Epidemiology and Burden of Treatment-Resistant Depression in Colombia: Analysis of the TRAL Study

Abstract

Objectives

To estimate the prevalence of treatment-resistant depression (TRD) in Colombia, and characterize hospitalized major depressive disorder (MDD) patients for disability, health-related quality of life (HRQoL) and work-impairment.

Materials and methods

Adult MDD patients (n = 162) from 4 centers in Colombia, with a clinical diagnosis of MDD based on DSM-5 and MINI, were consecutively included. Patients with psychosis, schizophrenia, bipolar disorder, schizoaffective disorder, dementia, with severe chemical dependence or currently participating in another clinical trial were excluded. Patient reported outcomes and clinical assessment scales were used as outcomes with WPAI:D (work impairment), EQ-5D (QoL) and Sheehan Disability Scale, as well as healthcare resource utilization.

Results

Prevalence of TRD in Colombia among patients treated at psychiatric reference sites was 32.1% [24.9%; 39.3%], higher in private institutions. Most patients were females and single. 55% reported having moderate to severe problems with usual activities at the beginning of the study. Mean percent of working time missed due to depression was around 53%, and work impairment was roughly 67%. A mean of 2.85 days in the last seven were marked by lost school/work, with higher losses in TRD patients compared to non-TRD. Overall results for HRQoL were lower in TRD patients, but disability and work impairment were higher in non-TRD.

Conclusions

There is a high prevalence of mental disorders in Colombia, including TRD, impacting significantly HRQoL. The burden associated with them may be reduced by a most comprehensive and customized usage of therapies, including the innovations, and implementation of national mental health programs.

Resumen

Objetivos

Estimar la prevalencia de depresión resistente al tratamiento (DRT) en Colombia, y caracterizar a los pacientes con trastorno depresivo mayor (TDM) que requieren hospitalización con respecto a discapacidad, calidad de vida (HRQoL) e incapacidad laboral.

Materiales y métodos

Se incluyó una muestra consecutiva de adultos (n = 162) de 4 centros en Colombia con diagnóstico de TDM basado en escalas DSM-5 y MINI. Se excluyeron casos de psicosis, esquizofrenia, trastorno bipolar, trastorno esquizoafectivo, demencia, drogodependencia grave o participantes de otro ensayo. Se utilizaron escalas de evaluación clínica y resultados informados por el paciente como WPAI:D (incapacidad laboral), EQ-5D (QoL) y escala de discapacidad de Sheehan, así como utilización de recursos médicos.

Resultados

La prevalencia de DRT en Colombia fue del 32,1% (24,9%; 39,3%) más alta en las instituciones psiquiátricas privadas. En la muestra predominaron mujeres y solteras. El 55% de la muestra informó problemas moderados a graves con las actividades habituales al inicio. El tiempo laboral perdido fue del 53%, y la incapacidad laboral por depresión el 67%. Se perdieron 2,85 días de actividad escolar/laboral de los últimos 7 días, más en pacientes con DRT que en pacientes sin DRT. Los resultados de HRQoL fueron más bajos en pacientes con DRT, pero la discapacidad y la incapacidad laboral fueron más altas en pacientes sin DRT.

Conclusiones

Existe una alta prevalencia de trastornos mentales en Colombia, incluyendo TRD, impactando significativamente en la HRQoL. La carga asociada con ellos puede reducirse mediante el uso más integral y personalizado de terapias, incluidas las innovaciones y la implementación de programas nacionales de salud mental.

Introduction

MDD is a complex clinical presentation of depression with a very significant global impact on disability and health-related quality of life (HRQoL),¹ as well as in independent living and execution of instrumental activities of daily living (IADLs).² Treatment-resistant depression (TRD) is frequently (but not consensually) defined as a failure to respond to two or more antidepressants at therapeutic doses, over an appropriate period of time, within the current depressive episode.³

Worldwide prevalence of MDD is estimated at 6%,⁴ and about 30% of the patients with MDD fail to respond to standard therapies (with both augmentation and combination strategies), thus developing TRD. Previous data on specific Colombian regions showed a prevalence of MDD around 16.5%,⁵ higher than the lifetime prevalence of MDD for the adult Colombian population.⁶ Regional results obtained in emergency department patients⁷ are, however, much higher, which is in line with the patient profile in these services. On the other hand, epidemiological data on TRD in Latin America (LatAm) is very limited, but the available international publication indicates ranges from 20 to 33%.^8,9

TRD is therefore a burdensome disease with a broad impact in patients, caregivers and national health systems, which can be measured in economic resources, but also in disability, health related quality of life (HRQoL) and work impairment.^8,10–13 Moreover, the burden takes a significant toll in psychosocial and humanistic dimensions, significantly hindering instrumental activities of daily living.¹⁴ The lack of therapies with adequate clinical outcomes is still an issue, even though some recent developments suggest improved results in the near future.¹⁵

Available research also suggests that the burden of TRD is more significant compared to non-TRD patients,^13,16 which is also stated in the recent Treatment-Resistant Depression in America Latina (TRAL) global results.¹⁷ Previous studies in Colombia found that the burden of MDD is significant, regardless of TRD. Moreover, this condition and other forms of the depressive spectrum, are responsible for more hospitalizations and the consequent increased costs.^18,19 Depressive disorders are a worldwide concern,¹ but Colombia has a specific post-conflict context that may promote the development of such conditions.^20,21 Previous research suggests that suicidal tendencies may also represent an issue in Colombia, interlinked with a high prevalence of affective disorders, according to the 2015 national mental health survey.²² A study with specific assessment tools for men suggested that prevalence of depression and suicide risk in Colombian male is higher than previously reported.²³

This paper reports on the Colombian subsample of the TRAL study,¹⁷ a multinational study aiming for an epidemiological characterization of TRD among MDD patients routinely followed at a diversity of public and private healthcare settings in four Latin American countries. Beyond epidemiology, TRD patients were characterized in critical areas, such as disability, HRQoL and work-impairment, which are usually reported as being impacted by this condition. Descriptive comparisons between TRD and non-TRD patients are also reported.

Materials and methods

Study design and population

This was a multicenter, multinational, observational study conducted between October 2017 and December 2018 at psychiatric reference sites (clinic, ambulatory, hospital, day-hospital) from the following Latin American countries: Argentina, Brazil, Colombia and Mexico. The TRAL study had two components: one cross-sectional, for full characterization of psychiatric, disability, quality of life, work impairment dimensions, along with the collection of medical and treatment history, and healthcare resource utilization; and a second component, involving a 1-year prospective follow up of the TRD sample for assessing clinical and safety outcomes and changes in the dimensions abovementioned. The analysis reported hereby refers to the cross-sectional part of the Colombian subsample (corresponding to 4 centers, 1 General hospital, 1Private psychiatric institution, 1 Public hospital and 1 Private psychiatric clinical site), focusing on socio-demographic, clinical/psychiatric variables as well as health care resource utilization, between February and November of 2018.

Patients meeting eligibility criteria were consecutively enrolled in the study according to their routine scheduled appointment. Male or female adults diagnosed with MDD according to the Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-5 criteria and confirmed by the MINI International Neuropsychiatric Interview (MINI) – 7.02 version²⁴ were eligible. At the day of appointment, patients had to be experiencing a new or persistent episode of depression, irrespective of the treatment status. The diagnosis of TRD was also complemented with a thorough clinical assessment through an interview at the time of the routine medical appointment. Diagnosis of psychosis, schizophrenia, bipolar disorder, schizoaffective disorder, or dementia, and/or serious substance dependence (according to physician's judgment) were exclusion criteria. Patients uncapable of completing protocol assessments or participating in experimental studies were excluded.

Data and assessments

During study appointment the MINI was used to conduct psychiatric interview and ascertain MDD diagnosis, suicide behavior disorder, occurrence of suicidality and post-traumatic stress over the past month. Suicidal Behavior Disorder is a proposed DSM-5 diagnosis which would be given to individuals who have made a suicide attempt within the past two years. All patients were evaluated by a physician regarding their depression severity through the Montgomery-Åsberg Depression Rating Scale (MADRS),²⁵ a 10-item scale used in trials and clinical routine practice, which shows a good discrimination between responders and non-responders to antidepressants. TRD diagnosis was based on the following criteria: patients had to be followed up adequately and treated with ≥2 antidepressants in the current episode, with absence of complete response to treatment based on MADRS.³

Sociodemographic (age, sex, marital status and years of formal education) and clinical features (age at diagnosis, disease duration and comorbidities) were collected by the physician, as well as current/past psychiatric treatments and healthcare resource utilization (ambulatory care and total number of days, psychiatric, psychologist, primary care and emergency department visits), referring to the previous year (365 days). Specifically concerning the healthcare resource utilization, data was obtained from available medical charts and costs were derived from the responses of the patients in the routine medical appointment. Since this analysis is merely descriptive in nature, no pharmacoeconomic analysis was derived and cost analysis is merely indicative. Also, patients self-reported on their quality of life, with EuroQol-5 Dimension [EQ-5D-5L]),²⁶ work impairment (Work Productivity and Activity Impairment Questionnaire – WPAI),²⁷ and disability and functional impairment (Sheehan Disability Scale – SDS),²⁸ which includes three domains – work/school, social and family life, assessed individually through 10-point visual analog scale.

Written informed consent was obtained from all participants. The study was approved by local Independent Ethics Committee/Institutional Review Board.

Statistical analysis

The Colombian subset of the TRAL sample included 162 MDD patients. TRAL sample size was calculated to be representative of the region (LatAm) and not for Colombia. However, given the sample size and the relevance of the sites enrolled in the study, the authors believe this manuscript provides a robust epidemiological and burden of treatment characterization of MDD patients with a diagnosis of TRD being followed at reference institutions in Colombia.

Quantitative variables were summarized as mean, median, standard deviation minimum and maximum, and qualitative variables were summarized as absolute frequency and percentage, overall and by TRD and non-TRD subgroups.

The prevalence of TRD among MDD patients was summarized as absolute frequency and percentage with 95% Confidence Interval (95%CI). Chi-square or Fischer's exact test was used to compare categorical variables between TRD and non-TRD subgroups while t-test for independent samples or the Mann–Whitney non-parametric test, or ANOVA/Kruskal–Wallis tests were used for comparing continuous variables. There was no imputation of missing data. Statistical significance was set at 5%. Statistical analysis was performed using SAS® (version 9.4, SAS Institute Inc, Cary).

Results

General and Colombia epidemiological characterization

Overall, 1475 MDD patients were included across the four Latin American countries. Colombia represents roughly 11% (162) of the total sample, mostly patients under current treatment for MDD (86.4%). Prevalence of TRD in Colombia was 32.1% – [24.9%; 39.3%] in the overall sample, with a higher proportion in treated MDD patients −34.3% [26.4%; 42.1%] (supplemental Fig. 1). The proportion of treated patients was higher in the TRD sub-sample (92.3%).

Concerning the setting in which TRD was being treated, the highest proportion of TRD was found in private psychiatric institution [n = 25/37, 67.6% (95% C.I. −52.5%; 82.7%)], followed by general hospital [n = 4/18, 22.2% (C.I. _95% −3.0%; 41.4%)], public hospital [n = 8/34, 23.5% (C.I. _95% −9.3%; 37.8%)] and private psychiatric clinical site [n = 11/51, 21.6% (95% C.I. −10.3%; 32.9%).

Socio-demographic characterization of MDD and TRD patients

Mean age of overall MDD patients (41.0 ± 13.8) was similar within subgroups of TRD patients (41.9 ± 12.3) and non-TRD patients (40.6 ± 14.5) – Table 1. Most patients were females (81.5%), irrespective of diagnosis, and single (50.9%), which had a higher proportion in the TRD sample. Over 70% of all MDD patients had ≥10 years of formal education, with a non-significant trend suggesting that the years of formal education is lower in the TRD patients.

Table 1.

Socio-demographic characteristics of MDD patients overall and by TRD (TRD vs non-TRD patients) – Colombia.

	MDD(n = 162)	Non-TRD(n = 110)	TRD(n = 52)	p value
Age (years), mean ± SD	41.0 ± 13.8	40.6 ± 14.5	41.9 ± 12.3	0.539*
Female	132 (81.5%)	90 (81.8%)	42 (80.8%)	0.872^†
Marital status
Single	82 (50.9%)	54 (49.1%)	28 (54.9%)
Married/consensual union	57 (35.4%)	38 (34.5%)	19 (37.3%)	0.553^¥
Divorced/Separated	20 (12.4%)	16 (14.5%)	4 (7.8%)
Widower	2 (1.2%)	2 (1.8%)	0
Missing	1	0	1
Years of formal education
0	1 (0.6%)	1 (0.9%)	0
1–4 years	12 (7.4%)	7 (6.4%)	5 (9.6%)
5–9 years	35 (21.6%)	20 (18.2%)	15 (28.8%)	0.441^¥
10–12 years	36 (22.2%)	26 (23.6%)	10 (19.2%)
≥13 years	78 (48.1%)	56 (50.9%)	22 (42.3%)
Missing	0	0	0

Unless otherwise noted, data are expressed as number and percentage. p values express TRD vs non-TRD comparisons. SD: standard deviation; MDD – major depressive disorder; TRD – treatment-resistant depression; non-TRD – MDD patient without TRD.

*

Mann–Whitney.

†

Chi square.

¥

Fischer exact test.

No statistically significant differences were found between TRD and non-TRD patients regarding age, gender, marital status and years of formal education (p > 0.05).

Clinical characterization of depressive symptoms

Clinical characterization of depressive symptoms was assessed in Colombia. The sample of MDD patients had a mean MADRS score of 29.3 ± 7.6 and 23.5% were rated as severely depressed (MADRS ≥ 35) – Table 2. MADRS score was lower for non-TRD patients (28.6 ± 8.2) compared to TRD patients (30.8 ± 5.8). All TRD subjects (100.0%) had moderate to severe depression.

Table 2.

Depression severity and psychiatric comorbidities - Colombia.

	MDD(n = 162)	Non-TRD(n = 110)	TRD(n = 52)
Montgomery-Asberg Depression Scale (MADRS)
Mean score^a ± SD	29.3 ± 7.6	28.6 ± 8.2	30.8 ± 5.8
Cut-off scores
Symptom absent (0–6)	1 (0.6%)	1 (0.9%)	0 (0.0%)
Mild depression (7–19)	14 (8.6%)	14 (12.7%)	0 (0.0%)
Moderate (20–34)	109 (67.3%)	68 (61.8%)	41 (78.8%)
Severe depression (35–60)	38 (23.5%)	27 (24.5%)	11 (21.2%)

a

Range: 0–60. Higher values indicate a higher level of depression. MDD – Major Depressive Disorder; TRD – Treatment Resistant Depression.

Healthcare resource utilization in Colombia

MDD patients attended to ambulatory care in 83.3% of the cases (non-TRD: 80.0%; TRD: 90.4%). Overall, 34.8% of the patients had at least 90 days of ambulatory care, but the number is noticeably higher in TRD patients (non-TRD: 26.2%; TRD: 51.1%). Average number of emergency visits (previous year) was 0.71 ± 1.0, rising considerably in the TRD sample (0.87 ± 2.3). The number of appointments range from 0.13 ± 0.6 non-pharmaceutical treatments appointments to 7.03 ± 8.9 in psychiatry appointments.

Mean psychiatric appointments was higher among TRD patients versus non-TRD patients (8.94 vs. 6.01, respectively), as well as the mean number of psychologist appointments (5.83 vs. 4.05), which averages 4.67 ± 9.0 in the overall mean sample. In both psychiatric and psychologist appointments, average costs of consultation were higher in the TRD sample. However, costs were higher in the non-TRD sample for appointments with other specialists and other health professionals’ appointments (Table 2). Overall, the mean cost of psychiatrist appointments is 362.55 ± 479.2, while the mean cost drops considerably for psychologist appointments (224.08 ± 329.3) – Table 3.

Table 3.

Healthcare resource utilization in the previous year among MDD patients in Colombia broken down by TRD.

	MDD(n = 162)	Non-TRD(n = 110)	TRD(n = 52)
Ambulatory care, n (%)
No	27 (16.7%)	22 (20.0%)	5 (9.6%)
Yes	135 (83.3%)	88 (80.0%)	47 (90.4%)
Number of days (ambulatory care), n (%)
7 days	39 (28.9%)	35 (39.8%)	4 (8.5%)
30 days	33 (24.4%)	23 (26.1%)	10 (21.3%)
60 days	16 (11.9%)	7 (8.0%)	9 (19.1%)
90 days	41 (30.4%)	21 (23.9%)	20 (42.6%)
>90 days	6 (4.4%)	2 (2.3%)	4 (8.5%)
Total	135	88	47
Missing values	0	0	0
Number of visits to emergency department
N	132	86	46
Mean ± SD	0.71 ± 1.0	0.63 ± 1.0	0.87 ± 2.3
Number of psychiatric appointments
N	135	88	47
Mean ± SD	7.03 ± 8.9	6.01 ± 8.1	8.94 ± 10.1
Cost of consultations ($ USD)
N	82	62	20
Mean ± SD	362.55 ± 479.2	311.23 ± 415.7	521.65 ± 623.52
Number of Psychologist appointments
N	132	86	46
Mean ± SD	4.67 ± 9.0	4.05 ± 7.6	5.83 ± 11.2
Cost of consultations ($ USD)
N	65	50	15
Mean ± SD	224.08 ± 329.3	164.18 ± 233.4	423.73 ± 499.5
Number of appointments – other specialists
N	128	84	44
Mean ± SD	0.85 ± 3.4	1.11 ± 4.4	0.36 ± 0.8
Cost of consultations ($ USD)
N	20	15	5
Mean ± SD	165.35 ± 373.0	186.33 ± 426.8	102.40 ± 128.38
Number of Primary Care appointments
N	128	84	44
Mean ± SD	1.24 ± 2.5	1.06 ± 2.0	1.59 ± 3.27
Cost of consultations ($ USD)
N	34	28	6
Mean ± SD	41.21 ± 39.1	36.11 ± 35.8	65.00 ± 48.48
Number of other health professionals’ appointments
N	127	83	44
Mean ± SD	0.17 ± 0.6	0.17 ± 0.6	0.16 ± 0.5
Cost of consultations ($ USD)
N	10	6	4
Mean ± SD	65.80 ± 45.8	76.7 ± 50.8	49.50 ± 37.4
Number non-pharmaceutical treatments appointments
N	127	83	44
Mean ± SD	0.13 ± 0.6	0.12 ± 0.5	0.14 ± 0.8
Cost of consultations ($ USD)
N	7	5	2
Mean ± SD	168.57 ± 221.5	96.00 ± 30.3	350.00 ± 445.5

SD: standard deviation.

Quality of life (EQ-5D-5L), work disability (WPAI) and disability (SDS)

Mobility assessment in the EQ-5D-5L revealed that most of the sample had no problems walking (66.0%), which can also be stated for self-care, in which 61.1% of the sample had no issues washing/dressing themselves (61.1%) – Table 4. Almost 55% of the patients reported having moderate or severe problems doing their usual activities, while 16% of the patients reported severe or extreme pain. Concerning anxiety/depression, 30.3% of the sample reported feeling severely or extremely anxious/depressed, in a self-rated assessment. Median current health overall score was 55.0, with TRD patients displaying lower median current health score (50.0) compared to non-TRD patients (60.0). TRD sample consistently showed worst quality of life: higher proportion of severe problems in walking (3.8% vs. 1.8%), severe washing/dressing themselves (1.9% vs. 0.9%) and inability to perform usual activities. On the other hand, non-TRD patients reported higher proportion of extreme pain or discomfort (4.5% vs. 3.8%) and extreme anxiety or depression (15.5% vs. 7.7%).

Table 4.

Quality of Life (EQ-5D-5L) among participants with MDD, broken down by TRD – Colombia.

	MDD(n = 162)	Non-TRD(n = 110)	TRD(n = 52)
Quality of Life (EQ-5D-5L)
Mobility, n (%)
I have no problems walking	107 (66.0%)	71 (64.5%)	36 (69.2%)
I have slight problems walking	39 (24.1%)	27 (24.5%)	12 (23.1%)
I have moderate problems walking	12 (7.4%)	10 (9.1%)	2 (3.8%)
I have severe problems walking	4 (2.5%)	2 (1.8%)	2 (3.8%)
I am unable to walk	0	0	0
Self-care, n (%)
I have no problems washing or dressing myself	99 (61.1%)	69 (62.7%)	30 (57.7%)
I have slight problems washing or dressing myself	47 (29.0%)	29 (26.4%)	18 (34.6%)
I have moderate problems washing or dressing myself	14 (8.6%)	11 (10.0%)	3 (5.8%)
I have severe problems washing or dressing myself	2 (1.2%)	1 (0.9%)	1 (1.9%)
I am unable to wash or dress myself	0	0	0
Usual activities, n (%)
I have no problems doing my usual activities	28 (17.3%)	19 (17.3%)	9 (17.3%)
I have slight problems doing my usual activities	45 (27.8%)	31 (28.2%)	14 (26.9%)
I have moderate problems doing my usual activities	59 (36.4%)	38 (34.5%)	21 (40.4%)
I have severe problems doing my usual activities	26 (16.0%)	20 (18.2%)	6 (11.5%)
I am unable to do my usual activities	4 (2.5%)	2 (1.8%)	2 (3.8%)
Pain/discomfort, n (%)
I have no pain or discomfort	43 (26.5%)	32 (29.1%)	11 (21.2%)
I have slight pain or discomfort	55 (34.0%)	33 (30.0%)	22 (42.3%)
I have moderate pain or discomfort	38 (23.5%)	26 (23.6%)	12 (23.1%)
I have severe pain or discomfort	19 (11.7%)	14 (12.7%)	5 (9.6%)
I have extreme pain or discomfort	7 (4.3%)	5 (4.5%)	2 (3.8%)
Anxiety/depression, n (%)
I am not anxious or depressed	13 (8.0%)	9 (8.2%)	4 (7.7%)
I am slightly anxious or depressed	26 (16.0%)	23 (20.9%)	3 (5.8%)
I am moderately anxious or depressed	74 (45.7%)	44 (40.0%)	30 (57.7%)
I am severely anxious or depressed	28 (17.3%)	17 (15.5%)	11 (21.2%)
I am extremely anxious or depressed	21 (13.0%)	17 (15.5%)	4 (7.7%)
Currenthealth^a
Median (min–max)	55.0 (0.0–100)	60.0 (10–100)	50.0 (0.0–90)

Unless otherwise noted, data are expressed as number and percentage.

MDD – Major Depressive Disorder; TRD – treatment-resistant depression; non-TRD – MDD patient without TRD; SD – standard deviation;

a

The health at the day of study appointment was assessed through a visual analogic scale (range 0 = worst health to 100 = best health).

Results for WPAI, concerning work impairment, mean percent of work time missed due to depression was around 53%, with non-TRD patients showing higher values (56.26 vs. 46.13). The same result was found for mean percent of impairment while working due to depression (62.3% vs 51.0%), with an overall MDD value around 59%. Overall work impairment due to depression is around 67%, with activity impairment due to depression showing similar proportions. For both variables, non-TRD patients presented higher impairment proportions (work impairment – 71.18; activity impairment – 68.36) versus TRD (work impairment – 51.0; activity impairment – 55.71) – Table 5.

Table 5.

Work Productivity and Activity Incapacity (WPAI:D), and Disability (Sheehan Disability Scale – SDS) among participants with MDD, broken down by TRD – Colombia.

	MDD(n = 162)	Non-TRD(n = 110)	TRD(n = 52)
Work Productivity and Activity Incapacity (WPAI:D)
Percent work time missed due to depression (%)
N	62	46	16
Mean ± SD	53.64 ± 42.3	56.26 ± 41.4	46.13 ± 45.4
Median	53.24	55.84	27.64
Missing values	2	1	1
Percent impairment while working due to depression (%)
N	40	30	10
Mean ± SD	59.50 ± 24.5	62.3 ± 23.7	51.0 ± 26.0
Median	65.00	70.00	60.00
Missing values	24	17	7
Percent overall work impairment due to depression (%)
N	40	30	10
Mean ± SD	67.31 ± 26.1	71.18 ± 24.3	55.71 ± 29.0
Median	72.14	76.36	60.00
Missing values	24	17	7
Percent activity impairment due to depression (%)
N	162	110	52
Mean ± SD	67.04 ± 25.1	68.36 ± 25.7	64.23 ± 23.7
Median	70.00	70.00	60.00
Missing values	0	0	0
Sheehan Disability Scale – SDS
The symptoms have disrupted your work/school, n (%)
Not at all	2 (1.5%)	1 (1.2%)	1 (2.2%)
Mildly	9 (6.9%)	6 (7.1%)	3 (6.7%)
Moderately	41 (31.5%)	24 (28.2%)	17 (37.8%)
Markedly	44 (33.8%)	30 (35.3%)	14 (31.1%)
Extremely	34 (26.2%)	24 (28.2%)	10 (22.2%)
Median (min–max)	7.5 (0.0–10.0)	8.0 (0.0–10.0)	7.0 (0.0–10.0)
The symptoms have disrupted your social life/leisure activities, n (%)
Not at all	3 (1.9%)	2 (1.8%)	1 (1.9%)
Mildly	11 (6.8%)	9 (8.2%)	2 (3.8%)
Moderately	56 (34.6%)	36 (32.7%)	20 (38.5%)
Markedly	64 (39.5%)	41 (37.3%)	23 (44.2%)
Extremely	28 (17.3%)	22 (20.0%)	6 (11.5%)
Median (min–max)	7.00 (0.0–10.0)	7.00 (0.0–10.0)	7.00 (0.0–10.0)
The symptoms have disrupted your family life/home responsibilities, n (%)
Not at all	2 (1.2%)	1 (0.9%)	1 (1.9%)
Mildly	8 (4.9%)	6 (5.5%)	2 (3.8%)
Moderately	57 (35.2%)	33 (30.0%)	24 (46.2%)
Markedly	70 (43.2%)	49 (44.5%)	21 (40.4%)
Extremely	25 (15.4%)	21 (19.1%)	4 (7.7%)
Median (min–max)	7.0 (0.0–10.0)	7.0 (0.0–10.0)	6.0 (0.0–10.0)
SDS overall score
Median (min–max)	21.0 (0.0–30.0)	22.0 (0.0–30.0)	20.0 (0.0–30.0)
Missing values	32	25	7
On how many days in the past 7 days did your symptoms cause you to miss school or work or leave you unable to carry out your normal daily responsibilities
Mean ± SD	2.85 ± 3.1	2.97 ± 3.1	2.58 ± 3.1
Missing values	0	0	0
On how many days in the past 7 days did you feel so impaired by your symptoms, that even though you went to school or work or had other daily responsibilities, your productivity was reduced
Mean ± SD	1.98 ± 2.4	2.28 ± 2.5	1.35 ± 1.8
Missing values	0	0	0

Unless otherwise noted, data are expressed as number and percentage.

MDD – Major Depressive Disorder; TRD – treatment-resistant depression; non-TRD – MDD patient without TRD; SD – standard deviation.

Table 5 depicts also the results for disability and functional impairment. According to the Sheehan Disability Scale, an average of 2.85 days of the last seven days were marked by missed school or work or inability to carry out everyday responsibilities. Concerning the number of days in which tasks were fulfilled but with reduced productivity, the mean was around 2 days in the last 7 days. Higher averages were found in non-TRD patients compared to TRD patients for number of days lost or with reduced productivity. MDD patients reported that their symptoms markedly or extremely disrupted the school/work (60%) and their social life/leisure activities (56.8%). Similar proportion was found in family life/home responsibilities disruption, with 58.6% of MDD patients reporting markedly/extremely disrupted. The median overall score of SDS was 21.0 (Table 5).

Discussion

The importance of the TRAL project as an epidemiological regional characterization of TRD in LatAm has been highlighted in previous papers. This project aimed to fulfill an unmet need in the epidemiology of a complex and often overlooked condition in the region. These results concern the Colombian sample and suggest that the prevalence in Colombia, as previously mentioned, is in line with the mean TRD prevalence in MDD patients. The TRAL project suggested a prevalence in the region around 31.7%, which may suggest that the prevalence in Colombia is slightly higher (34.3%). Interestingly, values are very similar to Argentina, and substantially higher than Mexico and lower than Brazil. Previous research showed that MDD was the most prevalent mental condition in Colombian women, with substance abuse leading in the case of men.²⁹ Overall, epidemiological results from TRAL are aligned with previous benchmark large international studies.^9,30 Local studies point for a prevalence of MDD over 16%,⁵ lower than regional estimates from a study – based on a clinical sample which suggests higher prevalence – with 1835 participants from Argentina, Brazil, Chile, Colombia, and Mexico (23.0–35.0%).⁷ However, reported values are higher than the ones obtained in other studies which is probably justified by the active follow-up of MDD patients at the reference sites. Nevertheless, prior to the TRAL study, evidence was scarce and comparisons with previous research in Colombia should be performed with some caution.

Concerning setting, in the Colombian sample, the prevalence of TRD was higher in private psychiatric institution, which is in contrast with results from Brazil. However, the proportion of treated patients remained high.

Some relevant data pertaining mental health in Colombia is outdated,²⁹ but concerning the country's history of violence, associated with high prevalence of substance abuse, there is a clear need for up-to-date epidemiological data.³¹ Decades of widespread armed conflict had severe consequences in the mental health of Colombians.²⁰ The 2015 national mental health survey did however depict the impact of MDD in Colombia. MDD was identified as the most predominant mental disorder, with a 4.3% lifetime prevalence.⁶ The challenges from MDD and TRD must be addressed by more efficient therapeutic approaches, combining both psychotherapy and psychopharmacological elements. These can only be achieved with reliable and current data, which was the main goal of the TRAL study. Raising awareness to the prevalence of TRD is essential.

The high prevalence of TRD in Colombia brings forward the burden of the disease. These results from the Colombian subsample raise some questions concerning the burden of TRD compared to non-TRD MDD patients, contrasting available literature.^12,13 Even though the number of consultations in psychiatry and psychology is higher for TRD, that was not the case for other types of medical consultations. Moreover, the costs associated with these consultations were also reportedly higher. Concerning costs, the data from TRAL should be interpreted with caution since these were self-reported by the patients. As previously shown,^{8,10,11,13,16} the burden of MDD and TRD is considerable in numerous dimensions. From economic burden to a more humanistic perspective, TRD must be addressed to curb the issues associated with the condition. Literature suggests that TRD has a higher burden than non-TRD, but in this sample that conclusion is far from obvious. In Colombia, the limited available data on burden shows the economic straining of the condition due to hospitalizations.¹⁸ Globally, depression has been identified as a major factor for re-hospitalization in Colombia.¹⁹ There is consensus that all mental disorders must be addressed due to their increasing burden,⁶ and this should be included in a broad mental health national program.

TRD is associated with a more severe clinical representation, having a pronounced impact in HRQoL, disability and work impairment,³² though common in depression disorders.³³ The present results have a contrasting nature. HRQoL as measured by EQ-5D-5L was reportedly worse in TRD patients, but that was not the case with both work impairment and disability results. Non-TRD patients showed worse disability rations, as well as higher work impairment. This contrasts with previously published results^8,12,34 and should be considered in future research, since it is commonly accepted that burden increases in all dimensions with treatment resistance. Concerning the results obtained in MADRS, the clinical representation of depression was more evident in TRD patients.

An issue common to other studies in TRD concerns diagnosis, since the definition of TRD is still not consensual.³ In the current study, since all patients were already being followed in reference sites for MDD, the diagnosis of TRD was defined after failure to two antidepressants in the current depressive episode occurred. Nevertheless, it is reasonable to assume that the proportion of MDD patients with TRD is higher than the one reported here since some cases may not be properly diagnosed. Screening and diagnosis of TRD remains an issue that must be addressed in the near future.

Mental health should be at the center of the healthcare decisions in Colombia. The injuries from decades of conflict are still present and will persist.²¹ Although the Colombian subsample of the TRAL sample was not formally calculated to allow inferential analysis between TRD and non-TRD patients, results derived from the overall data of this real-world regional study (TRAL)¹⁷ should promote a more thorough discussion of this problem in the Latin America region by all relevant stakeholders. Also, and due to the same sampling constraints, generalizations must be performed with caution, as the sample is not representative of Colombia's general population and results are descriptive in nature. This sample provides a robust description of MDD patients with TRD diagnosis being followed at reference sites in Colombia, but no inferential analysis was performed – except for the sociodemographic characteristics at baseline between non-TRD and TRD patients among the overall sample of MDD – ssince the authors preferred a more conservative approach to the results. The sites enrolled in this study are references in the country but do not cover the whole country.

Over 2 million Colombians suffer from depressive disorders, making these conditions a mental health priority.¹ The ongoing acceptance and implementation of international guidelines in Colombia is a positive sign, but implementation is still short from the proposed outcomes.³⁵ The publication of clinical guidelines for both adults and adolescents^36,37 should also consider the most current and effective therapeutic solutions for TRD.

Financial support statement

TRAL (Treatment-Resistant Depression in America Latina) study was funded by Janssen Latin America. CTI provided medical writing and editorial support funded by Janssen Latin America.

Conflicts of interest

RNCR: Declares no conflict of interests.

EFR: Researcher for the lnstituto Colombiano del Sistema Nervioso -Clinica Montserrat, in which he is the principal investigator and sub-investigator of several protocols of original epidemiological research of the institution. Has received professional fees for the time spent on subject interviews in the present study by Janssen Research & Development, as approved by the Independent Research Ethics Committee.

LMAB: Declares no conflict of interests.

GK: Is currently an employee at Janssen Pharmaceutical.

PC: Is currently an employee at Janssen Pharmaceutical.

Acknowledgments

The authors would like to thank all TRAL investigators and site coordinators, as well as the patients and their families.

Clinical Trial & Consulting Services (CTI) provided statistical analysis support to the TRAL funded by Janssen LatAm.

Diogo Morais from Clinical Trial & Consulting Services (CTI) provided medical writing assistance and editorial support with this manuscript, funded by Janssen LatAm. Janssen LatAm participated in the interpretation of data, review, and approval of the content. All the authors had access to all relevant data and participated in writing, review, and approval of this manuscript.