Depression is considered the most frequent mental disorder among the older population
CLINICAL ISSUES
The effect of nurse–patient interaction on anxiety and depression in cognitively intact nursing home patients
Gørill Haugan, Siw T Innstrand and Unni K Moksnes
Aims and objectives. To test the effects of nurse–patient interaction on anxiety and depression among cognitively intact
nursing home patients.
Background. Depression is considered the most frequent mental disorder among the older population. Specifically, the
depression rate among nursing home patients is three to four times higher than among community-dwelling older people,
and a large overlap of anxiety is found. Therefore, identifying nursing strategies to prevent and decrease anxiety and depres-
sion is of great importance for nursing home patients’ well-being. Nurse–patient interaction is described as a fundamental
resource for meaning in life, dignity and thriving among nursing home patients.
Design. The study employed a cross-sectional design. The data were collected in 2008 and 2009 in 44 different nursing
homes from 250 nursing home patients who met the inclusion criteria.
Methods. A sample of 202 cognitively intact nursing home patients responded to the Nurse–Patient Interaction Scale and
the Hospital Anxiety and Depression Scale. A structural equation model of the hypothesised relationships was tested by
means of LISREL 8.8 (Scientific Software International Inc., Lincolnwood, IL, USA).
Results. The SEM model tested demonstrated significant direct relationships and total effects of nurse–patient interaction on
depression and a mediated influence on anxiety.
Conclusion. Nurse–patient interaction influences depression, as well as anxiety, mediated by depression. Hence, nurse–
patient interaction might be an important resource in relation to patients’ mental health.
Relevance to clinical practice. Nurse–patient interaction is an essential factor of quality of care, perceived by long-term nurs-
ing home patients. Facilitating nurses’ communicating and interactive skills and competence might prevent and decrease
depression and anxiety among cognitively intact nursing home patients.
Key words: anxiety, depression, nurse–patient interaction, nursing home, structural equation model analysis
Accepted for publication: 11 September 2012
Introduction
With advances in medical technology and improvement in the
living standard globally, the life expectancy of people is
increasing worldwide. The document An Aging World (US
Census Bureau 2009) highlights a huge shift to an older popu-
lation and its consequences. Within this shift, the most rapidly
growing segment is people over 80 years old: by 2050, the per-
centage of those 80 and older would be 31%, up from 18% in
1988 (OECD 1988). These perspectives have given rise to the
notions of the ‘third’ (65–80 years old) and the ‘fourth age’
(over 80 years old) in the lifespan developmental literature
(Baltes & Smith 2003). These notions are also referred to as
the ‘young old’ and the ‘old old’ (Kirkevold 2010).
Authors: Gørill Haugan, PhD, RN, Associate Professor, Faculty of
Nursing, Research Centre for Health Promotion and Resources,
Sør-Trøndelag University College, HIST, Trondheim; Siw T
Innstrand, PhD, Associate Professor, Research Centre for Health
Promotion and Resources Norwegian University of Science and
Technology, NTNU, Trondheim; Unni K Moksnes, PhD, RN,
Associate Professor, Faculty of Nursing, Research Centre for
Health Promotion and Resources, Sør-Trøndelag University
College, HIST, Trondheim, Norway
Correspondence: Gørill Haugan, Associate Professor, Research
Centre for Health Promotion and Resources, HIST/NTNU, NTNU,
SVT/ISH, 7491 Trondheim, Norway. Telephone:
+47 73 55 29 27.E-mail: gorill.haugan@hist.no
© 2013 Blackwell Publishing Ltd 2192 Journal of Clinical Nursing, 22, 2192–2205, doi: 10.1111/jocn.12072
For many of those in the fourth age, issues such as physi-
cal illness and approaching mortality decimates their func-
tioning and subsequently lead to the need for nursing home
(NH) care. A larger proportion of older people will live for
shorter or longer time in a NH at the end of life. This
group will increase in accordance with the growing popula-
tion older than 65, and in particular for individuals older
than 80 years. Currently, 1�4 million older adults in the USA live in long-term care settings, and this number is
expected to almost double by 2050 (Zeller & Lamb 2011).
In Norway, life expectancy by 2050 is 90�2 years for men and 93�4 years for women (Statistics of Norway 2010). Depression is one of the most prevalent mental health
problems facing European citizens today (COM 2005);
and, the World Health Organization (WHO 2001) has esti-
mated that by 2020, depression is expected to be the high-
est ranking cause of disease in the developed world.
Moreover, depression is described to be one of the most
frequent mental disorders in the older population and is
particularly common among individuals living in long-term
care facilities (Choi et al. 2008, Karakaya et al. 2009,
Lattanzio et al. 2009, Drageset et al. 2011, Phillips et al.
2011). A linear increase in prevalence of depression with
increasing age is described (Stordal et al. 2003); the three
strongest explanatory factors on the age effect of depression
are impairment, diagnosis and somatic symptoms, respec-
tively (Stordal et al. 2001, 2003). Worse general medical
health is seen as the strongest factor associated with depres-
sion among NH patients (Djernes 2006, Barca et al. 2009).
A review that included 36 studies from various countries,
reported a prevalence rate for major depression ranging
from 6–26% and from 11–50% for minor depression.
However, the prevalence rate for depressive symptoms ran-
ged from 36–49% (Jongenelis et al. 2003). Twice as many
women are likely to be affected by depression than men
(Kohen 2006), and older people lacking social and emo-
tional support tend to be more depressed (Grav et al.
2012). A qualitative study on successful adjustment among
women in later life identified three main areas as being the
main obstacles for many; these were depression, maintain-
ing intimacy through friends and family and managing the
change process associated with older age (Traynor 2005).
Significantly more hopelessness, helplessness and depres-
sion are found among patients in NHs compared with those
living in the community (Ron 2004). Jongenelis et al.
(2004) found that depression was three to four times higher
in NH patients than in community-dwelling adults. Moving
to a NH results from numerous losses, illnesses, disabilities,
loss of functions and social relations, and approaching mor-
tality, all of which increases an individual’s vulnerability
and distress; in particular, loneliness and depression are iden-
tified as risks to the well-being of older people (Routasalo
et al. 2006, Savikko 2008, Drageset et al. 2012). The NH
life is institutionalised, representing loss of social relation-
ships, privacy, self-determination and connectedness.
Because NH patients are characterised by high age, frailty,
mortality, disability, powerlessness, dependency and vulner-
ability, they are more likely to become depressed. A recent
literature review showed several studies reporting prevalence
of depression in NHs ranging from 24–82% (Drageset et al.
2011). Also, with a persistence rate of more than 50% of
depressed patients still depressed after 6–12 months, the
course of major depression and significant depressive symp-
toms in NH patients tend to be chronic (Rozzini et al.
1996, Smalbrugge et al. 2006a).
Moreover, studies in NHs report a large co-occurrence of
depression and anxiety (Beekman et al. 2000, Kessler et al.
2003, Smalbrugge et al. 2005, Van der Weele et al. 2009,
Byrne & Pachana 2010). A recent review concerning anxi-
ety and depression reports a paucity of findings on anxiety
in older people (Byrne & Pachana 2010). Hence, more
research is urgently required into anxiety disorders in older
people, as these are highly prevalent and associated with
considerable disease burden (ibid.).
Depression and anxiety in NH patients are associated
with negative outcomes such as poor functioning in
activities of daily living and impaired quality of life (QoL)
(Smalbrugge et al. 2006b, Diefenbach et al. 2011, Drageset
et al. 2011), substantial caregiver burden and worsened
medical outcomes (Bell & Goss 2001, Koenig & Blazer
2004, Sherwood et al. 2005), increased risk of hospital
admission (Miu & Chan 2011), a risk of increased demen-
tia (Devanand et al. 1996) and a higher mortality rate
(Watson et al. 2003, Ahto et al. 2007). Accordingly, efforts
to prevent and decrease depression and anxiety are of great
importance for NH patients’ QoL.
Social support and relations to significant others are
found to be a vital resource for QoL and thriving among
NH patients (Bergland & Kirkevold 2005, 2006, Drageset
et al. 2009a, Tsai et al. 2010, Tsai & Tsai 2011), as well
as the nurse–patient relationship (Haugan Hovdenes 2002,
Cox & Bottoms 2004, Franklin et al. 2006, Medvene &
Lann-Wolcott 2010, Burack et al. 2012). The perspective
of promoting health and well-being is fundamental in nurs-
ing and a major nursing concern in long-term care (Nakrem
et al. 2011, Drageset et al. 2009b). However, low rates of
recognition of depression by staff nurses is found (Bagley
et al. 2000, Volkers et al. 2004).
Through the last decades, the importance of establishing
the nurse–patient relationship as an integral component of
© 2013 Blackwell Publishing Ltd Journal of Clinical Nursing, 22, 2192–2205 2193
Clinical issues Nurse-patient-interaction, depression, and anxiety
nursing practice has been well documented (Nåden &
Eriksson 2004, Arman 2007, Carpiac-Claver & Levy-
Storms 2007, Granados Gámez 2009, Rchaidia et al. 2009,
Fakhr-Movahedi et al. 2011). Excellent nursing care is
characterised by a holistic view with inherent human values
and moral; thus, excluding the patient as a unique human
being should be regarded as noncaring and amoral practice
(Haugan Hovdenes 2002, Nåden & Eriksson 2004, Aust-
gard 2008, Watson 2008). NH patients are in general
extremely vulnerable and hence the nurse–patient relation-
ship and the nurse–patient interaction are critical to their
experience of dignity, self-respect, sense of self-worth and
well-being (Dwyer et al. 2008, Harrefors et al. 2009,
Heliker 2009). NH patient receiving self-worth therapy
showed statistically significantly reduced depressive symp-
toms relative to control groups members 2 months after
receiving the intervention (Tsai et al. 2008). Self-worth
therapy comprised establishment of a therapeutic relation-
ship offering feedback and focusing the patient’s dignity,
emotional and mental well-being (ibid.).
Caring nurses engage in person-to-person relationships
with the NH patients as unique persons. Good nursing care
is defined by the nurses’ way of being present together with
the patient while performing nursing activities, in which
attitudes and competence are inseparately connected. ‘Pres-
ence’, ‘connectedness’ and ‘trust’ are described as funda-
mental cores of holistic nursing care (McGilton & Boscart
2007, Potter & Frisch 2007, Carter 2009) in the context of
the nurse–patient relationship in which the nurse–patient
interaction is taking place. Trust is seen as a confident
expectation that the nurses can be relied upon to act with
good will and to secure what is best for the individuals
residing in the NH. Hence, trust is the core moral ingredi-
ent in nurse–patient relationships; even more basic than
duties of beneficence, respect, veracity, and autonomy
(Carter 2009).
Caring is a context-specific interpersonal process that is
characterised by expert nursing practice, interpersonal sen-
sitivity, and intimate relationships (Finfgeld-Connett 2008)
which increases patient’s well-being (Nakrem et al. 2011,
Hollinger-Samson & Pearson 2000, Cowling et al. 2008,
Rchaidia et al. 2009, Reed 2009). The relationship between
NH staff attention and NH patients’ affect and activity par-
ticipation have been assessed among depressed NH
patients, showing that positive staff engagement was signifi-
cantly related to patients’ interest, activity participating,
and pleasure (Meeks & Looney 2011). These results suggest
that staff behaviour and engagement could be a reasonable
target for interventions to increase positive affect among
NH patients (ibid.).
In summary, the literature suggests depression as a com-
mon mental disorder among older people characterised by
high age, impairment, and somatic symptoms. In addition,
a large overlap of anxiety is reported. The patients’ sense
of loss of independency and privacy, feelings of isolation
and loneliness, and lack of meaningful activities are risk
factors for depression in NH patients. Nurse–patient inter-
action might be a resource for preventing and decreasing
depression among NH patients. To the authors’ knowl-
edge, previous research has not examined these relation-
ships in NHs by means of structural equation modelling
(SEM).
Aims
The main aim of this study was to investigate the relation-
ships between nurse–patient interaction, anxiety and
depression among cognitively intact NH patients by means
of SEM. Based on the theoretical and empirical knowledge
of depression, anxiety and nurse–patient interaction our
research question was: ‘Does the nurse–patient interaction
affect anxiety and depression in cognitively intact NH
patients?’ The following hypotheses were formulated:
� Hypothesis 1 (H1): nurse–patient interaction positively affects anxiety.
� Hypothesis 2 (H2): nurse–patient interaction positively affects depression.
� Hypothesis 3 (H3): depression negatively affects anxiety.
Methods
Design and ethical considerations
The study employed a cross-sectional design. The data was
collected in 2008 and 2009 in 44 different NHs from 250
NH patients who met the inclusion criteria: (1) local
authority’s decision of long-term NH care; (2) residential
time six months or longer; (3) informed consent compe-
tency recognised by responsible doctor and nurse; and (4)
capable of being interviewed. Two counties comprising in
total 48 municipalities in central Norway were selected,
from which 25 (at random) were invited to contribute in
this study. In total, 20 municipalities were partaken. Then,
all the NHs in each of the 20 municipalities was asked to
participate. A total of 44 NHs took part in the study. To
include as many participants from rural and central NHs,
respectively, the NHs was one by one invited to participate,
until the minimum of n = 200 was reached. The NH
patients were approached by a head nurse they knew
well. The nurse presented them with oral and written
© 2013 Blackwell Publishing Ltd 2194 Journal of Clinical Nursing, 22, 2192–2205
G Haugan et al.
information about their rights as participants and their
right to withdraw at any time. Each participant provided
informed consent. Because this population has problems
completing a questionnaire independently, three trained
researchers conducted one-on-one interviews in the patient’s
room in the actual NH. Researchers with identical profes-
sional background were selected (RN, MA, trained and
experienced in communication with older people, as well as
teaching gerontology at an advanced level) and trained to
conduct the interviews as identically as possible. Inter-rater
reliability was assessed by comparing mean scores between
interviewers by means of Bonferroni-corrected one-way
ANOVAs. No statistically significant differences were found
that were not accounted for by known differences between
the areas in which the interviewers operated.
The questionnaires relevant for the present study were part
of a questionnaire comprising 130 items. The interviews
lasted from 45–120 minutes due to the individual partici-
pant’s tempo, form of the day, and need for breaks. Inter-
viewers held a large-print copy of questions and possible
responses in front of the participants to avoid misunder-
standings. Approval by the Norwegian Social Science Data
Services was obtained for a licence to maintain a register
containing personal data (Ref. no. 16443) and likewise we
attained approval from The Regional Committee for
Medical and Health Research Ethics in Central Norway
(Ref. no. 4.2007.645) as well as the directory of the 44 NHs.
Participants
The total sample comprised 202 (80�8%) of 250 long-term NH patients representing 44 NHs. Long-term NH care was
defined as 24-hour care; short-term care patients, rehabilita-
tions patients, and cognitively impaired patients were not
included. Participants’ age was 65–104, with a mean of
86 years (SD = 7�65). The sample comprised 146 women (72�3%) and 56 men (27�7%), where the mean age was
87�3 years for women and 82 years for men. A total of 38 (19%) were married/cohabitating, 135 (67%) were widows/
widowers, 11 (5�5%) were divorced, and 18 (19%) were single. Duration of time of NH residence when interviewed
was at mean 2�6 years for both sexes (range 0�5–13 years); 117 were in rural NHs, while 85 were in urban NHs. In
all, 26�1% showed mild to moderate depression, only one woman scored >15 indicating severe depression, 70�4% was not depressed, and nearly 88% had no anxiety disor-
der. Missing data was low in frequency and was handled
by means of the listwise procedure; for the nurse–patient
interaction 4�0% and for anxiety and depression 5�0% had some missing data.
Measures
The Nurse–Patient Interaction Scale (NPIS) was developed
to identify important characteristics of NH patients’ experi-
ences of the nurse–patient interaction. The NPIS comprises
14 items identifying essential relational qualities stressed in
the nursing literature (Watson 1988, Martinsen 1993,
Eriksson 1995a,b, Nåden & Eriksson 2004, Nåden &
Sæteren 2006, Levy-Malmberg et al. 2008). Examples of
NPIS-items include ‘Having trust and confidence in the staff
nurses’; ‘The nurses take me seriously’, ‘Interaction with
nurses makes me feel good’ as well as experiences of being
respected and recognised as a person, being listened to and
feel included in decisions. The items were developed to
measure the NH patients’ ability to derive a sense of well-
being and meaningfulness through the nurse–patient inter-
action (Haugan Hovdenes 1998, 2002, Hollinger-Samson
& Pearson 2000, Finch 2006, Rchaidia et al. 2009). The
NPIS has shown good psychometric properties with good
content validity and reliability among NH patients;
(Haugan et al. 2012). The NPIS is a 10-points scale from 1
(not at all)–10 (very much); higher numbers indicating
better nurse–patient interaction (Appendix 1). Cronbach’s
Table 1 Means (M), standard deviations (SD), Cronbach’s alpha, and correlation coefficients for the study variables
Construct M SD Cronbach’s alpha NPIS HADS-A HADS-D
NPIS (10 items) 8�19 1�73 0�92 – HADS-A (5 items) 0�40 0�50 0�79 �0�114 – HADS-D (5 items) 0�74 0�58 0�66 �0�294* 0�340* – HADS (14 items) 2�85 0�34 0�78
*p < 0�01. NPIS, Nurse–Patient Interaction Scale; HADS, Hospital Anxiety and Depression Scale; HADS-A, Hospital Anxiety and Depression Scale –
Anxiety; HADS-D, Hospital Anxiety and Depression Scale – Depression.
© 2013 Blackwell Publishing Ltd Journal of Clinical Nursing, 22, 2192–2205 2195
Clinical issues Nurse-patient-interaction, depression, and anxiety
a = 0�92 (Table 1) and composite reliability = 0�92 (Table 2) of the NPIS construct was good.
Anxiety and depression were assessed by the Hospital
Anxiety and Depression Scale (HADS), comprising 14 items
(Appendix 2), with subscales for anxiety (HADS-A; seven
items) and depression (HADS-D seven items). Each item is
rated from 0–3, where higher scores indicate more anxiety
and depression. The maximum score is 21 on each subscale.
The ranges of scores for cases are as follows: 0–7 normal,
8–10 mild disorder, 11–14 moderate disorder, and 15–21
severe disorder (Snaith & Zigmond 1994). HADS has been
tested extensively and has well-established psychometric
properties (Herrmann 1997). To increase acceptability and
avoid individuals feeling as though they are being tested for
mental disorders, symptoms of severe psychopathology
have been excluded. This makes HADS more sensitive to
milder psychopathology (Stordal et al. 2003). HADS is
translated into Norwegian and found to be valid for older
people (Stordal et al. 2001, 2003).
Examples of sample-items are for depression: ‘I still enjoy
the things I used to enjoy’, ‘I can laugh and see the funny side
of things’, ‘I feel cheerful’, ‘I have lost interest in my appear-
ance’, and ‘I look forward with enjoyment to things’, and for
anxiety: ‘I feel tense and wound up’, ‘I get a sort of frightened
feeling as if something awful is about to happen’, ‘Worrying
thoughts go through my mind’, ‘I get a sort of frightened feel-
ing like ‘butterflies’ in the stomach’, and ‘I get sudden feeling
of panic’. The items were scored on a four-point scale ranging
from totally disagrees to totally agree. The internal consis-
tence of the anxiety and depression constructs (Table 1) was
satisfactory; a = 0�79 and a = 0�66, respectively. Composite reliability (qc) displayed values between 0�70–0�92 (Table 2); values >0�60 are desirable, whereas values >0·70 are good (Diamantopolous & Siguaw 2008, Hair et al. 2010).
Statistical analysis
A structural equation model (SEM) of the hypothesised
relations between the latent constructs of depression and
self-transcendence was tested by means of LISREL 8.8 (Scien-
tific Software International Inc., Lincolnwood, IL, USA)
(Jøreskog & Sørbom 1995). Using SEM accounts for ran-
dom measurement error and the psychometric properties of
the scales in the model are more accurately derived. Since
the standard errors are estimated under non-normality, the
Satorra–Bentler scaled chi-square statistic was applied as a
goodness-of-fit statistic, which is the correct asymptotic
mean even under non-normality (Jøreskog et al. 2000). In
line with the rules of thumb of conventional cut-off criteria
(Schermelleh-Engel et al. 2003), the following fit indices
were used to evaluate model fit: chi-square (v2); a small v2
and a non-significant p-value corresponds to good fit
(Jøreskog & Sørbom 1995). Further we used the root mean
square error of approximation (RMSEA) and the standar-
dised root mean square residual (SRMS) with values below
0�05 indicating good fit, while values smaller than 0�08 are interpreted as acceptable (Hu & Bentler 1998, Schermelleh-
Engel et al. 2003). The comparative fit index (CFI) and the
non-normed fit index (NNFI) with an acceptable fit at 0�95, and good fit at 0�97 and above were used, and the normed fit index (NFI) with an acceptable fit at 0�90, while a good fit was set to 0�95 (ibid.). Before examining the hypothesised relationships in the
SEM analysis, the measurement models were tested by con-
firmatory factor analysis (CFA). The CFA provided a good
fit to the observed data for the nurse–patient interaction
construct comprising ten items (v2 = 92�32, df = 77,
Table 2 Measurement models included in Model 1: nurse–patient
interaction (NPIS) to anxiety (HADS-A) and depression (HADS-D)
Items Parameter Lisrel estimate t-value R2
NPIS
NPIS1 kx1,1 0�63 6�04** 0�39 NPIS2 kx2,1 0�74 8�99** 0�55 NPIS3 kx3,1 0�74 10�41** 0�55 NPIS4 kx4,1 0�81 12�84** 0�65 NPIS5 kx5,1 0�66 6�16** 0�43 NPIS7 kx6,1 0�72 8�25** 0�51 NPIS9 kx7,1 0�77 14�39** 0�60 NPIS11 kx8,1 0�77 11�36** 0�59 NPIS12 kx9,1 0�69 8�18** 0�47 NPIS13 kx10,1 0�78 9�45** 0�61 HADS-A
HADS1 ky5,2 0�62 – 0�39 HADS3 ky7,2 0�73 7�04** 0�53 HADS5 ky11,2 0�62 4�65** 0�39 HADS9 ky13,2 0�69 5�60** 0�40 HADS13 kx14,2 0�66 6�00** 0�43 HADS-D
HADS2 ky1,1 0�74 – 0�54 HADS4 ky2,1 0�67 7�43** 0�45 HADS6 ky3,1 0�65 5�86** 0�42 HADS10 ky5,1 0�20 2�33* 0�04 HADS12 ky6,1 0�51 4�94** 0�26 qc NPIS 10 items qc 0�92 – – qc HADS-A 5 items qc 0�80 – – qc HADS-D 5 items qc 0�70 – –
Standardised factor loadings and t-values. Squared multiple correla-
tions (R2). †Composite reliability, qc ¼
P kð Þ2P
kð Þ2þP hð Þ � � (Hair et al. 2010).
*p < 0�05; **p < 0�01. HADS, Hospital Anxiety and Depression Scale; NPIS, Nurse–
Patient Interaction Scale.
© 2013 Blackwell Publishing Ltd 2196 Journal of Clinical Nursing, 22, 2192–2205
G Haugan et al.
p < 0�0110, RMSEA = 0�032, SRMR = 0�045, NFI = 0�97, NNFI = 0�99, CFI = 1�00) and the two-factor construct (HADS) of anxiety and depression comprising 10 items
(v2 = 54�22, df = 34, p < 0�015, RMSEA = 0�056, SRMR = 0�071, NFI = 0�93, NNFI = 0�96, CFI = 0�97). All parameter estimates were significant (p < 0�05) and loaded positively and clearly on their intended latent vari-
able with standardised factor loadings between 0�20–0�81. For scaling, the first factor loadings of each dependent
latent variable were set to 1.
Results
Descriptive analysis
Table 1 displays the means (M), standard deviations (SD),
Cronbach’s a and Pearson’s correlation matrix for the con-
structs of nurse–patient interaction, anxiety and depression.
The correlations between the measures were in the expected
direction. Moderate correlations were found between the
latent constructs included in the SEM model (Table 1). The
a-levels for the various measures indicate an acceptable
level of inter-item consistency in the measures (Nunally &
Bernstein 1994) with Cronbach’s a coefficients of 0�66 or higher.
Structural equation modelling (SEM)
To investigate how the nurse–patient interaction related to
anxiety and depression, model-1 was estimated. Figure 1
shows Model-1 with its measurement and structural
models, while Table 2 displays the factor loadings, R2 and
t-values. All estimates were significant (p < 0�05) and the
factor loadings ranged between 0�51–0�81 (except from item HADS10 ‘I have lost interest in my appearance’ with
factor loading = 0�20 and R2 = 0�04) and R2 values between 0�26–0�65. Model-1 fit well with the data: v2 = 211�44, p = 0�011, df = 167, RMSEA = 0�037, p- value = 0�92, NFI = 0�94, NNFI = 0�99, CFI = 0�99, and SRMR = 0�060. Table 3 shows the standardised regression coefficients of
the directional relationships and the total and indirect
effects between the latent constructs in Model-1. As
hypothesised, the directional paths from nurse–patient
interaction to depression displayed a significant negative
relationship (c1,1 = �0�37). The path between nurse– patient interaction and anxiety was not significant
(c1,2 = �0�09); however, a significant path from depression to anxiety (b1,2 = 0�55) was found, indicating a mediated effect (by depression) on anxiety (Table 3).
A scrutiny of the total effects of nurse–patient interaction
revealed statistical significant total effects on depression
(�0�37), as well as a significant total effect on anxiety from depression (0�55). Also, a significant indirect (mediated) effect from nurse–patient interaction on anxiety (�0�20) was displayed (Table 3).
Discussion
The aim of this study was to explore the associations
between nurse–patient interaction, anxiety, and depression
in cognitively intact NH patients. By doing so we sought to
contribute to a holistic nursing perspective of promoting
well-being in NH patients in three ways:
1 This study supplies empirical knowledge to the growing
body of nurse–patient interaction knowledge by exploring
Figure 1 SEM Model-1. Measurement models
and directional relationships for nurse–patient
interaction (NPIS) to anxiety (HADS-A) and
depression (HADS-D).
© 2013 Blackwell Publishing Ltd Journal of Clinical Nursing, 22, 2192–2205 2197
Clinical issues Nurse-patient-interaction, depression, and anxiety
cognitively intact NH patient’s’ experiences of the nurse–
patient interaction.
2 This study provides empirical insight to the associations
between, nurse–patient interaction, anxiety, and depres-
sion in a NH population, and
3 By means of advanced statistical analysis such as struc-
tural equation modelling (SEM), the result from this
study suggests a guideline for clinical nursing strategies
promoting well-being and decreasing depression and anx-
iety in NH patients. Finding ways to improve NH staff
nurses’ way of being present, connecting, and interacting
with the patients might be beneficial in this matter.
More specifically, three hypotheses were tested, from
which two were supported (H1, H3). We found that the
hypothesised relationship between nurse–patient interaction
and depression was fully supported; good nurse–patient
interaction was negatively associated with depression; the
better nurse–patient interaction the less depression. The
path from nurse–patient interaction to anxiety was not sig-
nificant; however, a significant relation between depression
and anxiety was found. Accordingly, also an indirect effect
of nurse–patient interaction on anxiety was displayed, med-
iated by the influence on depression. Hence, the model
tested indicates that nurse–patient interaction influences
both depression and anxiety. These findings are consistent
with previous research demonstrating significantly decrease
in depression for NH patients receiving self-worth therapy
and positive attention from NH staff (Tsai et al. 2008,
Meeks & Looney 2011).
Consequently, nursing approaches facilitating NH
patient’s confidence and trust in the staff nurses might pre-
vent and decrease depression in NH patients. In accordance
with former research, trust is a core moral ingredient in
helping relationships (Carter 2009). Therefore, facilitating
patients’ confidence that the staff nurses make all possible
effort to relieve ones’ plagues appear to be crucial for pre-
serving dignity (Cochinov 2002) and prevent depression.
Professional nursing care is determined by the way nurses
are using their knowledge, attitudes, behaviour and com-
munication skills to appreciate the uniqueness of the person
being cared for (Warelow et al. 2008). Accordingly, nurse–
patient interaction fostering experiences of being respected
and recognised as a person, being listened to and taken seri-
ously are positively associated with lower depression scores
among NH patients.
Previous research underlines that the nurse–patient
relationships and the nurse–patient interaction are critical
to patients’ sense of dignity, self-respect, feelings of self-
worth, meaning in life, and well-being (Haugan Hovdenes
2002, Dwyer et al. 2008, Harrefors et al. 2009, Heliker
2009). Moreover, dignity in NH patients has been differen-
tiated into intrapersonal dignity and relational dignity,
socially constructed by the act of recognition (Pleschberger
2007). Thus, nurse–patient interaction facilitating patients’
sense of being taken seriously and recognised as a unique
person might provide a sense of dignity, self-worth, and
thereby prevent and decrease depression among NH
patients.
Consequently, taking time for interestingly listening to
the NH patient appears as vital for preventing and decreas-
ing depression. Former research has pointed to continuity
of care provider to be critical for developing relationships
with patients’ overtime (McGilton 2002). Moreover, mutu-
ality in individuals’ relationships confirming women’s exis-
tence and value has been described as a major influence on
depression in women, whereas depressive symptoms results
Table 3 Structural equation modelling analysis: Model-1, standar-
dised gamma, total and indirect effects of nurse–patient interaction
on nursing home patients’ anxiety and depression
Construct Parameter Lisrel estimate t-value
NPIS to HADS-A c 1,1 �0�09 �0�84 NPIS to HADS-D c 1,2 �0�37 �4�58** HADS-D to HADS-A b 1,2 0�55 4�05**
NPIS t-value
Total effects of nurse–patient interaction on anxiety and depression
HADS-D
HADS2 �0�23 �4�58** HADS4 �0�21 �4�32** HADS6 �0�18 �3�62** HADS10 �0�08 �2�03* HADS12 �0�18 �3�48** HADS-A
HADS1 �0�05 �1�30 HADS3 �0�06 �1�34 HADS5 �0�05 �1�27 HADS9 �0�05 �1�25 HADS13 �0�05 �1�36
Indirect effect of nurse–patient interaction on anxiety and
depression
Anxiety �0�20 �3�11** Depression – –
*Significant at the 5% level.
**Significant at the 1% level.
Model 1: comprising six HADS-variables and 10 NPIS-variables.
Standardised gamma: standardised regression coefficients represent-
ing directional relationships between NPIS, anxiety, and depression.
Total effects: represents the total influence of the explanatory
variable on anxiety and depression.
Indirect effects: mediated influence.
HADS, Hospital Anxiety and Depression Scale; NPIS, Nurse–
Patient Interaction Scale.
© 2013 Blackwell Publishing Ltd 2198 Journal of Clinical Nursing, 22, 2192–2205
G Haugan et al.
from violating their sense of worthiness (Hedelin & Jonsson
2003). Consequently, nurses must be aware that their atti-
tude, appearance and behaviour are interpreted as a confir-
mation of the patient’s worthiness or worthlessness (ibid.).
In a recent study investigating the concept of receiving care,
one main theme was identified; ‘being of value despite any
potential disadvantages’ (Lundgren & Berg 2011). NH
patients are particularly vulnerable and dependent, thus
there are not many choices available. Receiving care high-
lights the human mode of being, which includes experiences
of being exposed resulting in an increased sense of vulnera-
bility; in turn, this motivates a seeking for valued and
appreciated mutual interactions within a caring process
(ibid.). Thus, taking time, ensuring continuity, and being
educated in interactional skills are not enough to enhance
well-being, a sense of meaning in life, and decrease depres-
sion; the care provider must be engaged in some way, such
as learning about the person through life histories (McGil-
ton & Boscart 2007, Walent 2008, Heliker 2009, Heliker
& Hoang Thanh 2010, Medvene & Lann-Wolcott 2010,
Wright 2010). The NH patient needs to feel understood,
acknowledged, confirmed, and valued, all of which provides
a sense of meaning in life, self-worth, and alleviates suffer-
ing (Haugan Hovdenes 2002, Medvene & Lann-Wolcott
2010).
Nursing homes are unique social environments; tradition-
ally, they offer limited privacy opportunities. Accordingly
NH patients may have infrequent contact with friends and
family members. Thus, NH staff nurses are particularly the
most important providers of social reinforcement (Haugan
Hovdenes 2002, Drageset et al. 2012). However, research
illustrates that NH staff rarely engages in social interac-
tions during mealtimes and does not appreciate this as an
important part of their duties (Pearson & Fitzgerald 2003),
as well as hardly responds to patients’ social engagement,
and seldom displays engagement-supportive behaviour
(Stabell et al. 2004, Meeks & Looney 2011). Caregiving
relationships involve all kinds of social interaction during the
course of which the patient’s sense of self-worth can either
be enhanced or thwarted (Haugan Hovdenes 2002, Hedelin
& Jonsson 2003, Halldorsdottir 2008). The nurse–patient
relationship has been designated as a sense of spiritual con-
nection which is experienced as a bond of energy (Hall-
dorsdottir 2008); a life-giving nurse–patient interaction
which is greatly empowering for the patient. By confirma-
tion, recognising, and empowering the older individuals’
views of who they are and would like to be, NH staff
nurses can positively influence NH patients well-being
(Randers et al. 2002, Tsai et al. 2008, Haugan et al.
2012), thriving (Bergland & Kirkevold 2005), and
consequently depression (Haugan & Innstrand 2012) and
anxiety.
Strengths and limitations
A notable strength of this research is the empirical exami-
nation of associations that have not been tested previously.
This study expands previous studies by testing the asso-
ciations between nurse–patient interaction, anxiety, and
depression among NH patients by using structural equa-
tion modelling. Using SEM accounts for random measure-
ment error and the psychometric properties of the scales
in the model are more accurately derived. The study builds
on a strong theoretical foundation with use of question-
naires demonstrating good psychometrical properties. Nev-
ertheless, the findings of this study must be discussed with
some limitations in mind.
First, Model-1 comprises 20 variables, indicating a desir-
able n = 200, while in the present study, n = 191. Informa-
tion input to the SEM estimation increases both with more
indicators per latent variable, as well as with more sample
observations (Westland 2010). The latent variables in the
model are measured by five and ten indicators that
strengthen the reliability. In this respect, the sample size in
the present study is suitable. Nevertheless, a larger sample
would significantly increase statistical power of the tests.
The present sample included fewer men than women,
reflecting the gender composition among the population of
that age in NHs.
Second, the cross-sectional design does not allow us to
determine conclusion regarding causality. A longitudinal
design would have strengthened the study by allowing
changes to be assessed and compared over time.
A third limitation concerns the use of self-reported data,
which implies a certain risk that the findings are based on
common-method variance (Podsakoff et al. 2003).
The fact that the researchers visited the participants to
help fill in the questionnaires might have introduced some
bias into the respondents’ reporting. The questionnaires
were part of a battery of questionnaires comprising 130
items. Thus, frail, older NH patients might tire when com-
pleting the questionnaires; this represents a possible bias to
their reporting. To avoid such a bias, experienced research-
ers were carefully selected and trained in conducting the
interviews following a standardised procedure including
taking small breaks on specific points during the process.
This procedure worked out very well; in just three cases the
interview had to be completed the next day due to respon-
dent’s fatigue. Actually, most participants were even more
vigorous after completing the interview.
© 2013 Blackwell Publishing Ltd Journal of Clinical Nursing, 22, 2192–2205 2199
Clinical issues Nurse-patient-interaction, depression, and anxiety
Relevance to clinical nursing
According to the European Commission’s Green Paper on
mental health (COM 2005), depression is one of the most
prevalent mental health problems facing European citizens
today. Taking into account the highly chronic nature of these
psychological states, we consider our findings noteworthy in
their suggestion that nurse–patient interaction might be an
important resource in relation to NH patients’ mental health.
Knowledge of how nurse–patient interaction, anxiety and
depression relate to each other in this respect is important for
researchers, nurses, nursing educators and clinicians.
This study demonstrates that nurse–patient interaction
influences depression as well as anxiety mediated by depres-
sion. Accordingly, facilitating nurse–patient interaction to
provide patients’ sense of worthiness, meaning in life, self-
acceptance and adjustment to the life situation and one’s
disabilities would promote integrity and well-being and pre-
vent despair and depression.
Due to a combination of factors such as patients’ commu-
nication impairment, clinicians’ focus on treating medical
conditions, normalisation of depression in later life and a
lack of training in mental health among staff in NHs, depres-
sion can easily go undetected among the NH population
(Bagley et al. 2000, Martin et al. 2007). Therefore, facilitat-
ing nurse–patient interaction and the staff nurses’ awareness
in assessing patients’ mood and connectedness resources
appear to be crucial. Offering connectedness might be a cen-
tral aspect of NH care (Lundman et al. 2010); enhancing
inner strength by acceptance of the self, death and one’s life
situation might prevent and decrease depression among NH
patients (Haugan & Innstrand 2012).
The interpersonal relationship in nurse–patient interac-
tions has been found to be an essential factor of quality of
care, as perceived by long-term care patients (Haugan Hovd-
enes 2002, Bergland & Kirkevold 2006, Brown Wilson &
Davies 2009). Nurse–patient interaction can enhance both
intrapersonal and interpersonal self-transcendence (Haugan
et al. 2012) and help NH patients preserve their dignity,
identity and integrity (Coughlan & Ward 2007, Tsai et al.
2008, Burack et al. 2012). By means of listening to the
patients, communicating and treating the patients with
respect, by using empathic understanding, and acknowledg-
ing him/her as a person who is to be taken seriously and
attended to, staff nurses might positively influence depres-
sion, anxiety, and well-being (Hollinger-Samson & Pearson
2000, Haugan Hovdenes 2002, Asmuth 2004, Finch 2006,
Jonas-Simpson et al. 2006, Haugan et al. 2012).
Therefore, NH staff nurses should be given more time
available interacting with their patients. A philosophical
shift from care and protection of the body to a person-cen-
tred care would be beneficial (Medvene & Lann-Wolcott
2010, Wright 2010, Jones 2011). In addition, some factors
seem crucial regarding quality of nurse–patient interaction;
in general, staffing levels are low while staff turnover is
high (Baughman & Smith 2010). Further, staff members
are generally poorly trained in nurse–patient interaction
providing well-being, and often they perceive a lack of
autonomy in job performance, feeling that they are not
respected for management (Castle & Engberg 2007, Caspar
& O’Rourke 2008, Bishop et al. 2009). To become a car-
ing caregiver, one must first be treated in a caring way
(Sikma 2006, Tellis-Nayak 2007). Hence, moving from the
traditional institutional model to a responsive, patient-cen-
tred homelike approach might have benefits for both NH
patients and staff (Jones 2010). Educational nursing curric-
ula should underline and facilitate nurse–patient interaction
in order of advancing staff nurses’ presence to assess,
prevent and decrease depression and anxiety among NH
patients. Also, essential is for nurses to develop confidence
in using the therapeutic tools available to create the best
mental health outcomes for the older person.
Acknowledgements
The authors wish to acknowledge the Sør-Trøndelag
University College, Faculty of Nursing, Trondheim, Nor-
way, for supporting this study, as well as the older patients
who voluntarily participated in the study.
Contributions
Study design: GH; data collection and analysis: GH and
manuscript preparation: GH, STI, UKM.
Conflict of interest
No conflict of interest has been declared by the authors.
References
Ahto M, Isoaho R, Puolijoki H, Vahlberg
T & Kivelä S (2007) Stronger symp-
toms of depression predict high coro-
nary heart disease mortality in older
men and women. International Journal
of Geriatric Psychiatry 22, 757–763.
Arman M (2007) Bearing witness: an exis-
tential position in caring. Contemporary
© 2013 Blackwell Publishing Ltd 2200 Journal of Clinical Nursing, 22, 2192–2205
G Haugan et al.
Nurse: A Journal for the Australian
Nursing Profession 27, 84–93.
Asmuth MV (2004) Developing and Sus-
taining Interpersonal Relationships
Between Certified Nursing Assistants
and Residents in Long-Term Care
Facilities. University of South Florida,
Sarasota, FL.
Austgard KI (2008) What characterises
nursing care? A hermeneutical philo-
sophical inquiry. Scandinavian Journal
of Caring Science 22, 314–319.
Bagley H, Cordingey L, Burns A, Mozley C,
Sutcliff C, Challis D & Huxley P (2000)
Recognition of depression by staff in
nursing and residential homes. Journal
of Clinical Nursing 9, 445–450.
Baltes PB & Smith J (2003) New frontiers in
the future of aging: from successful aging
of the young old to dilemmas of the
fourth age.Gerontology 49, 123–135.
Barca ML, Selbæk G, Laks J & Engedal K
(2009) Factors associated with depres-
sion in Norwegian nursing homes.
International Journal of Geriatric Psy-
chiatry 24, 417–425.
Baughman R & Smith K (2010) The effect
of Medicaid wage pass-through pro-
grams on the wages of direct care.
Medical Care 48, 426–432.
Beekman A, de Beurs E, van Balkom A,
Deeg D, van Dyck R & van Tilburg
W (2000) Anxiety and depression in
later life: co-occurance and commu-
nality of risk factors. American Jour-
nal of Psychiatry 157, 89–95.
Bell M & Goss A (2001) Recognition,
assessment and treatment of depres-
sion in geriatric nursing-home resi-
dents. Clinical Excellence for Nurse
Practitioners 5, 445–450.
Bergland A & Kirkevold M (2005) Resident-
caregiver relationships and thriving
among nursing home residents.
Research in Nursing & Health 28,
365–375.
Bergland A & Kirkevold M (2006) Thriv-
ing in nursing homes in Norway: con-
tributing aspects described by
residents. International Journal of
Nursing Studies 43, 681–691.
Bishop C, Squillance M, Meagher J,
Anderson W & Wiener J (2009) Nurs-
ing home work practices and nursing
assistants’ job satisfaction. Gerontolo-
gist 49, 611–622.
Brown Wilson C & Davies S (2009) Devel-
oping relationships in long-term care
environments: the contribution of
staff. Journal of Clinical Nursing 18,
1746–1755.
Burack OR, Weiner AS, Reinhardt JP &
Annunziato RA (2012) What matters
most to nursing home elders: quality
of life in the nursing home. Journal of
the American Medical Directors Asso-
ciation 13, 48–53.
Byrne G & Pachana N (2010) Anxiety and
depression in the elderly: do we know
any more? Current Opinion in Psychi-
atry 23, 504–509.
Carpiac-Claver ML & Levy-Storms L
(2007) In a manner of speaking: com-
munication between nurse aides and
older adults in long-term care settings.
Health Communication 22, 59–67.
Carter M (2009) Trust, power, and vulner-
ability: a discourse on helping in nurs-
ing. Nursing Clinics of North America
44, 393–405.
Caspar S & O’Rourke N (2008) The influ-
ence of care provider access to struc-
tural empowerment on individualized
care in long-term care facilities. Jour-
nals of Gerontology, Series B, Psycho-
logical Sciences and Social Sciences
63, S255–S265.
Castle N & Engberg J (2007) The influ-
ence of staffing characteristics on qual-
ity of care in nursing homes. Health
Services Research 45, 1822–1847.
Choi NG, Ransom S & Wyllie RJ (2008)
Depression in older nursing home resi-
dents: the influence of nursing home
environmental stressors, coping, and
acceptance of group and individual
therapy. Aging and Mental Health 12,
536–547.
Cochinov H (2002) Dignity conserving
care: a new model for palliative care.
Journal of American Medical Associa-
tion 287, 2253–2260.
COM (2005) Improving the Mental Health of
the Population: Towards a Strategy on
Mental Health for the European Union.
EuropeanCommunities, Brussels.
Coughlan R & Ward L (2007) Experiences
of recently relocated residents of a
long-term care facility in Ontario:
assessing quality qualitatively. Journal
of Nursing Studies 44, 47–57.
Cowling WR, SmithMC&Watson J (2008)
The power of wholeness, consciousness,
and caring a dialogue on nursing
science, art, and healing. Advances in
Nursing Science 31, E41–E51.
Cox J & Bottoms R (2004) Public health
and the older person. Journal of Com-
munity Nursing 18, 10–12.
Devanand DP, Sano M, Tang M, Taylor S,
Gurland B, Wilder D, Stern Y & May-
eux R (1996) Depressed mood and the
incidence of Alzheimer’s disease in the
elderly living in the community.
Archives of General Psychiatry 53,
175–182.
Diamantopolous A & Siguaw JA (2008)
Introducing LISREL. SAGE Publica-
tions Ltd, Los Angeles, CA.
Diefenbach GJ, Tolin DF & Gilliam CM
(2011) Impairments in life quality
among clients in geriatric home care:
associations with depressive and
anxiety symptoms. International Jour-
nal of Geriatric Psychiatry 27, 828–
835.