Article, Psychiatry

Effects of seasonality and daylight savings time on emergency department visits for mental health disorders

a b s t r a c t

Objectives: Emergency Department (ED) utilization accounts for a large portion of healthcare services in the US. Disturbance of circadian rhythms may affect mental and behavioral health (MBH) conditions, which could result in increased ED visits and subsequent hospitalizations, thus potentially inducing staffing shortages and increas- ing ED wait time. Predicting the burden of ED admissions helps to better plan care at the EDs and provides sig- nificant benefits. This study investigates if increased ED visits for MBH conditions are associated with seasonality and changes in daylight savings time.

Methods: Using ED Encounter data from a large academic medical center, we have examined univariate and mul- tivariate associations between ED visits for MBH conditions and the annual time periods during which MBH con- ditions are more elevated due to changes in the seasons. We hypothesize that ED visits for MBH conditions increase within the 2-week period following the daylight savings time changes.

Results: Increased MBH ED visits were observed in certain seasons. This was especially true for non-bipolar de- pressive illness. We saw no significant changes in MBH visits as associated with changes in the daylight savings time.

Conclusions: Data do not provide conclusive evidence of a uniform seasonal increase in ED visits for MBH condi- tions. Variation in ED MBH visits may be due to secular trends, such as socioeconomic factors. Future research should explore contemporaneous associations between time-driven events and MBH ED visits. It will allow for greater understanding of challenges regarding Psychiatric patients and opportunities for improvement.

(C) 2018

Introduction

Background

Mental health is one of the major challenges faced in the United States. Nearly one in five adults currently lives with a Mental illness (44.7 million in 2016) [1] and emergency department (ED) visits for mental and Substance use disorders have been increasing annually [2] making it one of the fastest growing components of emergency medical practice [3]. During 2006-2013, ED visits for depression, anxiety or stress reactions have increased by 56% and for psychoses or bipolar dis- orders by 52% [4]. ED visits in the US for behavioral and mental health (MBH) disorders (ICD9 codes 290-319) have increased from 3.9% (5.3 million) in 2011 to 4.1% (5.7 million) in 2015 [5,6]. Moreover, between 2006 and 2014, total number of ED visits in the US has increased by 15% and the rate of visits per 100,000 persons has increased by 7% [7], which

* Corresponding author.

E-mail addresses: [email protected] (V. Heboyan), [email protected] (S. Stevens), [email protected] (W.V. McCall).

not only imposes high economic costs and hardship on self-paying indi- viduals and insurance companies, but also stretches ED resources. Predicting the burden of ED admissions over time could contribute to better planning of care at the EDs and thus provide significant benefits. MBH patients can consume a disproportionate share of ED resources by increasing length of stay pending transfer to a psychiatric facility or mental health clearance. Between 1992 and 2001, the number of docu- mented mental health-related ED visits increased by 38%, compared to an 8% increase in overall ED usage [8]. Since 1950s the number of psy- chiatric Hospital beds has dropped by over 96% and N17% since 2010 [9]. As a result, Boarding times for psychiatric patients has been signifi- cantly longer averaging over 6.8 h compared to non-psychiatric patients [9,10] and these patients have been reported to experience extremely long length of stay compared to non-MBH patients [11]. Boarding times for psychiatric patients in Georgia EDs has been found to be as high as 34 h [10]. Furthermore, according to Zeller, chief of psychiatric emergency services at Alameda Health Systems in Oakland, CA, “Boarding psychiatric patients in an emergency department is both poor medicine and expensive“, resulting in over $2000 in additional costs for treating psychiatric patients at the EDs [9,10]. As a result, these delays

https://doi.org/10.1016/j.ajem.2018.10.056

0735-6757/(C) 2018

can induce staffing shortage, increased ED wait time, and even diver- sion. Long length of stay for mental health patients can also result in in- creased threats to provider safety due to agitation and anxiety, as well as reduced access to coping mechanisms and possibly delayed medications for psychosis [11]. It not only compromises the level of psychiatric care required for these patients while being boarded in the ED, but also re- sults in delays and diverts ED resources for other ED patients due to overcrowding.

Changes in seasonal and circadian patterns are pathogenic factors for affective disorders [12], which manifest at higher rates and intensities during certain times throughout the year. Seasonality may have an ef- fect on MBH ED visits due to seasonal characteristics associated with many mental and behavioral health disorders. Literature suggests that seasons influence the levels of mental and behavioral symptoms and a significant portion of such patients experience seasonal vulnerability [13,14]. Furthermore, daylight savings time change has been found to affect the photoperiod and circadian rhythm [12] and, as a result, dis- rupt sleep patterns [15,16], influence mood [16,17], and even result in traffic accidents [18-20]. Hansen et al. [21] found that in Denmark tran- sition from summer time to standard time during 1995-2012 was asso- ciated with an 11% increase in the incidence rate of acute hospital contacts for unipolar depressive episodes. Abreu [22] suggests that the literature has shown that as a result of changes in sleep duration, a large number of patients experience mood changes on the same day or the day after. Also, sleep duration decrease was found to be followed by a change in the direction of hypomania or mania and an increase of sleep duration was followed by a shift towards depression. Theoreti- cally, a bipolar patient who experiences less sleep during a spring time change would more likely experience a manic episode while the same patient who would experience more sleep during the fall time change would experience a depressive episode [23]. These exacerbate symp- toms that could lead to a mental health crisis. Such manifestations have the potential to significantly disrupt healthcare delivery opera- tions including that of EDs by significantly increasing the volume of visits related to affective disorders and resources needed to service such patients. Understanding how seasonality and changes in daylight saving times impact ED visits for MBH patients will enable healthcare decision-makers to develop appropriate strategies to improve quality of care for these patients.

Importance

Impact of seasonality and daylight savings time on triggering MBH conditions has been studied extensively. Several studies have also ex- plored the relationship between multiple hospital outcomes, such as length of stay, and MBH visits. However, there is a gap in the literature examining the relationship between ED MBH visits and seasonality and daylight savings time. We are studying whether there is MBH visit variation due to theoretical desynchronization of circadian rhythms caused by daylight savings time and seasonality. This study helps to fill this gap by exploring the statistical impact of seasonality and day- light savings time on ED MBH visits using a multi-year robust and rich dataset from a major southeastern hospital. Results of this study are of importance in that they can be used by healthcare providers to better serve patients likely to be affected by seasonality and daylight savings time changes. These results are also of importance to the healthcare ad- ministrators to better plan ED operations, allocate Healthcare resources, and design alternative healthcare service delivery methods.

Goals of this investigation

This study focuses on affective disorders of bipolar and non-bipolar Mood disorders. Its objective is to estimate the impact of seasonality and daylight savings time on ED visits for MBH conditions. We are ex- ploring if EDs experience statistically significantly higher visits for MBH conditions during the time (season) identified in the literature as

the annual time periods during which MBH conditions are more ele- vated due to changes in the seasons. Furthermore, we hypothesize that ED visits for MBH conditions increase during the time period imme- diately following the spring and fall daylight savings time changes.

Methods

Data collection

We use secondary de-identified data from the Augusta University Medical Center (AUMC) ED located in Augusta, GA at a latitude of 33. AUMC ED is the principal provider for emergency mental health services in catchment area of approximately 750,000 persons and serves both an urban and rural population. Data represents each ED adult (18 years of age or older) visit from January 2013 to October 2015. This study has been reviewed and approved by the Augusta University Institutional Review Board (ID 1255557-1).

Primary variables

MBH visit outcome variables

We construct MBH outcome variables in two ways. First we generate a dichotomous MBH variable (icd_mental), which takes on a value of one if a patient visits the ED for a MBH condition as classified in accor- dance with ICD-9 code ranges 290-319. Then we categorize MBH condi- tions into multiple categories. These categories are manic disorders (ICD9 codes 296.00-296.06, 296.10-296.16, 296.40-296.46, 296.50-296.56, 296.60-296.66, 296.7, 296.80-296.89), depressive disor- ders (ICD9 codes 296.20-296.26, 296.30-296.36, 311.00), dementia,

drugs, and schizophrenia (ICD9 codes 290.00-290.39, 290.40-295.99, 303.00-303.93, 304.00-305.96) and other mental health conditions (ICD9 codes 296.90-296.99, 297.00-302.99, 306.00-310.99,

312.00-319.99).

The diagnosis of bipolar depression could be categorized either pri- marily on the basis of its bipolar features or on the basis of its depressive features. Therefore, we conducted our analyses separately under both assumptions by considering ICD9 diagnoses 296.1 (manic disorder re- current episode), 296.3 (major Depressive disorder recurrent episode), and 296.5 (bipolar I disorder, most recent episode (or current) de- pressed) as ‘manic disorders and ‘depressive disorders’.

Seasonality variable

We have reviewed the literature [13,14] to identify the seasons when MBH conditions are especially elevated. Based on earlier findings we constructed categorical seasonality variable (season) with the fol- lowing categories: November-March is classified as Fall/Winter, April- July is classified as Spring/Summer and August-October is classified as off-season.

Daylight savings time variable

During 2013-2015, the United States has observed the following daylight savings times: March 10-November 3, 2013, March 9- November 2, 2014, and March 8-November 1, 2015 [24]. We have con- structed a categorical variable that corresponds to the 2-week after the beginning date and 2-weeks after the end date of each year in our sam- ple. Remaining days were groups into the ‘non-season’ category. In total we had 6 categories: 2013 spring, 2013 fall, 2014 spring, 2014 fall, 2015 spring, and off-season. We hypothesize that MBH conditions should el- evate in the aftermath of time changes and will manifest within the two weeks of the time change [19]. We also constructed another categorical variable that corresponds to 2-weeks before and after these daylight savings time beginning and end dates. Here we explore if MBH condi- tions start to manifest as daylight savings time nears and afterwards.

We also control for patients gender (male, female), race (white, black, other), age (18-24, 25-30, 31-40, 41-50, 51-65, and over 65), and health insurance status (private, public, self-pay, and other).

Empirical analyses

To examine univariate and multivariate associations between sea- sonality and daylight savings time and ED MBH visits, we first examine the statistical relationship between MBH visits and time periods of in- terest (i.e. seasons, daylight savings time). We use Pearson’s Chi- squared or Fisher’s Exact statistics to explore if statistical differences exist among categories.

Then, we constructed multivariate regression models to estimate the impact of time periods of interest and socio-demographic factors on MBH visits. The model for the dichotomous MBH variable is estimated using a logistic regression and odds ratios are reported. The model for categorical MBH variable is estimated using a multinomial logistic re- gression and the Relative Risk Ratio (RRR) is reported. RRR is similar to the Odds Ratio and indicates how the risk of the outcome falling in the comparison group compared to the risk of the outcome falling in the referent group changes with the variable in question. All analyses were conducted using STATA 14.

Results

Study cohort description

Throughout the study period, January 2013 to October 2015, a total of 139,598 encounters were registered, of which, 6932 (5%) were for mental and behavioral health conditions (ICD9 codes 296-319). About one fifth of these visits were for manic (6.8%, n = 472) and depressive disorders (13.7%, n = 951), while rest were for dementia/drugs/schizo- phrenia (37.9%, n = 2626) and other Mental disorders (41.6%, n = 2883). Majority of MBH patients were male (54%), white (51%), and al- most equally distributed across age ranges.

Main results

Univariate associations“>Univariate associations

Table 1 summarizes univariate associations between ED visit status (mental and behavioral health (MBH) vs. non-MBH) by major demo- graphics. Results indicate that MBH and non-MBH visits rates were sta- tistically different for all demographics. Compared with non-MBH visits,

men (59%), whites (25-40%), 18-24 year olds, and those with public in- surance or self-paying had higher odds of visiting the ED for a mental or behavioral health diagnosis. Furthermore, MBH visits in triage 2 level (triage 1 is the most severe and triage 5 is the least life-threatening) were 20 times higher compared top non-MBH visits.

Additional analyses (not reported) found significant difference in visit rates during Winter 2013 and Spring/Summer 2013 seasons (iden- tified in the literature for elevated levels of MBH issues) compared to non-season visits. No significantly different associations were reported for other seasons/years. We did not find any significant associations re- lated to the daylight savings time (DST; 2 and 3 weeks after DST; 1 and

2 weeks before and after DST; 2 weeks before DST vs 2 weeks after DST;

3 weeks before DST vs 3 weeks after DST).

Table 2 reports distribution of ED visits for non-MBH conditions and separate MBH conditions during each season identified in the literature for elevated levels of MBH issues. Results show that the greatest season- to-season variability of visits are for depressive and dementia/drugs/ schizophrenia disorders. These two also exhibit annual downward trends, which is more prominent in depressive disorders. Test results also indicate significant differences in the groups when grouped by season-year but not in aggregate seasons. We also examined the associ- ations by DST variables (indicated with Table 1 discussion), but no sig- nificant associations were found. The top ten ICD9 conditions in the other mental disorder category were 300.9 (n = 936), 300.0 (n = 770),

298.9 (n = 277), 307.81 (n = 124), 296.9 (n = 97), 312.9 (n = 81),

310.2 (n = 72), 307.9 (n = 51), 300.01 (n = 49), and 301.3 (n = 49).

Multivariate associations

Table 3 reports partial results from multivariate regression models examining associations between MBH visits (binary dependent vari- able) and major demographics, seasonality, and DST variables. Using stepwise regression approach, we estimated 4 models. Model 1 (not re- ported) includes only major demographic variables. Consistent with re- sults from Table 1, results in this model indicate that men, whites, 18-24 year olds, and those with Public insurance or self-paying had higher odds of visiting ED for MBH diagnosis. Model 4 (not reported) in- cluded a variable controlling for the effect of ‘2 weeks before and after DST‘. It did not improve the model (Akaike Information Criterion test) and was not statistically significant.

Table 1

Descriptive statistics and univariate associations with mental and behavioral health (MBH) visits in the emergency department, N = 139,598.

Variables

Non-MBH visit

MBH visit

Pearson’s X2

Univariate odds ratio

95% CI

p-Value

N

%

N

%

p-value

Gender

Female

80,436

57.62

64,299

46.06

0.000

1.00

Male

59,162

42.38

75,299

53.94

1.59

1.52-1.67

0.000

Race

White

53,675

38.45

70,678

50.63

0.000

1.00

Black

80,604

57.74

63,643

45.59

0.60

0.57-0.63

0.000

Other

5,319

3.81

5,277

3.78

0.75

0.66-0.86

0.000

Age

18-24

19,432

13.92

26,928

19.29

0.000

1.00

25-30

22,908

16.41

22,210

15.91

0.70

0.65-0.76

0.000

31-40

29,483

21.12

30,251

21.67

0.74

0.69-0.80

0.000

41-50

23,369

16.74

22,489

16.11

0.69

0.64-0.75

0.000

51-65

30,321

21.72

28,157

20.17

0.67

0.62-0.72

0.000

Over 65

14,071

10.08

9,562

6.85

0.49

0.44-0.55

0.000

Triage

Level 1

3,085

2.21

852

0.61

0.000

1.00

Level 2

16,905

12.11

93,098

66.69

20.06

14.77-27.26

0.000

Level 3

64,913

46.50

34,941

25.03

1.96

1.44-2.67

0.000

Level 4

48,287

34.59

8,613

6.17

0.65

0.47-0.89

0.008

Level 5

6,422

4.60

2,094

1.50

1.19

0.83-1.71

0.342

Insurance

Private

32,582

23.34

25,700

18.41

0.000

1.00

Public

52,182

37.38

59,148

42.37

1.44

1.34-1.54

0.000

Self-pay

40,972

29.35

47,910

34.32

1.48

1.38-1.59

0.000

Other

13,862

9.93

6,840

4.90

0.63

0.55-0.71

0.000

Table 2

Univariate Associations with Mental and Behavioral Health Emergency Department Visits, N = 139,529; row percent.

Variables

Non-mental

Manic disorder

Depressive disorder

Other mental disorders

Dementia Drugs Schizophrenia

Pearson’s X2 p-value

Ratio of MBH to non-MBH visits

Seasonality by year

Non-season

95.10

0.33

0.74

1.93

1.89

0.051

2013 Winter

94.65

0.33

0.65

2.37

2.01

0.057

2013 Spring/Summer

94.53

0.46

0.90

2.01

2.10

0.058

2013/14 Fall/Winter

94.89

0.36

0.84

2.07

1.84

0.000

0.054

2014 Spring/Summer

95.20

0.28

0.66

2.05

1.81

0.050

2014/15 Fall/Winter

95.40

0.34

0.48

2.06

1.72

0.048

2015 Spring/Summer

95.27

0.27

0.44

2.17

1.86

0.050

Seasonality, aggregate Non-season

95.10

0.33

0.74

1.93

1.89

0.052

Fall/Winter

95.03

0.34

0.66

2.14

1.83

0.445

0.052

Spring/Summer

94.99

0.34

0.67

2.08

1.93

0.053

Reported results indicate that during Winter and Spring/Summer of 2013 there were statistically higher visits for MBH conditions compared to the off-season. Higher visits were also reported 2 weeks after the DST in the Spring and Fall of 2013. Data did not show any statistically signif- icant differences during 2014 or 2015 in both variables.

Table 4 reports partial results from multivariate multinomial logistic regression models associations between MBH visits (multi-categorical dependent variable) and major demographics and seasonality and DST variables. Results for each MBH disorder is compared to non-MBH visits (referent group). Using stepwise regression approach, we estimated 4 models. Model 1 (not reported) includes only major demographic vari- ables. Model 2 adds seasonality variable to the list of demographic var- iables described in Table 1, and Model 3 adds additionally a variable controlling for the possible effect of increased MBH visits 2 weeks after DST. Here we hypothesized that MBH visits may increase during the 2 week period after the DST time change. For the latter (not re- ported), we saw select statistically significant effects only for the other mental disorders (OMD) and dementia, drugs, and schizophrenia (DDS) visits compared to non-MBH visits. For example, in the OMD visits were statistically higher 2 weeks after 2013 Fall (RRR = 1.44) DST time change and were lower 2 weeks after 2013 Spring (RRR = 0.68) DST time change compared to non MBH visits. Moreover, DDS visits were statistically different only during 2014 Spring DST time change (RRR = 0.63).

Table 3 Multivariate associations with mental and behavioral health emergency department visits, odds ratios. Dependent variable: visit for MBH condition (binary).

Model 4 (not reported) included a variable controlling for the effect of ‘2 weeks before and after DST‘. It did not improve the model (AIC test) and was not statistically significant. Model 2 was preferred using the AIC.

Table 4

Multivariate associations with specific categories of mental and behavioral health emer- gency department visits, relative risk ratioa. Dependent variable: visit for different MBH conditions (multicategorical).

Variable Model 2 Model 3

Manic disorders

Seasonality (ref. off-season)

2013 Winter 1.011 1.064

2013 Spring/Summer 1.440?? 1.440??

2013/14 Fall/Winter 1.108 1.05

2014 Spring/Summer 0.862 0.862

2014/15 Fall/Winter

1.061

1.009

2015 Spring/Summer

0.811

0.811

Depressive disorders

Seasonality (ref. off-season)

2013 Winter

0.888

0.924

2013 Spring/Summer

1.235??

1.235??

2013/14 Fall/Winter

1.16

1.176

2014 Spring/Summer

0.898

0.898

2014/15 Fall/Winter

0.651???

0.656???

2015 Spring/Summer

0.589???

0.589???

Other mental disorders

Seasonality (ref. off-season)

2013 Winter 1.253??? 1.321???

Variable

Model 2

Model 3

2013 Spring/Summer

1.065

1.065

Odds ratio

Odds ratio

2013/14 Fall/Winter

1.087

1.048

Seasonality (ref. off-season)

2013 Winter 1.110??

1.148

???

2014 Spring/Summer 2014/15 Fall/Winter 2015 Spring/Summer

2013 Spring/Summer 1.126??? 1.126???

2013/14 Fall/Winter 1.056 1.046

2014 Spring/Summer 0.99? 0.99

Dementia, drugs, and schizophrenia

Seasonality (ref. off-season)

1.069

1.069

1.07

1.097

1.127?

1.127?

2

2014/15 Fall/Winter

0.952

0.958

2013 Winter

1.067

1.073

2015 Spring/Summer

0.974

0.974

2013 Spring/Summer

1.092

1.092

weeks after DST (ref. before DST)

2013/14 Fall/Winter

0.973

0.99

2013 Spring

0.796??

2014 Spring/Summer

0.966

0.966

2013 Fall

1.237??

2014/15 Fall/Winter

0.928

0.923

2014 Spring

0.878

2015 Spring/Summer

0.997

0.997

2014 Fall

0.917

N

139,529

139,529

2015 Spring 1.01

Constant 0.061??? 0.061???

N 139,598 139,598

2

Log likelihood -32,218.35 -32,200.04

AIC 64,580.701 64,584.08

* p b 0.1.

Pseudo R

0.028 0.028

?? p b 0.05.

Significance of model test (p) 0.000 0.000

Log likelihood -26,799.823 -26,794.482

AIC 53,635.645 53,634.963

* p b 0.1.

?? p b 0.05.

??? p b 0.01.

??? p b 0.01.

a This table reports the Relative Risk Ratio (RRR) from a multivariate logistic regression, which, similar to the Odds Ratio, indicates how the risk of the outcome falling in the com- parison group compared to the risk of the outcome falling in the referent group changes with the variable in question. Results for Each Disorder are compared to non-Mental and Behavioral Health (MBH) visits (referent group).

When evaluating Manic Disorder visits relative to non-MBH visits, we did not see any statistically significant differences due to DST (2 weeks before; 2 weeks before and after). Meanwhile, only Spring/ Summer 2013 visits were 44% statistically higher (RRR = 1.44) com- pared to visits during ‘off-season’ days.

Results for Depressive Disorders did not show any statistically sig- nificant differences due to DST (2 weeks before; 2 weeks before and after). However, there were statistically significant results due to sea- sonality. For example, visits were higher during 2013 Spring/Summer (RRR = 1.24) and lower during 2014/15 Fall/Winter (RRR = 0.65) and 2015 Spring/Summer (RRR = 0.59) seasons.

Visits for Other mental disorders were higher only during 2013 Winter (RRR = 1.25) and 2015 Spring/Summer (RRR = 1.13) seasons. There were no statistically significant seasonal difference for Dementia, drugs, and schizophrenia.

We have also conducted additional analyses (not reported). First, we re-coded the ICD9 diagnoses 296.1 (manic disorder recurrent episode),

296.3 (major depressive disorder recurrent episode), and 296.5 (bipolar I disorder, most recent episode (or current) depressed) as ‘depressive disorders‘, which were previously coded as manic disorders. However, our results did not change. Second, we have combined seasons together (i.e. instead of non-season and 6 year/seasons, we have used non- season, fall/winter, and spring/summer). Results did not show statisti- cally significant variations due to seasonality using the modified season- ality variable. Furthermore, the ratio of MBH visits to non-MBH visits (Table 2, last column) remains relatively stable during the seasons and non-season. We hypothesize that non-medical, perhaps secular, factors are influencing variations in ED visits overall.

Discussion

This study used a 3-year population based de-identified dataset to examine changes in ED MBH visits due to seasonal variations and changes in daylight savings time. There is currently a major public health challenge regarding visit of psychiatric patients in the ED due to historically increasing ED visits levels and significant decreases in psychiatric visit capacity of EDs in the United States. Better understand- ing the contemporaneous association between time-driven events (e.g. seasons and changes in daylight savings time) and variations in ED visits for MBH conditions will provide valuable insight to healthcare adminis- trators and policymakers to further improve healthcare service delivery for the psychiatric patients and improve overall healthcare manage- ment. To our knowledge, this study is the first to examine the associa- tion between seasonality and daylight savings time and ED visits for MBH conditions.

Our results indicate of the existence of significant variations in the ED MBH visit patterns during certain seasons, however these variations were not consistent. Our data is derived from a hospital in State of Georgia, which is located relatively closer to the equator (latitude of 33) with less dramatic year around changes in day light and seasons. Compared to the northern regions, such as the Scandinavia,1 where there are more dramatic changes in day lights and seasons, we would have expected much weaker or no effects in Georgia. Therefore, our findings become even more interesting and a future research is needed to further explore contemporaneous associations between time-driven events and MBH ED visits. It will allow for greater understanding of challenges regarding psychiatric patients and opportunities for improvement.

Future research should also explore the economic and financial im- plications of elevated MBH visits due to seasonality and daylight savings time to inform hospital administrators and healthcare providers of best strategies to meet the medical needs of such patients. Future research should be expanded to include other geographic areas and different

1 [25].

types of medical centers. For example, a comparative study of EDs ser- vicing urban and Rural patients will provide better policy advice.

There may be other potential drivers of ED visits that were not cap- tured in our dataset. For example, we may hypothesize that hotter tem- peratures may be conductive of harmful behaviors, such as gun violence, that will result in increased ED visits. Furthermore, increased ED visits may be driven in part by other environmental and behavioral factors, such as previous primary care provider visits/referrals, behav- ioral health problems for which they saw a psych/social worker, or inad- equate housing that may elevate such symptoms. Future research should incorporate these potential determinants to generate a more complete model of MBH ED visits.

Finally, the Affordable Care Act came into effect in late 2010, which aimed to increase the number of individuals with health insur- ance and the access to regular physician services that many lacked due to the absence of health insurance. Therefore, it is warranted to ex- amine the impact of the ACA on MBH visits. We can hypothesize that there should be a downward trend in ED visits for MBH conditions, since increased health insurance coverage should result in a better care for MBH conditions. Because this study utilizes data for 2013-2015, we cannot suggest any pre- and post-differences from this data explicitly. However, results reported in Table 2 indicate of an existence of downward trend for depressive and dementia/drugs/ schizophrenia disorders, which may be due to the continuation of the secular trend based on the ACA impact. Furthermore, in another study

[26] we have used ED data from 2009 to 2013 to examine trends in ED visits and average financial charges. Results showed that there is a downward trend in post-ACA period in the number of annual ED visits and an upward trend in average per visit financial charge in the post- ACA period. We assume that these trends are due to patients gaining more coverage in post-ACA period, which results in less ED visits, espe- cially for low acuity reasons. As for the average per visit financial charges, we think that due to increased insurance coverage, the ED visits are being dominated by high acuity emergency visits, which require more Hospital resources for treatment, and, hence, higher costs.

Conclusion

Results show that increased MBH ED visits are registered in certain seasons, however these patterns are not consistent across all years and seasons of this study, providing no conclusive evidence of increased ED visits for MBH conditions during seasons where MBH conditions are elevated. Furthermore, we saw no significant changes in MBH visits due to changes in the daylight savings time.

The year-to-year variation in seasonal effects for ED mood disorder visits may be related to secular trends, such as changing socioeconomic factors and increased access to health insurance in post-ACA period [27]. It is apparent that future research is required to further examine con- temporaneous associations between time-driven events and MBH ED visits and economic and financial implications for hospital management at the rural and urban health centers. From a policy and economic per- spective, there is a need to explore the implication of ACA on ED visits for MBH conditions and how such waves of increased MBH visits impact staffing decisions, especially by trained behavioral health professionals.

Limitations

This study was retrospective and restricted to single facility. This sample was unique to one county in the southeastern United States.

Previous presentation

This study has not been presented before.

Financial support

This study has not received financial support from any source.

Disclosures

The authors report no conflict of interests.

Author contributions statement

VH has obtained data from the Augusta University Medical Center. VH and SS conceived the study and designed and conducted data anal- yses. WM provided subject matter expertise on the development of the study and interpretation of the results. All authors contributed substan- tially to drafting and finalizing the manuscript. VH takes responsibility for the manuscript as a whole.

References

  1. National Institute of Mental Health. Mental illness. [cited 2018 June 20]; Available from: https://www.nimh.nih.gov/health/statistics/mental-illness.shtml; 2018.
  2. Owens PL, Mutter R, Stocks C. Mental health and substance abuse-related emer- gency department visits among adults, 2007. HCUP statistical brief #92. Rockville, MD: U.S. Agency for Healthcare Research and Quality; 2010.
  3. Larkin GL, et al. Mental health and emergency medicine: a research agenda. Acad Emerg Med 2009;16(11):1110-9.
  4. Weiss AP, et al. Patient- and practice-related determinants of emergency depart- ment length of stay for patients with psychiatric illness. Ann Emerg Med 2012;60

    (2) (p. 162-71.e5).

    Centers for Disease Control and Prevention (CDC). National Hospital Ambulatory Medical Care Survey: 2011 emergency department summary tables. Atlanta, GA: Centers for Disease Control and Prevention (CDC); 2011.

  5. Centers for Disease Control and Prevention (CDC). National Hospital Ambulatory Medical Care Survey: 2015 emergency department summary tables. Atlanta, GA: Centers for Disease Control and Prevention (CDC); 2015.
  6. HCUPnet. Healthcare cost and utilization project. Rockville, MD: Agency for Healthcare Research and Quality; 2018.
  7. Stone A, et al. Impact of the mental healthcare delivery system on California emer- gency departments. West J Emerg Med 2012;13(1):51-6.
  8. American Psychiatric Association. Emergency room visits for mental health condi- tions: expect long waits. [cited 2018 June 21]; Available from: https://www.

    psychiatry.org/news-room/apa-blogs/apa-blog/2016/11/emergency-room-visits- for-mental-health-conditions-expect-long-waits; 2016.

    Zeller S, Calma N, Stone A. Effects of a dedicated regional psychiatric emergency ser- vice on boarding of psychiatric patients in area emergency departments. West J Emerg Med 2014;15(1):1-6.

  9. Stephens RJ, et al. Factors associated with longer length of stay for mental health emergency department patients. J Emerg Med 2014;47(4):412-9.
  10. Shapiro CM, et al. Daylight saving time in psychiatric illness. J Affect Disord 1990;19 (3):177-81.
  11. Geoffroy PA, et al. Seasonality and bipolar disorder: a systematic review, from ad- mission rates to seasonality of symptoms. J Affect Disord 2014;168:210-23.
  12. Rumble ME, et al. The relationship of person-specific eveningness chronotype, greater seasonality, and less rhythmicity to suicidal behavior: a literature review. J Affect Disord 2018;227:721-30.
  13. Monk TH, Folkard S. Adjusting to the changes to and from Daylight Saving Time. Na-

    ture 1976;261(5562):688-9.

    Monk TH, Aplin LC. Spring and autumn daylight saving time changes: studies of ad- justment in sleep timings, mood, and efficiency. Ergonomics 1980;23(2):167-78.

  14. Taub JM, Berger RJ. The effects of changing the phase and duration of sleep. J Exp Psychol Hum Percept Perform 1976;2(1):30-41.
  15. Monk TH. Traffic accident increases as a possible indicant of desynchronosis. Chronobiologia 1980;7(4):527-9.
  16. Hicks GJ, Davis JW, Hicks RA. Fatal alcohol-related traffic crashes increase subse- quent to changes to and from daylight savings time. Percept Mot Skills 1998;86(3 Pt 1):879-82.
  17. Varughese J, Allen RP. Fatal accidents following changes in daylight savings time: the American experience. Sleep Med 2001;2(1):31-6.
  18. Hansen BT, et al. Daylight savings time transitions and the incidence rate of unipolar depressive episodes. Epidemiology 2017;28(3):346-53.
  19. Abreu T, Braganca M. The bipolarity of light and dark: a review on Bipolar Disorder and circadian cycles. J Affect Disord 2015;185:219-29.
  20. Wehr TA, Sack DA, Rosenthal NE. Sleep reduction as a final common pathway in the genesis of mania. Am J Psychiatry 1987;144(2):201-4.
  21. Time and Date AS. Daylight savings time changes, New York. Available from: https://

    www.timeanddate.com/time/change/usa/new-york?year=2013; 2018 January 12.

    Roislien J, Sovik S, Eken T. Seasonality in trauma admissions – are daylight and weather variables better predictors than general cyclic effects? PLoS One 2018;13 (2):e0192568.

  22. Vernon M, et al. Costs and trends of emergency department utilization pre- and post-ACA: evidence from a rural Georgia hospital in American Public Health Associ- ation (APHA) annual meeting; 2017 (Atlanta, GA).
  23. Bin YS, Marshall NS, Glozier N. Secular trends in adult sleep duration: a systematic review. Sleep Med Rev 2012;16(3):223-30.

Leave a Reply

Your email address will not be published. Required fields are marked *