You are on page 1of 8

JONA

Volume 50, Number 6, pp 355-362


Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.

THE JOURNAL OF NURSING ADMINISTRATION

Exploring Practices of Bedside Shift


Report and Hourly Rounding. Is There an
Impact on Patient Falls?
Carolyn Sun, PhD, RN, ANP-BC Kenrick D. Cato, PhD, RN, CPHIMS, FAAN
Caroline J. Fu, MPH Lauren Stoerger, MSN, RN, NEA-BC, CCRN, CNRN
Jessica O'Brien, MS, RN, AGCNS-BC, PCCN Alan Levin, MSN, MBA, RN, CPHQ, NEA-BC

OBJECTIVE: To describe the relationship of inpa- While healthcare costs in the United States rank lower
tient falls to bedside shift report (BSR) and hourly compared with other high-resource countries, many
rounding (HR). Americans lack high-quality care and suffer poor out-
BACKGROUND: Falls are a major healthcare con- comes.1 In 2005, in an effort to improve healthcare
cern. Although measures such as BSR and HR are quality and concomitantly reduce costs, Congress
reported to reduce falls, studies are often based on identified hospital-acquired conditions that would
self-reported data related to nurse compliance with no longer be paid for by the Centers for Medicare &
protocols for HR and bedside report. Medicaid Services.2 Falls were among the list of
METHODS: Observational data were collected on nonreimbursable hospital-acquired conditions citing
nursing tasks, including BSR and HR. falls as a “serious preventable event.”2 Worldwide,
RESULTS: Nine thousand six hundred ninety-three falls are the 2nd leading cause of death from acciden-
observations were recorded on 11 units at 4 hospitals tal injury, and in the United States alone, the cost of
over 281 shifts. Falls were associated with shift and falls is estimated at $50 billion annually.3,4 Thus,
day of the week but not BSR, HR, or the frequency there is constant pressure on hospitals to improve pa-
of encounters with the patient. The regression model tient safety and reduce costs from adverse events.
included frequency with patient, shift, day of week, Nurse leaders are charged with cost containment
and HR. but are challenged with fall-prevention programs that
CONCLUSIONS: Increased nurse frequency with pa- may be more costly than potential cost savings.5 Inter-
tient may signal increased fall risks. Bedside shift ventions including hourly rounding (HR), bedside
report and HR may require robust and sustained in- shift report (BSR), gait belts, and the use of chair or
terventions to provide lasting effects. bed alarms have been cited as effective.6-10 It has been
suggested that program development, including HR,
without leadership engagement, is not an effective fall
Author Affiliations: Assistant Professor (Dr Sun), Hunter-Bellevue
School of Nursing, Hunter College; Adjunct Faculty (Dr Sun), Research prevention strategy.7
Coordinator (Ms Fu), and Assistant Professor (Dr Cato), Department While there has been a plethora of fall-prevention
of Research, Columbia University School of Nursing; and Nurse Re- strategies globally in a variety of settings,11,12 evidence
searcher (Dr Sun), Clinical Nurse Specialist (Ms O'Brien), and Director
of Nursing (Ms Stoerger), New York–Presbyterian, New York; Assis- suggests that nursing presence is directly related to the
tant Professor of Clinical Informatics (Dr Cato), Columbia University incidence of falls, with each additional RN hour
Department of Emergency Medicine; and Vice President and Chief per patient day associated with a reduced fall rate.13
Nursing Officer (Mr Levin), New York–Presbyterian Queens, Depart-
ment of Nursing, Flushing, New York. Whereas increased time with patients has been linked
Conflicts: None to declare. with improved patient and nurse satisfaction and re-
Correspondence: Dr Sun, Hunter-Bellevue School of Nursing, duced falls, lack of nursing time and interaction with
Hunter College, 425 E 25th Street, New York, NY 10010
(cjs.cumc@gmail.com). patients have been associated with higher risks of ad-
DOI: 10.1097/NNA.0000000000000897 verse events such as patient falls and medication errors

JONA  Vol. 50, No. 6  June 2020 355

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.


both in the United States and abroad.14-18 Nurses are whether HR was completed was self-reported data
critical to improving healthcare quality and reducing by nurses. Less is known about the effects of these in-
costs19,20 as nurses spend more time with patients than terventions after the implementation period has ceased
any other healthcare worker, and the quantity of that to assess whether these changes are sustained and the
time spent is directly correlated with patient safety.14,15 ultimate impact on patient-related outcomes.10
Nurses can play a critical role in preventing falls and Previously, methodologies such as time-and-motion
other negative patient outcomes by quickly identifying and work sampling have been used to assess nurse and
patients at risk as they care for patients, evaluating patient interactions.35-38 While time-and-motion studies
these risks, and acting upon the changing status of involve constant, prolonged observation of participants,
the patient, a process termed, “nurse surveillance.”21 work sampling is a method that has been used in
Nurse surveillance has been seen as integral in reduc- health-services research to more cost-effectively utilize
ing errors and improving patient outcomes, and research staff by recording observations at intervals
nurses are seen as key contributors to overall patient (rather than constant observation) and analyzing these
surveillance.22,23 observations with statistical modeling to allow a study
HR is the process by which a member of the nurs- to capture longitudinal workflow habits.39 To date,
ing staff performs proactive, intentional rounds on ev- there is a paucity of observational nursing studies of
ery patient, every hour to help identify and tend to the these types focused on BSR and HR.40
patients' needs. The main components of HR include Therefore, the purpose of this study was to de-
addressing patients' needs for pain control, position- scribe the frequency with which BSR and HR are car-
ing, personal hygiene, and possessions, as well as ried out and describe their relationship to patient falls
assessing the environment for safety and ensuring all in a setting where these practices have been in place
patient needs are met prior to leaving; however, the for a period and to assess the sustained effect of these
exact definition of components and timing vary interventions. This was to be accomplished through
widely.24 HR is purported to reduce patient anxiety, the following aims:
set patient and family expectations, and decrease pa-
tient falls.25 1. Aim 1: Describe the frequency of BSR and HR
BSR is the transfer of patient information, ac- in inpatient medical/surgical units using work-
countability, and authority from nurse to nurse at a sampling technique.
patient's bedside to encourage patient participation 2. Aim 2: Explore the relationship between nurse
and continuity of care.26 A main objective of BSR is surveillance including BSR and HR and pa-
to provide structured communication, which offers tient falls.
additional benefits such as patient satisfaction and
improved outcomes, including reductions in patient H2: BSR and HR may be negatively associated
falls.26,27 with patient falls.
With a push from organizations such as the Agency
for Healthcare Research and Quality and the Joint
Commission to involve the patient and the family dur- Methods
ing BSR,28,29 as well as many studies recommending Design
HR by nurses, there has been little research to assess Work sampling was used to quantify nursing tasks
how or whether BSR or HR is actually carried out, and detect relationships between those tasks.41 Nurses
or their subsequent effects on patient falls.8,30 A sys- were recruited to the study via direct verbal communi-
tematic review on the utility of HR to decrease falls re- cation, and they verbally consented to participating in
ported only 3 of 16 studies demonstrated a statistically the study. Each nurse was observed in intervals of
significant reduction in falls, and the method and defi- 15 minutes at all hours of the day and all days of the
nition of HR varied widely from study to study.31 A week; their activities were recorded at these intervals
scoping review on BSR reported that 18 of 22 studies using a previously published instrument, adapted for
were about the implementation of these measures, the needs of this study.39 The nurses were not told spe-
rather than looking at the longitudinal outcomes that cifically which tasks were being recorded, but that they
such interventions would affect.32 This could be due were common tasks one would perform in the nursing
to the difficulty assessing nurses' activities; direct ob- profession. Supplemental training on BSR and HR was
servation is often costly, but methods such as video not provided for this study.
surveillance can present ethical difficulties.33 The sem-
inal study by Meade et al34 suggested HR could im- Instrument
prove patient falls, decrease call bell use, and increase Using a previously published and validated tool for
patient satisfaction; however, all information about work sampling in a nursing setting for observation,39

356 JONA  Vol. 50, No. 6  June 2020

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.


researchers conducted tests for face validity among a Setting
group of 7 RNs who worked in inpatient settings with Institutional review board (IRB) and site approval were
at least 1 year of full-time experience. The instrument obtained at all study sites. The study was conducted at
lists the tasks that nurses routinely perform through- 4 hospitals: 2 large urban medical centers and 2 com-
out their shift on an inpatient medical/surgical unit; munity hospitals. Two to 5 medical/surgical units at
however, not all activities of interest in this study were each site were selected. All 4 hospitals were a part of
listed. Therefore, adaptations to the instrument were the same healthcare system and followed the same pol-
guided by their input and existing literature to meet icies, procedure guidelines, and patient experience ex-
the needs of the current study with the authors' per- pectations; at the time of data collection, HR policies
mission.39 We also included a single question about were in place for at least 5 years and BSR for at least
the nurses' self-ranked level of expertise according to 3 years.
Benner's42 “Novice to Expert” model to allow for ad-
justment for variations of experience level between Variables
nurses in the statistical model. The dependent variable was the total number of falls
on the shift (inclusive of all types of falls). Indepen-
Conceptual Definitions dent variables included nurse activities (as recorded
Hourly Rounding by the observer including BSR and HR), census, and
The healthcare system in which the study took place staffing data (Table 1).
does not have organizational policies that communi-
cate the expectation of staff to perform HR per proto-
col; however, patient experience guidelines are in Analysis Plan
place that suggest nursing staff complete hourly The sampling plan was consistent with strategies rec-
rounds during each shift. There are no regulations ommended by Sittig41 for work sampling studies. A
set to define when HRs are expected to occur during sample size of 2100 observations (“observations” re-
the hour, the format by which they should occur, or ferring to the number of times an activity is docu-
the specific elements that should be included in HR. mented rather than a unit of time) was calculated,
Because of this and the fact that previous studies have using 30% of nursing time spent on direct care activ-
depicted a wide range of definitions of HR, we opted ities based on historical studies, to detect a small effect
to make our definition as broad as possible to give the size using the methods described by Sittig.41 Observa-
greatest chance of capturing HR. Therefore, we de- tional data collected by the research assistant (RA) were
fined HR as anytime a nurse was observed with the entered into IBM SPSS Statistics for Macintosh, version
patient at least once in an hour period. 25.0 (IBM Corp, Armonk, New York) for analysis.
Bedside Shift Report Aim 1: Describe the Frequency of BSR and HR in
The healthcare system the study took place has orga- Inpatient Medical/Surgical Units Using
nizational policies that outline expectations for nurs- Work-Sampling Technique
ing staff to participate in BSR for every patient, Descriptive statistics were calculated to assess the fre-
every shift. During the transfer of care, nurses use a quency of BSR and HR. HR was a categorical vari-
standardized handoff process to communicate rele- able (completed = yes or no) counted as any time a
vant patient information, introduce the oncoming nurse was observed with the patient during 1 of the
nurse, answer any patient and/or family questions 4 observations per hour (therefore, if a nurse was ob-
and concerns, and perform an environmental safety served with a patient at 8:15, HR was recorded as
check. The environmental scan includes ensuring the “yes” for the 8:00, 8:15, 8:30, and 8:45 observa-
patients' call bell is within reach, any equipment is tions). For the purpose of this study, we counted
functioning properly, the bed is in the lowest position BSR as any observation where a nurse was giving re-
with appropriate side rails up, and all clinical alarms port to another nurse at the bedside (regardless of
are functioning and properly set (ie, bed alarm). Sim- time of shift).
ilarly, the nurse going off duty is responsible for com-
municating patients' fall risk factors and fall prevention Aim 2: Explore the Relationship Between Nurse
interventions specific for each patient. However, as Surveillance Including BSR and HR and Patient Falls
above, we opted to make our definition as broad as H2: BSR and HR May Be Negatively Associated
possible to give the greatest chance of capturing BSR. With Patient Falls
Therefore, for the purpose of this study, we counted Data on falls collected by the hospital were obtained
BSR as any observation where a nurse was giving re- and tested for associations with BSR and HR via the
port to another nurse at the bedside (regardless of time 2-sample t test (or Wilcoxon rank-sum test) and the
of shift) during a shift. w2 test (or Fisher exact test), as appropriate.

JONA  Vol. 50, No. 6  June 2020 357

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.


Table 1. Variable Sources and Definitions
Variable Functional Definition Variable Type

Source: Observational data collected using work-sampling technique


Nurses' activities Nursing tasks recorded by observer (see SDC 1) Continuous (no. per
12-h shift)
BSR by nurses Two nurses observed at bedside giving report on the patient at least Continuous (no. per
once per 12-h shift. 12-h shift)
HR by nurses Nurse observed with patient at least once per hour Continuous (no. per
12-h shift)
Frequency of nurse interactions No. of times nurse is at bedside during 12-h shift Continuous (no. per
with patients 12-h shift)
Source: Existing hospital quality data collected through NYP Quality Department, reported according to national reporting
standards
Patient falls Existing assisted and unassisted falls data as defined by National Categorical (yes or no) per
Database of Nursing Quality Indicators 12-h shift
Census data No. of patients on the unit during the shift Continuous (number
per 12-h shift)
Source: Existing hospital data collected through Hospital Personnel Department
Nurse staffing data No. of unique nurses during the shift Continuous (no.
Staffing ratios on the unit per shift per 12-h shift)

Because the fall data were at the shift level, corre- November 2018 over 281 shifts. The RAs consisted
lations between BSR and HR and falls were con- of 3 RNs, 2 certified nurse specialists at the sites
ducted at the shift level. A score to represent level of (but who did not work on the units), and 1 RA with
nursing expertise was created by using the average a master's degree in public health; 2 were hired, and
of the nurse level of expertise per shift. Identified pre- the others were volunteers. All hours of the day and
dictor variables with P < .30 on univariate analysis days of the week were observed. To ensure that data
(ie, for univariate relationship between predictor var- were collected consistently, we trialed the data collec-
iable of interest and fall binary outcome) were entered tion process by having the PI and all the RAs at each
into a multivariable logistic regression model.43,44 site collect data simultaneously on the same nurses
Collinearity between predictors in the model were to compare results for interrater reliability between
evaluated prior to the formulation of the final multi- data collectors. We discussed differences and pro-
variable model, and outliers were removed from the ceeded until we achieved 100% agreement among
final model. A Hosmer and Lemeshow test was per- raters (because the observations occur at an exact mo-
formed to test goodness of fit of the final model.45 Ad- ment in time, and nurses move quickly from one task
justed odds ratios (ORs) and 95% confidence intervals to the next, there was some variability between ob-
(CIs) for all predictor variables of interest were esti- servers because of differences in timepoints, but no
mated from the multivariable model. All P values were variability in assessment of what the activities were).
2-sided with statistical significance evaluated at the .05 We subsequently retrialed the data collection process
a level.46,47 Ninety-five percent confidence intervals for and compared results for interrater reliability between
all parameters were calculated to assess the precision data collectors; 78% agreement was achieved, which
of the obtained estimates. describes strong agreement.48 The vast majority of
the observations were collected by a single observer
Recruitment
(one of the paid RAs, n = 8626 [88.9%]) over a period
Before the observation, the RA consented and en- of about a year.
rolled nurses to participate in the study. The purpose
of the study and the study procedures were explained, Description of Nurse Surveillance
and the RA answered any questions. A handout was
Nurses conducted HR in 57.7% (n = 5595) of the pos-
provided with information about the study and con-
sible instances to do so. BSR was observed in 18.9%
tact information of the principal investigator (PI)
(n = 53) of the shifts. The nurse was observed with
and IRB; nurses were asked to self-rank their level of
the patient in 23.1% of the observations (n = 2244).
expertise.
Associations With Falls
Results Tests for associations between falls and predictor var-
Six RAs collected a total of 9696 observations from iables revealed statistically significant differences be-
11 units and 168 nurses between November 2017 to tween day and night shift associations between falls

358 JONA  Vol. 50, No. 6  June 2020

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.


Discussion
Table 2. w2 Categorical Variable Association
With Falls In this study, we observed HR in 57.7% of the obser-
Variable 2
w Statistic P vations; because of the paucity of direct observational
studies of this type, it is unknown if this represents
Unit 7.49 .68 higher or lower than average HR. Likewise, BSR
Day shift/night shift 5.34 .02
Day of the week 17.26 .01 was observed in 18.9% of the observations, but it is
Weekend/weekday 0.07 .79 unknown how this compares nationally because there
Site 3.55 .31 are few objective data on this topic. In addition to
BSR 0.01 .91
these metrics, we found the nurse was observed pro-
Significant at P = .05 level (2-sided). Those with a P < .3 were en- viding direct patient care in 23.1% of the observations.
tered into the logistic regression.
This is consistent with but slightly higher than previous
studies, which have suggested nurses spend between
and day shift/night shift (w2 = 5.34, P = .02) with falls 19.3% and 20.4% on direct patient care in medical/
more often occurring on night shift; and day of the surgical units.49,50
week (w2 = 17.26, P < .01) with falls most often occur- Interestingly, this study suggested that while HR
ring on Monday (Table 2). Other predictor variables could potentially reduce patient falls, the frequency
were not statistically significant (Tables 2 and 3). of interactions at the bedside alone may not reduce
Because census and staffing levels were found to patient falls. Intuitively, nurses at the bedside more
have high multicollinearity, these factors were not in- frequently may signal that the patient is a higher fall
cluded in the final model; retesting after removal of risk; future studies could adjust for this through the
these factors resulted in a variance inflation factor use of electronic health records to identify diagnoses,
(VIF) <2 for all independent variables. A binomial lo- comorbidities, and so on, to elucidate this phenomenon.
gistic regression was performed to ascertain the effects While this study included both community and
of variables with P < .3 (shift, day of the week, fre- academic medical center hospitals, falls rates were
quency with patient and HR) on the likelihood of a fall comparable between the 2 settings. It is important to
during the shift; 8 cases were identified as outliers and note that all hospitals in this study are in urban areas.
removed from the final logistic regression. The final Further investigation of the relationship of nurse
model was statistically significant w2 = 41.802, P < rounding and falls in the rural versus urban setting
.01). The Hosmer and Lemeshow test for goodness is needed. Researchers have observed regional differ-
of fit was w2 = 0.439, P > .99, indicating a good fit. ences in falls51 that have been attributed to differences
The Nagelkerke R2 test was 0.666, indicating 66.6% in the environment, hospital characteristics, and local
of the variance in falls can be explained by the model. practices.52 A study of 3 rural hospitals in Australia
In the final model, only the number of times the nurse found mixed results in patient falls following imple-
was observed with the patient remained significant mentation of BSR.10 More research focusing on the
(B = 0.564, P = <0.01), indicating for 1-unit increase setting-specific factors for falls in urban versus rural
in falls we expect a 0.564 increase in the number of hospitals is also needed.
observations with the patient (OR, 1.76; CI, 1.148– This study is limited by the fact that the data were
2.690). However, HR was near significance; for each collected in 15-minute intervals rather than through
reduction in the number of times the patient had HR, constant observation; these intervals may not have
the odds of a fall increased by a factor of 1.20 (OR, captured every BSR or incidence of HR. The use of
0.831; CI, 0.674–1.024; P = .08) (Table 4). technology to conduct studies to provide continuous

Table 3. Independent t Test for Mean of Continuous Variables With Number of Falls
95% CI
Sig (2-Tailed)
Variable t Test Statistic P Lower Upper

Average level of nursing expertise 0.45 0.65 −0.44 0.71


No. of observations with patient −1.88 0.08 −11.62 0.74
No. of times observed doing HR −1.33 0.21 −18.94 4.43
Average census −1.70 0.11 −5.78 1.46
Average staffing −0.87 0.41 −0.74 0.34

Significant at P = .05 level (2-sided). Those with P < .3 were entered into the logistic regression.

JONA  Vol. 50, No. 6  June 2020 359

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.


Table 4. Statistically Significant or Near-Significant Variables in Final Logistic Regression Model With
Falls as the Outcome Variable
Variables in the Equation
95% CI for Exp(B)
B SE Wald df P Exp(B) Lower Upper

No. of times nurse observed with patient 0.564 0.217 6.734 1 0.009 1.757 1.148 2.690
No. of times HR = yes −0.185 0.107 3.015 1 0.083 0.831 0.674 1.024
Constant −42.114 5982.930 0.000 1 0.994 0.000

Test for significance of model w2 = 41.802, P < .01, indicating a statistically significant model. Hosmer and Lemeshow test w2 = 0.439, P > .99, indicating
a good fit. Nagelkerke R2 = 0.666, indicating 66.6% of the variance in falls can be explained by the model.

monitoring (eg, obfuscated computer vision or radio- improving patient outcomes. Continued studies to illu-
frequency identification) could improve this study minate the utility of measures such as BSR and HR
and further clarify the utility of BSR and HR, and could help understand the most cost-effective use of
the overall influence of nurse surveillance. Moreover, time and resources, as well as how to maintain the qual-
this study did not assess the quality of the interactions ity of nurse-patient interactions. Institutions should ex-
with patients; future studies could examine this aspect amine the possibility that factors such as day of the
of nurse-patient interactions. Finally, because of the week and shift may play a role in falls, and nursing
low number of falls during the data collection period, management should also consider ways to increase
a longer study or sites with higher fall rates may be the sustainability of BSR and HR after rollout.
able to better capture the effects of BSR and HR on While there have been an increasing number of
patient falls. Future studies could also consider adding studies conducted on the implementation of BSR
patient characteristics to better describe the nature of and HR, less is known about the long-term sustain-
the falls and associated characteristics. While this study ability of these measures or subsequent effects on re-
objectively examines the frequency of nursing interac- ducing falls. This study contributes to the literature
tions of patients, further research is needed to improve by examining the effects of these measures long after
the understanding the quality of the interactions, as the initial rollout and suggests hospital leadership
well as the component of interactions that actually af- should consider how to ensure sustainability of these
fect patient outcomes in order to justify the cost for practices to increase the long-term benefits.
nursing management to implement measures such as
HR and BSR. Acknowledgments
The authors thank the reviewers and editor for their
Conclusion comments, which helped them refine and strengthen
Reducing inpatient falls is a major concern of hospi- their manuscript. They also acknowledge with grati-
tals nationally and globally. There may be continued tude funding that supported this study: Columbia
opportunities to redistribute nursing workload or to University School of Nursing Intramural Grant and
redesign workflow such that the nurse has increased a gift from Inspiren, Inc. Finally, they thank all of
opportunities for direct patient contact, including the nurses who participated in the study, not only
BSR and HR; future studies could help elucidate for helping them with this study but also for all the
whether these measures are necessary or helpful in work these nurses do every day.

References

1. McGlynn EA, Asch SM, Adams J, et al. The quality of health and nonfatal falls in older adults [published online March 7,
care delivered to adults in the United States. N Engl J Med. 2018]. J Am Geriatr Soc. 2018;66:693-698. doi:10.1111/jgs.15304.
2003;348(26):2635-2645. 5. Spetz J, Brown DS, Aydin C. The economics of preventing hos-
2. Centers for Medicare & Medicaid Services. (2007). CMS- pital falls: demonstrating ROI through a simple model. J Nurs
1533-FC. Medicare program; changes to the hospital prospec- Adm. 2015;45:50-57. doi:10.1097/NNA.0000000000000154.
tive payment system and fiscal year 2008 rates. (pp. 352-357). 6. Chan DKY, Sherrington C, Naganathan V, et al. Key issues to
3. Centers for Medicare & Medicaid Services. Falls. https://www.who. consider and innovative ideas on fall prevention in the geriatric
int/news-room/fact-sheets/detail/falls. Accessed February 10, 2020. department of a teaching hospital. Australas J Ageing. 2018.
4. Florence CS, Bergen G, Atherly A, et al. Medical costs of fatal doi:10.1111/ajag.12528.

360 JONA  Vol. 50, No. 6  June 2020

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.


7. Goldsack J, Bergey M, Mascioli S, Cunningham J. Hourly handoffs a systematic review of the literature. J Nurs Care Qual.
rounding and patient falls: what factors boost success? Nursing 2016;31:54-60. doi:10.1097/NCQ.0000000000000142.
(Lond). 2015;45:25-30. 28. Agency for Healthcare Research and Quality. Strategy 3: nurse
8. Wakefield DS, Ragan R, Brandt J, Tregnago M. Making the bedside shift report implementation handbook. Guide to Patient
transition to nursing bedside shift reports. Jt Comm J Qual Pa- and Family Engagement in Hospital Quality and Safety. 2013.
tient Saf. 2012;38(6):243-253. doi:10.1016/s1553-7250(12) https://www.ahrq.gov/sites/default/files/wysiwyg/professionals/
38031-8. systems/hospital/engagingfamilies/strategy3/Strat3_Implement_
9. Venema DM, Skinner AM, Nailon R, et al. Patient and system Hndbook_508.pdf. Accessed February 10, 2020.
factors associated with unassisted and injurious falls in hospi- 29. The Joint Commission. Sentinel Event Alert 58: Inadequate
tals: an observational study. BMC Geriatr. 2019;19:1-10. doi: hand-off communication. 2017. https://www.jointcommission.
10.1186/s12877-019-1368-8. org/-/media/tjc/documents/resources/patient-safety-topics/
10. Bradley S, Mott S. Handover: faster and safer? Aust J Adv Nurs. sentinel-event/sea_58_hand_off_comms_9_6_17_final_(1).
2012. pdf. Accessed February 10, 2020.
11. Morello RT, Soh SE, Behm K, et al. Multifactorial falls preven- 30. Benson E, Rippin-Sisler C, Jabusch K, et al. Improving nursing
tion programmes for older adults presenting to the emergency shift-to-shift report. J Nurs Care Qual. 2007;22:80-84. doi:
department with a fall: systematic review and meta-analysis. 10.1097/00001786-200701000-00015.
Inj Prev. 2019;25:557-564. doi:10.1136/injuryprev-2019- 31. Mitchell MD, Lavenberg JG, Trotta RL, Umscheid CA. Hourly
043214. rounding to improve nursing responsiveness: a systematic re-
12. Gulka HJ, Patel V, Arora T, et al. Efficacy and generalizability view. J Nurs Adm. 2014;44:462-472. doi:10.1097/NNA.
of falls prevention interventions in nursing homes: a systematic 0000000000000101.
review and meta-analysis. J Am Med Dir Assoc. 2020;33(5): 32. Bressan V, Cadorin L, Pellegrinet D, et al. Bedside shift hand-
413-425. doi:10.1016/j.jamda.2019.11.012. over implementation quantitative evidence: findings from a
13. Lake ET, Shang J, Klaus S, et al. Patient falls: association with scoping review. J Nurs Manag. 2019;27:815-832. doi:10.1111/
hospital Magnet status and nursing unit staffing. Res Nurs jonm.12746.
Health. 2010;33:413-425. doi:10.1002/nur.20399. 33. Bharucha AJ, London AJ, Barnard D, Wactlar H, Dew MA,
14. Aiken LH, Clarke SP, Sloane DM. Hospital staffing, organiza- Reynolds CF 3rd. Ethical considerations in the conduct of elec-
tion, and quality of care: cross-national findings. Nurs Out- tronic surveillance research. J Law. 2006;34:611-619. doi:10.
look. 2002;50:187-194. doi:10.1067/mno.2002.126696. 1111/j.1748-720X.2006.00075.x.
15. Aiken LH, Clarke SP, Sloane DM, Sochalski J, Silber JH. Hos- 34. Meade CM, Bursell AL, Ketelsen L. Effects of nursing rounds:
pital nurse staffing and patient mortality, nurse burnout, and on patients' call light use, satisfaction, and safety. Am J Nurs.
job dissatisfaction. J Am Med Assoc. 2002;288:1987-1993. 2006;106:58-70. doi:10.1097/00000446-200609000-00029.
doi:10.1001/jama.288.16.1987. 35. Kirkland KB, Weinstein JM. Adverse effects of contact isola-
16. Aiken LH, Sloane DM, Bruyneel L, et al. Nurse staffing and ed- tion. Lancet. 1999;354:1177-1178. doi:10.1016/S0140-6736
ucation and hospital mortality in nine European countries: a (99)04196-3.
retrospective observational study. Lancet. 2014;383:1824-1830. 36. Cohen B, Hyman S, Rosenberg L, et al. Frequency of patient
doi:10.1016/S0140-6736(13)62631-8. contact with health care personnel and visitors: implications
17. Fry L, Fry L, Walker M, et al. Use of robotic cats to reduce falls for infection prevention. Jt Comm J Qual Patient Saf. 2012;
in skilled nursing facility. J Am Med Dir Assoc. 2018;19:B22-B23. 38:560-565. doi:10.1016/S1553-7250(12)38073-2.
doi:10.1016/j.jamda.2017.12.070. 37. Huang A, Leafloor C, Lochnan H, et al. Time-motion studies of
18. Needleman J, Buerhaus P, Pankratz VS, et al. Nurse staffing internal medicine residents' duty hours: a systematic review and
and inpatient hospital mortality. N Engl J Med. 2011;364: meta-analysis. Adv Med Educ Pract. 2015. doi:10.2147/amep.
1037-1045. doi:10.1056/NEJMsa1001025. s90568.
19. Crisp N, Chen L. Global supply of health professionals. N Engl 38. Blay N, Duffield CM, Gallagher R, et al. Methodological inte-
J Med. 2014;370:950-957. doi:10.1056/NEJMra1111610. grative review of the work sampling technique used in nursing
20. Salmond SW, Echevarria M. Healthcare transformation and workload research. J Adv Nurs. 2014;70:2434-2449. doi:10.
changing roles for nursing. Orthop Nurs. 2017;36:12-25. doi: 1111/jan.12466.
10.1097/NOR.0000000000000308. 39. Pelletier D, Duffield C. Work sampling: valuable methodology
21. Kutney-Lee A, Lake ET, Aiken LH. Development of the hospi- to define nursing practice patterns. Nurs Health Sci. 2003;5:
tal nurse surveillance capacity profile. Res Nurs Health. 2009; 31, 38. doi:10.1046/j.1442-2018.2003.00132.x
32:217-228. doi:10.1002/nur.20316. 40. Finkler SA, Knickman JR, Hendrickson G, et al. A comparison
22. Series QC. Keeping patients safe: transforming the work envi- of work-sampling and time-and-motion techniques for studies
ronment of nurses. J Healthc Qual. 2004;26:56. doi:10.1111/ in health services research. Health Serv Res. 1993.
j.1945-1474.2004.tb00487.x. 41. Sittig DF. Work-sampling: a statistical approach to evaluation
23. Dresser S. The role of nursing surveillance in keeping patients safe. of the effect of computers on work patterns in healthcare. In:
J Nurs Adm. 2012;42:361-368. doi:10.1097/NNA.0b013e3182619377. Anderson JG, Aydin CE, eds. Evaluating the Organizational
24. Hutchinson M, Higson M, Jackson D. Mapping trends in the Impact of Healthcare Information Systems. New York: Springer:
concept of nurse rounding: a bibliometric analysis and research 2005. 174-88. https://link-springer-com.ezproxy.cul.columbia.edu/
agenda. Int J Nurs Pract. 2017;23(6). doi:10.1111/ijn.12584. content/pdf/10.1007/0-387-30329-4.pdf. Accessed June 18, 2019.
25. Hicks D. Can rounding reduce patient falls in acute care? An in- 42. Benner P. From novice to expert. Am J Nurs. 1982;82:402-407.
tegrative literature review. Medsurg Nurs. 2015. doi:10.1097/00000446-198282030-00004.
26. Walsh J, Messmer PR, Hetzler K, et al. Standardizing the bed- 43. Hilbe JM. Data analysis using regression and multilevel/
side report to promote nurse accountability and work effective- hierarchical models. J Stat Softw. 2009;30. doi:10.18637/jss.
ness. J Contin Educ Nurs. 2018;49:460-466. doi:10.3928/ v030.b03.
00220124-20180918-06. 44. Cohen J, Cohen P, West SG, et al. Alternative regression
27. Mardis T, Mardis M, Davis J, et al. Bedside shift-to-shift models: logistic, Poisson regression, and the generalized linear

JONA  Vol. 50, No. 6  June 2020 361

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.


model. In: Cohen J, Cohen P, West SG, et al, eds. Applied medical-surgical nurses spend their time? Perm J. 2008;12:
Multiple Regression/Correlation Analysis for the Behavioral 25-34. doi:10.7812/tpp/08-021.
Sciences. Mahwah, New Jersey: Routledge; 2013. 509-65. 50. Westbrook JI, Duffield C, Li L, et al. How much time do nurses
doi:10.4324/9780203 774441-17 have for patients? A longitudinal study quantifying hospital
45. Kramer AA, Zimmerman JE. Assessing the calibration of mor- nurses' patterns of task time distribution and interactions with
tality benchmarks in critical care: the Hosmer-Lemeshow test health professionals. BMC Health Serv Res. 2011;11. doi:10.
revisited. Crit Care Med. 2007;35:2052-2056. doi:10.1097/01. 1186/1472-6963-11-319.
CCM.0000275267.64078.B0. 51. Bae SH, Yoder LH. Implementation of the centers for medicare
46. Tabachnick BG, Fidell LS. Using Multivariate Statistics. 3rd ed. and medicaid services' nonpayment policy for preventable
New York: Harper Collins; 1996. hospital-acquired conditions in rural and nonrural US hospi-
47. Tabachnick BG, Fidell LS. Using multivariate statistics, 2019. tals. J Nurs Care Qual. 2015;30:313-322. doi:10.1097/NCQ.
https://lccn.loc.gov/2017040173. Accessed February 7, 2020. 0000000000000119.
48. McHugh ML. Interrater reliability: the kappa statistic. Biochem 52. Girotra S, Cram P, Popescu I. Patient satisfaction at America's
Med. 2012;22:276-282. doi:10.11613/bm.2012.031. lowest performing hospitals. Circ Cardiovasc Qual Outcomes.
49. Hendrich A. A 36-hospital time and motion study: how do 2012;5:365-372. doi:10.1161/CIRCOUTCOMES.111.964361.

362 JONA  Vol. 50, No. 6  June 2020

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.

You might also like