Content uploaded by David Russell
Author content
All content in this area was uploaded by David Russell on Jan 21, 2018
Content may be subject to copyright.
Continuity in Home Health Care: Is
Consistency in Nursing Personnel
Associated with Better Patient Outcomes?
David Russell, Robert J. Rosati, Peri Rosenfeld, Joan M. Marren
Home health care encompasses a range of
professional and supportive services (e.g.,
skilled nursing, rehabilitation, home health
aides) that aim to help individuals manage
their conditions and remain at home, avoiding
admission to a hospital or an institutional long-
term care setting. As one of the most rapidly
expanding sectors of the health care system,
research and policy experts have begun paying
greater attention to the quality of these home
health services. Recent efforts to improve
home health care quality have largely focused
on preventing hospital readmissions (Rosati &
Huang, 2007). Far less attention has been paid
to the more proximal characteristics of home
health services, including whether visits to
patients are coordinated and delivered in a
consistent manner. In an attempt to address
these issues, we use electronic health records
and other staffing data to: (1) describe the
level of continuity in home health patient care,
and (2) examine the relationship between con-
tinuity of care and patient outcomes.
Background
The term continuity of care is used by health
services researchers to describe patient–
provider relationships that are connected and
coordinated across time and settings (Donald-
son, 2001; Sparbel & Anderson, 2000). Prior
research suggests that ensuring continuity in
the process through which patients receive
health care improves safety, reduces costs, and
lowers the rate of emergency department utili-
zation (Christakis, Wright, Koepsell, Emerson,
& Connell, 1999). While several studies have
focused on the relationship between continuity
of care and patient outcomes in the hospital or
psychiatric setting, relatively little research has
examined this issue as it relates to home health
services (Bell et al., 2009; Christakis et al., 1999;
Saultz & Lochner, 2005; Ware, Tugenberg,
Dickey, & McHorney, 1999). Home health
patients tend to have multiple chronic condi-
tions and cognitive impairments that present
challenges to care planning and management,
and that often lead to longer lengths of stay
(Murtaugh et al., 2009). Inconsistent or unco-
ordinated care can compromise rapport and
decrease the provider’s ability to observe and
interpret changes in a patient’s appearance or
behavior (Woodward, Abelson, Tedford, &
Hutchison, 2004). Lapses in observation, ap-
propriate knowledge and skills, and trust
between the patient and provider may affect
the quality of home health services, possibly
leading to poorer patient outcomes (Wood-
ward et al., 2004). The present study tests the
hypothesis that patients who received home
health care by as many of the same nursing
personnel as possible would have a lower risk of
being hospitalized or receiving emergent care,
and have a greater chance of improving func-
tioning in activities of daily living between
admission and discharge from home health
care.
Study Design and Methods
Patient Population
This study is based on retrospective data from a
large urban not-for-profit home health care
Abstract: A growing body of evidence suggests that patients who
receive coordinated and uninterrupted health care services have
better outcomes, more efficient resource utilization, and lower
costs of health care. However, limited research has considered
whether attributes of continuity in home health care service
delivery are associated with improved patient outcomes. The
present study examines the relationship between one dimension
of continuity of care, consistency in nursing personnel, and three
patient outcomes: hospitalization, emergent care, and
improvement in activities of daily living. Analyses of data from
a large population of home health patients (N559,854) suggest
that greater consistency in nursing personnel decreases the
probability of hospitalization and emergent care, and increases
the likelihood of improved functioning in activities of daily living
between admission and discharge from home health care. These
results provide preliminary evidence that efforts to decrease
dispersion of nursing personnel across a series of home visits to
patients may lead to improved outcomes. The implications of
these findings for clinical practice and further research are
discussed in the paper.
Keywords
continuity of care
emergent care
functional limitations
home health care
hospitalization
patient outcomes
Vol. 33 No. 6 November/December 2011 33
Journal for Healthcare Quality
Vol. 33, No. 6, pp. 33–39
&2011 National Association for
Healthcare Quality
Journal for Healthcare Quality
agency. The Outcomes and Assessment Infor-
mation Set and administrative records are
utilized from a population of home health
care patients with acute care needs who meet
the following eligibility criteria: (1) patients
who received two or more skilled nursing visits
and (2) patients who were admitted and dis-
charged during the 2008 calendar year. The
analysis was limited to nursing services pro-
vided by RNs and excluded other types of
professional and paraprofessional services
(e.g., physical therapy, clinical social work,
home health aides). These criteria produced
a sample of 59,854 cases. The total number of
nursing visits provided to patients in this pop-
ulation ranged from 2 to 461, with the mean
and median at nine and six visits, respectively.
Measuring Continuity of Care
The measure used to estimate continuity in
home health care services is based on a for-
mula that was initially developed to model
dispersion in patient–provider contact across a
number of sources (Bice & Boxerman, 1977;
Shortell, 1976). This measure is preferable to
other quantitative indicators of continuity of
care, because it takes into account the number
of providers, the number of interactions with
each provider, and the total number of inter-
actions with all providers. The formula was
adapted for use in the current study as a mea-
sure of consistency in nursing personnel across
a series of home health care visits. While pa-
tients receive their skilled nursing services from
a single source (i.e., a local home health
agency), they may receive visits from multiple
providers of care (i.e., different nurses).
The formula produces scores that range be-
tween 0 and 1. Higher scores are indicative of
greater continuity in home health care services,
while lower scores suggest less continuity. Fig-
ure 1 presents an illustration of how these
scores are computed for hypothetical patients
with a series of 15 home care visits (the full
computation of each patient’s score is shown in
the notes below the figure). In the first case, a
hypothetical patient (Patient A) received all 15
visits from the same nurse and is assigned a
perfect score of 1. A second hypothetical pa-
tient (Patient B), who had 11 home visits from
Nurse E and 4 visits from Nurse F, is assigned a
score of 0.58. Hypothetical Patients C and D
received even fewer visits from the same
nurse(s), and have scores (0.15 and 0.06, re-
spectively) that reflect the greater dispersion of
service providers across their series of home
health care visits.
COC ¼Ps
j¼1n2
jn
nðn1Þ
where nis the total number of home visits, n
j
is
the number of home visits by nurse j, and sis
the total number of nurses.
Patients in the analytic sample tended to fall
in the middle range of the scoring distribution
on this measure. The mean sample score is
0.54 (SD 50.34) and the median value is 0.50.
In comparison with the examples in Figure 1,
the average case in the analytic sample most
closely resembles hypothetical Patient B. How-
Case Nurse Dispersion Continuity Score
Patient A EEEEEEEEEEEEEEE 1.00 1
Patient B EEEEFFEEEFFEEEE 0.58 2
Patient C EEEEFFFGGHHHIII 0.15 3
Patient D EFGGHIJJKLLLMNN 0.06 4
Figure 1. Examples of Scores Based on Nurse Dispersion
Notes.
w
All patients received a total of 15 nursing visits. Patients and nurses are represented by different letters.
1
Patient A continuity score 5[(15
2
)15]/[15(15 1)] 5210/210 51.00;
2
Patient B continuity score 5[(11
2
14
2
)
15]/[15(15 1)] 5122/210 50.58;
3
Patient C continuity score 5[(4
2
13
2
12
2
13
2
13
2
)15]/[15(15 1)] 5
32/210 50.15; and
4
Patient D continuity score 5[(1
2
11
2
12
2
11
2
11
2
12
2
11
2
13
2
11
2
12
2
)15]/[15(15 1)] 5
12/210 50.06.
Journal for Healthcare Quality
34
ever, a substantial proportion of patients had
scores on opposite ends of the measurement
distribution. Twenty-six percent of patients re-
ceived visits from the same nurse (a score of 1)
over the course of their length of stay, while
10% of patients never saw the same visiting
nurse over their entire period of home care (a
score of 0). Based on the nonnormal distribu-
tion of the continuity measure, cases were
divided into three groups: low continuity of
care (scores between 0.00 and 0.39), moderate
continuity of care (scores between 0.40 and
0.79), and high continuity of care (scores be-
tween 0.80 and 1.00; reference category in
logistic regression analyses).
Measuring Patient Outcomes and
Statistical Controls
Three patient outcomes are measured, includ-
ing hospitalization, emergency room visits, and
improved functioning in activities of daily liv-
ing. In accordance with prior research on
home health episodes, hospitalizations and
emergency room visits are counted only if they
occurred within 60 days of admission to home
health care (Rosati & Huang, 2007). Improved
functioning in activities of daily living is de-
fined as a dichotomous variable representing
whether there is a difference between admis-
sion and discharge from home health care in
the number of activities of daily living that a
patient needed assistance with or to perform.
Patients who required assistance with fewer
activities of daily living at discharge from
home health care compared with admission
to home health care are assigned a value of 1.
Those who required assistance with the same
number of activities of daily living, or a higher
number of activities of daily living, at dis-
charge from home health care compared with
admission to home health care are assigned a
value of 0.
To assess the association between continuity
in home health care and patient outcomes, we
estimate three logistic regression equations
that include dummy variables for low and mod-
erate levels of continuity. The regression
equations control for a range of demographic
and clinical factors, including geographic re-
gion of service, age, gender, race/ethnicity,
living arrangement, length of stay, diagnosis
(based on the International Classification of
Diseases, 9th Revision; Iezzoni, 1997), comor-
bidity (measured as the number of diagnoses in
addition to the primary diagnosis), number of
activities of daily living that the patient requires
assistance, estimated life expectancy (measured
as a dichotomous variable where o6 months is
coded as 1) and overall prognosis (measured as
a dichotomous variable where poor is coded as
1). Descriptive statistics for each demographic
and clinical variable stratified by level of con-
tinuity in home health care (low, moderate,
and high) are presented in Table 1.
Results
Table 2 presents adjusted odds ratios that
estimate the likelihood of certain patient out-
comes (hospitalization, emergency room
visitation, and improved functioning in activi-
ties of daily living) across three levels of
continuity in home health care, controlling
for a wide range of demographic and clinical
characteristics. Patients with high levels of con-
tinuity in home health care are compared with
patients with moderate and low levels of con-
tinuity. The results in Table 2 are consistent
with previous research that links greater con-
tinuity of care to improved patient outcomes.
For instance, patients with a low level of con-
tinuity are 1.4 times more likely to be
hospitalized and 1.3 times more likely to visit
the emergency room compared with those with
a higher level of continuity. Patients with low
continuity of care are 0.8 times less likely (20%
less likely) to have improved functioning in ac-
tivities of daily living at discharge from home
health care compared with patients with high
continuity of care. Patients with a moderate
level of continuity are 1.1 times more likely
than those with high continuity in their pro-
viders of home care to be hospitalized or to
visit the ER, but are not significantly different
in their likelihood of improved functioning in
activities of daily living. Presented another way,
Figure 2 displays a graph of predicted proba-
bilities for each of the three patient outcomes
across values of the continuity of care measure.
The predicted probabilities, which include
controls for demographic and clinical charac-
teristics, indicate a statistically significant linear
trend in the relationship between consistency
of nursing personnel and patient outcomes.
Discussion
Researchers have argued that efforts to in-
crease the continuity of patient care are central
to improving patient outcomes across health
Vol. 33 No. 6 November/December 2011 35
care settings (Bell et al., 2009; Christakis et al.,
1999; Saultz & Lochner, 2005; Ware et al.,
1999). This study offers preliminary evidence
that providing home health care patients with
consistent nursing personnel may reduce the
rate of episodes ending in hospitalization or a
visit to the emergency room, and increase the
likelihood of improving functioning in activi-
ties of daily living between admission and
discharge from home health care. Given the
significant associations observed with patient
outcomes, consistency in nursing personnel
may be an important indicator for home health
agencies to track as part of their quality im-
provement efforts. While other measures of
continuity in home health care should be con-
sidered, our results suggest that the measure
used in the present study could be used as a
tool to drive aggregate changes in patient out-
comes.
Some lapses in continuity in the provision of
home health care services are expected due to
scheduling conflicts and other staffing issues
such as patient caseloads with daily or twice
daily nursing visits (e.g., for wound treat-
ments). However, a low level of consistency in
the provider of care clearly compromises the
stability and therapeutic benefits of the rela-
tionship between nurse and patient. Patients
are likely to become frustrated when they must
constantly reexplain their situation and service
delivery needs to a clinician with whom they do
Table 1. Demographic and Clinical Characteristics Stratified by Continuity of Care
Total Sample
(N559,854)
Low Continuity
(n523,816)
Mod Continuity
(n518,158)
High Continuity
(n517,880)
Age (SD) 70.4 (16.3) 68.8
(16.7) 71.7 (16.0) 71.1 (16.0)
Gender (% female) 61.4 60.6
62.3 61.6
Race (% White) 30.5 30.5 30.8 30.0
Payer (%)
Medicare fee-for-service 34.5 31.1
39.0 34.4
Medicaid fee-for-service 4.4 4.3
4.9 3.9
Dually eligible 12.0 10.5
14.4 11.7
Managed care 45.1 49.8
38.2 46.0
Other 3.5 4.0
3.0 3.5
Living arrangements (% alone) 32.4 30.3
34.7 32.9
Mean length of stay (SD) 36.1 (32.4) 32.9
(31.9) 44.7 (35.7) 31.4 (27.7)
Mean number of visits (SD) 9.3 (13.1) 9.5
(15.8) 11.9 (13.4) 6.4 (6.6)
ICD-9 classifications (%)
a
Congestive heart failure 14.8 13.6
15.8 15.2
Coronary heart disease 0.8 0.9 0.7 0.7
Dementia 8.4 7.5
9.0 9.1
Diabetes 2.3 2.4 2.4 2.0
Functional limitations 1.1 1.2 1.1 1.0
Liver disease 1.1 1.0 1.0 1.1
Lymphoma 2.0 2.1 1.9 1.9
Nutritional issues 0.2 0.2 0.2 0.2
Pulmonary disease 13.7 12.9
14.3 14.1
Renal failure 3.3 3.4 3.3 3.2
Vascular disease 4.5 4.4
4.9 4.0
Comorbidity (% 4 or more) 91.4 90.8
91.9 91.8
ADL (% 1 or more limitations) 91.4 90.9
91.9 91.6
Life expectancy (% o6 months) 3.4 3.6 3.2 3.3
Overall prognosis (% poor) 12.3 12.5 12.5 11.6
Outcomes
Hospitalization (%) 18.9 21.2
18.7 16.1
Emergency room visit (%) 15.3 16.8
15.4 13.3
Improvement in ADL (%) 53.4 51.1
54.7 55.1
Note. ADL 5activities of daily living.
a
International Classification of Diseases—9th Revision; Diagnostic groups are based on those described by Iezzoni (1997).
Significant difference (po.001) across groups stratified by continuity of care.
Journal for Healthcare Quality
36
not have an established level of rapport and
trust. The period of orientation reduces the
amount of time available for the clinician to
provide care (Woodward et al., 2004). Lapses
in the consistency of nursing personnel may
also effect the nurse’s ability to detect subtle
changes in a patient’s condition or care needs
(Woodward et al., 2004), potentially increasing
the likelihood of hospitalization or emergent
care. The results suggest that targeting a group
of patients with very low levels of consistency in
their nursing personnel, and intervening to in-
crease their level of consistency to a more
moderate level, would have the most measur-
able impact on patient outcomes.
Study Limitations
There are several study limitations that should
be noted. First, descriptive results indicated
variability in demographic and clinical charac-
teristics across patients stratified by continuity
of care. For example, patients with lower levels
of continuity of care tended to be younger and
have a shorter length of stay than patients with
higher continuity of care. However, these fac-
tors were adjusted for in logistic regression
equations when examining the impact of con-
tinuity of care on patient outcomes. Further,
this study focused on a single indicator of con-
tinuity in home health care—consistency in
nursing personnel. Ethnographic research of
Table 2. Adjusted Odds Ratios
a
for Selected Patient Outcomes
Hospitalization ER Visitation ADL Functioning
High continuity of care (0.80–1.00) (Reference) (Reference) (Reference)
Moderate continuity of care (0.40–0.79) 1.13 1.13 1.00
(1.07–1.20) (1.06–1.20) (0.95–1.04)
Low continuity of care (0.00–0.39) 1.43 1.33 0.80
(1.35–1.50) (1.25–1.41) (0.77–0.83)
Log likelihood ratio 3075.45
2319.99
5990.61
Note. ADL functioning 5improved functioning in activities of daily living.
a
Estimates are adjusted for geographic region, length of stay, level of functioning in activities of daily living,
comorbidity, life expectancy, overall prognosis, diagnosis, payer type, age, gender, race/ethnicity, and living
arrangement.
po.001; estimates are presented with 95% confidence intervals; significant odds ratios are highlighted in
bold text; for ease of interpretation, continuity of care scores were stratified into three groups: high continuity of
care (0.80 x1), moderate continuity of care (0.79 x0.40), and low continuity of care (0 x0.39).
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
1
Continuity of Care
Predicted Probability
ER Visit Hospitalization ADL Improvement
Figure 2. Predicted Probabilities for Patient Outcomes Across Continuity of Care
Note. The magnitudes of the relationships observed in this graph between continuity of care and patient outcomes
are statistically significant (po.001). Hospitalizations and visits to the emergency room are counted within 60 days of
admission to home health care. Improvement in functioning of activities of daily living is calculated between
admission and discharge from home health care. The predicted probabilities for each outcome are adjusted for
demographic and clinical characteristics.
Vol. 33 No. 6 November/December 2011 37
home health patients suggests that uninter-
rupted service delivery is only one component
of a continuous relationship between patient
and provider, while other characteristics such
as care planning and coordination are equally
important to patients’ experiences of continu-
ity (Ware et al., 1999; Woodward et al., 2004).
The level of coordination of patient transitions
into home health from other health care set-
tings (e.g., hospital, physician office) is also a
component in the continuum of care. Com-
munication between providers in different
health care settings is likely to be an impor-
tant determinant of outcomes among patients
who have complex health care needs and/or
challenging self-care responsibilities (Coleman,
2003; Kripalani, Jackson, Shnipper, & Cole-
man, 2007). More research is needed to
understand the ways in which these different
aspects of continuity overlap, and how they are
interrelated with outcomes among patients
who receive home health care services. Addi-
tional work is also needed to replicate these
findings in other home health care agencies,
and to determine if the level of continuity in
care is influenced by the size of the patient
population or nursing workforce. A related
question concerns whether continuity in home
care is influenced by the type of services pro-
vided (e.g., rehabilitation therapy compared
with skilled nursing or home health aide ser-
vices) and/or neighborhood characteristics.
Additional research is needed to understand
how scheduling conflicts or the frequency of
nursing visits influences consistency in the pro-
viders of care. As the number of nurses age 55
and older increases, a greater proportion of
the nursing workforce is likely to seek flexible
hours and weekly schedule changes, such as
6-hr shifts or 4-day work weeks (Hart, 2007).
Given these contextual factors, it will be im-
portant to identify a level of consistency in the
provision of home care services that can be
used by agencies as a target for their opera-
tions. Finally, further research should replicate
the findings observed in the present study and
continue to examine the relationship between
continuity of home-based nursing services and
other outcomes, such as patient satisfaction
with care.
References
Bell, C. M., Schnipper, J. L., Auerbach, A. D., Kaboli, P. J.,
Wetterneck, T. B., Gonzales, D. V., et al. (2009). Asso-
ciation of communication between hospital-based
physicians and primary care providers with patient out-
comes. Journal of General Internal Medicine, 24, 381–386.
Bice, T. W., & Boxerman, S. B. (1977). A quantitative mea-
sure of continuity of care. Medical Care, 15, 347–349.
Christakis, D. A., Wright, J. A., Koepsell, T. D., Emerson, S.,
& Connell, F. A. (1999). Is greater continuity of care
associated with less emergency department utilization?
Pediatrics, 103, 738–742.
Coleman, E. A. (2003). Falling through the cracks: Chal-
lenges and opportunities for improving transitional care
for persons with continuous complex care needs. Journal
of the American Geriatrics Society, 51, 549–555.
Donaldson, M. S. (2001). Continuity of care: A reconcep-
tualization. Medical Care Research and Review, 58, 255–290.
Hart, K. A. (2007). The aging workforce: Implications for
health care organizations. Nursing Economics, 25, 101–
102.
Iezzoni, L. I. (1997). Assessing quality using administrative
data. Annals of Internal Medicine, 127, 666–674.
Kripalani, S., Jackson, A. T., Shnipper, J. L., & Coleman, E.
A. (2007). Promoting effective transitions of care at
hospital discharge: A review of key issues for hospitalists.
Journal of Hospital Medicine, 2, 314–323.
Murtaugh, C., Peng, T., Totten, A., Costello, B., Moore, S.,
& Aykan, H. (2009). Complexity in geriatric home care.
Journal for Healthcare Quality, 31, 34–43.
Rosati, R. J., & Huang, L. (2007). Development and testing
of an analytic model to identify home healthcare pa-
tients at risk for a hospitalization within the first 60 days
of care. Home Health Care Services Quarterly, 26, 21–36.
Saultz, J. W., & Lochner, J. (2005). Interpersonal continuity
of care and care outcomes: A critical review. Annals of
Family Medicine, 3, 159–166.
Shortell, S. M. (1976). Continuity of medical care: Con-
ceptualization and measurement. Medical Care, 14, 377–
391.
Sparbel, K. J. H., & Anderson, M. A. (2000). Integrated
literature review of continuity of care: Part 1, conceptual
issues. Journal of Nursing Scholarship, 32, 17–24.
Ware, N. C., Tugenberg, T., Dickey, B., & McHorney, C. A.
(1999). An ethnographic study of the meaning of con-
tinuity of care in mental health services. Psychiatric
Services, 50, 395–400.
Woodward, C. A., Abelson, J., Tedford, S., & Hutchison, B.
(2004). What is important to continuity in home care?
Perspectives of key stakeholders. Social Science and Med-
icine, 58, 177–192.
Authors’ Biographies
David Russell, PhD, is an Evaluation Scientist at the
Center for Home Care Policy & Research, Visiting Nurse
Service of New York. Dr. Russell works with research and
administrative staff at the agency to design and implement
evaluation studies of new programs and initiatives. His
research interests include health services, medical sociology,
and aging.
Robert J. Rosati, PhD, is Vice President of Clinical
Informatics at the Center for Home Care Policy & Research,
Visiting Nurse Service of New York. At the Center,
Dr. Rosati is responsible for evaluating care delivery and
reporting on clinical outcomes for all patients served by
VNSNY. Dr. Rosati has over 20 years experience in
healthcare and is currently on the faculty of Weill Cornell
Medical College and Hofstra University. He also is a
member of the JHQ Review Panel and Editorial Board.
Peri Rosenfeld, PhD, is a Senior Evaluation Scientist at the
Center for Home Care Policy & Research, Visiting Nurse
Service of New York. Dr. Rosenfeld’s areas of research
Journal for Healthcare Quality
38
interest include aging, nursing workforce, access to care,
and information literacy. She has published widely in these
areas and is a review editor for several journals including
CIN: Computers, Informatics and Nursing; Politics, Policy
and Nursing Practice; and Medical Care. She is also a
faculty member at the New York University College of
Nursing.
Joan M. Marren, RN, MA, MEd, serves as Chief
Operating Officer of the Visiting Nurse Service of New
York and President, VNSNY Home Care. In her role, Ms.
Marren ensures consistency in, and accountability for,
implementation of agency strategy across the organization
and its subsidiaries; serves as primary advocate and
standard-bearer of quality; and directs service delivery,
organizational design change and QI efforts. Her leader-
ship is characterized by a strong focus on innovation and
practice improvement and she plays a significant role in
shaping public policy. Ms. Marren was a 2004–2007
Robert Wood Johnson Executive Nurse Fellow.
For more information on this article, contact David Russell,
at david.russell@vnsny.org.
Vol. 33 No. 6 November/December 2011 39
