Mathers CD, Loncar D. Projections of global mortality and burden of disease
Additional file 3. Newcastle Ottawa Scale (NOS) from 2002–2030. PLoS Med. 2006;3:e442.
Additional file 4. Data extraction on Excel sheet 6. Cramer JA, Rosenheck R. Compliance with medication regimens for mental
and physical disorders. Psychiatr Serv. 1998;49(2):196–201.
Additional file 5. Funnel plot for exploration of publication bias (for 7. Farooq S, Naeem F. Tackling nonadherence in psychiatric disorders: current
overall pooled and subgroup analysis) opinion. Review. Dovepress Neuropsychiatr Dis Treat. 2014;10:1069–77.
8. Colom F, Vieta E. Non-adherence in psychiatric disorders: misbehaviour or
clinical feature? Acta Psychiatr Scand. 2002;105:161–3.
Abbreviations 9. WHO. Adherence to long-term therapies; Evidence for action. Geneva:
AOR: Adjusted odds ratio; CI: Confidence interval; MA: Medication adherence; World Health Organization; 2003. https://apps.who.int/iris/bitstream/
MNA: Medication non-adherence; NRR: No response rate; PRISMA: Preferred handle/10665/42682/9241545992.pdf;jsessionid.
Reporting Items for Systematic Review and Meta-Analysis; 10. Rekha R, Masroor J, Sushma K, et al. Reasons for drug non-compliance of
PROSPERO: Prospective Register of Systematic Reviews; RR: Response rate; psychiatric patients. J Indian Acad Appl Psychol. 2005;31(1):24–8.
TDR: Tropical Disease Research; US: United States; WHO: World Health 11. Taj F, Tanwir M, Aly Z, Khowajah AA, Tariq A, Syed FK, et al. Factors
Organization associated with non-adherence among psychiatric patients at a Tertiary
Care Hospital, Karachi, Pakistan: a questionnaire based cross-sectional study.
Acknowledgements J Pak Med Assoc. 2008;58(432):432–6.
We thank TDR/WHO international postgraduate scholarship program and 12. Smith F, Clifford S. Adherence to medication among chronic patients in Middle
School of Public Health, University of Ghana, for financial support. We thank Eastern countries: review of studies. East Mediterr Heal J. 2011;17(4):356–63.
College of Health and Medical Sciences, Haramaya University, for office 13. Lindstrom E, Bingefors K. Patient compliance with drug therapy in
arrangement. schizophrenia: economic and clinical issues. Pharmacoeconomics. 2000;
18(2):106–24.
14. Dunn LB, Leckband S, Dolder CR, Pharm D, Leckband SG, Ph R, et al.
Authors’ contributions Prevalence of and risk factors for medication nonadherence in patients with
AS, KT, AM, and AA conceived and designed the study. AS drafted the schizophrenia. J Clin Psychiatry. 2002;63(10):892–909.
manuscript and is the guarantor of the systematic review and meta-analysis. 15. Burton WN, Chen C-Y, Conti DJ, Schultz AB, Edington DW. The association
AS and GT developed the search strings and performed searching the stud- of antidepressant medication adherence with employee disability absences.
ies, selection of the studies, data extraction, and synthesis. AS, KT, AM, NA, Am J Manag Care. 2007;13(2):105–13.
GT, and AA extensively reviewed the manuscript and incorporated intellec- 16. Semahegn A, Torpey K, Manu A, Assefa A, Tesfaye G, Ankomah A.
tual inputs. All authors read, provided feedback, and approved the final ver- Psychotropic medication non-adherence and associated factors among
sion of the manuscript. adult patients with major psychiatric disorders: a protocol for systematic
review and meta-analysis. PROSPERO. 2017. CRD42017067436. https://www.
Funding crd.york.ac.uk/prospero/display_record.php?ID=CRD42017067436.
This work has been funded by TDR, the Special Programme for Research and 17. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al.
Training in Tropical Diseases, which is hosted at the World Health Preferred reporting items for systematic review and meta-analysis protocols
Organization and co-sponsored by UNICEF, UNDP, the World Bank, and (PRISMA-P) 2015 statement. Syst Rev. 2015;4(1):1–9 Available from: http://
WHO. The grant number for the University of Ghana is B40300. www.systematicreviewsjournal.com/content/4/1/1.
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 17 of 18
18. Semahegn A, Torpey K, Manu A, Assefa N, Tesfaye G, Ankomah A. 37. Teferra S, Hanlon C, Beyero T, Jacobsson L, Shibre T. Duplicated-perspectives
Psychotropic medication non-adherence and associated factors among on reasons for non-adherence to medication in persons with schizophrenia
adult patients with major psychiatric disorders: a protocol for a systematic in Ethiopia: a qualitative study of patients, caregivers and health workers.
review. BMC Syst Rev. 2018;7:10. BMC Psychiatry. 2013;13:168.
19. Rutgers. RUTGERS the state University of New Jersey. In: George F, editor. 38. Sher I, McGinn L, Sirey JA, Meyers B. Effects of caregivers’ perceived stigma
Smith Library of the Health Sciences. EndNoteX8.0.1(BId 10444), Cite While and causal beliefs on patients’ adherence to antidepressant treatment.
You Write TM Patented technology U.S patent number 8,092,241: University Psychiatr Serv. 2005;56(5):564–9.
of Ghana; 2017. 3091850168. Available from: http://www.libraries.rutgers. 39. Mohamed S, Rosenheck R, Mcevoy J, Swartz M, Stroup S, Lieberman JA.
edu/sites/default/files/smith/pdf/EndNoteBasic.pdf. Cross-sectional and longitudinal relationships between insight and attitudes
20. Wells G, Shea B, O’Connell D, Peterson J, Welch V, Losos M, et al. The toward medication and clinical outcomes in chronic schizophrenia.
Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised Schizophr Bull. 2009;35(2):336–46.
studies in meta-analyses. Ottawa: Ottawa Health Research Institute; 2010. 40. Savaş HA, Unal A, Vırıt O. Original papers/Araştırmalar treatment adherence
21. Ibrahim AW, Yahya S, Pindar SK, Wakil MA, Garkuwa A, Sale S. Prevalence in bipolar disorder. J Mood Disord. 2011;1(3):95–102.
and predictors of sub-optimal medication adherence among patients with 41. Sirey JA, Bruce ML, Alexopoulos GS, Perlick DA, Raue P, Friedman SJ, et al.
severe mental illnesses in a tertiary psychiatric facility in Maiduguri, North- Discontinuation in young and older outpatients perceived stigma as a
eastern Nigeria. Pan Afr Med J. 2015;21:39. predictor of treatment with depression. Am J Psychiatry. 2001;158(3):479–81.
22. Alene M, Wiese MD, Angamo MT, Bajorek BV, Yesuf EA, Wabe NT. Duplicate- 42. Sajatovic M, Valenstein M, Blow F, Ganoczy D, Ignacio R. Treatment
adherence to medication for the treatment of psychosis: rates and risk adherence with lithium and anticonvulsant medication among patients
factors in an Ethiopian population. BMC Clin Pharmacol. 2012;12:10. with bipolar disorder. Psychiatr Serv. 2007;58(6):855–63.
23. Eticha T, Teklu A, Ali D, Solomon G, Alemayehu A. Factors associated with 43. Sajatovic M, Valenstein M, Fc B, Ganoczy D, Rv I. Treatment adherence with
medication adherence among patients with schizophrenia in Mekelle, antipsychotic medications in bipolar disorder. Bipolar Disord. 2006;8:232–41.
Northern Ethiopia. PLoS One. 2015;10(3):e0120560. Laks J, editor; [cited 2017 44. Johnson FR, Özdemir S, Manjunath R, Hauber AB, Burch P, Thompson TR, et al.
Jun 18]. https://doi.org/10.1371/journal.pone.0120560. Approach linked references are available on JSTOR for this article: factor
24. Tesfay K, Girma E, Negash A, Tesfaye M. Non-adherence among adult treatments a stated-preference approach. Med Care. 2007;45(6):545–52.
psychiatric out patients in Jimma University Specialized Hospital, Southwest 45. Iseselo MK, Kajula L, Yahya-malima KI. The psychosocial problems of families
Ethiopia. Ethiop J Health Sci. 2013;23(3):227–38. caring for relatives with mental illnesses and their coping strategies: a
25. Hibdye G, Dessalegne Y, Debero N, Bekan L, Sintayehu M. Prevalence of qualitative urban based study in Dar Es Salaam. BMC Psychiatry. 2016;16:1–12.
drug non-adherence and associated factors among patients with bipolar 46. Olivares JM, Rodriguez-morales A, Diels J, Povey M. Long-term outcomes in
disorder at outpatient unit of Amanuel Hospital, Addis Ababa, Ethiopia. J patients with schizophrenia treated with risperidone long-acting injection
Psychiatry. 2015;1:003. https://doi.org/10.4172/2378-5756.S1-003. or oral antipsychotics in Spain: results from the electronic schizophrenia
26. Sirey JA, Bruce ML, Alexopoulos GS, Perlick DA, Friedman SJ, Meyers BS. treatment adherence registry (e-STAR ) *. Eur Psychiatry. 2009;24(5):287–96.
Perceived stigma and patient-rated severity of illness as predictors of 47. Åkerblad A, Bengtsson F, Holgersson M, Von Knorring L. Identi fi cation of
antidepressant drug adherence. Psychiatr Serv. 2001;52(12):1615–20. primary care patients at risk of nonadherence to antidepressant treatment.
27. Hill M, Crumlish N, Whitty P, Clarke M, Browne S, Kamali M, et al. Patient Prefer Adherence. 2008;2:379–86.
Nonadherence to medication four years after a first episode of psychosis 48. Adeponle AB, Thombs BD, Adelekan ML, Kirmayer LJ. And medication
and associated risk factors. Psychiatr Serv. 2010;61(2):189–92. [cited 2017 Jun adherence at a Nigerian psychiatric hospital family participation in
18]. https://doi.org/10.1176/ps.2010.61.2.189. treatment, post-discharge appointment and medication adherence at a
28. Moritz S, Favrod J, Andreou C, Morrison AP, Bohn F, Veckenstedt R, et al. Nigerian psychiatric hospital. Br J Psychiatry. 2009;194:86–7.
Beyond the usual suspects : positive attitudes towards positive symptoms is 49. Rashid AK, Rahmah MA. Treatment related risk factors associated with the
associated with medication noncompliance in psychosis. Schizophr Bull. default of depression treatment among the elderly: a case-control study.
2013;39(4):917–22. Open Geriatr Med J. 2010;3:11–6.
29. Mert DG, Turgut NH, Kelleci M, Murat S. Perspectives on reasons of medication 50. Roy R, Jahan M, Kumari S. Reasons for drug non-compliance of psychiatric
nonadherence in psychiatric patients. Patient Prefer Adherence. 2015;9:87–93. patients: a centre-based study. J Indian Acad Appl Psychol. 2005;31(1):24–8.
30. Novick D, Montgomery W, Treuer T, Aguado J, Kraemer S, Haro JM. 51. Omranifard V, Yazdani M, Yaghoubi M, Namdari M. Noncompliance and its
Relationship of insight with medication adherence and the impact on causes resulting in psychiatric readmissions. Iran J Psychiatry. 2008;3(1):37–42.
outcomes in patients with schizophrenia and bipolar disorder: results from 52. Burra TA, Chen E, Mcintyre RS, Grace SL, Blackmore ER, Stewart DE, et al.
a 1-year European outpatient observational study. BMC Psychiatry. 2015; Predictors of self-reported antidepressant adherence predictors of self-
15(1):189. [cited 2017 Jun 18]. https://doi.org/10.1186/s12888-015-0560-4. reported antidepressant. Behav Med. 2007;32(4):127–34.
31. Odo HO, Agbonile IO, Esan PO, Jeffrey S, BOJ S. Assessment of 53. Banerjee S. Factors affecting non-adherence/compliance among patients
adherence to psychotropic medications among out-patients at the diagnosed with unipolar depression in a psychiatric department of a
Pharmacy Department of a Psychiatric Hospital, in Benin City, Nigeria. general hospital in Kolkata, India. Trivandrum: Achutha Menon Centre for
Asian J Pharm. 2014;8:211–5. Health Science Studies, Sree Chitra Tirunal Institute fo; 2012.
32. Ibrahim A, Pindar SK, Yerima MM, Rabbebe IB, Shehu S, Garkuwa HA, et al. 54. Lacasta-tintorer D, García-lecina R, Flamarich-zampalo D, Font-canal T,
Medication-related factors of non adherence among patients with Martín MIF. Do depressed patients comply with treatments prescribed? A
schizophrenia and bipolar disorder: outcome of a cross-sectional survey in cross-sectional study of adherence to the antidepressant treatment. Actas
Maiduguri, North-eastern Nigeria. J Neurosci Behav Heal. 2015;7(5):31–9. Esp Psiquiatr. 2011;39(5):288–93.
33. Dibonaventura M, Gabriel S, Dupclay L, Gupta S, Kim E. A patient 55. Davé S, Classi P, Le TK, Maguire A, Ball S. Discontinuation of antidepressant
perspective of the impact of medication side effects on adherence: results therapy among patients with major depressive disorder. Open J Psychiatry.
of a cross-sectional nationwide survey of patients with schizophrenia. BMC 2012;2:272–80.
Psychiatry. 2012;12:20. 56. Mahaye S, Mayime T, Nkosi S, Mahomed FN, Pramlal J, Setlhabana O, et al.
34. Gurmu AE, Abdela E, Allele B, Cheru E, Amogne B. Rate of non-adherence Medication adherence of psychiatric patients in an outpatient setting.
to antipsychotic medications and factors leading to non-adherence among African J Pharm Pharmacol. 2012;6(9):608–12.
psychiatric patients in Gondar University Hospital, Northwest Ethiopia. 57. Sundell KA, Waern M, Petzold M, Gissler M. Socio-economic determinants of
Hindawi Publ Corp. Advances in Psychiatry.2014;2014:ID 475812. https:// early discontinuation of anti-depressant treatment in young adults. Eur J
www.hindawi.com/journals/apsy/2014/475812/. Pub Health. 2011;23:1–7.
35. Magura S, Rosenblum A, Fong C. Factors associated with medication 58. Akincigil A, Bowblis JR, Levin C, Walkup JT, Saira Jan P, Crystal S. Adherence
adherence among psychiatric outpatients at substance abuse risk. NIH to antidepressant treatment among privately insured patients diagnosed
Public Access Open Addict J. 2012;4:58–64. with depression. NIH Public Access Med Care. 2007;45(4):363–9.
36. Kikkert MJ, Schene AH, Maarten WJ, Robson D, Born A, Helm H, et al. 59. Prukkanone B, Vos T, Burgess P, Chaiyakunapruk N, Bertram M. Adherence
Medication adherence in schizophrenia: exploring patients’, carers’ and to antidepressant therapy for major depressive patients in a psychiatric
professionals’ views. Schizophr Bull. 2006;32(4):786–94. hospital in Thailand. BMC Psychiatry. 2010;10:64.
Semahegn et al. Systematic Reviews            (2020) 9:17 Page 18 of 18
60. Shigemura J, Ogawa T, Yoshino A, Sato Y, Nomura S. Predictors of
antidepressant adherence: results of a Japanese internet-based survey.
Psychiatry Clin Neurosci. 2010;64:179–86.
61. Bambauer KZ, Adams AS, Zhang F, Minkoff N, Grande A, Weisblatt R, et al.
Physician alerts to increase antidepressant adherence. Arch Intern Med.
2006;166:498–504.
62. Demyttenaere K, Adelin A, Patrick M. Six-month compliance with
antidepressant medication in the treatment of major depressive disorder.
Int Clin Psychopharmacol. 2008;23:36–42.
63. Mascha C, Bockting CLH, Schene AH. Adherence to continuation and
maintenance antidepressant use in recurrent depression. J Affect Disord.
2009;115:167–70.
64. Baldessarini RJ, Perry R, Pike J. Factors associated with treatment
nonadherence among US bipolar disorder patients. Hum Psychopharmacol
Clin Exp. 2008;23:95–105.
65. Nega M, Semahegn A, Demissie M. Psychotropic medication non-adherence
and associated factors among patients with severe psychiatric disorders
attending psychiatry clinics in selected hospitals, eastern Ethiopia: a cross-
sectional study. Unpubl Thesis 2016
66. Barendregt JJ, Doi SA, Lee YY, Norman RE, Vos T. Meta-analysis of
prevalence. J Epidemiol Community Health. 2013;67:974–8.
67. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al.
Meta-analysis of observational studies in epidemiology. A proposal for
reporting. JAMA. 2000;283(15):2008–12.
68. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis.
Stat Med. 2002;1558:1539–58.
69. Higgins JPT, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD. The
Cochrane collaboration’s tool for assessing risk of bias in randomised trials.
BMJ. 2011;343:d5928.
70. DerSimonian R, Laird N. Meta-analysis in clinical trials revisited: HHS Public
Access. Contemp Clin Trials. 2015;45(0 0):139–45.
71. Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected
by a simple, graphical test. BMJ. 1997;315:629–34.
72. Colom F, Vieta E, Tacchic M, Sanchez-Moreno J, Scott J. Identifying and
improving non-adherence in bipolar disorder Identifying and improving
non-adherence in bipolar disorders. Bipolar Disord. 2005;7(Suppl 5):24–31.
73. Chapman SCE, Horne R. Medication nonadherence and psychiatry: review.
Curr Opin Psychiatry. 2013;26(5):446–52.
74. Kreyenbuhl J, Record EJ, Palmer-bacon J. A review of behavioral tailoring
strategies for improving medication adherence in serious mental illness. Clin
Res. 2016;18(2):191–201.
75. Lanouette NM, Folsom DP, Sciolla A, Jeste DV. Psychotropic medication
nonadherence among United States Latinos : a comprehensive literature
review. Psychiatr Serv. 2009;60(2):157–74.
76. Smith D, Lovell J, Weller C, Kennedy B, Winbolt M, Young C, et al. A
systematic review of medication non-adherence in persons with dementia
or cognitive impairment. PLoS One. 2017;12(2):e0170651.
77. Aikens JE, Nease DE, Nau DP, Klinkman MS, Schwenk TL. Adherence to
maintenance-phase antidepressant medication as a function. Ann Fam Med.
2005;3(1):23–30.
78. Lilijana Š, Zvezdana M, Wahlbeck K, Haaramo P. Psychiatric readmissions
and their association with physical comorbidity: a systematic literature
review. BMC Psychiatry. 2017;17(2):1–17.
79. Masand PS, Roca M, Martin S, Turner JMK. Partial adherence to antipsychotic
medication impacts the course of illness in patients with schizophrenia: a
review. Prim Care Companion J Clin Psychiatry. 2009;11(4):148–55.
80. Barkhof E, Meijer CJ, De Sonneville LMJ, Linszen DH, De Haan L. Interventions
to improve adherence to antipsychotic medication in patients with
schizophrenia – a review of the past decade. Eur Psychiatry. 2012;27:9–18.
81. Higashi K, Medic G, Littlewood KJ, Diez T, Granström O, De Hert M.
Medication adherence in schizophrenia: factors influencing adherence and
consequences of nonadherence, a systematic literature review. Ther Adv
Psychopharmacol Orig. 2013;3(4):200–18.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.