Demographic and Clinical Factors Associated with Re-admission in Patients Admitted to Psychiatric Hospital and Psychiatric Ward of General Hospital in 2011: A 7-Year Follow-Up after Discharge

Document Type : Original article

Authors

1 Mental Health Research Center, Psychosocial Health Research Institute, Department of Psychiatry, School of Medicine, Iran University of Medical Sciences, Tehran, Iran

2 Department of Psychiatry, Sunnybrook Health Sciences Centre, Toronto, ON, Canada 3. Department of Psychiatry, University of Toronto, Toronto, ON, Canada

3 Preventive Medicine and Public Health Research Center, Psychosocial Health Research Institute, Community and Family Medicine Department, School of Medicine, Iran University of Medical Sciences, Tehran, Iran

Abstract

Background: The present study aimed to investigate the relationship between patients’ demographic/clinical factors and readmission.
Methods: This retrospective cohort study was conducted on all patients admitted to Iran psychiatric hospital and the Psychiatric Ward of Rasoul-e-Akram Hospital in Tehran in 2011. The study variable and post-discharge readmission status in a 7-years period until 2018 were collected by reviewing patients’ clinical charts, making telephone calls or studying patients’ records at Roozbeh Hospital. Thereafter, clinical factors such as Length of Hospital Stay (LOS), diagnosis, and medication compliance were evaluated by using statistical tests, including Meier-Kaplan survival analysis, Logrank test, and Cox regression analysis and SPSS software.
Results: The present study was conducted on 1410 inpatients, 434 women (30.8%) and 976 men (62.9%), whose most common disorder was mood disorder. The diagnoses of the patients were recorded based on DSM-IV-TR. Among the patients readmitted, the mean time to readmission was 24.94 months. Among the demographic and clinical factors, the longer length of stay, more previous hospitalizations, inpatient physical restraint, lack of therapeutic compliance, and diagnosis of psychotic disorders has increased the odds of readmission (all p<0.01).
Conclusion: To reduce the readmission rate among these patients, we need to use some methods to reduce readmission such as psychoeducation regarding nature of the disease, importance of medication adherence, and identifying warning signs that indicate the possibility of recurrence of the disease, especially more severe such as psychotic disorders. It is also recommended to avoid physical restraint, a factor of readmission risk, during hospitalization as much as possible.

Keywords


Abstract
Background: The present study aimed to investigate the relationship between patients’ demographic/clinical factors and readmission.
Methods: This retrospective cohort study was conducted on all patients admitted to Iran psychiatric hospital and the Psychiatric Ward of Rasoul-e-Akram Hospital in Tehran in 2011. The study variable and post-discharge readmission status in a 7-years period until 2018 were collected by reviewing patients’ clinical charts, making telephone calls or studying patients’ records at Roozbeh Hospital. Thereafter, clinical factors such as Length of Hospital Stay (LOS), diagnosis, and medication compliance were evaluated by using statistical tests, including Meier-Kaplan survival analysis, Logrank test, and Cox regression analysis and SPSS software.
Results: The present study was conducted on 1410 inpatients, 434 women (30.8%) and 976 men (62.9%), whose most common disorder was mood disorder. The diagnoses of the patients were recorded based on DSM-IV-TR. Among the patients readmitted, the mean time to readmission was 24.94 months. Among the demographic and clinical factors, the longer length of stay, more previous hospitalizations, inpatient physical restraint, lack of therapeutic compliance, and diagnosis of psychotic disorders has increased the odds of readmission (all p<0.01).
Conclusion: To reduce the readmission rate among these patients, we need to use some methods to reduce readmission such as psychoeducation regarding nature of the disease, importance of medication adherence, and identifying warning signs that indicate the possibility of recurrence of the disease, especially more severe such as psychotic disorders. It is also recommended to avoid physical restraint, a factor of readmission risk, during hospitalization as much as possible.
Keywords: Demography, Inpatients, Patient readmission, Psychiatric hospitals

Introduction
Psychiatric disorders are among the important public health issues worldwide as well as in Iran. The mental health pattern in Iran is similar to the western countries, but it seems that the prevalence of psychiatric disorders in Iran may be lower than these countries (1). It is estimated that at least about 7 million of Iranian population suffer from one or more psychiatric disorders (1). In addition, one systematic review on the prevalence of psychiatric disorders in Iran included 35 studies and showed that the prevalence estimates varied in the range of 1.9-58.8% (2). Iran’s 2003 national study on burden of diseases reported that neuropsychiatric disorders had the highest burden only after injuries and disasters and were contributing to 16.04% of the burden of diseases (3). It represents the importance of the role of the psychiatric disorders in providing preventive and management programs in Iran (1).
Admission is one of the most important modalities in the treatment of psychiatric disorders, which is considered by some authors as an appropriate short-term crisis intervention that can be used periodically during episodes of acute psychiatric disorders (4,5). Re-referrals to hospitals due to various reasons, such as failure of the earlier hospital admission (6), lead to higher bed occupancy rate, and increase the hospital costs imposed on both the health system and patients. Overall, inpatient treatment is an expensive way of caring for people with acute psychiatric disorders (7).
In addition, patients with serious psychiatric disorders are at higher risk of mortality (8) and morbidity (9) (30% higher) compared to the general population (10). The life expectancy of these patients is 15 years less than other people, and they are at risk of suicide.
Frequent admissions of psychiatric patients can reflect their psychological and social statuses, their support and social resources and the type of disease (11). Readmission has a major role in the reduction of the quality of life and increases the years of lost life (12) and also predicts the low quality of care (13), which can cause serious complications such as social and psychological dysfunctions. 
A study in the United States indicated the highest readmission rate within 30 days after discharge among adult patients with psychiatric disorders is in patients with schizophrenia and mood disorders (14). However, it was shown that the risk of readmission is high among patients with all psychiatric disorders.
Many factors were found to affect the readmission rate of psychiatric patients. In this study, several attempts were made to investigate the effect of risk factors on readmission. Accordingly, these factors are categorized into the following two groups: demographic and clinical (12). Demographic factors include age, sex, marital status, educational level, residence, and occupation. Additionally, clinical factors include diagnosis, medication type, route of administration, comorbidity, history of substance use disorder, history of suicide attempt, history of previous admissions, history of medication compliance, and Length Of Stay (LOS) (10).
Demographic factors have been investigated in several studies. Some of them reported the protective effect of employment on readmission in a bivariate analysis (15,16) and regarded unemployment as a risk factor (17). However, most of these studies showed no significant relationship between living conditions/educational level and readmission (18). Another study investigated the effect of demographic factors on readmission of depressed patients and then reported higher readmission rate among younger depressed patients (19).
Considering that clinical factors are among the important predictors of readmission, this issue has been known as the target in many studies. In most studies, the most important risk factor of the recurrence and readmission of psychiatric patients was found to be the history of frequent previous admissions (15,18). However, medication non-compliance has been cited as another important risk factor in this regard (20), and other factors such as the type of diagnosis, type of treatment and Electroconvulsive Therapy (ECT) (21), and self-injurious behaviors (19) have also been investigated.
Although willingness to perform studies in this field has increased in recent decades (22), little information is now available on the predictors of psychiatric readmission (19). On the other hand, there has been only a single similar study in Iran and few studies abroad. It should be noted that none of these studies has evaluated all demographic and clinical factors in a coherent manner.
Considering those previous studies which reported a relationship between the patients’ demographic/clinical characteristics and the readmission rates, since there is not much information in this field, especially in Iran, we aimed to investigate the relationship between these factors and readmission.
Therefore, several attempts were made to investigate the records of all patients admitted to Rasoul-e-Akram and Iran psychiatric Hospital in 2011. Furthermore, all their demographic and clinical factors were recorded. Thereafter, the included patients were followed up on their readmissions and other related factors for seven years, and finally, the relationship between the aforementioned factors and their readmissions was evaluated.

Materials and Methods
This was a retrospective cohort study, and the sample size was calculated by using a census of all the patients admitted to the psychiatric ward of Rasoul-e-Akram and Iran psychiatric Hospital in Tehran in 2011. Iran hospital is a single-specialty psychiatric hospital and includes 5 wards and each ward has 30 active inpatient beds, and psychiatric ward of Rasoul-e-Akram general Hospital also has 30 active beds. Additionally, Iran hospital has more facilities for patients like community-based care, specialized care for patients with substance use disorders, more post-discharge clinics, etc. After inquiring from the hospitals, we found out that approximately 2500 patients were admitted to these centers in 2011. The inclusion criteria included all patients with admission charts which were archived in the psychiatric ward of Rasoul-e-Akram and Iran psychiatric Hospital in 2011. Moreover, the exclusion criterion was short-term admission (less than seven days). Accordingly, significant numbers of medical patient records associated with a short-term emergency admission, which were not related to the purpose of this study, were excluded from the research. However, two other cases were also excluded due to their unwillingness to participate in the study and/or not responding to the phone call. The final sample size included 1410 patient records, of which 282 records were archived in Rasoul-e-Akram, and 1128 belonged to Iran psychiatric Hospital.
Considering the large data volume and the need for proper data management, a special application was designed to register the necessary patients’ specifications such as age, sex, and the number of readmissions. The output of this application was processed in SPSS software. Furthermore, an electronic database was designed to create the patients’ information registration system that allows future follow-ups, which multiple users can use to enter patients’ follow-up data.
At the first stage of data collection, a pilot assessment was conducted on 30 patient records by the project executors to fix the application bugs, which were detected and fixed by the application designer in several cases. 
At the second stage, when the system was completed, the patients’ records were evaluated. The number and date of readmissions in the above-mentioned hospitals until 2018 and all information related to variables were recorded in the designed application. The application’s information was completed as follows: yes and no options, several possible options for a particular variable, entering a date in the calendar or manually entering information in specific cases. People with medical education (interns) who were familiar with patients’ records and medical terms were hired for doing this stage.
At the third stage, all patients who did not have readmission record in the same hospital until 2018 (7 years) were followed-up as follows: To obtain more information, the researcher firstly contacted the patients by using phone numbers registered in the patients’ records, then they were asked about the frequency and date of their readmissions and outpatient referrals or their possible death. Accordingly, this stage was implemented twice in order to re-contact those patients who did not answer the previous phone call.
At the fourth stage, inquiring from another important psychiatric hospital in Tehran (Roozbeh Hospital) was implemented for cases who could not be contacted. For example, in the case of wrong number or unanswered phone call. Notably, in cases of having readmission, the related information was entered into SPSS. 
The final stage aimed to gather information from Civil Registration Office of Tehran to follow up on the patients’ survival status that remained unclear to the researchers until this stage. Unfortunately, the above-mentioned organization was unwilling to cooperate with us by emphasizing the confidentiality of information regarding dead people.

Data analysis 
Statistical calculations were performed using SPSS statistics 24.0 software. Central tendency and dispersion indices were also calculated. Thereafter, the readmission date was calculated using Kaplan-Meier survival analysis. The relationship between factors and readmission date was also determined by logrank test and Cox-Regression analysis.

Ethical considerations
During the phone calls with the participants, the researcher firstly introduced himself/herself to the patients and stated the purpose of the call, following up on the patients’ condition. It was noted that if there was no desire to cooperate, the patient’s record would be excluded from the study. The patients were aware that their participation or non-participation would have no effect on receiving medical care. The patients’ information was kept confidential, and after being registered in the system, no one had access to them at all.

Results
The present study’s results were obtained by evaluating the records of 1410 patients who were admitted to Iran psychiatric hospital (n=1128 cases, 80%) and the psychiatric ward of Rasoul-e-Akram general Hospital (n=282 cases, 20%) in 2011. The number of the admitted women and men was 434 (30.8%) and 976 (69.2%), respectively. At the time of data entry (2011), the mean ± Standard Deviation (SD) of the women’s and men’s age was 36.7±12.9 and 35.5±11.22 years old, respectively. The mean ± SD of the age of onset of the disorder was 26.96±10.7 and 27.4±10.13 years old for the women and men, respectively. Demographic and clinical characteristics of the included patients are presented in table 1.

 

Table 1. Demographic and clinical characteristics of the patients admitted to Iran and Rasoul-e-Akram psychiatric hospitals in 2011

Number (%)

Iran psychiatric hospital

Rasoul-e-Akram psychiatric hospital

Total

p-value

Number (%)

Number (%)

Sex

Female

294 (26.1)

148 (52.5)

442 (31.3)

0.001

Male

834 (73.9)

134 (47.5)

968 (68.7)

Marital status

Unmarried

558 (49.5)

109 (38.7)

667 (47.3)

0.001

Married

407 (36.1)

142 (50.4)

549 (38.9)

Divorced

155 (13.7)

25 (8.9)

180 (12.8)

Widow

8 (0.7)

6 (2.1)

14 (1.0)

Employment status

Employed

206 (18.3)

74 (26.2)

280 (19.9)

0.002

Unemployed

922 (81.7)

208 (73.8)

1130 (80.1)

Educational level

illiterate

36 (3.2)

20 (7.1)

56 (4.0)

0.001

High school

604 (53.5)

122 (43.3)

726 (51.5)

Diploma

364 (32.3)

79 (28.0)

443 (31.4)

Above diploma

124 (11.0)

61 (21.6)

185 (13.1)

Place of residence

Urban

99.40%

100.00%

99.50%

0.21

Rural

0.60%

0.00%

0.50%

Referral source

Family

97.90%

95.00%

97.30%

0.001

Doctor

0.00%

3.50%

0.70%

Legal authorities

2.10%

1.40%

2.00%

hospitalization restraint

Physical restraint

26.10%

6.40%

22.10%

0.001

Drug restraint

26.80%

0.40%

21.50%

Unknown

47.20%

93.30%

56.40%

Substance abuse

Yes

57.00%

31.90%

52.00%

0.001

No

42.50%

66.30%

47.20%

Unknown

0.50%

1.80%

0.80%

Therapeutic compliance

Yes

40.50%

28.70%

38.20%

0.001

No

59.50%

71.30%

61.80%

Type of discharge

Personal request

26.8

11.30%

23.70%

0.001

Doctor’s order

73.20%

88.70%

76.30%

Suicide history

Yes

19.10%

36.50%

22.60%

0.001

No

80.90%

63.50%

77.40%

Post-discharge visit

Registered

68.30%

47.90%

64.20%

0.001

Not registered

31.70%

52.10%

35.80%

 

The diagnoses of the patients were recorded based on DSM-IV-TR coding. The highest prevalence of psychiatric disorders belonged to mood disorders (n=588 cases, 41.7%). The results of chi-square showed that the prevalence rates of psychotic disorders (72.6%), mood disorders (58.2%), and substance use disorders (88.7%) were higher among men than women (p=0.001, df=4 and χ2=115.38). In contrast, the prevalence rates of anxiety disorders (61.5%) and other mental disorders (50.8%) were higher among women than men. Regarding the comorbid physical diseases, cardiovascular and neurological disorders, trauma, infectious disorders, and infectious/endocrine disorders were more common among men; however, endocrine disorders were more common among women (Table 2).

 

Table 2. Rate of psychiatric disorders and comorbid physical diseases in males and females

 

 

Female number (percentage)

Male number (percentage)

Total number (percentage)

χ2

p-value

Psychiatric disorders

Psychotic disorders

112 (27.4)

297 (72.6)

409 (29)

115.38

0.001

Mood disorders

246 (41.8)

342 (58.2)

588 (41.7)

Anxiety disorders

16 (61.5)

10 (38.5)

26 (1.8)

Substance use disorders

37 (11.3)

289 (88.7)

326 (23.1)

Others

31 (50.8)

30 (42.9)

61 (4.3)

Total

442 (31.3)

968 (68.7)

1410 (100)

Comorbidity

No comorbidity

275 (27.6)

720 (72.4)

995 (73.10)

72.35

0.001

Cerebrovascular

14 (42.4)

19 (57.6)

33 (2.4)

Endocrine

65 (66.3)

34 (33.7)

101 (7.4)

Neurology

28 (32.6)

58 (67.4)

86 (6.3)

Trauma

14 (20.9)

53 (79.1)

67 (4.9)

Infectious

2 (12.5)

14 (87.5)

16 (1.2)

Endocrine and infectious

19 (30.2)

44 (69.8)

63 (4.6)

Total

442 (31.3)

968 (68.7)

1410 (100)

 

There was a significant difference between these two hospitals in terms of the prevalence of psychiatric disorders (p=0.001 and χ2=101.04). The prevalence rates of psychotic and substance use disorders were higher among patients who were admitted to Iran hospital, and mood, anxiety, and other disorders were prevalent among the patients admitted to Rasoul-e-Akram general Hospital. Regarding the prevalence of comorbid physical diseases, neurological disorders, trauma, and infectious disorders also were more common among the patients admitted to Iran hospital, and cardiovascular and endocrine disorders were more common among the patients who were admitted to Rasoul-e-Akram general Hospital (Table 3).  

 

Table 3. Rate of psychiatric disorders and comorbid physical diseases in Iran and Rasoul-e-Akram general hospitals

 

 

Iran number (percentage)

Rasoul-e-Akram

number (percentage)

Total number (percentage)

p-value

Psychiatric disorders

Psychotic disorders

31.60%

18.80%

29.00%

0.001

Mood disorders

36.60%

62.10%

41.70%

Anxiety disorders

1.70%

2.50%

1.80%

Substance use disorders

27.00%

7.80%

23.10%

Others

3.20%

8.90%

4.30%

Comorbidity

No comorbidity

75.60%

62.30%

73.10%

0.001

Cardiovascular

1.60%

5.8

2.40%

Endocrine

5.30%

16.70%

7.40%

Neurology

6.30%

6.20%

6.30%

Trauma

5.70%

1.60%

4.90%

Infectious

1.40%

0.40%

1.20%

Endocrine and infectious

4.10%

7.00%

4.60%

 

Out of 1410 patients, 48 patients (3.4%) died due to physical illness (n=34 cases, 70.8%), accidents (n=1 case, 1.2%), and unspecified causes (n=13 cases, 27.1%). 
The results of one-way ANOVA showed that the mean LOS was significantly longer in the patients with psychotic disorders than in those with the mood and substance use disorders. These results also indicated that the mean LOS of the patients with mood disorders (p=0.001, F=20.22) and those with substance use disorder (p=0.001, F=13.52) in Iran hospital was significantly longer than those who were admitted to Rasoul-e-Akram Hospital (Table 4).

 

Table 4. Total patients and their length of stay(Days)

 

Number

Iran hospital

Rasoul-e-Akram

hospital

Total patients

p-value

Psychotic disorders

409

33.11 (±18.61)

23.77 (±13.47)

90.31 (±18.28)

0.06

Mood disorders

588

26.20 (±15.15)

20.58 (±10.06)

53.24 (±14.06)

0.001

Anxiety disorders

26

25.84 (±13.35)

20.00 (±11.25)

27.24 (±12.87)

0.31

Substance use disorders

326

27.12 (±14.70)

15.50 (±6.39)

33.26 (±14.58)

0.001

Others

61

18.22 (±9.98)

14.52 (±7.01)

70.16 (±9/01)

0.11

 

A total of 335 patients (23.8%) were readmitted (Table 5). Accordingly, there was a significant difference between different disorders and the number of the readmitted patients. It should be noted that among them, the highest and lowest readmission rates were seen in the patients with mood and anxiety disorders, respectively (p=0.014 and χ2=12.43).

 

Table 5. Percentage of readmission among the patients with major psychiatric disorders admitted to Iran and Rasoul-e-Akram psychiatric hospitals in 2011

 

 

Readmission

p-value

Yes

No

Psychotic disorders

Number

117

292

0.014

Percent

34.9

27.2

Mood disorders

Number

141

447

Percent

42.1

41.6

Anxiety disorders

Number

7

19

Percent

2.1

1.8

Substance use disorders

Number

58

268

Percent

17.3

24.9

Others

Number

12

49

Percent

3.6

4.6

Total

Number

335

1075

Percent

100

100

 

Among the readmitted patients, the mean/SD of the time interval between the two admission (mean time to readmission) was 24.94/23.19 months. A significant difference was observed between the two hospitals, in terms of the mean time to readmissions, in the patients with mood disorders (p=0.14), which was longer in Iran hospital compared to the other one. However, there was no significant difference in terms of the other disorders.
Logistic regression was used to investigate the relationship between demographic/clinical variables and readmission. At the first stage, sex, initial age of onset, age at the disease’s time, educational level, marriage status, length of hospital stay, number of previous hospitalizations, employment, history of substance use disorders, history of suicide attempt, comorbid physical illness, therapeutic compliance, and hospitalization physical restraint were put in the regression model. Subsequently, the results showed that longer LOS, frequent previous admissions, the presence of physical restraint, and lack of therapeutic compliance (do not take medications regularly) could increase the probability of readmission (p=0.001, p=0.002, p=0.001, and p=0.004, respectively). These factors were also examined by chi-square test. The results represented a significant relationship between the hospitals in question and readmission (p=0.001 and χ2=14.09). Among the readmitted patients, 87.2% were admitted to Iran hospital and 12.8% to Rasoul-e-Akram general Hospital. Testing the relationship between the route of drug administration and readmission by chi-square test showed that there was no significant relationship among the 4 groups of the route of drug administration (only oral, oral medicine and ECT and depot, oral and depot, oral and ECT) (p=0.091, df=3, and χ2=45.6).
Chi-square test was utilized to evaluate the relationship between the prescribed drugs and ECT with readmission. For this purpose, medications were divided into 15 groups as follows: atypical antipsychotics, typical antipsychotics, mood stabilizers, antidepressants, depot, buspirone, z-drugs, benzodiazepine, anticholinergic, anticonvulsant, B blocker, antihistamine, levothyroxine, ECT, and others. Among the prescribed drugs, depot and anticholinergics were found to be significantly associated with decrease readmission (p=0.03 and χ2=4.27) (p=0.001 and χ2=10.28). The patients receiving depot had no readmission in 70.1% of the cases and the patients receiving anticholinergics had no readmission in 73.5% of the cases.
According to the survival analysis, half of the patients were readmitted within 101 months after being discharged. According to the Kaplan-Meier test, a longer time interval between admissions was seen among the married patients. In contrast, earlier admission was seen among the unemployed patients, patients who were admitted to Iran hospital, patients with psychotic disorders, and patients with inadequate medication compliance (Figure 1).
The results of Cox-Regression analysis, regarding the time of readmission, showed a relationship among factors such as medication compliance, type of psychiatric disorder, employment, and the hospital in question with the time of readmission.  Medication non-compliance, admission in Iran hospital, unemployment, and psychotic disorders were reported as the risk factors for earlier admission.

Discussion 
The readmission of psychiatric patients is currently an important issue since it imposes a huge burden on both patients and society, indicating the importance of research in this field. The present study indicated that clinical factors such as hospital stay length, the number of previous admissions, and physical restraint during hospitalization significantly increase the readmission risk. In other words, longer hospital stay, frequent previous admissions, and physical restraint during hospitalization as well as therapeutic non-compliance increase the readmission risk, and there was no significant relationship between other variables and readmission.
The present study, consistent with Kent et al (23) and the study by Bernardo et al (15), who examined the relationship among different readmission factors, showed that therapeutic non-compliance predicts readmission. This relationship was also shown in the study by Li et al (24) and studies by Crosas (25) and Lien (26). In a retrospective cohort study on the readmission of psychiatric patients in Malawi, Barnett (27) also reported the effect of therapeutic non-compliance. It shows the importance of taking medications to control the disease and prevent a recurrence. 
Consistent with the present study, the positive history of previous admissions was reported in other studies such as Barekatain et al (12), Bernardo (15), Roque (10), and Donisi et al (18). In a study on patients discharged from a psychiatric hospital in Finland in 1990, Korkeila et al (28) showed the importance of previous admissions. They found that frequent prior admissions lead to readmission. Similarly, Donisi (18) and Bernardo (15) also referred to this factor as the most important risk factor of readmission. Frequent previous admissions indicate severe and possibly uncontrollable disease, which may lead to more admissions in the future and thus explain this relationship.
Although Bernardo (15) found a relationship among other factors such as male sex, younger age, unemployment, divorce, lower level of education, and drug abuse with readmission, Donisi et al (18) referred to old age, being married, and, less importantly, employment and higher education as possible factors which reduce the readmission risk. However, the number of previous admissions was the factor that was considered in all studies.
In their studies, Donisi et al (18) and Korkeila et al (28) indicated the effect of hospital length of stay on readmission rates. These two studies and the present study revealed that a longer hospital stay leads to a higher readmission rate. Therefore, it can be concluded that patients with long hospital stay are more likely to suffer from more severe form of disease, which requires longer hospital stays. As a result, the readmission risk is higher among these patients.
The present study showed a significant relationship between different disorders and readmission (p= 0.001). The highest readmission rate was reported among patients with mood disorders, psychotic disorders, and substance use disorders, respectively. However, the readmission risk was lower among the patients with other disorders and anxiety disorders. In a study, Barekatain et al (12) reported a relationship between Bipolar I Disorder (BID), Bipolar II Disorder (BIID), psychotic disorder, major depressive disorder, and anxiety disorder with readmission. Still, there was no relationship between cognitive disorder or substance use disorder with readmission. In Bernardo’s study (15), 200 patients who were discharged from a hospital in Canada in 1991 were randomly selected, and their readmission rates were evaluated. Among the readmitted patients, the most prevalent disorders included schizophrenia, followed by personality disorders, mood disorders, and schizoaffective disorder, respectively. This finding suggests that more severe disorders such as psychotic and mood disorders lead to more readmission rates than other milder disorders.
The present study also investigated the relationship among the variables of drug type and drug administration route with readmission. There was no significant difference among 4 drug categories: 
1. Oral only, 2. Oral, depot, ECT, 3. Oral and depot, 4. Oral and ECT. 
This finding is not consistent with Slade et al (21), which showed that ECT treatment reduced readmission rates. However, considering that all patients received different treatments in our study and, for example, did not receive only ECT treatment, it was not possible to calculate each treatment, and further studies are needed in the future.
However, there was a significant relationship among mood stabilizers, depot, and anticholinergic drugs with readmission. In another study, Pilon et al (29) found a significant relationship between paliperidone palmitate and readmission; that is, it reduced the readmission rate. Lien et al (26) found that second-generation antipsychotics reduced the readmission rates, which was not consistent with the present study.
The logrank test’s results showed a shorter time interval between discharge and readmission for widows and divorced and longer for married people, which may be because widows and divorced people do not receive adequate support and are more prone to recurrence. The results of Cox-Regression analysis, regarding the time of readmission, showed a shorter time interval between discharge and readmission for patients with therapeutic non-compliance, patients with psychotic disorder, unemployed patients, and patients who were admitted to Iran hospital. The shorter time interval between discharge and readmission, and the more admission rate of psychotic disorders in patients admitted to Iran hospital may be suggested that patients with more severe psychiatric disorders were admitted to Iran hospital since the mentioned hospital is a specialized psychiatric hospital.
Korkeila et al (28) found earlier readmission among those who had more than 30 days of hospitalization, among men compared to women, and like the present study, among those with psychotic disorders. Therefore, people with psychotic disorders need more attention due to the higher risk of recurrence.
Regarding the mentioned findings, we need a better understanding of the factors related to readmission to help better management and early intervention. Understanding these factors will decrease the number of readmissions and as a result it will decrease hospital costs imposed on the health system and patients.

Conclusion
This study’s results showed that longer hospital stay, frequent previous admissions, physical restraint during hospitalization, and therapeutic non-compliance could potentially increase the risk of readmission. Patients with psychotic disorders are also at high risk of readmission; therefore, the above issues should be focused on to reduce the readmission rates. Attempts should be made to perform psychoeducation for all patients, to explain the nature of disease, and focus on the importance of regular medications use to prevent recurrence. Also, education should be given to patients about identifying the warning signs that indicate the possibility of recurrence of the disease, especially more severe such as psychotic disorders, thus it helps patients to control their illness faster. It is also recommended to avoid physical restraint, a factor of readmission risk, during hospitalization as much as possible to reduce the readmission.
Other factors may also affect the readmission rate, such as compulsory hospital admission, financial status, types of medications, etc., which have not been studied in our study and are recommended to be studied in future studies.

Limitations
 By considering that the readmission of some patients could be in the other hospitals, where access to information was not practically possible, for example, in the case of admission in other cities, there was a restriction on the follow-up of all patients. However, as expected, the correspondence was made with Roozbeh hospital regarding the follow-up of patients’ readmission in those centers. There was also a similar restriction in cases where patients’ first visits were not recorded in the hospital file for reasons such as referral to the private office.
Another limitation was the access to complete information of patients in inpatient files, as there were incomplete files in one of the wards of Iran hospital. In these cases, attempts were made to complete the information by studying the nursing reports as much as possible. Another limitation was the wrong phone numbers or workplace phone numbers or the phone numbers of legal authorities registered in some files, which limited our access to information, although this limitation was eliminated to some extent by subsequent inquiries from other psychiatric centers. Another limitation was the problem of remembering the exact date of post-discharge visit or readmission to other centers by the patients or their companions, in which case the approximate date was recorded. In cases where patients recalled only the year of readmission, September 30 of the same year was recorded as the date of readmission. Since the Civil Registration Office of Tehran did not cooperate with us, we had no access to information on the deceased patients other than those died in the hospital.

Suggestions
In this study, the drugs were evaluated in groups; therefore, it is necessary to study the drugs separately in future studies to find their relationship with readmission.
Physical comorbidities were studied in general. This problem can be solved in a study that only examines the relationship between comorbidities and readmission.

Declaration
Ethics approval and consent to participate: Approved by IUMS Research Ethics Boards #IR.IUMS.FMD.REC.1397.259

Consent for publication 
Confirmed following IUMS REB requirements. 
Availability of data and material: the data for this study would be available in IUMS data center server.

Competing interests 
None.

Funding 
This study received a Mental Health Research Center research grant, Psychosocial Health Research Institute, Department of Psychiatry, School of Medicine, Iran University of Medical Sciences, Tehran, Iran (Grant #9511232692). There is no other financial support.

Acknowledgements 
The authors would like to express their gratitude to the management of Rasoul-e-Akram hospital and Iran psychiatric hospital for their cooperation and assistance in providing the information.

  1. Mohammadi MR, Davidian H, Noorbala AA, Malekafzali H, Naghavi HR, Pouretemad HR, et al. An epidemiological survey of psychiatric disorders in Iran. Clin Pract Epidemiol Ment Health 2005 Sep 26;1:16. https://pubmed.ncbi.nlm.nih.gov/16185355/
  2. Farhoudian A, Sharifi V, Amini H, Basirnia A, Mesgarpour B, Mansouri N, et al. Prevalence of psychiatric disorders in Iran: a systematic review. Iran J Psychiatry 2007;2:137-50.
  3. Naghavi M, Abolhassani F, Pourmalek F, Lakeh M, Jafari N, Vaseghi S, et al. The burden of disease and injury in Iran 2003. Popul Health Metr 2009 Jun 15;7:9. https://pubmed.ncbi.nlm.nih.gov/19527516/
  4. Bryson KK, Naqvi A, Callahan P, Fontenot D. Brief admission program: an alliance of inpatient care and outpatient case management. J Psychosoc Nurs Ment Health Serv 1990 Dec;28(12):19-23. https://pubmed.ncbi.nlm.nih.gov/2283597/
  5. Dott SG, Walling DP, Bishop SL, Bucy JE, Folkes CC. The efficacy of short-term treatment for improving quality of life: a pilot study. J Nerv Ment Dis 1996 Aug;184(8):507-9. https://pubmed.ncbi.nlm.nih.gov/8752082/
  6. Lyons JS, O'Mahoney MT, Miller SI, Neme J, Kabat J, Miller F. Predicting readmission to the psychiatric hospital in a managed care environment: implications for quality indicators. Am J Psychiatry 1997 Mar;154(3):337-40. https://pubmed.ncbi.nlm.nih.gov/9054780/
  7. Marshall M, Crowther R, Sledge WH, Rathbone J, Soares‐Weiser K. Day hospital versus admission for acute psychiatric disorders. Cochrane Database Syst Rev 2011 Dec 7;2011(12):CD004026. https://pubmed.ncbi.nlm.nih.gov/22161384/
  8. Laursen TM, Munk-Olsen T, Nordentoft M, Mortensen PB. Increased mortality among patients admitted with major psychiatric disorders: a register-based study comparing mortality in unipolar depressive disorder, bipolar affective disorder, schizoaffective disorder, and schizophrenia. J Clin Psychiatry 2007 Jun;68(6):899-907. https://pubmed.ncbi.nlm.nih.gov/17592915/
  9. Casey DE. Metabolic issues and cardiovascular disease in patients with psychiatric disorders. Am J Med 2005 Apr;118 Suppl 2:15S-22S. https://pubmed.ncbi.nlm.nih.gov/15903291/
  10. Roque AP, Findlay LJ, Okoli C, El-Mallakh P. Patient characteristics associated with inpatient psychiatric re-admissions and the utility of the READMIT clinical risk index. Issues Ment Health Nurs 2017 May;38(5):411-9. https://pubmed.ncbi.nlm.nih.gov/28448224/
  11. Machado V, Leonidas C, Santos MA, Souza J. Psychiatric readmission: an integrative review of the literature. Int Nurs Rev 2012 Dec;59(4):447-57. https://pubmed.ncbi.nlm.nih.gov/23134127/
  12. Barekatain M, Maracy MR, Hassannejad R, Hosseini R. Factors associated with readmission of patients at a university hospital psychiatric ward in Iran. Psychiatry J 2013;2013:685625. https://pubmed.ncbi.nlm.nih.gov/24236285/
  13. Durbin J, Lin E, Layne C, Teed M. Is readmission a valid indicator of the quality of inpatient psychiatric care? J Behav Health Serv Res 2007 Apr;34(2):137-50. https://pubmed.ncbi.nlm.nih.gov/17437186/

14.Vigod SN, Kurdyak PA, Dennis CL, Leszcz T, Taylor VH, Blumberger DM, et al. Transitional interventions to reduce early psychiatric readmissions in adults: systematic review. Br J Psychiatry Br J Psychiatry 2013 Mar;202(3):187-94. https://pubmed.ncbi.nlm.nih.gov/23457182/

  1. Bernardo AC, Forchuk C. Factors associated with readmission to a psychiatric facility. Psychiatr Serv 2001 Aug;52(8):1100-2. https://pubmed.ncbi.nlm.nih.gov/11474059/
  2. Saleh ES, El-Hadidy MA. Predictors of psychiatric early readmission in two Arab hospitals. Arab J Psychiatry 2012;23:12-21.
  3. Evans LJ, Harris V, Newman L, Beck A. Rapid and frequent psychiatric readmissions: associated factors. Int J Psychiatry Clin Pract 2017 Nov;21(4):271-6. https://pubmed.ncbi.nlm.nih.gov/28554237/
  4. Donisi V, Tedeschi F, Salazzari D, Amaddeo F. Pre-and post-discharge factors influencing early readmission to acute psychiatric wards: implications for quality-of-care indicators in psychiatry. Gen Hosp Psychiatry 2016 Mar-Apr;39:53-8. https://pubmed.ncbi.nlm.nih.gov/26804775/
  5. van Alphen NR, Stewart JG, Esposito EC, Pridgen B, Gold J, Auerbach RP. Predictors of rehospitalization for depressed adolescents admitted to acute psychiatric treatment. J Clin Psychiatry 2017 May;78(5):592-8. https://pubmed.ncbi.nlm.nih.gov/27529444/
  6. Cheng MY. Using king's goal attainment theory to facilitate drug compliance in a psychiatric patient. Hu Li Za Zhi 2006 Jun;53(3):90-7. https://pubmed.ncbi.nlm.nih.gov/16767629/ Chinese
  7. Slade EP, Jahn DR, Regenold WT, Case BG. Association of electroconvulsive therapy with psychiatric readmissions in US hospitals. JAMA Psychiatry 2017 Aug 1;74(8):798-804. https://pubmed.ncbi.nlm.nih.gov/28658489/
  8. Sfetcu R, Musat S, Haaramo P, Ciutan M, Scintee G, Vladescu C, et al. Overview of post-discharge predictors for psychiatric re-hospitalisations: a systematic review of the literature. BMC Psychiatry 2017 Jun 24;17(1):227. https://pubmed.ncbi.nlm.nih.gov/28646857/
  9. Kent S, Yellowlees P. Psychiatric and social reasons for frequent rehospitalization. Hosp Community Psychiatry 1994 Apr;45(4):347-50. https://pubmed.ncbi.nlm.nih.gov/8020919/
  10. Li C, Chen C, Qiu B, Yang G. A 2-year follow-up study of discharged psychiatric patients with bipolar disorder. Psychiatry Res 2014 Aug 15;218(1-2):75-8. https://pubmed.ncbi.nlm.nih.gov/24794029/
  11. Crosas JM, Cobo J, Ahuir M, Hernández C, García R, Pousa E, et al. Substance abuse and gender differences in first episode psychosis: impact on hospital readmissions. Rev Psiquiatr Salud Ment (Engl Ed) 2018 Jan-Mar;11(1):27-35. https://pubmed.ncbi.nlm.nih.gov/28648845/ Spanish.
  12. Lien L. Are readmission rates influenced by how psychiatric services are organized? Nordic J Psychiatry 2002 Jan 1;56(1):23-8. https://pubmed.ncbi.nlm.nih.gov/11869461/
  13. Barnett BS, Kusunzi V, Magola L, Borba CP, Udedi M, Kulisewa K, et al. Risk factors for readmission among a cohort of psychiatric inpatients in Lilongwe, Malawi. Int J Psychiatry Clin Pract 2020 Mar;24(1):25-30. https://pubmed.ncbi.nlm.nih.gov/31799886/
  14. Korkeila JA, Lehtinen V, Tuori T, Helenius H. Frequently hospitalised psychiatric patients: a study of predictive factors. Soc Psychiatry Psychiatr Epidemiol 1998 Nov;33(11):528-34. https://pubmed.ncbi.nlm.nih.gov/9803820/
  15. Pilon D, Amos TB, Kamstra R, Manceur AM, El Khoury AC, Lefebvre P. Short-term rehospitalizations in young adults with schizophrenia treated with once-monthly paliperidone palmitate or oral atypical antipsychotics: a retrospective analysis. Curr Med Res Opin 2019 Jan;35(1):41-9. https://pubmed.ncbi.nlm.nih.gov/30106313/