Article Text
PDF
Abstract
Objectives This study aimed to assess the prevalence of treatment-resistant depression (TRD) and associated factors among patients with major depressive disorder (MDD) on follow-up at Amanuel Mental Specialised Hospital, Addis Ababa, Ethiopia, 2021.
Design and setting An institution-based cross-sectional study design was employed using systematic random sampling techniques from 17 February to 26 March 2021.
Participants The study enrolled 412 participants with a response rate of 97.6%. The study population consisted of Saint Amanuel Mental Specialised Hospital follow-up patients with MDDs and all adult patients aged above 18.
Main outcome measures The main outcome of this study was TRD, which was measured using the Hospital Anxiety and Depression Scale-Depression (HADS-D). The collected data were entered into Epi-data software version 3.1 and exported to the statistical package for social science version 20 for analysis. Bivariate and multivariate logistic regression analyses were used to identify associated factors with TRD. The OR with a 95% CI was used to assess the strength of the association.
Results The prevalence of TRD was 41.5% (95% CI: 37.2 to 46.1). Female sex [AOR=2.43, 95% CI: 1.57 to 3.75], comorbid psychosis [AOR=1.89, 95% CI: 1.19 to 2.99], comorbid medical illness [AOR=1.67, 95% CI: 1.09 to 2.55] and family history of mental illness [AOR=2.27, 95% CI: 1.38 to 3.74] were factors significantly associated with TRD.
Conclusion and recommendation In this study, the prevalence of TRD among patients with MDDs on follow-up was high. Therefore, to improve outcomes, screening for TRD and creating specific diagnostic techniques are necessary. Additionally, preventive interventions against TRD must be established.
- Depression & mood disorders
- PSYCHIATRY
- Lipid disorders
- Burnout, Professional
- Caregiver Burden
- CLINICAL PHARMACOLOGY
Data availability statement
Data are available upon reasonable request. This research included pertinent data, and the corresponding author/s are available to offer further information upon reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
- Depression & mood disorders
- PSYCHIATRY
- Lipid disorders
- Burnout, Professional
- Caregiver Burden
- CLINICAL PHARMACOLOGY
STRENGTHS AND LIMITATIONS OF THE STUDY
This study used standardised and valid tools.
The study also used a large sample size, which decreased random error and increased the precision of the study.
Since the study was a cross-sectional study, it didn’t allow a temporal relationship between treatment-resistant depression and associated factors among patients with major depressive disorders.
Interviewer-administered data collection methods could magnify social desirability biases.
Treatment-resistant depression (TRD) occurs when a patient with major depressive disorder (MDD) does not respond to antidepressant medication. It is characterised by a unipolar depressive disorder that is unresponsive to an appropriate dosage of evidence-based treatment (40 mg of fluoxetine within the last 4 weeks) and duration (on treatment for at least 8 weeks; cognitive behavioural therapy or interpersonal psychotherapy should consist of 8–16 sessions over the same number of weeks). Additionally, within the first course of antidepressant medication, only half of patients with MDD achieve complete remission.1 Compared with patients with MDD who are not resistant to treatment, those with TRD have more comorbidities, poorer health-related quality of life, greater risk of suicide, greater direct and indirect healthcare resource utilisation and greater costs.2–4
MDDs account for 4.4% of the global disease burden. It is among the most prevalent non-communicable illnesses and one of the most urgent public health issues.1 5 Globally, around 250 million people are affected by MDD.6 An estimated lifetime prevalence of MDD among the Western community ranges from 12 to 16%.7 MDDs constitute 7.2% of the disease burden in the European Union and are the primary cause of illness and disability globally.8 In Africa, depression affects around 29 million people, and the prevalence of MDD is common among patients with chronic diseases.9 In Nigeria, the prevalence of MDD among patients with chronic disease was 32.5%.10
Another study in South Africa revealed that the prevalence of MDD was 34.8% among patients with chronic illness and 63.1% in traumatic patients.11 In sub-Saharan African countries, there is a dearth of evidence on the pooled prevalence of MDD,12 Single studies from Uganda,13 Sudan14 and Tanzania15 revealed the prevalence of MDD to be 29.3%, 31% and 42.4% among youths, traumatic adults and patients with chronic illness, respectively. The percentage of Ethiopians affected by depression is approximately 5%, although it varies from 15% to 30% in Sub-Saharan Africa (4). The prevalence of MDD among adult patients with neurolathyrism in Dawunt District, Ethiopia, was 38.7%.16 Another study was conducted in Ebinat town, Ethiopia, as that study had reported the prevalence of depression among Ethiopian adults to be 17.5%.17
MDD is a predominantly recurrent disorder, as 50–80% of patients who have received psychiatric care for an episode of major depression have at least one further episode and a median of four episodes in a lifetime.18 About 60% of individuals who have recovered from a depressive episode will have a recurrence within 5 years.19 Currently, over 20 antidepressant medications are available to treat depression.20 However, it is estimated that between 60% and 70% of people with major depressive illness do not completely recover from their symptoms, and over 11% of patients with depressive disorder experience a major depressive episode at some point during their treatment. Moreover, about 30% of patients with MDD do not achieve remission even after adequate trials of two antidepressant treatments.21 Depression that is resistant to treatment affects at least one-third of MDD patients.22
Inconsistent findings for each category of depression have been recorded, making it difficult to predict treatment outcomes in some cohorts.23 Predicting the duration of treatment for MDD is crucial nonetheless, as people with TRD experience increased rates of all-cause death.23 In addition to a marked decrease in their social and occupational functioning, the majority of patients do not experience remission after their initial round of therapy.24 Persistent symptoms in TRD often translate into exponential increases in work loss and medical costs compared with more responsive forms of illness.25
The study conducted in Thailand revealed a significant association between predicted TRD and variables like younger age of MDD patients, a younger age of onset of MDD, lower body mass index (BMI), a history of suicide attempts and self-harm and frequent smoking behaviours.26 Another study in Israel showed a significant association between predicted TRD and variables like increasing age and personality disorder.27 A study in Latin America identified that variables such as lower educational level, smoking, self-reported anxiety, chronic fatigue and back problems were significantly associated with MDD. Patients with MDD were more likely to be middle-aged, be women, be smokers, have a lower socioeconomic level and have a diagnosis of asthma or arthritis/rheumatism.28 Finding the factors that increase the likelihood of experiencing recurrent, persistent depressive episodes and a lack of response to depression treatment is crucial due to the significant health effects and financial burden that come with a poor longitudinal course of depressive illness. There is a dearth of data on TRD and associated factors in our setting. So, this study aimed to fill the information gap on TRD and associated factors among patients with MDDs on follow-up at St. Amanuel Mental Specialised Hospital, Addis Ababa, Ethiopia.
Materials and methods
Study setting, design and period
An institution-based cross-sectional study combined with a retrospective review of medical records was conducted at Amanuel Mental Specialised Hospital. It was established in 1930 by Italians to serve the nation of Ethiopia. The reason St. Amanuel Mental Specialised Hospital was chosen as the study’s location was because it is the only mental specialty hospital in Ethiopia that provides care for all patients with mental illnesses. Currently, it is one of the five federal hospitals in Addis Ababa. It is located in the western part of Addis Ababa in Addis Ketema sub-city, Kebele 08. The hospital has 259 beds, including 26 emergency beds and 15 outpatient departments. A 2020 mid-year service review reported that 1052 patients who received services had MDD. On average, a total of 902 depressive patients get follow-up services each month from the outpatient department of the hospital. The study was conducted from 17 February to 26 March 2021.
Patient and public involvement
Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this study.
Source of population and study population
All patients with MDDs on follow-up at Saint Amanuel Mental Specialised Hospital, Addis Ababa, Ethiopia, were the source population, and all adult patients aged >18 on follow-up at Saint Amanuel Mental Specialised Hospital, Addis Ababa, Ethiopia, were the study population.
Inclusion and exclusion criteria
The study included all adult patients with a MDD diagnosis who were over the age of 18. Patients who were deemed critically ill—that is, those who were severely depressed, catatonic or experiencing acute depression with psychosis or who had problems with their hearing, communication or cognitive function—were excluded. Therefore, the severity of the physical or mental condition may make it impossible to gather sufficient or thorough data.
Sample size determination
The sample size was calculated using the single population proportion formula with a 95% CI and a 5% margin of error and taking the prevalence of 50%:
The sample size of the second objective was smaller than the first objective. Therefore, the final sample size by taking 10% non-response was 422.
Sampling procedure/technique
A systematic sampling method was used to find study participants. By dividing the total number of study participants (n=902) who received an MDD diagnosis during follow-up during the data collection month by the total sample size, the sampling interval was calculated. Thus, two was the chosen skip interval. Using a lottery method, the first patient was chosen from the registered outpatient book.
Data collection method and instruments
Based on their experience with data collection and supervision, three health professional data collectors and one supervisor were assigned. The data were collected using a structured questionnaire adapted from previous studies. The questionnaire on substance-related factors was adapted from the WHO’s Alcohol, Smoking and Substance Involvement Screening Test (ASSIST). Each question on the ASSIST has a set of responses to choose from, and each response from questions 2 to 7 has a numerical score. The interviewer circles the numerical score that corresponds to the client’s response to each question. At the end of the interview, the scores from questions 2 to 7 are added together across each substance (tobacco, alcohol, cannabis, cocaine, amphetamine-type stimulants, inhalants, sedatives/sleeping pills, hallucinogens, opioids and ‘other’ drugs) to produce an ASSIST risk score for each substance. The ASSIST risk score for tobacco (range 0–31), alcohol (range 0–39), cannabis (range 0–39), cocaine (range 0–39), amphetamine-type stimulants (range 0–39), inhalants (range 0–39), sedatives or sleeping pills (range 0–39), hallucinogens (range 0–39), opioids (range 0–39) and ‘other’ drugs (range 0–39).29 Its Cronbach’s’ alpha is 80%, sensitivity is 0.80% and specificity is 71%. Social support was assessed using the Oslo three-item social support scale. The total scores of each item are summed up and range from 3 to 14 and have three broad categories: ‘poor social support’ (3–8), ‘moderate support’ (9–11) and ‘strong support’ (12–14). The internal consistency of the Oslo social support scale in this study was 0.83. It has been used in Ethiopia in different clinical settings.30
TRD was assessed using the Hospital Anxiety and Depression Scale (HADS). The HADS aims to measure symptoms of anxiety and depression and consists of 14 items: seven for the anxiety subscale (HADS Anxiety) and seven for the depression subscale (HADS-D). HADS Anxiety focuses mainly on symptoms of generalised anxiety disorder, and HADS-D focuses on anhedonia, the main symptom of depression.31 Each item is scored on a response scale with four alternatives ranging between 0 and 3. After adjusting for six items that are reverse-scored, all responses are summed to obtain the two subscales. Recommended cut-off scores according to Zigmond and Snaith32 are 8–10 for doubtful cases and ≥11 for definite cases. An optimal balance between sensitivity and specificity was found using a cut-off score of 8 or above for both HADS anxiety and HADS-D.33 Possible scores range from 0 to 21 for depression, with cut-off points greater than or equal to 8. In Ethiopia, it was validated, and its internal consistency was 0.76 for depression subscales.34 The data were collected by the face-to-face interview method and review of medical records.
Study variables
Dependent variable
TRD (Yes/No).
Independent variables
Sociodemographic factors: age, sex, residence, marital status, ethnicity, religion, educational level, jobs; clinical factors: age at the beginning of the illness, duration without treatment, duration of taking the medication, comorbid psychosis, comorbid anxiety, comorbid medical illness, family history of mental illness, history of admission, social support, and type of medication taking.
Operational definitions
Social support
It was assessed using the Oslo three-item social support scale. The total scores of each item are summed up and range from 3 to 14 and divided into three broad categories: ‘poor social support’ (3–8), ‘moderate social support’ (9–11) and ‘strong social support; (12–14).30
TRD
It was assessed using the HADS, which has seven-item questionnaire subscales for depression symptoms. Possible scores range from 0 to 21 for depression, with cut-off points of greater than or equal to 8.34
Data quality control
Data quality was assured before, during and after the data collection process. Before data collection, a standardised questionnaire was prepared. The questionnaire was pre-tested outside the study area on about 5% of the sample size. The questionnaire that was developed in English was translated to the local language (Amharic and Afan Oromo) and then translated back to English to see the consistency. Data collectors and supervisors received 1 day of instruction on how to use the questionnaire, sampling methodologies, ethical principles, data management and participant identification. During the data collection process, there was close day-to-day supervision, and the questionnaire was checked to ensure completeness and validity by supervisors and the principal investigator. After data collection, the collected data were coded and entered in EpiData software version 3.1. Double data entry into Epi Data software by two independent data clerks was done. Then the two data sets were validated in the software for consistency and mismatches.
Data processing and analysis
The data were checked, entered into Epi-data software version 3.1 and imported to SPSS version 20 for analysis. Sociodemographic characteristics and other factors were analysed by descriptive statistics. A bivariable logistic regression analysis was run to determine the association between the independent variables and the outcome variable. The P value and an OR with a 95% CI were computed. Then the variables with a p value ≤0.2 were taken into a multivariable model to control for all possible confounders. For model fitness, the Hosmer–Lemeshow model of fitness was checked, and the test fits the model. The strength of associations was evaluated using the adjusted OR with a 95% CI, and a P value of less than 0.05 was considered statistically significant.
Ethical approval and consent to participate
Ethical clearance was obtained from the Ethical Review Committee (ERC) of Saint Amanuel Mental Specialised Hospital with reference number (SOM/1540/2021). A formal letter of permission was obtained from Saint Amanuel Mental Specialised Hospital and submitted to the medical directorate. We received informed written consent from study participants. Confidentiality was maintained by omitting personal identification.
Results
Socio-demographic characteristics of the respondents
A total of 412 adult outpatients on follow-up were involved in this study, with a response rate of 97.63%. The mean age of the participants was 39.71 (± 14.94). More than half (53.6%) were male, and nearly three-fifths (53.2%) wererural residency. More than half (53.64%) of the study participants were orthodox Christians, and around two-fifths (40.30%) were married. More than two-fifths (43.2%) were Oromo by their ethnicity, and around one-third (30.6%) of the participants were farmers. Close to one-fourth (22.6%) of the participants were college or university graduates (table 1).
Distributions of sociodemographic characteristics of participants among patients with major depressive disorder on follow-up at Saint Amanuel Mental Specialised Hospital, Addis Ababa, Ethiopia, 2021 (n=412)
Distribution of clinical factors among respondents
The majority of respondents (83%) experienced the onset of their illness at an age greater than 11 years. Around two-thirds (66.5%) were living without treatment for less than or equal to 1 year. More than half (60.2%) of the study participants were taking the antidepressant medications for about 1 year. More than two-thirds (70.1%) of the respondents had comorbid psychosis (table 2).
Distribution of clinical factors of respondents among patients with major depressive disorders on follow-up at Saint Amanuel Mental Specialised Hospital, Addis Ababa, Ethiopia, 2021 (n=412)
Psychosocial and substance use factors among respondents
The majority of the respondents (70.1%) had poor social support. More than two-thirds of the respondents (69.4%) had a history of lifetime substance use, and the majority of them (68.20%) have been smoking cigarettes in their lifetime. Almost half of respondents (51.21%) had a history of current substance use (table 3).
Psychosocial and substance use characteristics of respondents among patients with major depressive disorder on follow-up at Saint Amanuel mental Specialised Hospital, Addis Ababa Ethiopia, 2021 (n=412)
Prevalence of TRD among respondents
In the current study, the prevalence of TRD was 41.5% (95% CI: 37.2 to 46.1).
Factors associated with TRD
To determine the association of independent variables with TRD, bivariable and multivariable logistic regression analyses were carried out. In the bivariate analysis, factors including female sex, comorbid psychosis, comorbid anxiety, a comorbid medical illness, family history of mental illness, history of admission and current cigarette smoking were associated with TRD at a P value of less than 0.2. These factors were entered into the multivariate logistic regression model to control confounding effects. Factors such as female sex, comorbid psychosis, comorbid medical illness and family history of mental illness were significantly associated with TRD at a p value of less than 0.05.
The result of multivariate analysis showed that the odds of developing TRD were 2.43 [AOR=2.43; 95% CI: 1.57 to 3.75] times higher among females compared with males. The odds of developing TRD were 1.89 [AOR=1.89; 95% CI: 1.19 to 2.99] times higher among patients with comorbid psychosis compared with those who did not have comorbid psychosis. The odds of developing TRD were 1.6 [AOR=1.67; 95% CI: 1.09 to 2.55] times higher among respondents with comorbid medical or physical illness compared with those who had no comorbid physical illness. The odds of developing TRD were 2.27 [AOR=2.27; 95% CI: 1.38 to 3.74] times higher among respondents who had a family history of mental illness compared with those who had no family history of mental illness (table 4).
Factors associated with treatment-resistant depression on bivariable and multivariable analysis among patients with major depressive disorder on follow-up at Saint Amanuel mental Specialised Hospital, Addis Ababa Ethiopia, 2021 (n=412)
Discussion
In this study, the prevalence of TRD and its possible association with various factors were assessed. The results revealed that a remarkable proportion of people on follow-up due to MDD on antidepressant treatment had TRD.
The prevalence of TRD among people with MDDs was found to be 41.5% [95% CI: 37.2, 46.1]. The findings of this study are in line with those of a study carried out among patients with MDDs in Brazil, which was 42.5%35 and a large multicenter study in Europe, 41%.4 However, this finding is lower than studies conducted in the UK, which was 46.9%36; 52.3% in China37; 50% in the USA, which was done among pregnant women; and 78% in Brazil.38 The possible reason for the variation of the above prevalence might be due to the variation in the study population contrast and the socio-cultural difference between Ethiopia and other countries. The variation may also be due to the measurement tools used in the study. In the current study, TRD was assessed using the HADS. But different types of measurement tools were used by other comparative studies. The variation in the prevalence of comorbidities among patients with MDDs in different study areas may also be the reason for the discrepancies.
On the other hand, the prevalence of the current study is higher than four studies done in the USA, which were 33.4%,39 29.4%,40 22%41 and 9.6%,42 and studies done in Italy, 11.5%43; Mexico, 21%44; Colombia, 32%44; Argentina, 33%44; Canada, 22%45; Taiwan, 21%46; Japan, 12%47; Israel, 24.4%48; and Poland, 25.2%.49 This difference might be due to the nature of public services provided in the current study setting. Amanuel Mental Specialised Hospital is a university-based research centre with a greater demand from higher complexity and more severe patients, which may predispose to increased prevalence of TRD in the current study setting. Another possible reason for the high prevalence of TRD might be due to low social support towards mental illness, stigmatisation of mental illness, low socioeconomic status and a high substance use habit in Ethiopia.
The variation may also be due to different criteria used for identifying TRD in different studies. The previous studies used distinct criteria for identifying TRD, such as the number of lines of treatment indicated in prescriptions, reduction in BDI score, or adherence to antidepressant medication.48 49
Regarding factors associated with TRD, the odds of developing TRD were 2.43 [AOR=2.43, 95% CI: 1.57 to 3.75] times higher among females compared with males. Which was supported by studies done in Boston,40 eastern Ethiopia,50 Dawunt District, Ethiopia,51 and in Addis Ababa, Ethiopia.52 The reason might be a common psychological disturbance in females and gender violence, which might lead to TRD, which is more common in females than males. The other possible reason might be due to cultural influence, in which females may not discuss their problems with others as males, and it might be due to females having greater vulnerability to other psychosocial stress. Biological factors might play a role for the differences.
The study also showed that the odds of developing TRD were 1.89 [AOR=1.89; 95% CI: 1.19 to 2.99] times higher among patients with comorbid psychosis compared with those who did not have comorbid psychosis. This is consistent with findings from the USA39 and Australia.53 But it is different from the study done in Brazil. This difference may be due to methodological differences, an assessment tool used, the presence of cultural concepts, values and beliefs and social withdrawal. The other reason for the variation might be due to the presence of psychiatric comorbidity associated with reduced symptomatic improvement,54 55 higher readmission rates,56 57 poorer functioning,54 lower quality of life,54 and increased suicidality.58
The current study revealed that the odds of developing TRD were 1.6 [AOR=1.67; 95% CI: 1.09 to 2.55] times higher among respondents with comorbid medical or physical illness compared with those who had no comorbid physical illness. This is supported by a study conducted in the USA,39 the UK36 and Australia.53 The reason might be due to the same effect of medical illness among human beings. Patients with comorbid medical or physical illness may worry about their disease condition, which affects their quality of life, limiting them from doing daily activity actively and by creating stress in their life. As a result, this may predispose them to MDD.
The findings of the study also showed that the odds of developing TRD were 2.27 [AOR=2.27; 95% CI: 1.38 to 3.74] times higher among respondents who had a family history of mental illness compared with those who had no family history of mental illness. This study is different from a study done in the USA,39 Brazil35 and the UK.36 This finding might be that the shared gene characteristics among those major depressive patients with family history of mental illness may predispose them to have higher TRD.
In this study, there was a high prevalence of TRD among adult outpatients on follow-up at Amanuel Mental Specialised Hospital. In order to prevent and treat TRD, healthcare providers should provide mindfulness-based therapies and meditation techniques that can help patients manage stress, reduce rumination and improve emotional regulation, all of which can be beneficial in TRD. In order to reduce TRD, it also needs supportive care from family and friends. Healthcare providers should regularly monitor the patient’s progress and adjust the treatment plan as needed.
This study used standardised and valid tools. The study also used a large sample size, which decreased the random error and increased the precision of the study. Since the study was a cross-sectional study, it did not allow a temporal relationship between TRD and associated factors among patients with MDDs. Interviewer-administered data collection methods could magnify social desirability biases.
Conclusion and recommendation
The prevalence of TRD among patients with MDD on follow-up treatment was high. TRD was significantly associated with female gender, comorbid psychosis, comorbid medical illness and family history of mental illness. Better TRD management should be warranted, which includes a multi-pronged approach of improved diagnosis, comprehensive evaluation of environmental and developmental factors driving TRD and development of specific targeted treatment strategies encompassing pharmacological, psychological, device-based and/or nutritional therapies. In order to determine the temporal relationship between TRD and its associated factors, longitudinal studies should be conducted in the future.
Strengths and limitations of the study
The tools employed in the current study were valid and standardised. Additionally, the study employed a large sample size, which improves study precision and reduces the random error. The study’s cross-sectional design precluded the possibility of a temporal correlation between TRD and related variables in MDD patients. Social desirability biases may be exacerbated by interviewer-administered data gathering techniques.
Data availability statement
Data are available upon reasonable request. This research included pertinent data, and the corresponding author/s are available to offer further information upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
The conduct of the study adhered to the Helsinki Declaration of Medical Research Ethics. Ethical clearance was obtained from the Ethical Review Committee (ERC) of Saint Amanuel Mental Specialised Hospital with reference number (SOM/1540/2021) and granted ethical clearance and permission. Participants in the study were informed of the importance and goal of the research, as well as the need to maintain the confidentiality of their information. Only after obtaining full, informed, voluntary, written and formally signed consent did data collecting start. The participants received guarantees that the information they provided would only be used for study and that their names would not be disclosed.
Acknowledgments
Our deepest gratitude goes ahead to the data collectors and study participants for their time and effort.
References
Footnotes
Contributors JD is the work’s guarantor, while MS is the primary investigator. Co-authors all made significant contributions to this work, including conception (MS, JD, SDD, EK, TF, EM, KA DBD and HA), study design (MS, EK and TF), execution (MS, JD, SDD and HA), methodology (MS, JD, HA and DBD), data analysis acquisition (MS, JD, SDD, EK, TF, EM, KA DBD and HA), writing the original draft (MS, JD, HA and SDD), review and editing (MS, EK,EM, HA and DBD). Every author committed to take responsibility for every portion of the work and took part in the article’s drafting, revision and critical evaluation. The final paper has been read and approved by all authors.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.