a b s t r a c t

Objective: Psychiatric Patient boarding in emergency department (ED) is a severe and growing problem. In July 2013, Minnesota implemented a law requiring jailed persons committed to state psychiatric facilities be trans- ferred within 48-h of commitment. This study aims to quantify the effect of this law on a large ED’s psychiatric patient flow.

Methods: A pre- and post- comparison of 2011-2015 ED length of stay for adult Psychiatric patients was performed using Electronic Medical Record data. Comparisons of the median LOS were assessed using a seg- mented regression model with time series error, and Risk differences (RD) were used to determine changes in the proportion of patients with LOS >=3 and >=5 days. Changes in patient disposition proportions were assessed using risk ratios.

Results: The median ED LOS for patients admitted for psychiatric care increased by 5.22 h from 2011 to 2015 (95% CI: (4.33, 7.15)), while the frequency of patient encounters remained constant. Although no significant difference in the rate of ED LOS increase was found pre- and post- implementation, the proportion of adults with LOS

>=3 days and >=15 days increased (RD 0.017 (95% CI: (0.013, 0.021)); 0.002 (95% CI: (0.001,0.004)), respectively). Conclusions: The proportion of ED adult psychiatric patients experiencing prolonged LOS increased following the implementation of a statewide law requiring patients committed through the criminal justice system be trans- ferred to a state psychiatric hospital within 48 h. Identifying characteristics of subsets of psychiatric patients disproportionally affected could suggest focused healthcare System improvements to improve ED psychiatric care.

(C) 2018

Introduction

In many hospitals, emergency department (ED) patients who re- quire inpatient psychiatric hospitalization have prolongED boarding times.[1] This can be challenging to ED operations, since timely disposi- tion of patients is necessary to provide effective care. Quality of care for ED patients is known to be compromised by longer Boarding times in

* Corresponding author at: Mayo Clinic, 200 1st St SW, Rochester, MN 55905, United States.

E-mail addresses: [email protected] (K.T. Maass), [email protected] (G.J. Melin), [email protected] (R.L. Campbell), [email protected]

(P.J. Novotny), [email protected] (J.J. Westphal), [email protected] (D.M. Nestler), [email protected] (K.S. Pasupathy).

the ED.[2-7] Several studies have described ED patient length of stay (LOS) as a function of patient arrival time, number of ED patients, med- ical acuity, hospital capacity, and emergency physician staffing.[1,8,9]

Studies have also shown increased operational burdens on EDs

when mental health resources are decreased.[10,11] To this point, a law in Minnesota known as the “48-h Rule” went into effect in July 2013 [12] and prioritizes state psychiatric hospital access to persons in jail over patients in other settings, such as other hospitals and healthcare institutions. Fig. 1 depicts patient movement throughout the mental health care system and displays the interactions between jailed and non-jailed patients who access such services. Whereas the law provides benefits for mental health patients in statewide jails, little is known about its potential consequences on the flow of mental health patients through EDs in Minnesota.

https://doi.org/10.1016/j.ajem.2018.03.016 0735-6757/(C) 2018

The state of Minnesota, like the United States as a whole, has sustained a precipitous decline in per capita psychiatric Hospital beds since the 1960s.[13,14] Minnesota is attempting to care for psychiatric patients in a community care model emphasizing patients’ rights to the least institutionalization necessary for their care.[15] Yet, much of the state is designated as a mental health professional shortage area.

[16] The state mental health care system suffers from higher demand than supply of multiple resources, including psychiatric acute care hos- pital beds, longer term state psychiatric hospital beds, housing for the mentally ill, and outpatient psychiatry.[15-18] Minnesota is far from unique; throughout the country, ED boarding of psychiatric patients is consuming ED resources and delaying mental health treatment.[19]

The objective of this study is to examine the effects of the 48-h Rule on ED psychiatric patient LOS. Understanding the effects of Minnesota state legislation on ED patient flow can benefit other states considering policy reform. Potential consequences include increased LOS regardless of patient volume and disproportionate effects on patients who can be admitted to state facilities but not private facilities.

Methods

Study design and setting

This is a retrospective cohort study of ED patients seen for behavioral health concerns between January 2011 and December 2015 at a major academic tertiary care center. This academic hospital’s ED has an annual volume of approximately 73,000 patient encounters. Overall, the admis- sion rate from the ED is approximately 30%, suggesting a high-acuity population. The ED has a four-bed area designed specifically for treating behavioral health patients, and staff use additional ED treatment rooms when behavioral health volume exceeds four simultaneous patients. A psychiatry resident staffs the ED 24 h a day, and has attending psychia- try physician support around the clock. Once the determination is made by a psychiatry physician that the patient requires inpatient manage- ment (whether voluntary or involuntary), a social worker assists with locating an inpatient bed for the behavioral health patient. The hospital campus has an associated psychiatric hospital, with 73 inpatient beds divided into a child and adolescent unit of 18 beds, an adult mood disor- ders unit of 16 beds, an acute adult unit of 25 beds, a medical psychiatry unit of 7 beds, and a geriatric unit of 7 beds.

Selection of patients

The specific population of interest includes adult patients seen spe- cifically for behavioral health concerns in the ED. Patients may register with any chief complaint, and may be given a psychiatric final diagnosis, but never undergo comprehensive psychiatric evaluation in the ED (ex- amples include severe overdose requiring medical management, acute intoxication, or a patient with a panic attack that is reassured and dismissed). For any patient to be admitted from the ED to the home institution’s psychiatric hospital or be transferred directly to another psychiatric facility, a Psychiatry consult must be performed in the ED. To best capture patients that underwent comprehensive psychiatric as- sessment in the ED, our cohort included all adult patients where a Psy- chiatry consult was performed while the patient was in the ED. Specifically, these patients had a Psychiatry consult documented in the electronic medical record (EMR) between the arrival time and depar- ture time of their ED stay. For patients who are admitted or transferred directly to another inpatient facility, ED staff enters the “decision to admit/transfer” time in the EMR when the decision is made.

As the 48-h Rule was implemented in July 2013, the pre-post cohorts included patients 30 months before and 30 months after the law’s im- plementation (January 1, 2011 to December 31, 2015). All adult ED pa- tients seen by psychiatry were considered for this study. The STROBE checklist for strengthening observational studies in epidemiology was observed in the reporting of results.

Data and analysis

Existing EMR data was queried, including patient age, date and time of ED arrival and departure, and final disposition. For each patient, his/ her final disposition was classified as: (1) discharge; (2) local psychiat- ric admission; (3) external transfer for psychiatric admission; or (4) other disposition. ED LOS was calculated by subtracting the date and time of patient arrival to the ED, from the date and time of patient de- parture from the ED. A segmented regression model with time series error of the lognormal transformation of the median monthly LOS was used to test whether the rate of change in LOS was affected by the im- plementation of the 48-h Rule. Changes in LOS were assessed for all ED adult psychiatric patients as well as for the subset of those patients requiring psychiatric admission.

Fig. 1. Patient flow diagram.

Fig. 2. ED psychiatric encounters pre- and post- intervention: pediatric and adult.

Fig. 3. ED adult psychiatric population: dispositions in pre- and post- intervention stages.

After observing a pattern of outliers in LOS post implementation of the 48-h Rule, risk differences (RD) were calculated to assess whether the proportion of patients experiencing extreme lengths of stay differed prior to and after implementation. Cutoff points for extreme lengths of stay were identified post-hoc as the maximum length of stay (in days) in 2011 and 2012 (rounded up): 3 days and 15 days, respectively. Individual record review was conducted to explore commonalities among patients with ED LOS exceeding 15 days.

The primary outcome of interest was median ED LOS for adult pa- tients with Psychiatry consults, before and after the 48-h Rule was implemented. A secondary outcome was the proportion of patients with protracted LOS (>=3 days and >=15 days) before and after implementation.

Results

Characteristics of lengths of stay in the emergency department

During the 5-year study period, a total of 14,689 Psychiatry consults were performed in the ED; 2856 pediatric patient encounters were ex- cluded from the study see Fig. 2. Of the 11,833 ED adult psychic encoun- ters, 6187 were performed prior to the 48-h Rule implementation, and 5646 were performed after. Most ED adult psychiatric consults had a final disposition of “local admit” (43.7%) or “discharge” (38.9%), while consults requiring “external transfer” (7.6%) or having “other disposi- tion” (9.8%) were less frequent see Fig. 3. Consults classified as “other disposition” included those with a lack of sufficient information in the disposition field or dispositions not determined by psychiatric status, such as medical hospitalizations or detoxification programs which

may or may not be followed by psychiatric hospitalizations. Detailed disposition information for consults classified as “other disposition” is listed in Table 1.

The proportion of adult psychiatric patients presenting to the ED that were discharged (Risk Ratio (RR) 0.89; 95% CI: (0.85, 0.93)) or transferred to another facility (RR 0.82; CI; (0.72, 0.93)) decreased after July 2013; the proportion admitted locally remained constant (RR 1.03; 95% CI: (0.99, 1.08)) and other dispositions increased (RR

1.64; 95% CI: (1.47, 1.84)) see Table 2.

Overall, ED adult psychiatric patients had a median LOS of 4.1 h (95% CI: (4.0, 4.1)) pre- 48-h Rule and 5.2 h (95% CI: (5.1, 5.3)) post- 48-h

Rule. Median LOS by disposition type is reported in Table 2 and in- creases each year for all Patient dispositions. Fig. 4 shows the yearly

Table 1

Dispositions classified as “Other Disposition”.

Disposition Number of patient encounters

Medical hospitalization 451

Detoxification programs 209

“To be determined” 157

Psychiatric hospital waiting list 95

law enforcement custody 68

Skilled nursing facility 64

Inpatient rehabilitation program 56

Documented as unknown 22

Left without being seen 22

Not documented 8

Hospice 2

ED observation unit 1

Table 2

ED adult psychiatric patient volumes and LOS.

2011	2012	2013	2014	2015	Pre	Post
Unique ED adult psychiatric 1931 patients	1850	1808	1733	1749	4176	3894

similar conclusion was found when the analysis was restricted to pa- tients who were psychiatrically admitted (slope change 0.83; 95% CI: (-0.16, 1.83)).

While there were no significant differences in the median LOS pre- and post- 48-h Rule implementation, the proportion of patients

Mean visits per patient	1.3	1.3	1.3	1.3	1.3	1.5	1.4	experiencing LOS >=3 days and >=15 days increased (>=3: RD 0.017, 95%
Total ED adult psychiatric	2548	2420	2343	2271	2251	6187	5646	CI: (0.013,0.021); >=15: 0.002, 95% CI: (0.001, 0.004)) see Table 3. Review
encounters Patient encounters (by disposition) Discharge	998	1051	921	794	842	2548	2058	of patient charts for patient encounters with protracted LOS >= 15 days revealed that, of thirteen such stays, twelve were by patients with a his- tory of violent crime, and the remaining one was by a patient with a his-
External transfer	221	186	183	181	128	514	385	tory of an unspecified sex offense.
Local admit	1108	1019	1085	988	973	2664	2509

Limitations

Other	221	164	154	308	308	461	694
Median patient LOS (hours; by
disposition)
Discharge	3.5	3.7	4.0	4.2	4.5	3.7	4.2
External transfer	9.4	10.6	12.2	13.4	18.6	10.3	14.4
Local admit	3.5	3.9	4.2	4.8	7.7	3.8	5.1
Other	4.8	5.4	5.9	6.4	8.4	5.2	7.2

total adult psychiatry patient hours, mean LOS in hours, and the number of adult Psychiatry encounters over the five-year study period.

Median monthly LOS by for all ED adult psychiatric patients over the five-year period is shown in Fig. 5. A segmented regression model with time series error demonstrates there was no significant change in the rate at which median monthly LOS increased pre- and post- implemen- tation of the 48-h Rule (slope change 0.51; 95% CI: (-0.14, 1.11)). A

The observational nature of this study does not allow us to isolate causal factors in changes of ED LOS. The history of violence common to the patients with the longest LOS was discovered in post-hoc review and couldn’t be treated statistically because obtaining criminal back- ground information on the entire population of interest was not practi- cable. A dedicated follow-up study is planned to assess whether patients with criminal histories experience significantly longer ED LOS than pa- tients without such histories.

Generalizability of results is limited, as this is a single site study. Trends could be different for EDs not associated with psychiatric hospi- tals, those not affiliated with academic institutions, or those in different states or geographical regions.

Fig. 4. ED LOS, total patient hours, and encounters of adult psychiatric patients.

Fig. 5. Monthly median LOS segmented regression model with time series error.

Conclusions that can be drawn from comparing LOS before and after the implementation of the July 2013 law are limited by the presence of unknown confounding factors. Annual LOS increased throughout the study period before and after implementation of the law for reasons that can’t be fully ascertained with the available data. Such factors in- clude, but are not limited to, changes in ED procedures and staffing, changes in psychiatric resources at any level of care, changes in the ap- plications of psychiatric commitments, changes to social safety net pro- grams and demographic changes. One external factor that is known is that state hospital LOS increased steadily throughout the period under investigation.[20] The overall trend towards greater LOS later in the study period biases our study away from the null hypothesis that the 48 h rule did not affect LOS.

Discussion and conclusion

In this study of psychiatric patients in a large academic ED, imple- mentation of a law prioritizing jailed patients over ED patients was as- sociated with significant increases in psychiatric patient LOS in the ED. While the overall number of psychiatric patients in our ED did not in- crease in the 30 months after the law’s implementation and no signifi- cant difference in the rate of increase in the median LOS for adult psychiatric patients was identified, the proportion of patient encounters resulting in extremely long LOS increased after implementation of the 48-h Rule.

To understand why patients were in the ED for days or weeks at a time, we evaluated the medical records of all encounters where the length of stay was greater than fifteen days. Before the law’s implemen- tation, no patient was in the ED for this long. After implementation, 13

Table 3

Protracted lengths of stay.

patients stayed in the ED for more than two weeks. Of these Prolonged ED stays, 12 were by patients with a history of violent crime, and the thirteenth was by a patient with a history of an unspecified sexual of- fense. Patients with such criminal history are typically refused admis- sion to private psychiatric hospital beds, but are not excluded from state hospital beds. These patients accordingly had to stay in the ED until a direct transfer to a state psychiatric hospital bed was available. Analyzing the proportion of psychiatric patients with histories of violent crimes and/or sexual offenses that experience extended ED LOS is planned as a future study.

The causes of psychiatric hospital overflow are multifactorial and in- clude psychiatric hospital capacity (a function of the number of hospital beds and the average LOS in psychiatric hospitals), hospital exclusion criteria, access to outpatient psychiatric care, and the use of institutions that delay or prevent the Need for hospitalization (such as detoxifica- tion, chemical dependency programs, crisis beds, and group homes). Data obtained from correspondence with the Department of Health and Human Services, as seen in Table 4, shows increasing length of stay in Minnesota’s state psychiatric hospitals. Increasing psychiatric in- patient LOS causes decreased access to psychiatric beds for patients not already admitted. These data suggest that throughput and/or output factors (e.g. in-hospital factors and/or the community resources needed for an acceptable discharge plan) are decreasing the efficiency of hospital treatment, or that the Acuity of patients receiving inpatient treatment is increasing. This may suggest a root problem of which a shortage of psychiatric hospital beds is a downstream effect.

This study used a before-and-after design, and direct causation can- not be declared. However, our findings raise the question of whether improving access to inpatient psychiatric services for a subset of pa- tients has decreased access for other patients. Several factors, when con- sidered together, lead to this assertion. First, many private psychiatric

	Pre-48 hour rule: Jan. 2011-Jun. 2013	Post- 48 hour rule: Jul. 2013-Dec. 2015	Risk difference (95% CI)	hospitals will not accept patients with a violent crime history. In addi- tion, state psychiatric hospitals have long waiting lists. Increasing LOS in Minnesota’s state psychiatric hospitals see Table 4 implies decreased
LOS b 3 days	6177	5540	0.017 (0.013, 0.021)	bed availability for patients seeking inpatient services in a given amount
LOS >= 3 days	10	106		of time. Finally, when a state psychiatric hospital bed becomes available,

persons committed from jail are prioritized over committed patients in private psychiatric hospitals.

LOS b 15 days	6187	5633	0.002 (0.001, 0.004)
LOS >= 15 days	0	13

Table 4 Minnesota state psychiatric hospital average lengths of stay.
	FY2011	FY2012	FY2013	FY2014	FY2015
Anoka metro regional treatment center	83.8 days	93.8 days	105.4 days	110.5 days	126.1 days
Community behavioral health hospitals (Aggregated)	20.4 days	22.8 days	28.6 days	37.4 days	42.6 days

Linear regression of AMRTC LOS = 73.53 + 10.13?Year R^2 = 0.982. Linear regression of CBHH LOS = 12.66 + 5.9?Year R^2 = 0.970.

* Data provided by correspondence with Minnesota Department of Human Services [20] and reflects the state psychiatric hospitals in Minnesota except Minnesota Security Hospital for which LOS was unavailable.

The sum of these flow issues leads to several logical conclusions: (1) the responsibility of housing committed psychiatric patients has shifted from jails to state psychiatric hospitals; (2) private psychiatric hospitals will have a harder time placing patients in state psychiatric facilities, leading to longer ED LOS; (3) psychiatric patients presenting to the ED may have a long wait before finding a bed, and is exaggerated in pa- tients that can only be placed in a state psychiatric facility. Therefore, ED patients needing state psychiatric facility placement may wait sev- eral weeks before successful placement.

This study highlights the need to address an increasing problem of prolonged ED waits for patients needing psychiatric hospitalization, and suggests this may be especially true for patients with a history of vi- olence or otherwise meeting exclusion criteria from non-state hospitals. Many approaches to decreasing the need for hospital beds should be considered, including interventions to reduce the need for hospitaliza- tion, improve the efficacy of hospital care, and to open up the “back door” of the hospitals to more timely discharges. Additionally, EDs themselves can improve their ability to provide psychiatric care, poten- tially averting the need for many hospitalizations.[21]

At first glance it may appear that the solution to a shortage of psychi- atric hospital capacity would be to add more hospitals and more hospi- tal staff. However, this will not address the root of the problem. The inpatient mental health care system is backed up due to insufficient sub-acute to chronic care and other resources necessary to properly dis- charge patients, resulting in longer psychiatric hospital stays.[15-18] Hospitals are thereby rendered less efficient at stabilizing crises, and are increasingly becoming places where people with Mental illness spend time without receiving hospital level care, while others in need of those same services go without.

Minnesota’s 48-h Rule is intended to improve access to mental health services for jailed patients, but may have unintended conse- quences on access to care for other subsets of patients. Jails expedited the transfers of mentally ill people to hospitals in the absence of new Hospital resources to serve the increased demand. One consequence is that state psychiatric hospitals, private psychiatric and medical hospi- tals, and EDs took custody of mentally ill people in excess of their ability to provide treatment. Presently, the hospital medical system is making up for missing community resources at great monetary and human cost. A properly functioning healthcare system needs each department and each institution to do what it does best, and to maintain a capacity proportionate to the needs for its services. Further studies investigating the longitudinal needs of those who have been psychiatrically hospitalized, best practices in the rehabilitation and long term care of mentally ill people with and without criminal histories, and the barriers to psychiatric hospital discharge would be helpful.

Source of support

Robert D. and Patricia E. Kern Center for the Science of healthcare delivery through Emergency Department, Clinical Engineering Learning Laboratory.

References

1. Barratt H, Rojas-Garcia A, Clarke K, et al. Epidemiology of mental health attendances at emergency departments: systematic review and meta-analysis. PLoS ONE 2016; 11(4):e0154449 [Electronic Resource].
2. Bernstein SL, Aronsky D, Duseja R, et al. The effect of emergency department crowding on clinically oriented outcomes. Acad Emerg Med 2009;16(1):1-10. https://doi.org/10.1111/j.15532712.2008.00295.x [published Online First: Epub Date].
3. Sun BC, Hsia RY, Weiss RE, et al. Effect of emergency department crowding on out- comes of admitted patients. Ann Emerg Med 2013;61(6) 605-11 e6 [published On- line First: Epub Date] https://doi.org/10.1016/j.annemergmed.2012.10.026.
4. Pines JM, Hollander JE. Emergency department crowding is associated with poor care for patients with severe pain. Ann Emerg Med 2008;51(1):1-5.
5. Hoot NR, Aronsky D. Systematic review of emergency department crowding: causes, effects, and solutions. Ann Emerg Med 2008;52(2):126-36. https://doi.org/10.1016/ j.annemergmed.2008.03.014 [published Online First: Epub Date].
6. Timm NL, Ho ML, Luria JW. Pediatric emergency department overcrowding and im- pact on patient flow outcomes. Acad Emerg Med 2008;15(9):832-7. https://doi.org/ 10.1111/j.1553-2712.2008.00224.x [published Online First: Epub Date].
7. Pines JM, Pollack Jr CV, Diercks DB, Chang AM, Shofer FS, Hollander JE. The associa- tion between emergency department crowding and adverse cardiovascular out- comes in patients with chest pain. Acad Emerg Med 2009;16(7):617-25.
8. Wiler JL, Griffey RT, Olsen T. Review of modeling approaches for emergency depart- ment patient flow and Crowding research. Acad Emerg Med 2011;18(12):1371-9. https://doi.org/10.1111/j.1553 2712.2011.01135.x [published Online First: Epub Date].
9. Wargon M, Casalino E, Guidet B. From model to forecasting: a multicenter study in emergency departments. Acad Emerg Med 2010;17(9):970-8. https://doi.org/10. 1111/j.1553-2712.2010.00847.x [published Online First: Epub Date].
10. Kalucy R, Thomas L, King D. Changing demand for mental health services in the emergency department of a public hospital. Aust N Z J Psychiatry 2005;39(1-2): 74-80. https://doi.org/10.1111/j.14401614.2005.01512.x [published Online First: Epub Date].
11. Nesper AC, Morris BA, Scher LM, Holmes JF. Effect of decreasing county mental health services on the emergency department. Ann Emerg Med 2016;67(4): 525-30. https://doi.org/10.1016/j.annemergmed.2015.09.007 [published Online First: Epub Date].
12. Huntley M, Ward JE, Isaacson Loeffler, Lourey. HF 1233: Status in the house for the 88th legislature. Secondary HF 1233: Status in the House for the 88th Legislature; 2013-2014https://www.revisor.mn.gov/bills/bill.php?b=House&f=HF1233&y= 2013&ssn=0.
13. Chow WS, Priebe S. Understanding psychiatric institutionalization: a conceptual re- view. BMC Psychiatry 2013;13:169. https://doi.org/10.1186/1471-244X-13-169 [published Online First: Epub Date].
14. Fuller Torrey E, Fuller D, Geller J, Jacobs C, Ragosta K. No room at the inn: trends and consequences of closing public psychiatric hospitals 2005-2010. Secondary No Room at the Inn: Trends and Consequences of Closing Public Psychiatric Hospitals 2005-2010; 2012http://www.treatmentadvocacycenter.org/storage/documents/ no_room_at_the_inn-2012.pdf.
15. Putting the Promise of Olmstead into Practice. Minnesota’s Olmstead Plan. Second- ary Putting the Promise of Olmstead into Practice: Minnesota’s Olmstead Plan Au- gust 10, 2015 Revision. http://www.dhs.state.mn.us/main/idcplg?IdcService=GET_ FILE&RevisionSelectionMethod=LatestReleaed&Rendition= Primary&allowInterrupt=1&noSaveAs=1&dDocName=dhs16_196300; 2013.
16. Trangle M. Psychiatric flow. Secondary Psychiatric Flow; 2017http://www. rarereadmissions.org/documents/MH-Crisis-Clinic-Paramedics-20150120.mp3.
17. Minnesota Olmstead Subcabinet Report to the Court; 2014.
18. Report on program waiting lists. Services MDoH, ed.; 2014.
19. Alleviating ED boarding of psychiatric patients. Quick Safety, 19. ; 2015. p. 3.
20. Jacobsen S. In: Pasupathy DK, ed.; 2016.
21. 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. https://doi.org/10.5811/westjem.2013.6.17848 [pub- lished Online First: Epub Date].