Received: 8 January 2021

DOI: 10.1002/pon.5648

ORIGINAL ARTICLE
- Accepted: 1 February 2021

Stronger therapeutic alliance is associated with better

quality of life among patients with advanced cancer

Teresa Thomas1,2 | Andrew Althouse3 | Lauren Sigler2 | Robert Arnold2,3 |

Edward Chu4 | Douglas B. White2,5 | Margaret Rosenzweig1,2 | Kenneth Smith3 |
Thomas J. Smith6 | Yael Schenker2,3

1
School of Nursing, University of Pittsburgh,
Pittsburgh, Pennsylvania, USA Abstract
2
Palliative Research Center (PaRC), University Objective: Patient–oncologist therapeutic alliance is a foundation of quality cancer
of Pittsburgh, Pittsburgh, Pennsylvania, USA
care, although there is limited research demonstrating its relationship with patient
3
Division of General Internal Medicine,
University of Pittsburgh Medical Center,
outcomes. We investigated the relationship between therapeutic alliance and
Pittsburgh, Pennsylvania, USA patient quality of life with a secondary goal of determining whether the association
4
Albert Einstein Cancer Center, Albert Einstein varied by patients' baseline level of psychological distress.
College of Medicine, Bronx, New York, USA
5
Methods: Cross‐sectional analysis of baseline data from a randomized clinical trial
Department of Critical Care Medicine,
Program on Ethics and Decision Making in of 672 patients with advanced cancer participating in a primary palliative care
Critical Illness, The CRISMA Center, University intervention trial. Patients completed baseline self‐reported measures of thera-
of Pittsburgh School of Medicine, Pittsburgh,
Pennsylvania, USA peutic alliance (The Human Connection Scale, range: 16–64), overall quality of life
6
Department of Medicine Section of General (Functional Assessment of Cancer Therapy—Palliative Care, range: 0–184), and
Internal Medicine, Johns Hopkins Sidney
psychological distress (Hospital Anxiety and Depression Scale, range: 0–42). First,
Kimmel Comprehensive Cancer Center,
Baltimore, Maryland, USA we determined the relationship between therapeutic alliance and quality of life
using multivariable regression adjusting for confounders. We then examined if
Correspondence
Teresa Thomas, School of Nursing, University
psychological distress was an effect modifier in this relationship by adding inter-
of Pittsburgh, 3500 Victoria St, Pittsburgh, PA, action effects of depression and anxiety symptoms on therapeutic alliance into the
15213, USA.
Email: t.thomas@pitt.edu
regression model.
Results: Patients reported high levels of therapeutic alliance (56.4 ± 7.4) and
Funding information
moderate quality of life (130.3 ± 25.5). Stronger therapeutic alliance was associated
American Cancer Society Mentored Research
Scholar, Grant/Award Number: Grant MSRG‐ with better quality of life after adjusting for other confounding factors (β = 3.7, 95%
18‐051‐51; National Palliative Care Research confidence interval = 2.1, 5.3, p < 0.01). The relationship between therapeutic
Center Career Development Award; NCI
R01CA197103; P30CA047904 alliance and quality of life was generally consistent regardless of psychological
distress.
Conclusions: Collaborative, trusting relationships between patients with advanced
cancer and their oncologists are associated with better patient quality of life. Future
research should investigate the causal, longitudinal nature of these relationships.

KEYWORDS
advanced cancer, affect, cancer, distress, oncology, palliative care, quality of life, therapeutic
alliance, psycho‐oncology

Psycho‐Oncology. 2021;1–9. wileyonlinelibrary.com/journal/pon © 2021 John Wiley & Sons Ltd.

- 1
2
- THOMAS ET AL.

1 | BACKGROUND among a cohort of patients with advanced cancer and (2) determine if
psychological distress modifies the effect of therapeutic alliance on
Quality of life is a highly relevant patient outcome in the advanced quality of life. We hypothesized that higher therapeutic alliance is
cancer setting as patients strive to maximize their well‐being in the significantly related to higher quality of life and that this association
presence of an incurable illness.1 Quality of life encompasses a would be weaker for patients with high levels of psychological
holistic view of patients' physical, social, emotional, and functional distress compared to patients with low levels of distress.
health and is directly linked to other patient outcomes, including
cancer‐ and treatment‐related symptom severity and disability.2
While quality of life depends on several factors, an important factor 2 | METHODS
is the care provided by oncology clinicians. Clinicians directly assess
and manage patient’s quality‐of‐life concerns, and indirectly support This study analyzes baseline data from the Care management by
patients in self‐managing their quality‐of‐life needs. In addition, the Oncology Nurses to address supportive care needs (CONNECT)
nature of the relationship between patients and oncology clinicians cluster randomized trial (NCT02712229). The primary focus of this
3–5
likely impacts the quality of life. clinical trial evaluated the effect of a nurse‐led primary palliative care
Therapeutic alliance, defined as the collaborative bond between intervention on patient and caregiver outcomes. The details of the
patients and their healthcare providers, is the cornerstone of quality CONNECT trial design have been published, and no prior paper has
cancer care.6,7 Within cancer care, patients who believe their addressed this research question.20
oncologist listens to them, responds to their concerns, and is trust- Patients were recruited from 17 oncology clinics within the
worthy and caring, are thought to experience improved medical care University of Pittsburgh Medical Center Hillman Cancer Center
than patients without such a bond.8 Having a strong therapeutic network in Western Pennsylvania. Eligibility criteria included age
alliance may improve patient outcomes, by first enhancing commu- ≥21 years; metastatic solid tumor; Eastern Cooperative Oncology
nication between patients and oncologists, and second improving Group performance status ≤2, indicating that the patient is ambu-
understanding and shared goals and decision‐making about patient latory and capable of all self‐care; receiving ongoing care from an
needs and priorities.9,10 Ultimately, the goal of therapeutic alliance is oncologist participating in the trial; and for whom the oncologist
to enhance patient‐centered care and clinical services that improve “would not be surprised if the patient died in the next year.”21 Patient
patient quality of life.11 exclusion criteria included inability to read and write in English;
Research investigating therapeutic alliance in patients with cognitive impairment or inability to consent to treatment; inability to
advanced cancer shows inconsistent results with patient outcomes.12 complete baseline data collection; and hematologic malignancies.
Understanding this relationship is important because if therapeutic
alliance is associated with positive patient outcomes, then oncolo-
gists should be trained in the behaviors and attitudes that promote 2.1 | Variables
therapeutic alliance. Some studies have validated the positive asso-
ciation between therapeutic alliance and patient psychological 2.1.1 | Therapeutic alliance
symptoms, well‐being, coping strategies, and treatment adher-
ence.8,9,13 In contrast, a study by Bar‐Sela et al.14 demonstrated that We assessed patient perceptions of their therapeutic alliance with
a strong therapeutic alliance with nurses—but not oncologists—was their primary oncologist using the Human Connection Scale (HCS).9
significantly associated with patients' ability to manage their cancer This scale was originally evaluated amongst a sample of patients with
symptom burden. More alarming, another study found that patients advanced cancer and demonstrated strong internal consistency
with advanced cancer who reported a strong therapeutic alliance (Cronbach's α = 0.90) and concurrent validity with other outcome
were less likely to express their concerns to their oncologist about measures.9 The HCS consists of 16 items with 4‐point Likert‐type
15
treatment and disease status. response options assessing patient‐perceived levels of therapeutic
Patients experiencing psychological distress, including depressive alliance. For example, items assess patient trust in their doctor,
and anxiety symptoms, may have difficulty building therapeutic alliance respect for their doctor, and perception of their doctor's concern
16
due to higher symptom burden and difficulty building rapport with with their quality of life. The HCS total score ranges from 16 to 64,
oncologists.17 Without support from their oncologists, patients in with higher scores reflecting stronger patient‐oncologist therapeutic
distress may lack the necessary motivation and self‐efficacy to address alliance. In our sample, the HCS had strong internal consistency
their quality‐of‐life concerns and needs.18,19 Therefore, we aimed to (Cronbach's α = 0.91).
clarify the role of therapeutic alliance as it relates to patient psycho-
logical distress and quality of life. If positively associated, targeted
attention can be given to supporting the aspects of patient–oncologist 2.1.2 | Quality of life
therapeutic alliance that improve patient outcomes.
Therefore, our aims are to (1) describe the relationship between We used the Functional Assessment of Chronic Illness Therapy—
patient–oncologist therapeutic alliance and patient quality of life Palliative Care (FACIT‐Pal) to assess patient overall quality of life.22
THOMAS ET AL.
- 3

The 46‐item scale includes five domains of quality of life—physical, weaker for patients with higher levels of psychological distress. We
social, emotional, functional, and palliative care related quality of life included all of the previous confounders in the model, except for the
that were attained through in‐depth interviews with patients with HADS, which was now entered into the model as a potential effect
advanced cancer. This measure was validated in a large sample of modifier. We repeated the same model for patients above and below
patients with advanced cancer22 and had high internal consistency in the respective thresholds of HADS‐depression and HADS‐anxiety.
our sample (Cronbach's α = 0.93). The FACIT‐Pal total score ranges Finally, we tested interaction terms between (1) HADS‐depression
from 0 to 184, with higher scores reflecting higher quality of life. and HCS and (2) HADS‐anxiety total scores and HCS in the multi-
variable regression models to formally test whether depression and
anxiety were effect modifiers of the relationship between HCS and
2.1.3 | Psychological distress FACIT‐Pal.

We used the Hospital Anxiety and Depression Scale (HADS) to assess

patient psychological distress. Unlike the FACIT‐Pal, which assesses 3 | RESULTS
general quality of life, the HADS strictly focuses on patient depres-
sive and anxiety symptoms. This scale is frequently used with This manuscript includes baseline data from all 672 participants
advanced cancer patient populations.23 Scores for the anxiety and enrolled in the CONNECT trial (Table 1). In this study, patients had a
depression subscales range from 0 to 21, with higher scores indi- mean age of 69.3 (SD 10.2) years, were balanced by gender (54%
cating higher levels of psychological distress. The internal consis- female), and were mainly Caucasian (94%) and non‐Hispanic (99%).
tency of the depression and anxiety subscales in our sample were Patients reported a variety of solid tumor types, the most frequent of
Cronbach's α = 0.78 and α = 0.82, respectively. When indicated in which were lung (36%), breast (13%), colon (10%), and pancreatic
our analytic plan, we used previously developed cut‐off scores for (10%) cancer. Nearly half (44%) of patients were within a year of
patients with cancer on the HADS‐depression (≥5) and HADS‐ their first cancer diagnosis while 20% of patients were diagnosed ≥5
anxiety (≥7) subscales to compare patients with high and low levels years ago. Nearly half of patients (45.3%) had been receiving care
of psychological distress.24 from their oncologist for a year or longer. Most patients (69%) were
receiving chemotherapy treatment at baseline.
Patients rated high levels of therapeutic alliance (HCS mean total
2.1.4 | Additional clinical and demographic variables score 56.4 ± 7.4) and moderate levels of quality of life (FACIT‐Pal
mean total score: 130.3 ± 25.5) (Table 1). Therapeutic alliance was
We accounted for variables that are known to relate to therapeutic significantly associated with quality of life (Pearson R = 0.30,
alliance and quality of life in our analyses. We considered the p < 0.001) indicating that patients who reported higher levels of
following patient self‐reported variables as confounders: age, race, therapeutic alliance tended to report higher overall quality of life
ethnicity, time since diagnosis, and current chemotherapy treatment. (Figure 1).
Table 2 reports the results of our multivariable regression ana-
lyses. In the adjusted model, therapeutic alliance remained signifi-
2.2 | Analysis cantly associated with quality of life (β = 3.7, 95% confidence interval
[CI] = 2.1, 5.3, p < 0.01), meaning that a 10‐point increase on the HCS
We describe our sample using descriptive statistics including counts, was associated with a 3.7‐point increase in the FACIT‐Pal total score.
frequencies, and measures of central tendency. We used Pearson The stratified analysis demonstrated that the relationship be-
correlations and scatterplots to visually assess the relationships be- tween therapeutic alliance and quality of life was generally consistent
tween therapeutic alliance, quality of life, and psychological distress. regardless of psychological distress. There was a positive association
We conducted multivariable regression analyses to estimate the between HCS and FACIT‐Pal for patients with HADS‐depression ≥5
relationship between therapeutic alliance and quality of life while (β = 8.6, 95% CI = 5.8, 11.4, p < .01) and for patients with HADS‐
adjusting for potential confounding variables. We selected con- depression <5 (β = 5.9, 95% CI = 2.2, 9.6, p < 0.01). Similarly, there
founding variables a priori that are thought to impact both thera- was a positive association between HCS and FACIT‐Pal for patients
peutic alliance and quality of life: age, race, ethnicity, time since with HADS‐anxiety ≥7 (β = 7.0, 95% CI = 4.1, 9.9, p < 0.01) and <7
diagnosis, current chemotherapy treatment, and psychological (β = 6.4, 95% CI = 3.6, 9.3, p < 0.01). The magnitude of the effect of
distress.9,25–27 We used linear mixed‐effects models with all con- HCS on quality of life was slightly greater for patients with higher
founders treated as fixed effects. We rescaled the HCS by a factor of levels of distress in both models. However, the interaction terms of
10 to ensure the interpretability of the regression model results. HADS‐depression and HCS and HADS‐anxiety and HCS were
Therefore, each reported beta coefficient indicates the change in nonsignificant (p = 0.36 and p = 0.36, respectively), providing little
expected value of FACIT‐Pal per 10‐point increase in the HCS. evidence that the relationship between therapeutic alliance and
We followed the main effects model with a stratified analysis to quality of life is modified by psychological distress. Of note, the beta
determine if the relationship between HCS and FACIT‐Pal was coefficients in the stratified analyses are larger than the beta
4
- THOMAS ET AL.

TABLE 1 Patient baseline characteristics, therapeutic alliance (HCS), quality of life (FACIT‐Pal), psychological distress (HADS) (N = 672)

Patient characteristics n (%)

Age, years (mean ± standard deviation) 69.3 ± 10.2

Sex

Male 312 (46.4)

Female 360 (53.6)

Race

Caucasian/White 632 (94.0)

African American/Black 33 (4.9)

Asian 5 (0.7)

Other 2 (0.3)

Ethnicity

Latino 9 (1.3)

Education

High school degree or less 335 (49.9)

Some college or 2‐year degree 201 (29.9)

4‐year degree 88 (13.1)

Master's degree, professional school, or doctoral degree 41 (6.1)

Declined to answer 6 (0.9)

Cancer type

Bladder 15 (2.2)

Brain 3 (0.4)

Breast 85 (12.6)

Colon 69 (10.3)

Endometrial 6 (0.9)

Esophageal 34 (5.1)

Gallbladder 2 (0.3)

Gastric 17 (2.5)

Head and neck 13 (1.9)

Kidney 27 (4.0)

Liver 8 (1.2)

Lung 243 (36.2)

Melanoma 9 (1.3)

Ovarian 22 (3.3)

Pancreatic 69 (10.3)

Prostate 29 (4.3)

Rectal 11 (1.6)

Sarcoma 6 (0.9)

Urethral 2 (0.3)

Other 2 (0.3)

Time since diagnosis

<1 month 23 (3.4)

THOMAS ET AL.
- 5

TABLE 1 (Continued)

Patient characteristics n (%)

≥1 month but <6 months 174 (25.9)

≥6 months but <1 year 96 (14.3)

≥1 year but <2 years 109 (16.2)

≥2 years but <5 years 131 (19.5)

≥5 years 136 (20.2)

Declined to answer 1 (0.1)

How long have you been receiving care from your current oncologist?

Less than 1 month ago 42 (6.3)

1–6 months ago 208 (31.0)

6 months to 1 year ago 114 (17.0)

1–2 years ago 118 (17.6)

2–5 years ago 119 (17.7)

>5 years ago 67 (10.0)

Declined to answer 3 (0.4)

Receiving chemotherapy at study enrollment 463 (68.9)

a
Patient‐reported outcomes
Mean ± SD Median (range)

Therapeutic alliance (HCS)

Total score (16–64) 56.4 ± 7.4 58 (19–64)

Quality of life (FACIT‐Pal)

Total score (0–184) 130.3 ± 25.5 131 (41–184)

Psychological distress (HADS)

Depression (0–21) 5.4 ± 3.8 5 (0–19)

Anxiety (0–21) 5.8 ± 3.9 5 (0–20)

Abbreviations: FACIT‐Pal, Functional Assessment of Chronic Illness Therapy—Palliative Care; HADS, Hospital Anxiety and Depression Scale; HCS,
Human Connection Scale.
a
Each measure is listed along with its possible range of scores.

coefficient in the overall analysis, likely because the HADS subscales depression and anxiety symptoms indicating that patients experi-
were removed as confounders. encing psychological distress also have a strong relationship between
Figure 2 illustrates the relationship between HCS and FACIT‐Pal therapeutic alliance with their oncologist and better quality of life.
stratified by HADS‐depression and HADS‐anxiety. The slope of The results suggest that a large change in therapeutic alliance is
regression lines for patients above and below the respective needed to improve quality of life, though a lack of published data
depression and anxiety thresholds are similar. regarding minimally important differences for either variable make
the clinical significance of this association difficult to determine.28
This result provides support to the report by Mack et al.9 that
4 | DISCUSSION therapeutic alliance is positively associated with several aspects of
quality of life in patients with advanced cancer.
Our results demonstrate that stronger patient–oncologist thera- This study highlights the importance of therapeutic alliance in
peutic alliance is associated with moderately better quality of life the advanced care setting. Patients in this study reported similar
among patients with advanced cancer. This finding supports our levels of therapeutic alliance compared to other studies in advanced
hypothesis that when patients perceive that their oncologist un- cancer populations.9,15 Efforts to enhance therapeutic alliance should
derstands their priorities and cooperates with them in managing their target both patients and oncologists as both groups contribute to the
care, patients experience positive health outcomes including better overall bond. Recent studies that tested combined patient and
quality of life. The association is consistent across levels of oncologist interventions to improve communication and shared
6
- THOMAS ET AL.

F I G U R E 1 Relationship between therapeutic alliance (HCS)

and quality of life (FACIT‐Pal) (N = 672). FACIT‐Pal, Functional
Assessment of Chronic Illness Therapy—Palliative Care; HCS,
Human Connection Scale

T A B L E 2 Adjusted relationships between therapeutic alliance

(HCS) and quality of life (FACIT‐Pal), overall and stratified by
degree of psychological distress (HADS‐Depression and HADS‐
Anxiety)

Adjusted

Β (95% CI) p Value

a
Entire cohort (n = 672) 3.69 (2.11, 5.27) <0.01
b
Stratified (Depression)

HADS‐D ≥ 5 (n = 369) 8.40 (5.53, 11.27) <0.01

HADS‐D < 5 (n = 303) 6.35 (2.68, 10.02) <0.01

b
Stratified (Anxiety)
F I G U R E 2 Relationship between therapeutic alliance (HCS)
HADS‐A ≥ 7 (n = 263) 6.99 (4.09, 9.90) <0.01 and quality of life (FACIT‐Pal) Stratified by (A) Baseline HADS‐
Depression and (B) HADS‐Anxiety. FACIT‐Pal, Functional
HADS‐A < 7 (n = 409) 6.77 (3.94, 9.60) <0.01
Assessment of Chronic Illness Therapy—Palliative Care; HADS,
Note: Beta coefficients scaled such that each represents the expected Hospital Anxiety and Depression Scale
change in FACIT‐Pal for an additional 10 points on the Human
Connection Scale.Abbreviations: CI, confidence interval; FACIT‐Pal,
Functional Assessment of Chronic Illness Therapy—Palliative Care; holistically, and encouraging open discussion of patient concerns.
HADS, Hospital Anxiety and Depression Scale. Once established, this may improve subsequent patient outcomes
a
Including fixed covariates for age, race, ethnicity, HADS‐Depression, including quality of life8 and end‐of‐life care.9
HADS‐Anxiety, time since diagnosis, and current chemotherapy Contrary to our hypothesis that psychological distress would
treatment.
b
weaken the relationship between therapeutic alliance and quality of
Including fixed covariates for age, race, ethnicity, time since diagnosis,
and current chemotherapy treatment.
life, patients with higher levels of depressive and anxiety symptoms
had a similar, if not stronger, relationship than patients with lower
levels. Although patients with higher psychological distress reported
decision‐making among cancer patients can assist researchers in lower therapeutic alliance overall, this finding suggests that patients
developing therapeutic alliance interventions.29,30 As oncologists experiencing distress benefit from perceiving their relationship with
engage in communication skills training, they will gain skills to their oncologist as trusting and collaborative and that this is associated
support patients in improving patient outcomes.31,32 As patients with better quality of life. This may occur because patients experi-
actively participate in their care, oncologists will likely engage in encing depressive and anxiety symptoms, when they are able to build
more relationship‐building efforts to support their patients' needs therapeutic alliance with their oncologist, have considerable oppor-
33
including their quality of life concerns. Therapeutic alliance tunity to improve their quality of life that further reinforces their
interventions should focus on building trust, recognizing the patient strong, collaborative relationship with their oncologist.34,35 Clinical
THOMAS ET AL.
- 7

interventions to support distressed patients connecting with caring, and lacked racial/ethnic diversity, which may limit the generalizability
responsive, reliable oncologists may mitigate the poor outcomes of our findings. Additional work is needed to assess therapeutic alliance
among patients with high distress.19,36 in more diverse patient populations. Fourth, our data focused on
There are several potential mechanisms linking strong thera- patient perceptions of their therapeutic alliance but lacked clinician
peutic alliance with better quality of life. We proposed that high factors that likely impact therapeutic alliance, including age, gender‐
levels of patient–oncologist therapeutic alliance lead to higher and racial‐concordance with patients, and communication style.8,40
quality of life through means of communication, collaboration, and
shared goals and understanding. Patients with a strong therapeutic
alliance will (a) work with their oncologists to discuss their concerns, 4.2 | Clinical implications
(b) participate as an active member in decisions and options, and
(c) develop a shared understanding with the oncologist about prior- Patients with a strong therapeutic alliance reported better quality of
ities for treatment and well‐being. Through each of these actions, the life across varying levels of psychological distress. Oncologists should
patient and oncologist co‐create pathways by which they can discuss cultivate a therapeutic alliance with their patients, recognizing that
and manage the patients' quality‐of‐life concerns. Patients who this bond can support the patients' well‐being.
report high therapeutic alliance may feel comfortable discussing their This study underscores the possible benefits of building a caring,
quality‐of‐life goals with their oncologist, who in turn incorporates trusting relationship between patients with advanced cancer and
these concerns into the patients' clinical care.37 Oncologist behaviors their oncologists. Future research should consider the longitudinal
likely contribute to patients' perceptions of therapeutic alliance and relationships between these variables to understand how therapeutic
the relationship between therapeutic alliance and quality of life. alliance, quality of life, psychological distress, and other patient
Oncologists who proactively assess and manage patients' quality of outcomes interact and evolve over time.
life likely also engage in behaviors that build therapeutic alliance (e.g.,
building a holistic understanding of the patient, clarifying patients' A C K NO W L E D G M EN T S
38 The following funding sources supported this study:(Schenker) NCI
goals, and eliciting patients' needs). Oncologists and patients may
exert a mutual influence over each other in which oncologists R01CA197103; (Schenker) Palliative Research Center (PaRC) at the
actively build a therapeutic alliance with patients who are more likely University of Pittsburgh; (UPMC Hillman Cancer Center)
to experience high quality of life.39 Understanding the key oncologist P30CA047904; (Thomas) American Cancer Society Mentored
behaviors that support therapeutic alliance and are associated with Research Scholar Grant MSRG‐18‐051‐51; and (Thomas) National
high quality of life will be particularly helpful in identifying targets for Palliative Care Research Center Career Development Award.
oncologist communication‐ and skills‐training.
CO N F L I C T O F I N T E R ES T S
The authors declare that there are no conflict of interests.
4.1 | Study limitations
E T HI CS S T A T E M E NT
This analysis has several limitations. First, given the cross‐sectional All study procedures involving human participants were performed in
nature of the data, we were unable to determine the directionality of accordance with the ethical standards of the University of Pittsburgh
the associations between therapeutic alliance and quality of life. It is and with the 1964 Helsinki Declaration and its later amendments or
conceivable that patients experiencing high quality of life may attri- comparable ethical standards. This study received ethical approval
bute this to their oncologist's care, creating a sense of trust and com- from the University of Pittsburgh Institutional Review Board
fort with their oncologist or vice versa. Despite this potential (STUDY19090204). The trial is registered at clinicaltrials.gov
relationship, it seems more plausible that the patient‐oncologist rela- (NCT02712229). Informed consent was obtained from all individual
tionship, characterized as a dynamic, ongoing collaboration, would participants included in the study.
produce ways to improve patient quality of life rather than patient
quality of life positively or negatively impacting the relationship. D AT A A V A I L A B I L I T Y S T A T EM E N T
Longitudinal research is needed to (a) determine how the relationship The data that support the findings of this study are available from the
between these constructs evolves over time, including changes in pa- corresponding author upon reasonable request.
tients' disease status which likely impact quality of life, and (b) test the
hypothesized mechanisms connecting therapeutic alliance to quality of OR C I D
life. Second, there may be alternative explanations clarifying the pos- Teresa Thomas https://orcid.org/0000-0001-7403-7314
itive association between therapeutic alliance and quality of life that
relate to unmeasured confounders that jointly impact both variables
R E F E R E NC E S
(endogeneity). This is a common limitation of cross‐sectional data and
1. DeMartini J, Fenton JJ, Epstein R, et al. Patients’ hopes for advanced
can be addressed in longitudinal research capable of testing causality cancer treatment. J Pain Symptom Manag. 2019;57(1):57‐63. https://
and mechanisms. Third, our patient population was quite homogeneous doi.org/10.1016/j.jpainsymman.2018.09.014.
8
- THOMAS ET AL.

2. Park SA, Chung SH, Lee Y. Factors influencing the quality of life of 19. Bisseling E, Cillessen L, Spinhoven P, et al. Development of the ther-
patients with advanced cancer. Appl Nurs Res. 2017;33:108‐112. apeutic alliance and its association with internet‐based mindfulness‐
https://doi.org/10.1016/j.apnr.2016.10.002. based cognitive therapy for distressed cancer patients: secondary
3. Samuel CA, Mbah O, Schaal J, et al. The role of patient‐physician analysis of a multicenter randomized controlled trial. J Med Internet
relationship on health‐related quality of life and pain in cancer Res. 2019;21(10):e14065. https://doi.org/10.2196/14065.
patients. Support Care Cancer. 2020;28(6):2615‐2626. https://doi. 20. Becker CL, Arnold RM, Park SY, et al. A cluster randomized trial of a
org/10.1007/s00520‐019‐05070‐y. primary palliative care intervention (CONNECT) for patients with
4. Farin E, Nagl M. The patient‐physician relationship in patients with advanced cancer: protocol and key design considerations. Contemp
breast cancer: influence on changes in quality of life after rehabili- Clin Trials. 2017;54:98‐104. https://doi.org/10.1016/j.cct.2017.
tation. Qual Life Res. 2013;22(2):283‐294. https://doi.org/10.1007/ 01.005.
s11136‐012‐0151‐5. 21. Moss AH, Lunney JR, Culp S, et al. Prognostic significance of the
5. Ernstmann N, Wirtz M, Nitzsche A, et al. Patients’ trust in physician, “surprise” question in cancer patients. J Palliat Med. 2010;13(7):837‐
patient enablement, and health‐related quality of life during colon 840. https://doi.org/10.1089/jpm.2010.0018.
cancer treatment. J Cancer Educ. 2017;32(3):571‐579. https://doi. 22. Lyons KD, Bakitas M, Hegel MT, Hanscom B, Hull J, Ahles TA.
org/10.1007/s13187‐017‐1166‐y. Reliability and validity of the Functional Assessment of Chronic
6. Epstein RM, Street, Jr., RL. Patient‐Centered Communication in Cancer Illness Therapy‐Palliative care (FACIT‐Pal) scale. J Pain Symptom
Care: Promoting Healing and Reducing Suffering. Bethesda, MD: Manage. 2009;37(1):23‐32. https://doi.org/10.1016/j.
National Cancer Institute; 2007. https://cancercontrol.cancer.gov/ jpainsymman.2007.12.015.
brp/docs/pcc_monograph.pdf. Accessed August 6, 2020.. 23. Zigmond AS, Snaith RP. The hospital anxiety and depression scale.
7. Rogers AI. The cornerstone of medicine: the physician‐patient Acta Psychiatr Scand. 1983;67(6):361‐370. https://doi.org/10.1111/
relationship. Am J Gastroenterol. 2007;102(8):1594‐1595. https://doi. j.1600‐0447.1983.tb09716.x.
org/10.1111/j.1572‐0241.2007.01160.x. 24. Singer S, Kuhnt S, Götze H, et al. Hospital anxiety and depression
8. Hillen MA, De Haes HCJM, Smets EMA. Cancer patients’ trust in scale cutoff scores for cancer patients in acute care. Br J Cancer.
their physician—a review. Psychooncology. 2011;20(3):227‐241. 2009;100(6):908‐912. https://doi.org/10.1038/sj.bjc.6604952.
https://doi.org/10.1002/pon.1745. 25. Gonzales FA, Sangaramoorthy M, Dwyer LA, et al. Patient‐clinician
9. Mack JW, Block SD, Nilsson M, et al. Measuring therapeutic alliance interactions and disparities in breast cancer care: the equality in
between oncologists and patients with advanced cancer. Cancer. breast cancer care study. J Cancer Surviv. 2019;13(6):968‐980.
2009;115(14):3302‐3311. https://doi.org/10.1002/cncr.24360. https://doi.org/10.1007/s11764‐019‐00820‐7.
10. Trice ED, Prigerson HG. Communication in end‐stage cancer: review 26. Maly RC, Frank JC, Marshall GN, DiMatteo MR, Reuben DB.
of the literature and future research. J Health Commun. 2009;14: Perceived efficacy in patient‐physician interactions (PEPPI): valida-
95‐108. https://doi.org/10.1080/10810730902806786. tion of an instrument in older persons. J Am Geriatr Soc. 1998;46(7):
11. Hoerger M, Epstein RM, Winters PC, et al. Values and options in 889‐894..
cancer care (VOICE): study design and rationale for a patient‐ 27. Fuertes JN, Mislowack A, Bennett J, et al. The physician‐patient
centered communication and decision‐making intervention for working alliance. Patient Educ Counsel. 2007;66(1):29‐36. https://doi.
physicians, patients with advanced cancer, and their caregivers. BMC org/10.1016/j.pec.2006.09.013.
Cancer. 2013;13:188. https://doi.org/10.1186/1471‐2407‐13‐188. 28. Ringash J, O’Sullivan B, Bezjak A, Redelmeier DA. Interpreting clin-
12. Lelorain S, Brédart A, Dolbeault S, Sultan S. A systematic review of ically significant changes in patient‐reported outcomes. Cancer.
the associations between empathy measures and patient outcomes 2007;110(1):196‐202. https://doi.org/10.1002/cncr.22799.
in cancer care. Psychooncology. 2012;21(12):1255‐1264. https://doi. 29. Epstein RM, Duberstein PR, Fenton JJ, et al. Effect of a patient‐
org/10.1002/pon.2115. centered communication intervention on oncologist‐patient
13. Ma Y, Xu W, Liang Z, et al. Patient‐oncologist alliance and psycho- communication, quality of life, and health care utilization in
social well‐being in Chinese society strongly affect cancer manage- advanced cancer the VOICE randomized clinical trial. JAMA Oncol.
ment adherence with cancer of unknown primary. Psychooncology. 2015;1(5):92‐100. https://doi.org/10.1001/jamaoncol.2016.4373.
2017;26(7):991‐998. https://doi.org/10.1002/pon.4245. 30. Henselmans I, Laarhoven HWM, Maarschalkerweerd P, et al. Effect
14. Bar‐Sela G, Yochpaz S, Gruber R, Lulav‐Grinwald D, Mitnik I, of a skills training for oncologists and a patient communication aid
Koren D. The association between the working alliance and on shared decision making about palliative systemic treatment: a
sharing concerns by advanced cancer patients. J Clin Oncol. 2015; randomized clinical trial. Oncologist. 2020;25(3):e578. https://doi.
33(15_suppl):e20553. https://doi.org/10.1200/jco.2015.33.15_suppl. org/10.1634/theoncologist.2019‐0453.
e20553. 31. Moore PM, Rivera S, Bravo‐Soto GA, Olivares C, Lawrie TA.
15. Freytag J, Street RL, Xing G, et al. The ecology of patient and caregiver Communication skills training for healthcare professionals working
participation in consultations involving advanced cancer. Psychooncol- with people who have cancer. Cochrane Database Syst Rev. 2018;
ogy. 2018;27(6):1642‐1649. https://doi.org/10.1002/pon.4710. 2018(7):CD003751. https://doi.org/10.1002/14651858.CD003751.
16. Grotmol KS, Lie HC, Loge JH, et al. Patients with advanced cancer pub4.
and depression report a significantly higher symptom burden than 32. De Vries AMM, de Roten Y, Meystre C, Passchier J, Despland J‐N,
non‐depressed patients. Palliat Support Care. 2019;17(2):143‐149. Stiefel F. Clinician characteristics, communication, and patient
https://doi.org/10.1017/S1478951517001183. outcome in oncology: a systematic review. Psychooncology.
17. Hay CM, Courtney‐Brooks M, Lefkowits C, Hagan TL, Edwards RP, 2014;23(4):375‐381. https://doi.org/10.1002/pon.3445.
Donovan HS. Symptom management in women with recurrent 33. Street RL, Krupat E, Bell RA, Kravitz RL, Haidet P. Beliefs about
ovarian cancer: do patients and clinicians agree on what symptoms control in the physician‐patient relationship: effect on communica-
are most important? Gynecol Oncol. 2016;143(2):367‐370. https:// tion in medical encounters. J Gen Intern Med. 2003;18(8):609‐616.
doi.org/10.1016/j.ygyno.2016.08.235. https://doi.org/10.1046/j.1525‐1497.2003.20749.x.
18. White LL, Cohen MZ, Berger AM, Kupzyk KA, Bierman PJ. Self‐ 34. Demark‐Wahnefried W, Aziz NM, Rowland JH, Pinto BM. Riding the
efficacy for management of symptoms and symptom distress in crest of the teachable moment: promoting long‐term health after
adults with cancer: an integrative review. Oncol Nurs Forum. the diagnosis of cancer. J Clin Oncol. 2005;23(24):5814‐5830.
2019;46(1):113‐128. https://doi.org/10.1188/19.ONF.113‐128. https://doi.org/10.1200/JCO.2005.01.230.
THOMAS ET AL.
- 9

35. Bernacki R, Paladino J, Neville BA, et al. Effect of the serious illness 39. Street RL, Gordon H, Haidet P. Physicians’ communication and
care program in outpatient oncology: a cluster randomized clinical perceptions of patients: is it how they look, how they talk, or is it just
trial. JAMA Intern Med. 2019;179(6):751‐759. the doctor? Soc Sci Med. 2007;65(3):586‐598. https://doi.org/
36. Applebaum AJ, Duhamel KN, Winkel G, et al. Therapeutic 10.1016/j.socscimed.2007.03.036.
alliance in telephone‐administered cognitive‐behavioral therapy 40. Pollak KI, Arnold RM, Jeffreys A, et al. Oncologist communication
for hematopoietic stem cell transplant survivors. J Consult about emotion during visits with patients with advanced cancer.
Clin Psychol. 2012;80(5):811‐816. https://doi.org/10.1037/a00 Artic J Clin Oncol. 2008;25:5748‐5752. https://doi.org/10.1200/
27956. JCO.2007.12.4180.
37. Bar‐Sela G, Yochpaz S, Gruber R, Lulav‐Grinwald D, Mitnik I, Koren
D. The association between the strength of the working alliance and
sharing concerns by advanced cancer patients: a pilot study. Support
Care Cancer. 2016;24(1):319‐325. https://doi.org/10.1007/s00520‐ How to cite this article: Thomas T, Althouse A, Sigler L, et al.
015‐2794‐6. Stronger therapeutic alliance is associated with better quality
38. Back AL, Arnold RM, Baile WF, Tulsky JA, Fryer‐Edwards K.
of life among patients with advanced cancerTHOMAS et al..
Approaching difficult communication tasks in oncology. CA Cancer
Psycho‐Oncology. 2021;1–9. https://doi.org/10.1002/pon.5648
J Clin. 2005;55(3):164‐177. https://doi.org/10.3322/canjclin.55.
3.164.