Management of psychiatric treatments of patients diagnosed with bipolar disorder in the COVID-19 pandemic: A one-year evaluation in the pandemic

Abstract

Objective

The course of bipolar disorder (BD) is sensitive to factors that may disrupt biological and social rhythms. It is important for patients diagnosed with BD to continue their follow-up and treatment during the pandemic due to personal and social effects. This study aimed to evaluate the disease course and treatment compliance of individuals diagnosed with BD during the pandemic.

Methods

A total of 267 patients with BD were included in the study. The scales were applied by phone calls. A sociodemographic data form was filled out during the phone interviews. Diagnostic criteria for hypomanic, manic, and depressive episodes in DSM-5 were questioned and recorded through the created form.

Results

During the first of the pandemic, a total of 72 (27.0%) patients had a mood episode, of which 56 (21.0%) were manic/hypomanic episodes and 16 (6.0%) depressive episodes. Also, 54.7% of the patients were able to obtain their medications thanks to the extended medication reports. Being unable to use their medications regularly, having a seasonal pattern of disease, and using an increased number of psychotropics were significant predictors of a new episode. While 74.5% of the patients wanted to talk to their psychiatrists online, only 1.1% could reach the psychiatrist online.

Discussion

The effects of the COVID-19 pandemic are particularly evident in patients with a seasonal pattern. Telepsychiatry practices should be actively included in clinical practice, and government policies developed for treatment compliance seem important.

Introduction

A new type of coronavirus, caused by the pandemic known as coronavirus disease-2019 (COVID-19), continues to affect the world in waves.¹ In the pandemic, which is an extraordinary situation, many arrangements have been made regarding health services, and risk groups have been determined.^2,3 In planning these services, local and international mental health associations tried to create a consensus on approaches to individuals with mental illness, and suggestions were made.^4,5 The patients were informed about the importance of planning out their daily lives to practice regular maintenance, sleep hygiene, and healthy nutrition patterns. On the other hand, plans were tried to be made for the professionals working in the field of mental health, and suggestions were given to the managers who made the arrangements related to the health system to plan out the treatment of the patients, easy supply of drugs, and safe continuation of these services in cases where emergency or inpatient treatment was required.^4,6

The course of bipolar disorder (BD) is sensitive to factors that can disrupt biological and social rhythms.⁷ In addition, it is crucial for patients diagnosed with BD to continue their psychiatric treatment and to reach the center where they are followed so that the patient does not go through episodes. Risk factors such as disruption of daily routines, mental effects of the infection itself, and changes related to alcohol use have been reported for patients diagnosed with BD, which may cause them to have an episode during the pandemic.⁸ Situations and rules that were applied to reduce the spread of COVID-19 such as quarantine, curfews, and social distancing might disrupt the usual sleep and wake patterns, as well as the number and quality of social contacts and activities. In addition, potential factors such as general concerns about the pandemic and reduced access to treatment may trigger a new bipolar episode. Irregularities in biological rhythm can trigger either depressive or manic episodes.⁹ Patients with a manic/hypomanic episode may fail to observe social distancing or other hygienic measures, unintentionally contributing to the spread of COVID-19 and undermining infection control efforts. This puts them and their immediate surroundings at a higher risk of infection. It is of great importance to evaluate how patients diagnosed with BD continue their treatment during the pandemic, how their disease course is, and which factors have adverse effects on the course of the disease since individuals have wide-ranging consequences that can affect themselves, their immediate environment, and the society.⁸

Predictions have been made about how patients diagnosed with BD will be affected by the pandemic, but there are limited studies in which patients are evaluated qualitatively by conducting online or face-to-face interviews.¹⁰ In a study by Karantonis et al. in which 43 patients diagnosed with BD and 24 healthy controls were evaluated, no intra-group differences were found in age, sex, living situation, and job loss or reduced work hours due to COVID-19.¹¹ Mutlu and Yagcioglu evaluated the relapse rate according to the treatment of the patients in their study, in which they evaluated a total of 155 patients with serious mental illness, including a limited number of patients (n:24) diagnosed with BD, and reported that long-acting injections had no control in preventing relapse.¹² In these studies, patients diagnosed with a psychotic disorder were followed up, and there was no evaluation of how individuals diagnosed with BD met their daily life needs, how they continued their psychiatric treatments, and how they obtained their medications during the pandemic.

Our study aimed to investigate the factors affecting individuals diagnosed with BD, their disease course, and their treatment compliance; and to determine how they met their daily needs and how they continued their psychiatric follow-up in the first year of the pandemic.

To the best of our knowledge, our study is the first to be conducted with a relatively large group of patients with a detailed discussion of treatment adherence and social adjustment of patients with BD during the first year of the pandemic. It is thought that the results of our research can guide the planning and maintenance of psychiatric treatments of patients diagnosed with BD and psychosocial support interventions in extraordinary situations such as pandemics.

Methods

Ethical considerations

The study was approved by the Ethics Committee of the University of Health Sciences, Ankara City Hospital (numbered E1-20-1161 and dated September 30, 2020) and was conducted according to the criteria set by the Declaration of Helsinki. When the participants were called, the researcher’s name and the name of the institution they were working in were given at first. It was stated right after that they were contacted regarding a study. Knowledge of the study, as well as all possible benefits and risks were explained. The participants were included in the study when they verbally confirmed that they understood and agreed to participate in the study. The forms were completed only by those who declared that they consented to participate in the study.

Study design and participants

This research had a cross-sectional design in which the scales were applied by phone calls. Patients who had the diagnoses of bipolar affective disorder (ICD 10-F31) and any of its sub-diagnoses in Ankara City Hospital were included in the study. The diagnoses that are considered valid for this study are the ones entered into the national health system and made by the psychiatrists who monitored the patients through clinical consultation. Patients that were determined to be in remission according to the notes that were taken during examinations within six months before the onset of the pandemic in Turkey (March 2020) were included in the study. Patients with a score of three or less according to the “Clinical Global Impression (CGI)—Severity Scale” and those who were stated to be in remission in outpatient notes were determined to be in remission. Noncompliance to the treatment was defined as the patient's complete discontinuation of their medication for 10 consecutive days or taking less than 75% of the prescribed daily dosage in the previous 30 days.¹³

Individuals who could be reached with the “convenience sampling approach” were included in the study. At the end of the first year of the pandemic, the phone calls were started on March 1, 2021, and were completed on June 5, 2021. The phone numbers of the patients were obtained from the hospital information registry system. Two experienced psychiatrists made 30-minute phone calls to take answers for the sociodemographic data-form questions that were previously prepared. Patients that were not in remission or were hospitalized in a psychiatry clinic in the six months before the pandemic, as well as patients with comorbid organic mental disorders or intellectual disabilities, were excluded from the study. Diagnostic criteria for hypomanic, manic, and depressive episodes in DSM-5 were questioned and recorded in detail through the created form to determine the mood of the patients during the telephone interview. Patients with a CGI score of 4 and above and meeting the diagnostic criteria for an episode according to the DSM-5 diagnostic criteria were considered in the episode period and as in recurrence. A total of 736 patients were followed up in our outpatient unit with the specified diagnoses within six months before the onset of the pandemic, and 267 people who agreed to participate in the study were interviewed. The flow chart of the patients included in the study is shown in Figure 1.

Figure 1.

Flow chart of participants included in the study.

Measurement tools

During the telephone interviews, sociodemographic data were collected from the participants, including the following information: the patient’s age, marital status, education level, the place they were living before the pandemic, the place they lived during the study period, history of their disorder (number and type of episodes, hospitalizations, etc.), the treatments they used, how they continued their treatment during the pandemic, whether they were infected with COVID-19, how they contacted their doctor, and whether they wanted to meet with their doctor through telepsychiatry applications.

Statistical evaluation of data

The Kolmogorov–Smirnov test was used for the assumption of normality distribution. Accordingly, it was accepted that the variables with p < 0.05 did not comply with the normal distribution, and the variables with p > 0.05 showed a normal distribution. The Student's t-test was used to determine the differences between patients in remission and patients who have an episode in terms of continuous variables. The relationship between categorical variables was tested using the Chi-squared test and Fisher's exact test according to expected frequency requirements. Logistic regression analysis was performed to predict the episode development risk and the characteristics of an episode using the variables found significant in two-group comparisons (recurrences/remissions) and further statistical analyses. Mean ± SD values were given as descriptive statistics. The SPSS for Windows version 22.0 package program was used for statistical analysis, and p < 0.05 was considered statistically significant.

Results

Of the 267 patients participating in the study, 121 (45.3%) were men, and 146 (54.7%) were women. 190 (71.2%) were interviewed alone, and 77 (28.8%) with both themselves and their families. The mean age was 44.55 ± 11.37 (min = 20; max = 72) years, and the mean age of disease onset was 28.27 ± 8.81 (min = 13; max = 62) years. The mean disease duration of the patients was 193.98 ± 118.91 (min = 12; max = 588) months. The mean number of hospitalizations was 2.87 ± 2.52 (min = 1; max = 17). The disease of 44 (16.5%) patients showed rapid cycling, and 108 (40.5%) showed a seasonal pattern. It was found that 72 (27.0%) of the patients had a mood episode during the first year of the pandemic, and 195 (73.0%) were in remission. Also, 56 (21.0%) patients had manic/hypomanic episodes, and 16 (6.0%) patients had depressive episodes.

To understand whether the variables differentiate into patients in remission and patients who have episodes, a Chi-squared analysis was performed for categorical variables. Also, the Student's t-test was used to compare the continuous variables between patients in remission and patients who have episodes, since the data showed a normal distribution. The results were shown in Table 1. As seen in Table 1, 14 (19.4%) of 72 patients who had an episode and 9 (4.6%) of 195 patients who were in remission did not use their medications regularly, and this difference was statistically significant (χ² (1, N = 267) = 14.572, p < 0.01).

Table 1.

Comparison of patients diagnosed with bipolar disorder in recurrence and remission in terms of sociodemographic and clinic variables.

	Patients in recurrence (n:72)	Patients in remission (n:195)	Total (n:267)	Statistical analysis*
Age (years, mean ± SD)	44.91 ± 11.63	44.42 ± 11.29	44.55 ± 11.37	p:0.752 t:0.316
Gender
Female	42 (58.3%)	104 (53.3%)	146 (54.7%)	p:0.466
Male	30 (41.7%)	91 (46.7%)	121 (45.3%)	X2: 0.530
Marital status
Single	35 (48.6%)	71 (36.4%)	106 (39.7%)	p:0.239
Married	31 (43.1%)	101 (51.8%)	132 (49.4%)	X2: 4.218
Widow/divorced	6 (8.3%)	23 (11.8%)	29 (10.9%)
Working status
Employment	21 (29.2%)	71 (36.4%)	92 (34.5%)	p:0.089
Retired	9 (12.5%)	39 (20.0%)	48 (18.0%)	X2: 4.849
Unemployment	42 (58.3%)	85 (43.6%)	127 (47.5%)
Education years (mean ± SD)	9.62 ± 4.22	10.04 ± 4.24	9.92 ± 4.23	p:0.478 t:−0.711
Living place
Alone	17 (23.6%)	49 (25.1%)	66 (24.7%)	p:0.589
Spouse/children	36 (50.0%)	106 (54.4%)	142 (53.2%)	X2: 1.058
Mother/father/siblings	19 (26.4%)	40 (20.5%)	59 (22.1%)
Change in living place
Yes	1 (1.4%)	6 (3.1%)	7 (2.6%)	p:0.678
No	71 (98.6%)	189 (96.9%)	260 (97.4%)
Access to basic needs
Yes	71 (98.6%)	191 (98.0%)	262 (98.1%)	p:1.000
No	1 (1.4%)	4 (2.0%)	5 (1.9%)
How to reach basic needs?
Himself	43 (59.7%)	131 (67.2%)	174 (65.2%)	p:0.256
Family	29 (40.3%)	64 (32.8%)	93 (34.8%)	X2: 1.288
Duration of illness (months, mean ± SD)	192.70 ± 119.78	194.45 ± 118.90	193.98 ± 118.91	p:0.916 t:−0.106
Age of onset (years, mean ± SD)	28.36 ± 8.81	28.23 ± 8.83	28.27 ± 8.81	p:0.918 t:0.103
The mean number of hospitalizations	3.04 ± 2.49	2.78 ± 2.53	2.87 ± 2.52	p:0.490 t:−0.692
Rapid cycling
Yes	15 (20.8%)	29 (14.9%)	44 (16.5%)	p:0.266
No	57 (79.2%)	166 (85.1%)	223 (83.5%)	X2: 1.358
Seasonal pattern
Yes	38 (52.8%)	70 (35.9%)	108 (40.5%)	p:0.013
No	34 (47.2%)	125 (64.1%)	159 (59.5%)	X2: 6.220
Depot antipsychotic use
Yes	11 (15.3%)	29 (14.9%)	40 (15.0%)	p:0.947
No	61 (84.7%)	166 (85.1%)	227 (85.0%)	X2: 0.004
History of hospitalization
Yes	63 (87.5%)	133 (68.2%)	196 (73.4%)	p:0.002
No	9 (12.5%)	62 (31.8%)	71 (26.6%)	X2: 10.029
Phone call by the institution
Yes	7 (9.7%)	14 (7.2%)	21 (7.9%)	p:0.608
No	65 (90.3%)	181 (92.8%)	246 (92.1%)	X2:0.469
Going to outpatient visits regularly
Yes	30 (41.7%)	56 (28.7%)	86 (32.2%)	p:0.044
No	42 (58.3%)	139 (71.3%)	181 (67.8%)	X2:4.038
The mean number of psychotropic use	2.09 ± 0.75	1.83 ± 0.64	1.89 ± 0.69	p:0.005 t:2.851
Multiple psychotrop use
Yes	57 (79.2%)	142 (72.8%)	199 (74.5%)	p:0.291
No	15 (20.8%)	53 (27.2%)	68 (25.5%)	X2:1.116
Treatment compliance
Yes	58 (80.5%)	186 (95.4%)	244 (91.4%)	p < 0.001
No	14 (19.5%)	9 (4.6%)	23 (8.6%)	X2:14.689
Using mood stabilizers
Yes	62 (86.1%)	167 (85.6%)	229 (85.8%)	p:0.922
No	10 (13.9%)	28 (14.4)	38 (14.2%)	X2: 0.010
Diagnosed with COVID-19
Yes	4 (5.6%)	10 (5.1%)	14 (5.2%)	p:1.000
No	68 (94.4)	185 (94.9%)	253 (94.8%)
Request for online interview
Yes	61 (84.8%)	138 (70.8%)	199 (74.5%)	p:0.020
No	11 (15.2%)	57 (29.2%)	68 (25.5%)	X2: 5.390

n: Number of persons; %: Percent. *Chi-squared test was used for categorical variables; if the assumptions were not met, Fisher’s exact test was used, Student t-test was used for continuous variables.

Forty of the patients (15.0%) were using depot antipsychotics, and there was no statistically significant difference between depot users and non-users in terms of having a mood episode (χ² (1, N = 267) = 0.004, p = 0.947). As also seen in Table 1, when those who has a seasonal pattern of disease and those who did not were compared in terms of having a mood episode, the rate of having an episode was statistically significantly higher in those with a seasonal pattern (χ² (1, N = 267) = 6.220, p = 0.013). The comparison of the other variables in the recurrence and remission groups is also given in Table 1.

In descriptive analyzes to assess how patients obtain their medication, 244 (91.4%) of the patients were found to be using their drugs regularly, 146 (54.7%) of the patients could take their drugs directly from the pharmacy with a medication report without the need for a prescription, and only 3 (1.1%) patients were prescribed their drugs after the telepsychiatry interview. The way how patients obtained their medicines during the pandemic period is shown in Figure 2.

Figure 2.

How did individuals with bipolar disorder get their medicines during the pandemic period?

Descriptive analyzes were used to evaluate the oral and depot psychotropics used by the patients. When the treatments used by the patients were evaluated, the average number of psychotropic use was 1.89 ± 0.69 (min = 1; max = 5). While the mean number of psychotropics for those who had an episode was 2.09 ± 0.75, it was 1.83 ± 0.64 for those who remained in remission (t(265) = 2.851, p = 0.005). A total of 68 patients used single psychotropics, 34 of which were mood stabilizers and 34 antipsychotics. It was found that 199 patients used multiple psychotropics. As seen in Table 2, the most commonly used mood stabilizer in patients was valproic acid (n:145, 54.3%) followed by lithium (n:69, 25.8%). Quetiapine (n:94, 35.2%) was the most commonly used antipsychotic, followed by olanzapine (n:57, 21.3%). The oral and depot antipsychotics and the type of combinations used by the patients are shown in Table 2.

Table 2.

The mostly used psychotrophs and the type of combinations in treatment of the patients.

Oral antipsychotic drugs	(n)
Quetiapine	94
Olanzapine	57
Aripiprazole	25
Risperidone	22
Amisulpride	7
Haloperidol	7
Paliperidone	4
Chlorpromazine	2
Mood stabilizers
Valproic acid	145
Lithium	69
Carbamazapine	17
Lamotrigine	2
Depot antipsychotic drugs
Long-acting injectable risperidone	14
Depot aripiprazole	13
Paliperidone palmitate	11
Zuclopentixol decanoate	2
Haloperidol decanoate	1
Others
SSRI	9
SNRI	6
Benzodiazepine	4
The most commonly used drug combinations
Valproic acid—Quetiapine	44
Valproic acid—Olanzapine	24
Lithium—Quetiapine	19
Lithium—Olanzapine	12
Lithium—Aripirazole	9
Valproic acid—Aripirazole	9
Valproic acid—Aripirazole long-acting injection	7
Valproic acid—Risperidone long-acting injection	6
Valproic acid—Paliperidone long-acting injection	5
Valproic acid—Lithium	4
Valproic acid—Risperidone	3
Others	57

n: Number of persons.

Chi-squared analysis was performed to find out whether there was a difference in terms of episode occurrences between the use of single/multiple psychotropics or the use of any mood stabilizers. Fifteen (22.1%) of 68 patients using single psychotropics and 57 (28.6%) of 199 patients using multiple psychotropics were found to have had a mood episode, and this difference was not statistically significant (χ² (1, N = 267) = 1.160, p = 0.291). While 229 (85.8%) of the patients used at least one mood stabilizer, 38 (14.2%) did not. It was found that 62 (27.1%) of 229 patients who used mood stabilizers and 10 (26.3%) of 38 patients who did not use mood stabilizers had a mood episode, and there was no statistically significant difference between the two groups (χ² (1, N = 267) = 0.010, p = 0.922).

It was found that 143 (64.7%) of 221 patients using either valproic acid, lithium, or carbamazepine did not have their therapeutic drug monitoring checked for one year. When questioning the status of patients having COVID-19, 14 (5.2%) of the 267 patients were found to be diagnosed with COVID-19. It was found that 11 (4.1%) individuals could continue their medication regularly during the period when they were treated for COVID-19. Out of 14 patients infected with COVID-19, 1 patient was hospitalized and there were no deaths.

When the online interview status of the patients is evaluated, it was found that 246 (92.1%) of the patients were not called for an online interview by the center they were followed up during the first year of the pandemic, even though 199 (74.5%) wanted to continue their follow-up online with their psychiatrist. It was found that 61 (84.8%) of 72 people who had an episode and 138 (70.8%) of 195 people who were in remission wanted to make an online psychiatric interview, and the difference was statistically significant (χ² (1, N = 267) = 5.489, p = 0.019).

In the further statistical evaluation, an episode development risk was tried to be predicted by logistic regression analysis. For this reason, variables that were found to be statistically significant between patients who remained in remission and those who entered an episode were included in the analysis (as an independent variable). The two universal confounders (age and gender) and work status were also tested in the binary logistic regression model. In the model, noncompliance with treatment, going to the outpatient visits regularly, increase in the number of psychotropics used per day, and the seasonal pattern of the disease were significant predictors for the episode (Omnibus Tests of Model Coefficients: χ²(3) = 36.646, p < 0.001). This is also demonstrated by the fact that the model predicts remission with 95.9% (n:187 of 195) accuracy, and recurrence with 25.0% accuracy (n:18 of 72), with an overall percentage of 76.8%. When the employment status was examined, being unemployed increased the rate of having an episode compared to being retired. This difference very close at the level of significance when compared to the reference category (retired status) in the model (p: 0.059). The logistic regression analysis of the variables predicting having an episode and classification table with accuracy percentages are given in Table 3.

Table 3.

Logistic regression analysis of predictors for recurrence and classification table with accuracy percentages**.

	B	S.E.	Wald	df	Sig.	OR	CI 95%
Treatment compliance
*** Yes
No	1.693	0.484	12.213	1	<0.001	5.435	2.103–14.044
The mean number of psychotropic use per day (Average value according to the number of psychotropics each patient uses per day)	0.588	0.217	7.321	1	0.007	1.800	1.176–2.755
Going to outpatient visits regularly
***No
Yes	0.877	0.319	7.535	1	0.006	2.404	1.285–4.496
Seasonal pattern
***No
Yes	0.701	0.308	5.185	1	0.023	2.015	1.103–3.684
Age (Average value)	0.016	0.014	1.234	1	0.267	1.016	0.988–1.044
Gender
***Female
Male	0.01	0.312	0.001	1	0.975	1.010	0.548–1.862
Working status
***Retired			5.675	2	0.059
Unemployment	1.028	0.467	4.852	1	0.028	2.794	1.120–6.972
Employment	0.499	0.486	1.052	1	0.305	1.647	0.635–4.271
Constant	−2.294	1.002	5.240	1	0.022	0.101
Classification Table with accuracy percentages
Observed	Predicted
	Percentage Correct Having an episode (n) (%)
	None	Yes
Having an episode (n)
None	187	8				95.9
Yes	54	18				25.0
	Overall percentage					76.8

n: Number of persons; %: Percent; Omnibus Tests of Model Coefficients: χ²(3) = 36.646, p < 0.001, Nagelkerke R²: 0.186; **The cut value is 0.500; ***Reference category.

Discussion

This study is a retrospective evaluation of the disease course and treatment compliance of individuals with BD during the first year of the pandemic.

It has previously been reported that quarantine itself or quarantine-related stress and daily rhythm disturbances associated with the COVID-19 pandemic cause an increased risk of having a first-episode manic episode and a manic episode in individuals currently diagnosed with BD.^14–16

Koenders et al. prospectively evaluated 36 patients diagnosed with BD during the COVID-19 pandemic and monitored their mood changes. They reported a significant increase in manic symptoms in the early stages of the pandemic compared to the pre-pandemic period, and these symptoms decreased in the following months. This decrease was interpreted to be associated with the easing of quarantine measures and positive coping strategies with the fear of COVID-19.¹⁷ There are no data on the timing of recurrences in our study. Nevertheless, during the one-year review, there were new waves, curfews, and precautions for the COVID-19 pandemic. Adaptation and coping strategies are, therefore, in a state of constant change.

There is no information about how a stressful life event, which could affect the whole world, affects patients with BD, and there are limited studies in this area regarding COVID-19. Pre-pandemic and post-pandemic episode rates of the same patients cannot be compared, since our study group did not have longer retrospective follow-ups regarding the pre-pandemic period. When the studies examining the rate of episodes and the types of these episodes are evaluated, the research of Vázquez et al., which included both naturalistic and controlled studies, draws attention. They found the annual recurrence rate to be 26.3% in 10 naturalistic studies (n: 3904) and 21.9% in 15 controlled studies (n: 4825). Considering the polarity of the episode in this study, the rates of depressive episodes were found to be higher than hypomanic/manic/mixed episodes.¹⁸

In the multinational, multicenter, and observational study of Vieta et al., 2896 patients diagnosed with BD were followed for 12 to 27 months. At least one recurrent episode was detected in 40.5% (n: 1173) of the patients. These episodes were hypomanic/manic/mixed in 21.4% of patients and depressive episodes in 23.1% of the patients. Totally, 369 patients from Turkey were included in this study; 11.9% of them developed hypomanic/manic/mixed episodes, and 8.1% had depressive episodes.¹⁹ In both of these studies, the data were obtained before the COVID-19 pandemic. Although a complete comparison cannot be made because there was no sample group, the recurrence rates of the patient group in our study did not show a significant difference from previous studies. However, the incidence of manic/hypomanic episodes was significantly higher than the rate of depressive episodes. With the significantly higher rates of manic/hypomanic episodes, we think that quarantine, restriction of social activities, stress related to COVID-19, and disrupted daily rhythms due to the pandemic might be a risk for the development of a new manic/hypomanic episode. While these conditions are also a risk factor in the development of depressive episodes, the fact that manic episodes are more frequent may be because many variables, including biological factors, play a role in the course of the disease.

In a meta-analysis study examining the association of recurrence with psychotropics in patients diagnosed with BD, naturalistic and randomized controlled studies were compared. With randomized controlled trials, annual recurrence rates with specific treatments were reported as follows: combinations of antipsychotic and mood-stabilizing agents (14.6%/year) < antipsychotics alone (15.8%/year) < single per lithium (23.8%/year) < anticonvulsants (29.9%/year) < placebo (31.9%/year). Moreover, between-treatment differences in recurrence rates were significant (p = 0.01) in randomized controlled trials. However, in the same meta-analysis, there was no relationship between these drug groups and recurrence rates according to 10 naturalistic studies.¹⁸ Our study found no relationship between recurrence in patients diagnosed with BD and treatment-related changes such as single/multiple psychotropic uses, depot antipsychotic use, and mood stabilizer alone/in combination therapy.

In our study, there were no differences about sociodemographic characteristics of the patients, such as age, gender, marital status, and employment status between recurrence and remission patients. We have general information that sociodemographic characteristics such as social support, being married and having a job are protective in the course of mental illnesses. However, in an important review in which many studies were evaluated, it was seen that these features were not among the factors associated with recurrence rates. It was stated that the most important factor associated with recurrence was the follow-up period.¹⁸ In our study, there was no relationship between sociodemographic characteristics and recurrence. The follow-up period is not long as it includes the COVID-19 period. In further studies, it was thought that it would be better to re-evaluate these variables associated with recurrence when the follow-up period is longer.

Not using drugs regularly, going to outpatient visits regularly, showing a seasonal pattern of the disease, and increasing the number of psychotropics used were found to be predictors for a new episode in our study. Going to outpatient visits regularly can be seen as an unexpected finding for recurrence. However, this result may have been caused by the frequent and regular visits to the outpatient unit due to the complaints of the patients who had episodes. Increasing the number of psychotropic drugs used may complicate drug compliance. In addition, as patients’ complaints increase, clinicians add psychotropics to treatment, which may cause recurrence rates to increase as the number of psychotropics used increases.

The higher rate of recurrence in those with a seasonal disease pattern is thought to be important information that has not been included in the literature before. It has been stated that when BD has a seasonal pattern, changes/differences in photic stimuli and circadian rhythm may cause fluctuations in mood.²⁰ The circadian rhythm changes that occurred with the pandemic may have affected patients with seasonal patterns more and caused more recurrences in this group.

With a decision taken by the Ministry of Health in Turkey during the COVID-19 pandemic, the medication reports of patients, who had chronic diseases, were extended. In addition, thanks to the extended medication reports, these people were able to buy their drugs directly from the pharmacy without the need for a prescription. In our study, it was found that more than half of patients diagnosed with BD were able to take their medication in this way. It was also found that this practice is essential in continuing drug treatment in patients with schizophrenia in our country.²¹ This situation shows us the importance of maintaining health services management interdisciplinary with state policies during the pandemic. Also, therapeutic drug monitoring for mood stabilizers was not measured in 143 (67.8%) of patients over a one-year period. Since it was not among the aims of the study, the maintenance doses of the psychotropics were not evaluated. Although there are discussions about therapeutic drug monitoring for mood stabilizers in the literature, we also think that blood level is important for the protective efficacy and safety of the drug. In such cases, we think that it is important to follow these patients in places that provide primary health care and to be able to ask for therapeutic drug monitoring.²²

Online consultations with individuals with chronic psychiatric diseases are recommended in the literature during the COVID-19 period.²³ With the pandemic, a significant development was observed in telepsychiatry practices. By conducting online interviews with patients, it may be possible to evaluate their mental state, plan treatment, and refer the patients to the hospital when necessary. Although most of our patients (74.5%) wanted to continue their outpatient visits online, only a small portion (1.1%) could be reached online. In the first year of the pandemic, the intensive work of healthcare professionals and the lack of sufficient technical support for online interviews may have caused telepsychiatry applications to not be used actively, especially in public hospitals. The inclusion of telepsychiatry practices in the management of health services can contribute to the follow-up and treatment of patients.

This study had some limitations. The data of our study were obtained by retrospectively questioning the patients’ symptoms through telephone interviews, which may have caused episodes to be missed in patients with low anamnesis reliability. Conducting a prospective follow-up study with a larger sample would provide more comprehensive information on the potential impact of the pandemic on patients diagnosed with BD.

Secondly, some patients may not have been reached because they were in an episode, and these individuals may have been excluded from the study. Thirdly, the patients included in our study were patients with BD who were followed up in our institution. The medication reports of almost all patients followed up in our institution are regularly renewed. This may have caused drug compliance to be higher than normal. The effects of these differences can be better understood with multicenter studies. Finally, the first year of the pandemic was evaluated in our study. This period of the pandemic was the most stressful timeline because of the unknowns of the infection, lockdowns, and curfews. The problems or stress of the patients during this period of the pandemic should not be generalized as that of the whole pandemic.

Conclusion

It was seen that most patients diagnosed with BD continued their pharmacological treatments during the pandemic period, were able to meet their basic needs, and there was no significant change in their family life in Turkey. Most patients demand telepsychiatry practices, and the accessibility and development of clinical practices in this area seem important. We think that the results of our study can guide the planning in the psychiatric management of patients diagnosed with BD.