Introduction
Compassion can be defined as “a virtuous response that seeks to address the suffering and needs of a person through relational understanding and action” (Sinclair et al. Reference Sinclair, McClement and Raffin-Bouchal2016a; Reference Sinclair, Beamer and Hack2017a). Compassion differs from sympathy and empathy (Sinclair et al. Reference Sinclair, Beamer and Hack2017a). Sympathy is a pity-based state that involves an emotional reaction to the suffering of another person who is often perceived as weak and inferior and is rooted in a lack of understanding (Perez-Bret et al. Reference Perez-Bret, Altisent and Rocafort2016; Strauss et al. Reference Strauss, Lever Taylor and Gu2016; Sinclair et al. Reference Sinclair, Beamer and Hack2017a). Empathy refers to the capacity to connect with and understand the inner state of another person emotionally and cognitively (Perez-Bret et al. Reference Perez-Bret, Altisent and Rocafort2016; Strauss et al. Reference Strauss, Lever Taylor and Gu2016; Sinclair et al. Reference Sinclair, Beamer and Hack2017a). Compassion differs from empathy as it involves, “feeling for and not feeling with the other” (Singer and Klimecki Reference Singer and Klimecki2014) that is coupled with a sensitivity to others’ suffering that leads to a prosocial response to take action to alleviate or prevent it (Singer and Klimecki Reference Singer and Klimecki2014; Gilbert et al. Reference Gilbert, Catarino and Duarte2017; Brito-Pons and Librada-Flores Reference Brito-Pons and Librada-Flores2018). Accordingly, both the personalized and comprehensive nature of compassion has been recognized as salient and pivotal component of delivering high-quality holistic care to individuals (Sinclair et al. Reference Sinclair, McClement and Raffin-Bouchal2016a; Brito-Pons and Librada-Flores Reference Brito-Pons and Librada-Flores2018) particularly in palliative care as it promotes a care model that addresses the psychological, social, emotional, and existential needs of individuals at the end of their life (Sinclair et al. Reference Sinclair, Norris and McConnell2016b; Brito-Pons and Librada-Flores Reference Brito-Pons and Librada-Flores2018; Kang et al. Reference Kang, Go and Bruera2018). Recent research has shown that compassion is positively associated with patient quality of life and satisfaction with the care received, as well as with reduced healthcare worker burnout, increased job satisfaction, and workplace well-being (Sinclair et al. Reference Sinclair, Norris and McConnell2016b; Sinclair et al. Reference Sinclair, Beamer and Hack2017a; Lown et al. Reference Lown, Dunne and Muncer2017; Tehranineshat et al. Reference Tehranineshat, Rakhshan and Torabizadeh2019; Malenfant et al. Reference Malenfant, Jaggi and Hayden2022; Pavlova et al. Reference Pavlova, Paine and Sinclair2023).
Despite the importance of providing compassionate care being well recognized by healthcare organizations and standards of practice (Riddick Reference Riddick2003; Canadian Medical Association 2004; Commissioning Board Chief Nursing Officer and DH Chief Nursing Adviser 2012), patients still report compassionate care as being inadequate in healthcare (Lown et al. Reference Lown, Dunne and Muncer2017), while healthcare providers desire to provide compassionate care, they recognize the need for specific training to improve outcomes for patients and families (Coffey et al. Reference Coffey, Saab and Landers2019; Bovero et al. Reference Bovero, Adriano and Di Girolamo2023). Furthermore, there are several barriers that hinder the implementation of compassionate care practice, including individual, professional, organizational, and educational factors (Christiansen et al. Reference Christiansen, O’Brien and Kirton2015; Sinclair et al. Reference Sinclair, Norris and McConnell2016b; Dev et al. Reference Dev, Fernando and Kirby2019). Notably, the lack of a valid and reliable measure of patient experience of compassion has limited research in clinical practice and the development of evidence-based training approaches (Sinclair et al. Reference Sinclair, Russell and Hack2017c; Reference Sinclair, Kondejewski and Hack2022).
A recent systematic review, assessing the design and psychometric attributes of various compassion measures, identified the Sinclair Compassion Questionnaire (SCQ) as the “gold standard” measure of compassion (Sinclair et al. Reference Sinclair, Kondejewski and Hack2022). The SCQ is a 15-item patient-reported measure of compassion (Sinclair et al. Reference Sinclair, Hack and MacInnis2021) developed according to the Patient Compassion Model, an empirical model developed by Sinclair and his team, based on direct patient reports, which: defines and distinguishes compassion from related constructs; outlines the different components of compassion and their mutual relationships (Sinclair et al. Reference Sinclair, McClement and Raffin-Bouchal2016a; Brito-Pons and Librada-Flores Reference Brito-Pons and Librada-Flores2018); and, importantly, is transferable across care settings and patient populations. The SCQ has been validated in a sample of patients living with an incurable, life-limiting illness, demonstrating excellent internal consistency (Cronbach’s alpha of 0.96) and test–retest reliability (ranging from 0.74 to 0.89). The confirmatory factor analysis supported a single factor of compassion and item response theory analyses indicated that the SCQ precisely measures compassion across the wide range of patient experiences of healthcare providers (Sinclair et al. Reference Sinclair, Hack and MacInnis2021, Reference Sinclair, Kondejewski and Hack2022). Convergent validity was assessed using the 12-item Schwartz Center Compassionate Care Scale (Lown et al. Reference Lown, Muncer and Chadwick2015), showing a significant positive correlation (r = 0.75, p < 0.001), which confirms that the SCQ effectively assesses the construct of compassion. Divergent validity was assessed using the PICKER Patient Experience Questionnaire (Jenkinson Reference Jenkinson2002) and the Edmonton Symptom Assessment Scale (ESAS-r) (Watanabe et al. Reference Watanabe, Nekolaichuk and Beaumont2012). The results showed a moderately strong positive correlation between SCQ and the PICKER Patient Experience Questionnaire (r = 0.60, p < 0.001), and a weak negative association with depression (r = −0.13) and poor well-being (r = −0.17), as measured by the ESAS-r (p < 0.001), indicating that the SCQ is related to, but sufficiently distinct from, patient satisfaction and symptom distress (Sinclair et al. Reference Sinclair, Hack and MacInnis2021, Reference Sinclair, Kondejewski and Hack2022).
To date, the SCQ has been validated in English, Mandarin, and Spanish patient populations (Sinclair et al. Reference Sinclair, Russell and Hack2017c; Chu et al. Reference Chu, Jaggi and Louis2023; Soto-Rubio et al. Reference Soto-Rubio, Andrew and Beatriz2024). The aims of the present study were to validate the SCQ in Italian in order to provide a valid and reliable measure of patient-reported compassion to apply in both clinical practice and research in Italy, by assessing its psychometric properties using a study sample of patients diagnosed with cancer and with a life expectancy of less than 4 months.
Methods
The research protocol was divided into 2 different phases: (1) translation and back-translation of the SCQ, involving input from a panel of experts, (2) data collection from cancer patients at the end-of-life. The forward translation and subsequent back-translation procedure aimed to obtain semantic and linguistic equivalence in the Italian version with respect to the original English. Permission for translation was obtained from the original authors in advance.
A native Italian speaker fluent in both languages and familiar with the measure translated the SCQ from English to Italian in consultation with 2 members of the authorship team (A.B. and S.S.). The measure was back translated into English by an external translator who was not familiar with the measure. The resulting English version was checked against the original English measure and discrepancies were resolved by a panel of experts.
Data collection
A sample of terminal cancer patients were recruited from February 2022 to October 2023 from “Città della Salute e della Scienza di Torino” University Hospital and “Vittorio Valletta” Hospice in Turin, Italy. Inclusion criteria were: ≥ 18 years old; having a diagnosis of cancer; being able to provide informed consent; and meeting eligibility criteria to access palliative care (National Law on Palliative Care and Pain Treatment, No. 38/2010). The criteria to access palliative care include being terminally ill with no available or appropriate curative treatment, having an unfavorable prognosis with a presumed life expectancy of 4 months or less, and scoring 50 or lower in the Karnofsky Performance Status (KPS). Exclusion criteria included having a diagnosis of a severe psychiatric disorder and/or cognitive impairment with an inability to provide informed consent or to complete the study procedures. Patients were approached by a research assistant to complete the self-report scales and sociodemographic information, with clinical data being collected from medical records and clinical charts. Ethical approval of the study was obtained from the Hospital’s and Hospice’s Ethics Committee (#0034403). All patients included in the study were informed about the aims of the research and provided informed written consent.
Measures
The SCQ
Compassion was assessed using the Italian version of the Sinclair Compassion Questionnaire (SCQit), a 15-item scale developed to assess patients experiences of compassion across a variety of settings (Sinclair et al. Reference Sinclair, Jaggi and Hack2020). Patients were asked to rate their experience of compassion from their healthcare providers using a 5-point Likert scale of agreement (1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree) (Sinclair et al. Reference Sinclair, Hack and MacInnis2021). The SCQit overall score is determined by calculating the mean score of all items, with higher scores indicating greater reported compassion.
The Edmonton Symptom Assessment Scale (ESAS)
Presence and levels of cancer-related physical distress was assessed using the Edmonton Symptom Assessment System (ESAS-Revised) (Bruera et al. Reference Bruera, Kuehn and Miller1991) in its validated Italian version (Ripamonti et al. Reference Ripamonti, Leporati and De Feo2022). It rates the severity of 9 common symptoms of advanced cancer (i.e., pain, tiredness, nausea, depression, anxiety, drowsiness, appetite, well-being, and shortness of breath). Each symptom’s severity is rated from 0 (no symptoms) to 10 (worst symptom severity).
The Functional Assessment of Chronic Illness Therapy-Treatment Satisfaction-Patient Satisfaction (FACIT-TS-PS) measurement system
Quality of care and patient satisfaction were assessed using the Italian version of the Functional Assessment of Chronic Therapy-Treatment Satisfaction-Patient satisfaction (Guglielmetti Reference Guglielmetti2013) tool. This 30-item scale is composed of 5 subscales, namely: physician communication; treatment staff communication; technical competence; nurse communication; confidence and trust. Higher scores indicate better patient satisfaction (Peipert et al. Reference Peipert, Beaumont and Bode2014).
The Trust in Oncologist Scale-Short Form (IT-TIOS-SF)
Trust in the oncologist was assessed using the Italian version of the Trust in Oncologist Scale-Short Form (IT-TIOS-SF) (Hillen et al. Reference Hillen, Postma and Verdam2017; Bani et al. Reference Bani, Rossi and Cortinovis2021). It is a 5-item scale, using a 5-point Likert answering scale ranging from 1 (strongly disagree) through to 5 (strongly agree).
Statistical analysis
Descriptive statistics of demographic characteristics are reported as means and standard deviations for continuous variables, and frequencies and percentages for categorical variables. We computed descriptive statistics on the 15-item SCQit, including: means, SD, skewedness, kurtosis, and range, together with absolute frequency and response percentages. The internal reliability of the questionnaire was assessed by calculating Cronbach’s alpha. To ensure that data were suitable for factor analysis the Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy and Bartlett’s test of sphericity were performed. Exploratory factor analysis (EFA) was performed using the principal axis factor method and oblimin rotation. A scree plot and parallel analysis were performed to choose the number of factors, and the factor structure identified by EFA was tested through confirmatory factor analysis (CFA) and item response theory (IRT). To assess the divergent validity FACIT-TS-PS, IT-TIOS-SF, and ESAS questionnaires were considered. Finally, the relationships between SCQit and demographic variables were assessed by means of univariate linear regression. All the statistical analyses were performed using the statistical software R version 4.2.1.
Results
A total of 131 patients were enrolled on the study, and their characteristics are reported in Table 1. Descriptive statistics on the 15-item SCQit are reported in Table 2, and the frequencies and percentages of responses are shown in Figure 1. A Cronbach’ alpha value of 0.98 was obtained for the 15-item SCQit. Cronbach’s alpha remained stable when removing one item at time, thus all the items were considered for the following analyses. The criteria for running EFA were confirmed; indeed, the KMO measure of sample adequacy was 0.96 and Bartlett’s test statistic significant, with χ2 = 1692.22, df = 105, p < 0.001. The scree plot (Figure S1) indicated a single latent compassion factor that was determined with parallel analysis and that explained 79% of the variance. Factor loadings and communalities are reported in Table 3. To confirm the EFA results, we performed CFA and IRT, and to evaluate the model’s fit we computed the χ2, comparative fit index (CFI) and root-mean-square error of approximation (RMSEA) (Table 4). CFA factor loadings ranged between 0.81 and 0.92, and correlated (r = 0.97) with IRT factor loadings (see Supplementary materials for IRT results and Figures S2 and S3) ranging from 0.86 and 0.96.
Percentage of responses for each item.
Descriptive characteristics of the patients (n = 131)
Descriptive statistics of SCQit items
Factor loadings and communalities from EFA
Model fit metrics for CFA and IRT
In assessing divergent validity, as expected the SCQit correlated positively with both the FACIT-TS-PS (r = 0.81; 95% CI 0.71–0.88) and TIOS-SF (r = 0.75; 95% CI 0.65–0.83). This confirms that the SCQit is related to, but sufficiently distinct from, patient satisfaction and trust. Moreover, a slight negative correlation was revealed between the SCQit and ESAS (r = −0.17 95% CI −0.36–0.03).
The results of univariate linear regression models, performed to evaluate the relationship between SCQit and demographic characteristics, are shown in Table 5. Significant differences emerged between score of compassion and care location, with patients in hospitals experiencing less perceived compassion in comparison to patients in hospices (−0.42, p = 0.015). Religious patients reported slightly higher experiences of compassion (0.37, p = 0.043); while patients with high prognosis/diagnosis awareness reported significantly lower levels of compassion (−0.37, p = 0.023).
Linear Regression Model: effects of demographic variables on SCQ
Discussion and conclusion
Discussion
The aim of the present study was to evaluate the psychometrics properties of the SCQit in terminally ill cancer patients with a life expectancy of less than 4 months. As in the original validation (Sinclair et al. Reference Sinclair, Hack and MacInnis2021), a single factor model was identified, and excellent internal consistency was confirmed in an Italian context, with each of the 15 items mapping onto each of the domains of the patient informed Compassion Model (Sinclair et al. Reference Sinclair, McClement and Raffin-Bouchal2016a). Thus, this study provides evidence confirming that the Italian version of the SCQ is a valid and reliable measure of compassion, that is equally robust as the original English version of the SCQ.
Likewise, divergent validity analyses confirmed that the SCQit is related to, but sufficiently distinct from, patient satisfaction and trust. Furthermore, the positive association found between SCQit and FACIT-TS-PS and TIOS-SF suggests that patient experiences of compassion, enhances patient satisfaction and trust in their healthcare professionals. While future research is needed, we postulate that compassion is likely a predictor of patient satisfaction, while relating to patients as fellow human beings, understanding them as persons with unique needs, and actively attending to their suffering engenders deeper patient–provider trust. These results are important as patients who report greater trust in their physicians exhibit greater healthy behaviors, less symptoms, higher quality of life, and were more satisfied with their overall care (Sinclair et al. Reference Sinclair, McClement and Raffin-Bouchal2016a, Reference Sinclair, Norris and McConnell2016b). Notably, these findings further substantiate our previous findings (Chandra et al. Reference Chandra, Mohammadnezhad and Ward2018), that suggested compassionate healthcare providers are better able to meet patients’ needs in a more authentic manner while improving the patient–doctor therapeutic alliance. Moreover, the SCQit was slightly negatively correlated with the severity of distress related to overall symptoms (measured with the ESAS-r). This negative association implies that compassionate care was perceived as lower in participants reporting higher symptom distress, and consequently experienced a lower quality of life.
Secondary analysis aimed at evaluating the sociodemographic and care setting variables associated with SCQit showed no significant correlations between compassion and sex, age, marital status, employment, diagnosis, and stage of illness. However, lower levels of compassion were found among patients admitted to a hospital in comparison to those in hospice. A similar result was found also in Sinclair’s original validation, which found long-term care patients experienced lower levels of compassion in comparison to patients in hospice and an acute palliative care unit (Sinclair et al. Reference Sinclair, Russell and Hack2017c). These findings are in line with previous data which highlighted hospices as a preferential place of care for patients facing the end-of-life (Currow et al. Reference Currow, Agar and Phillips2020; Seale and Kelly Reference Seale and Kelly1997; Tatum Reference Tatum2020). Interestingly, our results revealed that religious patients perceive their healthcare providers as being more compassionate. While this may be due to a variety of factors such as lower expectations, differences in conceptualizations of compassion, additional sources of compassion (e.g. religious communities, God), other studies have demonstrated that religion, spirituality, and culture are external factors that affect individual’s perception of compassionate care (Sinclair et al. Reference Sinclair, Jaggi and Hack2018; Singh et al. Reference Singh, King-Shier and Sinclair2020).
Finally, our data also revealed that patients who with high prognostic awareness reported less compassion than those who were only partially aware of their condition. We are somewhat perplexed by this finding and recommend further research to investigate this phenomenon. Perhaps perceptions of compassion vary when the hope for a cure is no longer possible, or perhaps oncologists are less competent in expressing compassion at the end-of-life, or perhaps as has been reported in other studies (Hayuni et al. Reference Hayuni, Hasson‐Ohayon and Goldzweig2019; Sinclair et al. Reference Sinclair, Raffin-Bouchal and Venturato2017b), burnout associated with care at the end-of-life diminishes healthcare providers’ ability to provide compassion.
Strengths and limitations of the study
The main strength of the present study is that it has produced a valid, reliable, and easy-to-use tool for assessing the provision of compassionate care by healthcare providers toward Italian patients facing the end-of-life. Being able to better understand patients’ experiences of compassion can guide intervention research and facilitate the development of clinical cultures of compassion, which can enhance the patient’s quality of life and equipe healthcare providers with the necessary clinical skills to do so, while also mitigating burnout (Sinclair et al. Reference Sinclair, Norris and McConnell2016b; Reference Sinclair, Hack and MacInnis2021).
This study also has some limitations. First, future studies with larger samples are needed to confirm further the psychometric properties of the SCQit. Moreover, the generalizability of the SCQit to other patient populations is limited as this study was focused exclusively on end-of-life cancer patients. Another limitation is our inability to assess the convergent validity of the SCQit, as no other validated compassion measures are available in Italian, further highlighting the paucity of studies on this topic in this context. Finally, this study is limited in its cross-sectional design, which did not allow us to assess whether terminally ill patients experienced changes in the compassion they received during their final months – including changes in perceptions of compassion among patients with high prognostic awareness. Future studies could conduct a longitudinal assessment of compassion, although we recognize that the challenges associated with recruiting and losing patients before the end of a study is an inherent challenge in end-of-life research. Moreover, it could be interesting to investigate sociodemographic and clinical factors associated with compassion in other clinical populations, to determine if experiences of compassion vary across clinical settings, patient populations, and healthcare provider teams.
Conclusion
The results of the present study demonstrate that the SCQit is a valid and reliable measure of compassion in terminal cancer patients with a life expectancy of less than 4 months. The SCQit has the potential to improve patients’ perception of compassion during end-of-life care while providing healthcare providers with the necessary attitudes, knowledge, and skills to improve the compassion they provide in their practice. These findings underscore the important role that compassion plays in enhancing patient–provider relationships, in understanding patients’ personhood and unique needs, and in improving quality palliative care.
Practice implications
The SCQit can be used in both clinical practice and research in Italy to assess patients’ perception of compassion during end-of-life care and to evaluate the effectiveness of training to promote compassionate care in healthcare providers. Further research is needed to validate the SCQit in other populations and healthcare settings.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1478951524001202.
Acknowledgments
The authors would like to thank for the time given the individuals who participated in the study. The authors are also grateful to Stephanie Parsley for the proofreading of the manuscript.
Use of the SCQ, authored by Dr. Shane Sinclair, Dr. Tom Hack, Dr. Cara MacInnis, Harrison Boss, Priya Jaggi, Dr. Susan McClement, Dr. Aynharan Sinnarajah and Dr. Genevieve Thompson was made under license from UTI Limited Partnership.
The SCQit and other versions of the SCQ are available at www.compassionmeasure.com; by emailing the contact author directly (sinclair@ucalgary.ca) or by emailing ipm@innovatecalgary.com.
Author contributions
Conceptualization, A.B., A.F., S.S., and S.C.; data collection, A.F., F.C., and I.D.G.; statistical analysis: S.U. and P.B.; interpretation of data, all authors; writing – original draft preparation, A.B., A.F., and S.C.; figure preparation, S.U. and P.B.; writing – review and editing, all authors; supervision, A.B., L.O., and S.C. All authors have read and approved the final version of the manuscript.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Competing interests
The authors declare that there is no conflict of interest.
We confirm all patient identifiers have been removed or disguised so the patients described are not identifiable and cannot be identified through the details of the story.
Open access
