Rather a Competent Practitioner than a Compassionate Healer: Patients’ Satisfaction with Interactions in Psychiatric Inpatient Care

Abstract

Interactions with staff are important aspects in patients’ experiences of psychiatric inpatient care (PIC). This study aimed to evaluate patients’ satisfaction with their interactions with PIC staff and whether sociodemographic factors, depression and anxiety symptoms were associated with their perceptions of these interactions. In this cross-sectional study, we collected data from 84 patients receiving inpatient care in three psychiatric settings in Sweden. The patients’ perceptions of interactions with staff and self-reported degrees of depression and anxiety were evaluated through questionnaires. Overall, patients were satisfied with the patient–staff interaction. However, significantly higher scores were related to staffs’ practical competence than to their compassion. Older patients reported significantly more satisfaction than younger patients with their most recent meeting with staff. Tailored nursing interventions may improve staff’s compassionate capacity. Further research in larger samples is needed to improve our understanding of the factors associated with how patients perceive their interactions with staff.

Introduction

Psychiatric inpatient care (PIC) commonly includes pharmacological interventions and counseling, together with different types of behavioral and milieu therapies (Smith & Spitzmueller, 2016; Thomas et al., 2002). On their course to recovery, patients tend to express a need for support and encouragement in their progress (Gunasekara et al., 2014), highlighting the importance of staffs’ ability to help and strengthen patients in their recovery. Previous studies have shown that patients’ satisfaction with PIC hospitalization is associated with better long-term prognoses and a predictor of treatment compliance and further use of PIC (Chevalier et al., 2018; Kuosmanen et al., 2006). The most important relationships in the psychiatric unit are those between patients and staff, emphasizing the need for therapeutic relationships built on trust, empathy, and mutual respect (Walsh & Boyle, 2009). From the patients’ perspective, interactions with staff are vital to their recovery (Wyder et al., 2015).

Interactions between patients and staff are key to PIC patients’ satisfaction with their care. The patient–staff relationship seems to be a crucial aspect of the inpatient experience. When staff recognize patients as whole people, patients seem to find it easier to be physically and emotionally close to others and to themselves (Eldal et al., 2019). When staff spend time and do everyday activities with them, patients feel more satisfied with their care (Adnoy Eriksen et al., 2014; Molin et al., 2016a). It has also been shown that staff who do not spend time or talk with the patient are perceived to reflect a lack of interest and commitment (Stenhouse, 2011).

From a nursing perspective, PIC is generally considered a complex context. Staff need not only to manage patients’ rights to autonomy and acknowledge them in their interactions, but also to provide adequate medical care (Cleary, 2004). In the everyday clinical setting, staff need to manage medical versus personal recoveries (Tuffour, 2017). Studies show that staff consider personal interactions and relationships with patients as fundamental to PIC (Fourie et al., 2005; Moreno-Poyato et al., 2016; Shattell et al., 2008), but they generally do not have adequate time, support, or staffing to optimize their interactions with patients (Stenhouse, 2011; Walsh & Boyle, 2009). Meaningful interactions occur in intimate meetings between patients and staff members. Previous researchers have reported lack of time for quality interactions in PIC, which may cause stress when staff are hindered from working in line with their ideals (Graneheim et al., 2014; Molin et al., 2016b; Shattell et al., 2008). In addition, nurses express both stress and ethical issues in relation to protecting patients’ rights and autonomy (Ulrich et al., 2010). Moral distress as such, might negatively affect nurses’ interaction with patients as it could mean that they shift their focus from the patients’ best interest to their own survival (Gabrielsson et al., 2016).

Various patient characteristics have been associated with satisfaction with care in general, so there may be identifiable factors that affect interactions between patients and staff in PIC in particular. For example, researchers of a European multisite study suggest that higher patient satisfaction with interactions in PIC was associated with older age, having an occupation, living with others and having a close friend (Bird et al., 2020). However, quantitative studies on factors associated with patients’ satisfaction with PIC in Sweden are scarce, as is knowledge of the effect of patients’ sociodemographic factors (e.g., gender, age, level of education, and civil state) on these interactions. An improved understanding of such factors may not only contribute to better understanding of successful interactions, but also reveal aspects in PIC that may benefit from improvement.

This study aimed to evaluate patients’ satisfaction with their interactions with PIC staff and whether sociodemographic factors, depression and anxiety symptoms were associated with their perceptions of these interactions.

Materials and methods

This cross-sectional study, conducted at three psychiatric care units in two geographically close counties with both rural and urban areas in northern Sweden, was part of a larger project to evaluate the implementation of the Time Together nursing intervention in PIC (Molin et al., 2017). The catchment area for two of the units meant that some patients had to travel as much as 150 km to get to the hospital, while for the third hospital the maximum distance was approximately 60 km.

All three units had common rules and routines about locked doors, fixed times for meals, smoking breaks, and opportunities to go outdoors. In general, the work on the units was based on routines. Medical treatment was predominant, and opportunities for psychosocial interventions such as joint activities, planned dialogues, and psychoeducation were generally rare. The staff worked in interprofessional teams consisting of enrolled, registered and mental health nurses, unit managers, occupational therapists, physicians, and consulting psychiatrics. Other professionals could be consulted when necessary.

Study population

All admitted patients in the three units were offered participation in the study during the period of January to May, 2017. Patients were admitted both voluntarily and involuntarily and treated for various forms of mental ill health, addictions, or both. The inclusion criteria were 18 years or older and cared for in one of the three selected psychiatric inpatient units. The only exclusion criterion for patients was insufficient knowledge of the Swedish language to fill out the questionnaire or participate in interviews.

Procedure

During the study period, patients admitted to the units received information about the study, both in writing and verbally, from an assigned staff member, not part of the research team, and were invited to participate. Participants received the questionnaires during their first week of admission. The assigned staff member distributed the questionnaires to the eligible and consenting participants at each unit, together with instruction to fill in the questionnaires while thinking about their most recent staff interaction.

Questionnaires

A translated version of the Caring Professional Scale (CPS) was used to inquire into patients’ perceptions of their latest interaction with healthcare staff. The CPS is based on Swanson’s Caring Theory, which states that humane nursing aims to take care of people with actual or potential health disorders until they can take care of themselves. Swanson (1991) defines caring as ‘a nurturing way of relating to a valued other toward whom one feels a personal sense of commitment and responsibility’ (p. 162).

The CPS was developed as a Likert-type scale with 18 items scored from 1 to 5 to rate patients’ experiences of their interactions with staff (Swanson, 2000). The scales consist of two independent subscales, ‘competent practitioner’ and ‘compassionate healer’. The reliability of the instrument was deemed excellent (Cronbach’s alpha = 0.97) and criterion validity, satisfactory (Swanson, 2000). The translated (i.e. from English to Swedish) version of CPS, used in this study, consists of 15 separate items, resulting in a score ranging from 15 to 75. The lower the number, the higher rating quality of the interaction. Cronbach´s α for CPS was 0.95 in our sample.

We used the Visual Analog Scale (VAS) to measure participants’ satisfaction with their most recent interaction with staff in PIC (Huskisson, 1974). The scale is well-tested and often used in the social and behavioral sciences measuring subjective phenomena (Wewers & Lowe, 1990). It measures 10 cm in length and ranges from 0 (very unsatisfactory) to 10 (very satisfactory).

The valid Hospital Anxiety and Depression (HAD) scale was used to assess self-reported anxiety and depression (Zigmond & Snaith, 1983). All questions are answered on a four-point scale from 0 (not at all) to 4 (most of the time). The reliability of the Swedish version is considered satisfactory with Cronbach’s α 0.90 (Lisspers et al., 1997). Cronbach’s α for HAD was 0.86 in our sample.

Demographic data on gender, age, level of education, and living alone or not were also collected.

Statistical methods

Descriptive statistics were used to characterize the study population. The Kolmogorov-Smirnov method was used for the test of normality. Parametric statistics were applied on normally distributed data (independent sample t-test). Non-parametric statistics were used for comparisons and correlations between groups, considering the ordinal character of variables and no normal distribution (Chi² test, Mann-Whitney U, and Spearman’s Rho). The sum score of all items on CPS was calculated, as was a sum score for each of the subscales, ‘compassionate healer’ and ‘competent practitioner’. Wilcoxon signed rank test for related samples was used to analyze differences between the two subscales, and the sample’s median score was calculated. A multivariable logistic prediction model was planned to evaluate patient-related sociodemographic factors associated with their reported level of satisfaction with their healthcare provider. The patients’ level of satisfaction according to the CPS were categorized as either ‘equal or below median’ or ‘over median’ and then used as the dependent variable in the logistic multivariable regression model. To check for multicollinearity, a correlation matrix was conducted between the independent variables, which included level of education, and living alone or not (Field, 2013) and controlled for age and gender. Age was analyzed as a binary variable, with participants aged over 35 years denoted as ‘older’. Multicollinearity was considered positive at >0.7 and not present in the data. For all other analyses, a probability level of <0.05 was considered statistically significant. The statistical analyses were carried out using the Statistical Package for the Social Sciences (IBM SPSS Statistics for Macintosh, Version 26).

Ethical considerations

Data were anonymized in accordance with the Declaration of Helsinki to prevent the identification of any individual participant. Before their inclusion, all participants signed their written, informed consent. All participants were also informed of their right to withdraw their participation at any time without giving any reason. The assigned staff member was informed to administer the questionnaires to the participants when they were not in obvious emotional difficulties. The project was approved by the Regional Ethics Review Board in Umeå (Ref no: 2016/339-31) and registered at ClinicalTrials.gov (Study ID: NCT02981563).

Results

A total of 84 participants (42 women, 38 men, 4 missing data) were included (Table 1). There was no significant difference in age between women and men (Mean age = 37.7 yrs; SD = 12.6; t = 1.791; p = 0.077). The frequency of self-reported psychiatric diagnoses was 58.8%. There was no difference in distribution of psychiatric diagnoses among women and men (Chi² = 1.691; p = 0.429). Most participants qualified for possible or probable anxiety or depression (Table 1). The estimated degree of anxiety was positively associated with the degree of depression (Chi² = 18.39; df = 4; p = 0.001).

Table 1. Descriptive statistics of the participants in the study (n = 84).

Variable	n (%)
Gender Men Women	38 (45.2) 42 (50.0)
Age, years^a	37.7 (SD 12.6)
Level of education Compulsory education Intermediate education Higher education/university	28 (32.9) 37 (43.5) 15 (17.6)
Living alone	37 (44.0)
Self-reported psychiatric disease Yes No Don’t know	50 (58.8) 17 (20.0) 13 (15.3)
Self-reported level of anxiety None Possible Probable	11 7 60
Self-reported level of depression None Possible Probable	17 19 42

^a Mean age and standard deviation (SD).

On average, patients ranked their satisfaction with the recent meeting with healthcare staff as 6.84 (SD = 2.79) on the VAS (n = 78). Older patients reported significantly higher satisfaction with the meeting than younger patients (Mann-Whitney U = 436.50; p = 0.013). There was no significant difference in satisfaction between men and women (men: mean rank = 38.97, women: mean rank = 36.32; Mann-Whitney U = 628.00; p = 0.598). Individual estimated degree of satisfaction was not significantly correlated to self-reported degree of either depression (Spearman’s rho, correlation coefficient = −0.211; p = 0.071, 2-sided) or anxiety (Spearman’s rho, correlation coefficient = −0.096; p = 0.416, 2-sided).

The distribution of patients' answers on their perceptions of their interaction with health care staff according to the CPS is presented in Table 2. The median total sum score was 29 (min 5 and max 65). The participants reported significantly higher scores on their perception of interaction regarding the staffs’ competence (median = 13.0; IQR = 7) when compared to the compassionate subscale (median= 15.5 IQR= 11.25). Related samples Wilcoxon Signed Rank test p ≤ 0.05). In the multivariable regression model, controlled for age and gender, no patient characteristic factors were identified that were associated to the level of satisfaction (Table 3).

Table 2. Experiences of interaction with health care professionals according to a modified questionnaire based on the Caring Professional Scale (N = 84).

		n	Yes (1)	Mostly (2)	Neither or (3)	Sometimes (4)	No, absolutely not (5)	Irrelevant	Mean	IQR
Competent practitioner	Positive	79	36 (46)	25 (32)	9 (11)	4 (5)	5 (6)		2	1
	Informative	72	25 (35)	24 (33)	8 (11)	10 (14)	4 (6)	1 (1)	2	2
	Clinically competent	75	24 (32)	22 (29)	15 (20)	7 (9)	2 (3)	5 (7)	2	2
	An attentive listener	78	37 (48)	22 (28)	6 (8)	7 (9)	6 (8)		2	1
	Centered on you	75	23 (30)	25 (33)	8 (11)	11 (15)	6 (8)	2 (3)	2	2
	Technically skilled	77	13 (17)	20 (26)	20 (26)	10 (13)	2 (3)	12 (16)	2	2
	Respectful	77	37 (48)	31 (40)	2 (3)	3 (4)	4 (5)		1.5	1
Compassionate healer	Comforting	77	25 (32)	21 (27)	14 (18)	5(6)	8 (10)		2	2
	Understanding	80	35 (44)	27 (34)	4 (5)	10 (13)	4 (5)		2	1
	Personal	72	22 (31)	22 (31)	12 (17)	6 (8)	6 (8)	4 (6)	2	2
	Caring	75	29 (39)	21 (28)	11 (15)	7 (9)	6 (8)	1 (1)	2	2
	Supportive	77	28 (26)	30 (39)	6 (8)	7 (9)	5 (6)	1 (1)	2	1
	Aware of your feelings	72	14 (19)	19 (26)	18 (25)	9 (13)	10 (14)	2 (3)	2.5	1
	Visibly touched by your experience	73	14 (19)	20 (27)	24 (33)	5 (7)	7 (10)	3 (4)	3	1
	Able to offer you hope	76	21 (28)	26 (34)	13 (17)	9 (12)	6 (8)	1(1)	2	1

IQR = interquartile range.

Table 3. Associations between participants reported level of satisfaction with their care provider and their sociodemographic factors.

Independent variable	ß	Odds ratio (95% CI)
Gender	–0.309	0.73 (0.28–1.94)
Age^a	–0.889	0.41 (0.15–1.12)
Level of education^b	0.253	1.28 (0.38–4.37)
Living alone or not	0.518	1.17 (0.44–3.13)

Associations analyzed in a multivariable logistic regression model controlled for age and gender.

^a Age dichotomized to below or above the median 35.0.

^b Level of education dichotomized to primary/secondary school versus higher education/university .

Discussion

The main finding from this study is that the majority of the patients reported being primarily satisfied with staffs’ interaction at the three psychiatric inpatient units. Among the individual, patient-related factors analyzed, younger patients reported lower levels of satisfaction in their last interaction with a staff member. Interestingly, patients reported significantly higher on items related to the subscale competent practitioner than to items related to the subscale compassionate healer.

Most participants in this study were satisfied with their last interaction with staff, in line with previous quantitative studies suggesting that patient generally tend to be satisfied with PIC (Kuosmanen et al., 2006; Ratner et al., 2018). Previous studies have shown that the characteristics of healthcare staff that contribute most to satisfaction with care are respect, caring, and kindness (Hörberg et al., 2004; King et al., 2019; Wagoro et al., 2008) and that patients are generally satisfied with their relationships with healthcare staff (Wagoro et al., 2008). This is in line with our findings, showing that most participants perceived their healthcare staff as both competent and compassionate. However, the participants in our study reported significantly lower quality for the staffs’ compassionate skills than for their practical competence. This may indicate that staff focus on medical recovery before personal recovery. This has been previously discussed by other researchers. For example, Coffey et al. (2019) report that there is an ambivalence among staff about the relevance of personal recovery in PIC where patients are very unwell. Interestingly, the results from quantitative versus qualitative studies on patients’ perceptions of interaction with staff tend to diverge. Although quantitative data from the above-mentioned studies tend to show that patients are generally satisfied with PIC, qualitative results suggest that patients greatly prefer personal interactions with staff over medical interventions (Molin et al., 2016a; Waldemar et al., 2018). In addition, patients with mental ill health express a need to be treated with compassion in care (Gunasekara et al., 2014). Further efforts are therefore needed to improve interactions to promote patients’ personal recovery in PIC, particularly interactions in which staff show their compassion.

The relation between patient age and satisfaction with care has been thoroughly investigated. In general, and in line with our findings, older patients have reported more satisfaction than younger patients with interactions with staff (Kuosmanen et al., 2006; Zendjidjian et al., 2014). In medicine, different explanations have been suggested for why older people report more satisfaction (Crow et al., 2002); there may, for instance, be a generational difference in which older patients are more accepting overall than younger. Another explanation is that older patients may have lower expectations of healthcare staff based on their previous experiences when standards of care were lower. Older patients may also perceive a greater obligation not to complain about their care (Crow et al., 2002). In this study, except for the greater satisfaction of older patients, no patient-specific factors were associated with more or less satisfaction with interaction with staff. No other data were available to explore possible any further factors than those identified, so future studies in larger samples are needed to reveal the reasons behind this finding.

The results of our study also show that degree of depression is not correlated with satisfaction with the interaction with staff. Previous research, however, suggests that people with depressive symptoms generally report fewer positive interactions and more negative social interactions (Nezlek et al., 2000). Since our study was based on a relatively small sample, it may have been underpowered to return any significant findings. These results should be interpreted with some caution. In the relationship between depression and satisfaction with interactions with staff, however, a number of possible factors may be considered. First, the self-image and self-esteem of patients with depression seem to be negatively affected by their ill health. As a consequence, their expectations of interactions with others may be lower (Sowislo & Orth, 2013). It is therefore possible that perceptions of the care provided by healthcare staff are affected by the patient’s feelings of depression and unworthiness of the care. Also, based on a potential imbalance of power in the relationship between patients and staff (Molin et al., 2016b), it is possible that patients tend to overestimate their satisfaction with the interaction.

Taken altogether, even though patients report satisfaction with their interactions with staff, our findings highlight the necessity for staff to continuously reflect on their skills in patient interactions, especially their compassionate capacities.

Limitations

In this study, data were collected in a standardized manner and according to a predefined protocol (Molin et al., 2017). Validated instruments were used to assess the interactions and patient characteristics, but these instruments had been designed for use in different patient populations. The CPS instrument was originally developed to evaluate women’s experiences of caregivers’ interactions during and following a miscarriage. For unknown reasons, three items of 18, all negative, were removed from the original in the version translated into Swedish. This may have contributed to overestimations in the measured levels of satisfaction. The other instruments used in the study, VAS and HAD, are both regarded as reliable and valid instruments, and thus, these results can be considered reliable for their respective constructs. No information from the patients’ medical records were available for analysis. Consequently, it was not possible to carry out any analysis of other potential associations with levels of satisfaction. The three units studied had only small differences in organizational factors, routines, structures and factors such as the nursing staff’s levels of education, previous experience, and working conditions. Taken together, therefore, and our results may be generalized to comparable settings, although with some caution, considering the relatively small sample.

Conclusion and clinical implications

Overall, patients in the studied psychiatric inpatient care units were satisfied with their last interaction with healthcare staff, although younger patients reported lower levels of satisfaction. Further studies in larger samples are needed to increase understanding of the factors associated with patients’ perceptions of their interactions with healthcare staff.

Patient–staff interactions are core components in patient satisfaction with PIC and in the work of promoting patients’ personal recovery. Our study shows that patients in PIC are generally satisfied with these interactions, but they report higher scores for staffs’ practical competence than their compassionate capacities. Therefore, tailored nursing interventions are needed to increase compassionate competence among staff. Nursing interventions based on theoretical frameworks that assume fellowship and humanity and joint activities are promising directions for this improvement. Mental health nurses, as experts on relational work and associated nursing interventions, are in a key position to lead such a development.

Author contributions and authorship statement

Study design: JM, AR, BML, UHG; Data collection: JM; Analysis: AL, MV; Manuscript: JM, MV, AL, AR, BML, UHG. All authors are in agreement with the manuscript.

Acknowledgements

The authors would like to extend our sincere gratitude to all who contributed and enabled collection of data in this study: Sofie Press, Fredrik Boberg, Alf Eriksson, Malin Lundin, Ida Ingelsson, Ida Maria Barcheus, and Angela Pettersson.

Disclosure statement

The authors declare no conflicts of interest.

Additional information

Funding

Funding was granted by The Swedish Society of Nursing.