Report on Critical Incidents in Virginia’s

State Operated Mental Health Facilities October 1, 2017 - September 30, 2018

Prepared by The disAbility Law Center of Virginia

March 2019

INTRODUCTION

The disAbility Law Center of Virginia (dLCV) is a private non-profit organization, operating under the authority of

federal law and designated by state law to act as the protection and advocacy system for people with disabilities

in Virginia.

The Code of Virginia requires that all facilities operated by the Department of Behavioral Health and Developmental Services (DBHDS) must report to dLCV within 48 hours of a “critical incident.” DBHDS is then required to provide all other known information within 15 days. A “critical incident” is any event resulting in death or loss of consciousness or an event requiring medical treatment.

During federal fiscal year 2018 (FY 18), dLCV received a total of 318 Critical Incident Reports from mental health facilities operated by the Department.

EXECUTIVE SUMMARY

In 2018, following review of FY 17 data, dLCV contacted D BHDS about concerns arising from the Critical Incident

Reports. We noted an increase in deaths at DBHDS-Operated mental health facilities and a corresponding

increase in deaths occurring soon after admission. There appeared to be a correlation between the bed of last

resort legislation and the admission of individuals with higher healthcare (as opposed to strictly mental health

care) needs.

In FY18, deaths and the numbers of deaths within three months of admission continue to be major concerns. The number of deaths reported by DBHDS-operated mental health facilities increased for the 2nd consecutive year, from 57 in FY 17 to 59 in FY 18. FY 18 unseated the previous year to become the “deadliest” year at mental health facilities since at least FY 12. All adult hospitals reported at least one death during FY 18. Piedmont Geriatric Hospital, Eastern State Hospital, and Catawba Hospital reported the greatest number of deaths (16, 14, and 13, respectively).

DBHDS psychiatric hospitals are not designed to act as primary medical providers nor, generally, as nursing

facilities. DBHDS should identify any underlying reasons for the increased deaths of individuals recently admitted

– including the inappropriate admissions of individuals who would be more appropriately served in medical

hospitals, or an inability to provide healthcare within the DBHDS system. Once reasons are identified, corrective

action can commence and, hopefully, we will see a decrease in state hospital mortality.

BACKGROUND

The Virginia Department of Behavioral Health and Developmental Services (DBHDS) generates Critical Incident

Reports (CIRs) on occurrences in their institutions resulting in injuries requiring medical treatment, and on

occurrences resulting in loss of consciousness and death. As the populations the state facilities serve are

substantially different, dLCV evaluates data from State Hospitals and Training Centers separately. This report will

detail CIR trends in DBHDS-operated mental health (MH) facilities during the 2018 Federal Fiscal Year (FY 18).

dLCV’s MH CIR data is based on reporting from:

Catawba Hospital (CAT)

2

Central State Hospital (CSH)

Commonwealth Center for Children and Adolescents (CCCA)

Eastern State Hospital (ESH)

Northern Virginia Mental Health Institute (NVMHI)

Piedmont Geriatric Hospital (PGH)

Southern Virginia Mental Health Institute (SVMHI)

Southwestern Virginia Mental Health Institute (SWVMHI)

Western State Hospital (WSH)

dLCV regularly monitors conditions in state facilities and responds to questions and complaints from residents

and their families. dLCV reviews CIRs on a weekly basis and analyzes quantitative data from the reports to

identify trends. Qualitative and quantitative data from the reports inform dLCV’s work in the state facilities and

this report.

DEATHS AT MH FACILITIES

TOTAL DEATHS The number of deaths reported by DBHDS-operated MH facilities increased for the 2nd consecutive year, from 57 in FY 17 to 59 in FY 18. While FY 2017 saw the highest number of deaths in mental health facilities in many years, FY 18 is now the “deadliest” year at MH facilities since at least FY 12. All adult MH facilities reported at least one death during FY 18. PGH, ESH and CAT reported the greatest number of deaths (16, 14, and 13, respectively). The steady rate with which PGH deaths in particular h ave increased over the last 2 years raises significant concerns.

3

MH Total Deaths

70 57 59

60 47

50 42

40 32 34 32

30

20

10

0

FY12 FY13 FY14 FY15 FY16 FY17 FY18

20

18

16

14

12

10

8

6

4

2

0

FY12

CAT

FY13

CSH ESH

Hospital Deaths

FY14 FY15

NVMHI PGH

FY16

SVMHI

FY17

SWVMHI

FY18

WSH

40

30

20

10

0

FY 18 MH Deaths: Did Death Involve One or More of the "7 Medical Triggers?"

30

21

9

-Yes No Unknown

THE “7 MEDICAL TRIGGERS” Incidents preceding death at DBHDS-operated facilities have historically been primarily “Medical Condition.” Medical Conditions were the basis for 49 of the 59 deaths (83%) reported by MH facilities in FY 18. Facilities used the essentially meaningless “Loss of Consciousness” to explain 7 of the deaths (11.8%); this is the most that this category has been used since our recordkeeping began. While the number of unexplained incidents resulting in death dropped to just 1 in FY18, facilities generally did not provide much detail on the deaths, in spite of the statutory requirement to provide all known details in the 15 day report. As a result, we could not determine whether or not 30 of the FY 18 deaths (50.8%) involved the “7 Medical Triggers1.”

1 As defined by DBHDS: Aspiration/Aspiration Pneumonia, Bowel Obstruction/Constipation, Decubitus Ulcers, Dehydration, Seizures, Sepsis, and Urinary Tract Infections. (CVTC Community Provider Training)

4

DISCHARGE PRIOR TO D 2 EATH At both PGH and SWVMHI, the majority of reported deaths (62.5% and 50%, respectively) occurred after the

decedent had been discharged from the hospital. As a geriatric facility, PGH discharged many of its residents to

Nursing Homes or hospice providers in anticipation of their decline or death. We do not question the

appropriateness of those discharges, but merely note it. By contrast, facilities like CSH and NVMHI do not house

geriatric residents and reported a low number of deaths (3 and 2, respectively). While some of the CSH and

NVMHI deaths were due to medical conditions, few of the deaths were expected.

The number of deaths occurring during a special hospitalization increased from 9 to 12 in FY 18.

FY 18 MH Deaths: Was the Individual Discharged Prior to Death?

30 28

25

20

15

19

12

10

5

0

Yes No Special Hospitalization

DEATHS BY AGE As expected, geriatric deaths at MH facilities have considerably outnumbered non-geriatric deaths since our

record-keeping began. In FY 18, the number of non-geriatric deaths increased slightly.

2 Facilities are required to submit CIRs for any death that occurred at their facility, or within 21 days of an individual’s discharge, per DI 401.

5

45

40

35

30

25

20

15

10

5

0

22

10

FY12

27

7

FY13

Hospital Deaths By Age 42

31 31

22

16 15

1110

FY14 FY15 FY16 FY17

MH Deaths Non Geri MH Deaths Geri

42

17

FY18

PROXIMITY TO ADMISSION For several years, a number of facility directors have expressed concern with the number of “inappropriate medical admissions” that they have received since revisions to Virginia Code § 37.2-809 went into effect in 2014,

designating the DBHDS hospitals as “facilities of last resort.” We reviewed the number of deaths occurring

within given time intervals after admission. The number of deaths occurring within 90 days of admission has

increased substantially since 2014 and reached an all-time high in FY 18. The proportion of MH deaths occurring

within 90 days of admission has also increased since 2014. Before the 2014 Legislation was in effect, deaths

within 90 days of admission did not exceed 22.7% (in FY 12), and reached their lowest point in FY 14 (15.6%).

From FY 15 through FY 18, the proportion of deaths occurring within 90 days did not dip below 30%, reaching

36.2% in FY 15, 31% in FY 16, 32.1% in FY 17, and 38.9% in FY 18.

6

MH Deaths Within 90 Days of Admission

FY18

FY17

FY16

FY15

FY14

FY13

FY12

0 5 10 15 20 25

■ 0 to 7 ■ 8 to 14 ■ 15 to 31 ■ 32 to 61 ■ 62 to 90

100%

Proportion of MH Deaths by Proximity to Admission

80%

60%

40%

20%

0%

FY12 FY13 FY14 FY15 FY16 FY17 FY18

■ MH Total 0 to 90 ■ MH Total 90+

OVERALL INCIDENT REPORTING

dLCV is concerned that specific mental health facilities may be failing to comply fully with reporting

requirements, resulting in underreporting. We are addressing issues with specific facilities as we identify the

issue. Inconsistent compliance necessarily impacts the reliability of our conclusions below. More importantly,

however, inconsistent compliance with reporting requirements may indicate the absence of quality

improvement and risk management oversight by DBHDS and by facility management.

In FY 18, DBHDS-Operated MH facilities reported 318 Critical Incidents—a very slight decrease from the 333

reported in FY 17. ESH reported the greatest number of incidents (66), followed by PGH (55). ESH is the largest

DBHDS-Operated MH facility in Virginia, so their status as highest reporter is expected. PGH, however, is a

relatively small facility with 123 beds and an entirely geriatric population. One might expect a geriatric MH

facility to have a higher rate of certain incidents, such as falls and deaths, compared to facilities serving a

younger population; still, it is concerning that PGH residents were subject to a greater number of serious injuries

than facilities with substantially more psychiatrically acute admissions.

For comparison, CSH, the second largest hospital, houses Virginia’s maximum security forensic unit. CSH was the

third-largest CIR reporter, with 44 serious injuries in FY 18—eleven fewer than PGH. FY 18 CIR data supports

staffs’ assertions that the populations of CSH and ESH have become far more acute and prone to self-injury and

peer assault, as evidenced by the fact that they reported the highest numbers of self-injurious behaviors (SIB)

(CSH reported 14 and ESH reported 12). These rates of injury may also be related to the transfer of individuals

from jails for restoration of competence. While numbers remain high, the number of SIB incidents at CSH and

ESH h ave decreased substantially in the last year; in FY 17, they reported 24 and 17 incidents of SIB, respectively.

7

CIR Incident Types at DBHDS-Operated MH Facilities FY18

70

4

60

12

6 50

2

2

1

9 15 40

7 2

14 4

1 8

30 13

11 4

8 3

1 2 20 3 1 1

1 28 2 1 4

12 1 2 17 3 13 1 7

1

10

0

1 12 2 1 1 1

3 1 10

2 3 8 6 1 2 2 1 4 1 1 4 3 3 1 2 2 1 1 1

2

6

2

CAT CSH CCCA ESH NVMHI PGH SVMHI SWVMHI WSH

■ Accident ■ Aggressive Behavior ■ Alleged Assault-Peer Alleged Assault-Staff □

■ Alleged Sexual Assault-Peer ■ Allergic Reaction ■ Aspiration/Choking ■ Fall

■ Loss of Conciousness ■ Medical Condition ■ Other ■ Recreational Activities

■ Self-Injurious Behavior ■ Unexplained

INCIDENT CATEGORIES

In previous years, falls and SIB made up the vast majority of MH CIRs. Falls are still the most prevalent incident

type, accounting for 28.93% of incidents. Meanwhile, SIB decreased substantially in the last year, dropping from

20.72% of incidents in FY17 to just 13.21% of incidents in FY18. Peer assaults and medical conditions continue to

make up a sizeable proportion of incidents.

8

Proportion of CIR Incident Types at All DBHDS-Operated MH Facilities FY18

Aggressive Behavior 3.77%

Accident Unexplained 5.35%

6.92%

Self-Injurious Behavior

Recreational Activities 13.21% Alleged Assault-Peer Alleged 2.20% Other 15.41% Assault-Staff

1.26% 0.63%

Alleged Sexual Assault-Peer

Medical Condition 2.20%

17.30% Allergic Reaction

0.31%

Aspiration/Choking 0.31%

Fall 28.93%

Loss of Conciousness 2.20%

We compared the 4 most prevalent incident types—peer assaults, falls, medical conditions, and SIB—across the

facilities within the context of population. Below are the per capita statistics for FY 17 and FY 18. Due to CCCA’s

uniquely small population and high bed turnover, including them in the per capita analysis can lead to

misinterpretation of the data. While the rate of CCCA incidents is correct within the context of dLCV’s analysis,

we do not believe it is a comparable statistic; it has been excluded from our per capita analysis of incidents.

The rate of incidents secondary to falls increased during FY 18, particularly at PGH, where the rate increased

from 0.175 per capita, or about 17% to 0.228 per capita, or about 23%. At SVMHI, the rate of falls remains

unusually high at 0.154 per capita, or about 15%. This rate of falls is more consistent with PGH and CAT, which

house geriatric residents, than it is with other, non-geriatric hospitals. The rate of SIB decreased at every facility.

9

0.250

0.200

0.150

0.100

0.050

0.000

Per Capita Rate of CIR-Reported Incidents at DBHDS-Operated MH Facilities FY 17

0.0

29

0.1

44

0.1

73

0.0

10

0.0

55

0.0

29

0.0

34

0.1

01

0.0

27

0.0

20

0.0

30

0

.05

7

I .. 1 •• 1 ••• , It I I I.. I ■ .111 ••• 0.0

34

CAT CSH ESH NVMHI PGH SVMHI SWVMHI WSH 0

.02

60

.00

0

■ Alleged Assault-Peer Per Capita ■ Fall Per Capita 0

.06

0

■ Medical Condition Per Capita ■ Self-Injurious Behavior Per Capita 0

.03

3

0.1

75

0

.10

0

0.0

08

0.0

63

0

.14

3

0.0

00

0

.01

6

0.0

24

0.0

66

0.0

54

0.0

66

0.0

00

0

.04

7

0.0

26

0.0

21

0.250

0.200

0.150

0.100

0.050

0.000

0.0

19

0

.16

2

0.1

24

0.0

00

0.0

50

0

.01

30

.01

3

0.0

59

0.0

44

0.0

37

0.0

31

0.0

41 0.0

66

0

.01

70

.01

70

.03

3

0.0

08

0.2

28

0

.12

2

0

.00

0 Per Capita Rate of CIR-Reported Incidents at DBHDS-Operated MH

Facilities FY18 0

.01

5

0.1

54

0

.00

00

.00

0 0.0

36 0

.07

10

.04

80

.04

8

0.0

26

0

.03

0

0.0

09

0

.00

4

• 1 •• 1 1111 l •• 1 _ • 1111 ••--CAT CSH ESH NVMHI PGH SVMHI SWVMHI WSH

■ Alleged Assault-Peer Per Capita ■ Fall Per Capita

■ Medical Condition Per Capita ■ Self-Injurious Behavior Per Capita

INJURY YPES TMost of the injuries reported by MH facilities in FY 18 were lacerations or skin tears (108 total, or 32.39%),

followed by deaths (59 total, or 18.55%) and fractures (56 total, or 17.61%). Overall, the proportion of primary

injuries reported in FY 18 was very similar to that from FY 17.

10

70

60

50

40

30

20

10

0

CIR Primary Injury Types at DBHDS-Operated MH Facilities FY18

1 2 -1

2

2

28

14 1 2

2

6 12 19 3 13

11

5 8 16

8 1 1 1 10 1

2 2 11 14 2

4 8 9 2 13 1

1 1 4 8 3 3 2 1 3 5 2 2 2 2 1 2 1 1 1 2 2 2 2 1 1 1 1

CAT CSH CCCA ESH NVMHI PGH SVMHI SWVMHI

■ Abrasion/Scratch Alleged Sexual Assault ■ Allergic/Adverse Reaction □

■ Bite (Human) ■ Bruise/Contusion/Hematoma Concussion/TBI □

■ Death ■ Eye Injury ■ Fracture

■ Ingestion of Foreign Object Insertion of Foreign Object ■ Laceration/Tear □

Loss of Consciousness Other ■ Pain □

■ Swelling/Redness ■ Tooth Injury ■ Unknown

3

1 1 1

3

5

1

4

2

2

WSH

11

.11 111 .I II 111 111 .11 ■ ■ ■

I .II II •• I I ■■I 111 -·· ■ ■ ■

For the reasons stated previously, we have not included CCCA in our per capita analysis of injuries. Fractures

decreased at all facilities except NVMHI during FY18.

The rate of deaths decreased slightly at CAT (from .173 per capita or about 17% to .124 per capita or about

12%). Deaths increased substantially at PGH, however, from .092 per capita or about 9% in FY 17 to .13 per

capita or about 13%. The increase at PGH is more dramatic when you consider that the FY 17 rate was already a

huge increase from FY 16, when the rate was only .041 per capita or about 4%. The high rate of deaths at CAT

and PGH is due, at least in part, to the large proportion of geriatric residents at both facilities. Still, as we

previously stated, DBHDS must endeavor to identify causal factors to this increase in deaths if they hope to

reduce mortality in the state hospitals.

0.200

0.180

0.160

0.140

0.120

0.100

0.080

0.060

0.040

0.020

0.000

0.1

73

0.0

96

0.1

44

0.0

17

0.0

55

0.0

88

0.0

37

0.0

40

Per Capita Rate of CIR-Reported Injuries at DBHDS-Operated MH 0

.04

7Facilities FY17

0.0

00

0.0

09

0.0

78

0.0

92

0.0

75

0.1

33

0.0

32

0.0

32

CAT CSH ESH NVMHI PGH SVMHI SWVMHI 0

.04

8

Death Per Captia Fracture Per Capita Laceration/Tear Per Capita 0

.04

8 0.0

72

0.0

78

WSH

0.0

13 0

.03

9

0.0

34

0.200

0.180

0.160

0.140

0.120

0.100

0.080

0.060

0.040

0.020

0.000

0.1

24

0.0

76

0.1

81

0.0

13

0.0

46

0.0

50

0.0

47

0.0

37

0.0

95

0.0

17

0.0

25

0.0

74

0.1

30

0.0

49

0.1

14

0.0

15

0.0

15 0.0

31 0.0

48

0.0

60 0.0

77

Per Capita Rate of CIR-Reported Injuries at DBHDS-Operated MH

Facilities FY18

CAT CSH ESH NVMHI PGH SVMHI SWVMHI

Death Per Captia Fracture Per Capita Laceration/Tear Per Capita

0.0

09

0.0

17

0.0

21

WSH

12

60

50

40

30

20

10

0

Total CIR-Reported Incidents Occurring on a Given Day of the Week at DBHDS-Operated MH Facilities FY18

50 47

44 41 41

38 40

Monday Tuesday Wednesday Thursday Friday Saturday Sunday

18

16

14

12

10

8

6

4

2

0

CIR-Reported Incidents Occurring on a Given Day of the Week at DBHDS-Operated MH Facilities FY18

Monday Tuesday Wednesday Thursday Friday Saturday

CAT CSH CCCA ESH NVMHI PGH SVMHI SWVMHI • •

Sunday

WSH

TIMING OF INCIDENTS Facilities reported a slight “bump” in incidents occurring on Saturdays (50 total incidents). Reporting seems to

be fairly consistent across other days of the week. The only real outlier among the “day of week” data is PGH,

which reported a disproportionate 16 incidents on Tuesdays. It is unclear what caused these “bumps” in the data, but dLCV advocates will use this information to guide monitoring in FY19.

As expected, most of the incidents were reported during periods of higher social activity, such as lunch and leisure times. That said, there are much higher number of incidents reported during the lunch hour (12pm-1:59pm) during FY 18 (46) than in FY 17 (24).

13

Total CIR Incidents Per Month at DBHDS-Operated MH Facilities FY18

40

30

20

3533

26 26 25 25

31

22 24 24 23 24

10

0

Oct. Nov. Dec. Jan. Feb. Mar. April May June July Aug. Sep.

Reported Time of CIR Incidents at DBHDS-Operated MH Facilities (Total) FY18

50 46

45 41 40

40 35 30

28 28

25 20

20 20 17 17 16 18

15 10

5 3

0

FY 18 CIR reporting at MH facilities peaked in July (35), with the fewest reports coming in October (22). Monthly

reporting during FY 18 was more consistent than in FY 17 (high o f 42, low of 18).

CONCLUSION

dLCV continues to have serious concerns about the ability of DBHDS mental health facilities to manage the

increasingly acute and medically fragile individuals admitted to their care. This is not necessarily an indictment

of the care provided, but may be a result of the inability to provide that care in a facility designed for psychiatric

purposes. In Federal Fiscal Year 2018 (FY 18), dLCV notified DBHDS of these concerns, particularly the increase

in deaths and the corresponding increase of deaths occurring soon after admission. We shared with the

Department our speculation that there is a correlation between the bed of last resort legislation and the

admission of individuals with higher healthcare (as opposed to strictly mental health care) needs.

14

RECOMMENDATIONS

The Department of Behavioral Health, together with other relevant state resources, should take immediate

steps to identify the causes of the increased death rates at state mental health facilities. Specifically, the

Department should

- Determine the source of admission for every death that occurred within 90 days of admission, in

every mental health facility operated by the state, to determine whether there are any obvious

problematic sources of admission,

- Conduct a complete analysis of each death, regardless of admission date, to determine whether any

death was related to one of the “7 Triggers.” In our experience, many deaths related to one of these

triggers (aspiration, bowel obstruction, decubitus ulcers, dehydration, seizure, sepsis, and urinary

tract infection) can be avoided with improved staff training and appropriate resources,

- Evaluate staffing levels at each facility, especially for medical staff,

- Increase staff training to be able to identify and respond to medical conditions, and

- Review the falls protocol in each of the state’s mental health facilities

As always, the disAbility Law Center of Virginia welcomes any opportunity to discuss these findings with the

Department of Behavioral Health and Developmental Services, with relevant policy makers or with other

concerned parties.

15