1

April

The Impact of Video Monitoring onPatient Falls and Cost of Care

Ellen Barrington DNP, RN-BC, NEA-BCMarie Foley-Danecker DNP, RN, CCRN, NE-BC

Cira Fraser PhD., RN, ACNS-BCJanet Mahoney, PhD., RN, APN-C, NEA-BC

Presentation Outline2

11

44

22

33

55

66

77

88

99

Introduction to Falls and Video Monitoring

Critique and Review of the Literature

Methods

Data Collection and Analysis

Results

Discussion

Implications

Limitations

Conclusions

The purpose of this study wasto examine the impact ofvideo monitoring on patientfalls and cost of care.

3

Background and Significance4

Cost of inpatient falls $54.9

billion by 2020 (CDC, 2013)

Video monitoring -

newest technology

to prevent patient

falls

Hundreds of thousands of

patients fall,

30-50% result in injury

annually (The Joint

Commission, 2015)

Prevention Strategies -

low-cost personal alarms,

costly 1:1 direct

observation (Rausch,

Bjorkland, 2010)

Since 2009, 465 falls

with injury, 63%

resulted in death (The

Joint Commission,

2015)

PICOT Question5

In hospitalized adults (P), how does the use of a patient video monitoringsystem (I) versus the usual care (C), impact patient falls and affect the cost

of care (O) over a six-month period in two hospitals (T)?

Literature Search6

Linkage

RevisedSearch

Keywords

Databases

Review

CINHAL, PubMed, EBSCO

Key words: video monitoring,

falls and return on investment

Linked terms: falls and video monitoring, and

falls with return on investment

Linked terms: fall prevention in hospitals and

cost of falls in hospitals

CINHAL, Google Scholar

Abstract review completed

Search criteria: English and peer reviewed

Key words: falls prevention in hospital since

2012 and cost of falls in hospitals since 2012

Search criteria: English and peer reviewed

71 abstracts reviewed

929 articles reviewed independently

for specificity to video monitoring.

Final yield 9 articles.

Researcher 2Researcher 1

Donabedian Patient Safety Model-Theoretical Framework7

Organizational Structure -

how care is organized and

applied by nursing services

Organizational

Process – how care

is provided; actions

measured to reduce

negative outcome

1

23

Evaluates howstructure &

process influenceoutcomes

(Donabedian,1980)

Berenholtz, S.M., Pustavoitau, A., Schwartz, S.J., & Pronovost, P.J. (2007). How safe is my intensive care unit? Methods for monitoring and measurement. CurrentOpinion in Critical Care 13, 703-708.

Donabedian, A. (1980). The definition of quality and approaches to assessment. Ann Arbor: MI, Health Administration Press.Reed, K.D. (2008). The American association of critical care nurse’s Beacon award: a framework for quality. Critical Care Nurse, 20, 383-391. doi:

10.1016/j.ccell.2008.08.002.

Clinical outcomes -

are the results

achieved

Target Audience and Participants8

Inclusion CriteriaAll adult admitted patients for a three month periodbefore and a six month period post implementation ofvideo monitoring

Study SampleInpatient adults, male and female, 18-110 years,hospitalized, April 1, 2015 through November 30, 2017Two sets of participants from each of the two sites

Sets ofParticipants

Sample Size

Patients who fell pre and post implementation of videomonitoringSet 1Systematic sampling of every third video monitoredpatient as listed chronologically on reports generated inthe video monitor software

Set 2

N=595, VM n=397, Falls n=197 in Non VM and n=1 in VMExclusion criteria - contraindicated for video monitoring (patientsnot able to be redirected verbally, suicidal or homicidal ideation),pregnant

Setting for Data Collection9

Acute care hospitals inthe same network

Both hospitals hadthe same number of

cameras, policiesand operating models

Hospital Two

315 bedcommunity hospitalrange of specialized

inpatient andoutpatient services

Magnet award fornursing excellence

Hospital One

224 bedcommunity hospitalrange of specialized

inpatient andoutpatient services

Magnet award fornursing excellence

Same procedure andform for data collection

used at both sites

Data Collection – How the VM System Works10

Data – no video footage is recorded, hence no videofootage was reviewed

Digital system – retains alarm event and demographicinformation (admission, transfer, discharge information,census)

Demographic information – entered by the videomonitor technicianAlarm events – verbal, automated or alarm in room

Data - saved in Excel format - available to download

Fall Data Collection11

OneLink – documentationsystem used by RN’s to recordfalls

Fall data - pre and postimplementation wascollected

Data Collection Form - usedto collect data from allsources (OneLink reports,medical records and VMreports)

Cost Data Collection12

Calculated total salary expense of patient observation - based on theknown median patient observer and patient care technician salary

Calculated total hours used for VM program (VM Tech and Rounder)

Calculated hours per month spent on non VM patient observation (1:1)

Calculated hours per month Patient Care Technicians performed patientobservation (time they could have been on the unit)

Human Resources provided salary informationNurse Information Resource Center (NIRC) provided worked hours

Cost of Falls to the Organization

Estimated cost of afall with injury

Financial Risk to theorganization

$30,000 X 41 =Financial risk

41

Falls with injury

In 9-month totalsample period

$1,230,000$30,000

Burns, E. R., Stevens, J. A., & Lee, R. (2016). The direct costs of fatal and non-fatal falls among older adults – United States. Journal of Safety Research, Sep (58), 99-103. doi:10.1016/j.jsr.2016.05.001.

Centers for Disease Control & Prevention. (2016, August 19). Costs of falls among older adults. Retrieved from https://www.cdc.gov/homeandrecreationalsafety/falls/fallcost.html

Descriptive Data Pre and Post Implementation

Hospital 1Monthly Mean

Hospital 2Monthly Mean

Pre Post Pre Post

Patient Days Total 10814 11014 19549 19454

1:1 Observation Hrs 5132.7 3243.93 4117.03 3802.93

VM Hours n/a 1500.1 n/a 1460.17

Morse Score 50.83 50.19 53.05 53.65

Falls 8.33 4.0 14.33 16.67

Falls/1,000 PatientDays 2.36 1.06 2.2 2.55

Fall Injury Rate 0.46 0.36 0.20 0.51

PCA Hrs asObserver 3656.42 2112.17 1916.17 2231.67

Analysis of Pre and Post Implementation

Hospital 2Hospital 1

Results by Gender

50.1% 50.5% 51.3%

SampleVideo

MonitoredFalls

49.1% 49.5% 48.7%

Women

Men

16

Characteristics of Patient Monitoring Group17

68.97 Mean Age

197Patients

Female 101Male 96

n/aMean DaysMonitored

55.05 Admission Morse

59.43Morse Prior to VM

or Fall66.33Morse Prior to VM

or Fall

60Admission Morse

03Mean DaysMonitored

398Patients

Female 197Male 201

77.96Mean Age

Video Monitoring Non Video Monitoring

Analysis of Patient Monitoring Group18

Change in MorseThe change in Morse score for the patients who fell in the non VM group (n=197)from admission to prior to fall was significant (t=3.279, p=0.001).

Relationship between falls and Morse scoreA Pearson's Correlation revealed a strong, positive relationship between falls bymonth and average Morse scores by month in 2016 (r=0.847, p=.000), whichexplained 72% of the variance. In 2017, 73% of the variance between falls bymonth and average Morse was explained with the Pearson’s Correlation(r=0.854, p=.000).

GenderSlightly more men were video monitored, and marginally more women fell

Relationship between volume and fallsIn 2016, Pearson’s Correlations between patient days by month and falls bymonth showed a strong positive correlation (r=0.762, p= 0.001). In 2017,Pearson’s Correlation between patient days by month and falls by monthshowed a strong positive correlation (r=.904, p=0.000).

Fall Severity Pre and Post Video Monitoring19

41 fallswith

injury

Fall injuryscale 1-3:1= mild2=moderate3=severe

Pre-videomonitored fall

severityaverage was

1.36.

Post videomonitored fall

severitydecreased to

1.24

Percentage and Number of Falls by Diagnosis Category20

14.713.6

24.6

7.6

2

37.9

0

5

10

15

20

25

30

35

40

Neuro CV Resp Ortho SubstanceAbuse

Other

Perc

enta

ge

Diagnosis Category

Percentage of Falls at Change of Shift21

60Morse

Admission

0.3% VMfall8.1%

83.5 fallsmid shift

Non Shift Change Falls

Shift Change Falls

8.1%Non VM pmshift change

Non VM amshift change

83.5%Non VM

falls

Percentage and Number of Falls by Time Period22

0

2

4

6

8

10

12

14

16

18

0001-0300 0301-0600 0601-0900 0901-1200 1201-1500 1501-1800 1801-2100 2101-0000

29

2220

3032

17

21

27

Perc

enta

ge

Time

Falls Weekend vs. Weekday23

0.3%

23.9%

76.1%

Non VM

Weekend Falls

VM Weekend Fall

Non VM

Weekday Falls

Falls by Location24

8 (4.1%) falls occurred in other areas

152 (77.2%) falls

occurred in a patient

room

37 (18.8%) falls occurred in the

bathroom

1 (0.3%) patient fell in their room while on video monitor

77.2%

18.8%

4.1%0.3%

Hospital 1 Fall Rate25

Video monitoring was implemented April 2016. The trend line for thedecrease in fall rate was noted through 2016 and 2017.

Hospital 2 Fall Rate26

Video monitoring was implemented November 2016. A slight increase in thefall rate trend was noted from 2016 through 2017.

Findings

Hospital 1

Reduced:

patient

falls/1,000

patient days

one to one

observation hrs

PCA as observer

hrs

1

Hospital 2

Reduced:

1:1 observation

hrs

Increased:

patient falls/1,000

patient days

PCA as observer

hrs

2

Significant

change:

Morse score

on admission

and Morse

score just

before fall

3

Reduced:

fall severity

from 1.36 to

1.24

4

Reduced:

1:1 observation

hrs 24%

cost per

observed hr

20%

5

Return on Investment

Annual salary costof staffing VM at

both hospitals peryear

The current staffing model at both hospitals costs an estimated

$428,890 in salary per year. While increasing the capacity of

video monitoring will also incur an increase in salary expense,

this expense is reasonable in relation to patient safety, and the

high costs of patient falls with injury.

$428,890

Reduction in costper observed hour

Related to the decrease

in one-to-one observation

hours and shifting of

resources

20%

Reduction in 1:1observation hours

24%

Discussion29

The study found a strongpositive correlation

between patient daysand number of patient

falls. Hospital 2 had 44%more patient days than

Hospital 1.

Both Hospital 1 and Hospital 2implemented the VM program with

the same amount of videomonitoring cameras. Not evidentin the research of the literature isthe number of video monitoringcameras based on hospital sizenecessary to impact patient falls.However, based on the difference

in size of the two hospitals, thecorrelation between patient daysand falls, and the impact of ten

cameras on the smaller hospital’sfall rate, further exploration is

warranted.

Hospital 2 used more PCAas observer hours in the

post implementation period.Removing frontline

caregivers to provide 1:1observation has been

shown to deplete staffingresources (Sand-Jecklin,Johnson, & Tylka, 2016)

and does not decrease falls(Boswell, Ramsey, Smith,

and Wagers, 2001;Harding, 2010; Tzeng &

Yin, 2007).

Consistent with thepublished research,

the increase in PCA asobserver hours in

Hospital 2 could be aconfounding factor to

explain why fallsincreased exclusive ofVM implementation.

Boswell, D. J., Ramsey, J., Smith, M. A., & Wagers, B. (2001). The cost-effectiveness of a patient sitter program in an acute care hospital: A test of the impact ofsitters on the incidence of falls and patient satisfaction. Quality Management in Healthcare, 10(1), 10-16.

Harding, A. D. (2010). Observation assistants: Sitter effectiveness and industry measures. Nursing Economics, 28(5), 330-336.Sand-Jecklin, K, Johnson, J.R., & Tylka, S. (2016). Protecting patient safety: Can video monitoring prevent falls in high-risk patient populations? Journal of Nursing

Care Quality, 31(2), 131-138.

Discussion30

The study found astrong positive

correlation betweenpatient days andnumber of patientfalls. Hospital 2had 44% more

patient days thanHospital 1.

Both Hospital 1 and Hospital2 implemented the VMprogram with the same

amount of video monitoringcameras. Not evident in theresearch of the literature is

the number of videomonitoring cameras based

on hospital size necessary toimpact patient falls.

However, based on thedifference in size of the two

hospitals, the correlationbetween patient days andfalls, and the impact of ten

cameras on the smallerhospital’s fall rate, furtherexploration is warranted.

Hospital 2 used morePCA as observer hours

in the postimplementation period.

Removing frontlinecaregivers to provide1:1 observation has

been shown to depletestaffing resources

(Sand-Jecklin,Johnson, & Tylka,

2016) and does notdecrease falls (Boswell,

Ramsey, Smith, andWagers, 2001; Harding,

2010; Tzeng & Yin,2007).

Consistent withthe publishedresearch, the

increase in PCAas observer

hours in Hospital2 could be aconfounding

factor to explainwhy fallsincreased

exclusive of VMimplementation.

Implications

Increasing the video monitoringcapacity will reduce falls. This ideais further supported by the smallerhospital realizing a reduction in fallswith ten cameras and the largerhospital not realizing an overallreduction in falls.

4

Findings of this study suggest that thenursing team is triaging use of videomonitoring to patients with the mostobvious risk of fall (average Morseprior to being put on VM 66.33). Thisstudy found benefit would occur forpatients with a Morse of 55 andabove.

2

The average admission Morse scorefor all patients at both hospitals for2016 and 2017 was 51.5. Theaverage Morse on admission forpatients who fell was 55.05,implicating that it is difficult todifferentiate on admission theaverage patient (Morse 51.5) from thepatients who will fall (Morse 55.05).

3

Results of this study indicate videomonitoring is effective in decreasingfalls in patients at risk for fall. Of 198falls reviewed only one occurredwhile the patient was videomonitored.

1

Limitations32

Limited DataMorse Score

The Morse score was takenin total; no investigation

variables that calculate theMorse score was performed.

While the Morse score is the majorindication for initiation of VM,

another element is the subjectiveopinion of the nurse in assessing

the need for VM. This can bedriven by patient behaviors such asimpulsivity, disruption of treatment

and lack of safety awareness.

VM is not an appropriate strategyfor all patients – at the time ofthis study exceptions includepatients who cannot be re-

directed, suicidal and homicidalpatients, and restrained patients.These patients were included in

the fall sample.

The limited number of VMcameras available results intriaging the use of this finiteresource. Further research is

needed to establish an appropriatenumber of VM cameras to impactthe fall rate of a hospital based on

acuity and patient volume.

Limited data wascollected in the pre

video monitoring samplefor patients who did not

fall.

Limitations and Diagnostic Categories33

This study used the

admitting

diagnosis which

could change for a

patient based on

clinical findings.

1

Co-morbidities

were not

considered, and

the clinical status

of the patient at

time of fall was not

assessed.

2

The diagnosis categories

were broad, and there is

room for overlap among

diagnoses. In this study,

the largest diagnostic

category for the patients

who fell was the “other”

category.

3

The reason for a

fall may not be

related to the

diagnosis.

4

Conclusions34

Findings of thisstudy suggestthat a videomonitoringprogram iseffective in

reducing fallsand falls with

injury.

Further studyis needed toestablish aminimum

number of VMcameras to

impact the fallrate based onpatient acuityand volume.

Additionalstudy is

needed toestablish acut point toefficiently

initiate VM,to have themaximumimpact onfall rates.

To effectivelymaximize costsavings, further

study shouldfocus on the

maximum ratioof patients tovideo monitor

technician, andto assess theimpact of the

patient rounderon the fall rate.

35

Any Questions ?

36

Thank you

References37

Aydin, C., Donaldson, N., Aronow, H. U., Fridman, M., & Brown, D. S. (2015). Improving hospital patient falls. Journal of Nursing Administration,45(5), 254-262.

Bayen, E., Jacquemot, J., Netscher, G., Agrawal, P., Noyce, L. T., & Bayen, A. (2017). Reduction in fall rate in dementia managed care throughvideo incident review: Pilot study. Journal of Medical Internet Research, 19(10), 1-16.

Berenholtz, S.M., Paustavoitau, A., Schwartz, S.J., & Pronovost, P.J. (2007). How safe is my intensive care unit? Methods for monitoring andmeasurement. Current Opinion in Critical Care 13, 703-708.

Borikova, I., Tomagova, M., Miertova, M., & Ziakova, K. (2017). Predictive value of the Morse fall score. Central European Journal of Nursingand Midwifery, 8(1), 588-595.

Boswell, D. J., Ramsey, J., Smith, M. A., & Wagers, B. (2001). The cost-effectiveness of a patient sitter program in an acute care hospital: A testof the impact of sitters on the incidence of falls and patient satisfaction. Quality Management in Healthcare, 10(1), 10-16.

Burns, E. R., Stevens, J. A., & Lee, R. (2016). The direct costs of fatal and non-fatal falls among older adults – United States. Journal ofSafety Research, Sep (58), 99-103. doi: 10.1016/j.jsr.2016.05.001.

Center for Disease Control. (2013). Costs of falls among older adults. Retrieved from http://balancengineering.com/wp-content/uploads/2011/08/cdc-costs-of-falls-among-older-adults.pdf

Centers for Disease Control & Prevention. (2016, August 19). Costs of falls among older adults. Retrieved fromhttps://www.cdc.gov/homeandrecreationalsafety/falls/fallcost.html

Center for Medicare and Medicaid. (2008). CMS improves patient safety for Medicare and Medicaid patients by addressing never events.Retrieved from https://www.cms.gov/Newsroom/MediaReleaseDatabase/Fact-sheets/2008-Fact-sheets-items/2008-08-042.html?DLPage=2&DLEntries=10&DLSort=0&DLSortDir=descending

References38

Cournan, M., Fusco-Gessick, B. and Wright, L. (2016). Improving patient safety through video monitoring. Rehabilitation Nursing.doi:10.1002/rnj.308

Davis, J., Kutash, M., & Whyte, J. (2017). A comparative study of patient sitters with video monitoring versus in-room sitters. Journal of NursingEducation and Practice, 7(3), 137-142.

Donabedian, A. (1980). The definition of quality and approaches to assessment. Ann Arbor: MI, Health Administration Press.

Dunne, T. J., & Gaboury, I., & Ashe, M. C. (2014). Falls in hospital increase length of stay regardless of degree of harm. Journal of Evaluation inClinical Practice, 20, 396-400.

Dunton, N., Gajewski, B., Klaus, S., & Pierson, B. (2007). The relationship of nursing Workforce characteristics to patient outcomes. OJIN: TheOnline Journal of Issues in Nursing, Vol. 12(3). Available:www.nursingworld.org/MainMenuCategories/ANAMarketplace/ANAPeriodicals/OJIN/TableofContents/Volume122007/No3Sept07/NursingWorkforceCharacteristics.aspx

Hackensack Meridian Health. (2017). Bayshore Medical Center. Retrieved October 19, 2017, fromhttp://www.bayshorehospital.org/B/aboutus/AwardsandRecognition.cfm

Hackensack Meridian Health. (2017). Ocean Medical Center. Retrieved October 20, 2017, fromhttp://www.oceanmedicalcenter.com/OMC/aboutus/AwardsandRecognition.cfm

Hardin, S. R., Dienemann, J., Rudisill, P., & Mills K. K. (2013). Inpatient fall prevention: Use of in-room webcams. Journal of Patient Safety, 9(1),29-35.

Harding, A. D. (2010). Observation assistants: Sitter effectiveness and industry measures. Nursing Economics, 28(5), 330-336.

Johnson, M., George, A., & Tran, D. T. (2011). Analysis of falls incidents: Nurse and patient preventive behaviours. International Journal ofNursing Practice, 17(1), 60-66.

References39

Klymko, K., Etcher, L., Munchiando, J., & Royce, M. (2016). Video monitoring: A room with a view, or a window to challenges in falls preventionresearch? MEDSURG Nursing, 25(5), 329-333.

Morse, J. M., Black, C., Oberle, K., & Donahue, P. (1989). A prospective study to identify the fall-prone patient. Social Sciences Medicine, 28(1),81-86.

Oliver, D., Healey, F., & Haines, T. P. (2010). Preventing falls and fall-related injuries in hospitals. Clinics in Geriatric Medicine, 26, 645-692.

Rausch, D. L., & Bjorklund, P. (2010). Decreasing the costs of constant observation. Journal of Nursing Administration, 40(2), 75-81.

Reed, K.D. (2008). The American association of critical care nurse’s Beacon award: a framework for quality. Critical Care Nurse, 20, 383-391.doi: 10.1016/j.ccell.2008.08.002.

Robinovitch, S. N., Feldman, F., Yang, Y., Schonnop, R., Leung, P. M., Sarraf, T., Sims-Gould, J., & Loughin, M. (2013). Video capture of thecircumstances of falls in elderly people residing in long-term care: an observational study. The Lancet, 381, 47-54.

Sand-Jecklin, K, Johnson, J.R., & Tylka, S. (2016). Protecting patient safety: Can video monitoring prevent falls in high-risk patient populations?Journal of Nursing Care Quality, 31(2), 131-138.

The Joint Commission. (2015, September 28). Preventing falls and fall-related injuries in health care facilities. Sentinel Event Alert, 55.Retrieved from https://www.jointcommission.org/assets/1/18/SEA_55.pdf

Tzeng, H., & Yin, C. (2007). Using family visitors, sitters, or volunteers to prevent inpatient falls. Journal of Nursing Administration, 37(7/8), 329-334.

Votruba, L., Graham, B., Wisinski, J., & Syed, A. (2016). Video monitoring to reduce falls and patient companion costs for adult inpatients.Nursing Economics, 34(4), 185-189.