Construction and psychometric evaluation of the Swedish language Person-centred Climate Questionnaire staff version

Transcription

1 Journal of Nursing Management, 2009, 17, Construction and psychometric evaluation of the Swedish language Person-centred Climate Questionnaire staff version DAVID EDVARDSSON R N, P h D 1,2, P.O. SANDMAN R N, P h D 3 and BIRGIT RASMUSSEN R N, P h D 4 1 Assistant Professor, Department of Nursing, Umea University, Umea, Sweden, 2 Adjunct Associate Professor, Australian Centre for Evidence-Based Aged Care, (ACEBAC), La Trobe University, Melbourne, Victoria, Australia, 3 Professor and 4 Associate Professor, Department of Nursing, Umea University, Umea, Sweden Correspondence David Edvardsson Department of Nursing Umea University SE Umea Sweden david.edvardsson@nurs. umu.se E D V A R D S S O N D., S A N D M A N P. O. & R A S M U S S E N B. (2009) Journal of Nursing Management 17, Construction and psychometric evaluation of the Swedish language Personcentred Climate Questionnaire staff version Aim The aim of the study was to construct and evaluate psychometric properties of the Swedish language Person-centred Climate Questionnaire staff version. Background Person-centred care is often quoted but ill defined, and the literature warrants the development of valid and reliable measurement tools. Methods During 2006, a questionnaire was constructed and distributed to a sample of Swedish hospital staff (n = 600). Questionnaire data was subjected to item analysis and reduction. Psychometrical properties of the questionnaire were evaluated. Results The 14-item Person-centred Climate Questionnaire showed satisfactory psychometric properties. Measures of validity were good, internal consistency was high, CronbachÕs alpha was satisfactory for the total scale (0.88) and test retest reliability was adequate. Conclusion The results indicate that the staff Person-centred Climate Questionnaire is a valid and reliable tool for assessing staff perceptions of the person centredness of hospital environments. Implications for nursing management This instrument makes it possible to study the degree of person-centredness in relation to different organizational systems, environments, staff groups and managerial styles. In addition, staff variables such as turnover rates, health outcomes and efficacy can be related to staff perceived person centredness of the organization. Keywords: health facility environment, nursing, organization and administration, patient-centered care, questionnaires, social environment Accepted for publication: 26 January 2009 Introduction Person-centred care is a concept that points towards the importance of including the life world of patients into caring (McCormack 2004). Person-centred care has been described to include: knowing the person and biography; recognizing personhood and respecting individuality; acknowledging the lived world and personal values; promoting relationships; and empowering patient decision making (Kitwood 1997, Brooker 2004, Kjeldman et al. 2006, McCormack & McCance 2006, Slater 2006, OÕDonovan 2007, Penrod et al. 2007). DOI: /j x 790 ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd

2 Evaluation of the PCQ-S Person-centred care can shift the power to the ill person, and is not to be confused with client-centred or patient-centred care, as the connotations of patient and client can contribute to shift the power back to the health care staff by focusing on the disabling illness rather than the able person (Slater 2006). However, the content and meaning of person-centred care is not unambiguously described or defined in the literature. It has, for example, been described that person centredness demands certain aspects of staff members; of the care environment; and of the activities comprising the care process (McCormack & McCance 2006). Person-centred staff were described as being competent and committed, as having interpersonal skills and knowing themselves. Person-centred care environments were described as consisting of systems that facilitate shared decision making and a sharing of power. Finally, person-centred processes were described as such activities that operationalize person-centred principles, for example, using biographical information and facilitating patient participation (McCormack & McCance 2006). Similarly, Brooker (2007) described that person-centred care demands valuing the person and staff caring for them, acknowledging individuality which includes prioritizing needs of patients rather than needs of staff, using the perspective of the person, and creating a positive social environment to promote wellbeing. Even though the concept is widely used in academia as well as in professional nursing forums, it has been criticized for being often quoted but ill defined (Nolan et al. 2004). Recent publications have asked for developments of valid and reliable measurement tools by which levels of person centredness can be explored and linked to patient and staff outcomes (McCormack & McCance 2006, Slater 2006). Person-centred care with its holistic and humanistic ideals is often used synonymously to good quality care even though this remains to be supported by evidence (McCormack 2004). Others have actually questioned if personcentred care can be achieved or if it is nothing more than a political slogan or evangelical ideal, and if demanding this from a strained everyday practice can contribute to staff stress (Packer 2000). It has also been suggested that person-centred care falls short of acknowledging the interdependency of humans. In their opinion, the individuality expressed through the concept of person-centred care does not give due weight to the importance of relationships in nursing care, and thus they propose an alternative model of relationshipcentred care (Nolan et al. 2004). However, Slater (2006) suggested that another concept only contributes to increased conceptual inconsistency. McCormack (2004) described that one of the core concepts of person-centred nursing, being in place, has great potential to influence well being even although, in his opinion, few studies have been performed to assess the influence of place on patientsõ experiences (McCormack 2004). Studies on the meaning and significance of place for patient and staff well-being have been performed and undergone operationalization into scales measuring to what extent the climate of health care settings are experienced as person centred. Two scales measuring the person centredness of environments have been developed, one for patients (Edvardsson et al. 2008) and one for staff. This article reports the psychometric properties of the Swedish language Person-centred Climate Questionnaire staff version (PCQ-S). Methods The aim of the present study was to evaluate psychometric properties of the Swedish language PCQ-S, a questionnaire measuring to what extent the climate of health care settings is experienced as person centred by staff. The study employed methods of scale construction and psychometric evaluation. Sample To achieve variation in physical environments and type of care, a sample of 25 hospital wards at three hospitals in northern Sweden were invited to participate. Four medical wards at one hospital declined participation, and the final sample consisted of 21 hospital wards. All health care staff (n = 1053) at these wards were considered eligible for participation and were provided a questionnaire. A total of 600 health care staff participated in the study, which represents a response rate of 57%. Written consent was not required and return of a completed questionnaire was taken to imply consent. Envelopes with completed questionnaires were collected in a sealed box at the site or sent directly to the principal investigator. A sub-sample of 37 staff members from a combined medical/surgical ward was recruited to participate in test retest evaluation in which the questionnaire was completed twice with 1 week in between. The ethical committee of the medical faculty at Umeå University approved the study (Dnr ). Construction of the PCQ A preliminary self-report 45-item Likert-type questionnaire was created from the theory conceptualizing supportive environments as places that maintain the ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd, Journal of Nursing Management, 17,

3 D. Edvardsson et al. personhood of the individual through providing a person-centred climate (Edvardsson et al. 2005, Edvardsson & Street 2007, Edvardsson 2008) and from research literature. The initial questionnaire was formulated as statements about the climate at the ward and covered the three main climate categories of the theory, namely safety, everydayness and hospitality. A six-point Likert scale was used for scaling purposes (1 = No, I disagree completely, to 6 = Yes, I agree completely). The layout of the scale followed the item order suggested by the factor analysis with items placed in factorial order with descending factor loadings. Psychometric evaluation To evaluate the validity and reliability of the instrument, item reduction and evaluation was performed, content and construct validity of the PCQ was evaluated, and reliability was assessed through estimates of internal consistency and test retest reliability. Item reduction and evaluation Item reduction involved principal component analysis with varimax orthogonal rotation to ensure independence of factors. We used KaiserÕs eigenvalue >1 criterion to decide on the number of factors to extract. Low factor loadings (<0.5) or items loading moderate to high (>0.4) in multiple factors were used as elimination criteria. Item performance was evaluated with squared multiple correlations, item-total correlations and CronbachÕs Alpha if item deleted (Nunnally & Bernstein 1994, Pett et al. 2003). Validity and reliability Content validity of the instrument was evaluated in two ways, comparison with literature on person-centredness; and consultations with an expert group of four senior nurse researchers with experience of scale development who were asked to evaluate and comment on the relevance of items for clinical practice and research, clarity of item formulation, and readability of the items and the scaling. The experts were also asked to evaluate and comment on the content validity of the scale in relation to their experiences of person-centred climates from clinical work and research. The content validity of the scale was regarded as satisfactory by the experts as it was deemed to reflect the most central aspects of the phenomenon under scrutiny, and the scale was considered to contain items that reflected dimensions described in the literature as being central aspects of person centredness. Construct validity was estimated using principal component analysis with varimax orthogonal rotation, and goodness-of-fit through confirmative factor analysis. Internal consistency for total and subscale scores was estimated using CronbachÕs coefficient Alpha and the criterion level for acceptable reliability was set to Alpha 0.7 (Nunnally & Bernstein 1994). Test retest reliability was estimated using intraclass correlation analysis, and the ability of the scale to detect true differences between units was explored through examining the dispersion of mean scores. Results Six hundred health care staff provided data for the psychometric evaluation. As shown in Table 1, the majority of staff were female (86%): 12% were between the ages of 18 and 25 years, 74% were between 25 and 55 years, and 14% were over the age of 56. Most participants were registered nurses (44%) or enrolled nurses (38%), and the mean health care work experience was 16 years (with a standard deviation of 13 years). Item reduction and evaluation Thirty-one items were statistically suggested for elimination from the item reduction procedures, and these were also theoretically examined to make sure that the elimination made sense in light of theory which it did in all cases. Item evaluation of the final 14 items comprising the scale showed satisfactory inter-item agreements and uni-dimensionality as evidenced by squared multiple correlations ranging from r = 0.32 to r = 0.64, corrected item-total correlations ranging from r = 0.56 Table 1 Description of the sample (n = 600) Variable n (%) Age (years) (12) (26) (24) (24) (14) Gender Female 498 (86) Male 82 (14) Position Registered nurses 256 (44) Enrolled nurses 222 (38) Physicians 15 (3) Nursing students 14 (2) Personal care attendants 9 (3) Physiotherapists 6 (1) Occupational therapists 5 (1) Orderlies 54 (8) 792 ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd, Journal of Nursing Management, 17,

4 Evaluation of the PCQ-S to r = 0.64, and CronbachÕs Alpha if the item deleted ranging from 0.87 to Validity and reliability As the expert group estimated the content validity of the questionnaire as satisfactory, the 14 items were also evaluated in light of contemporary research publications of person-centred care as described in the literature, and these analyses confirmed the content validity of the instrument. Construct validity was supported from factor analyses of the final 14-item questionnaire. As shown in Table 2, these resulted in a three-factor solution explaining 60.0% of the total variance. The first factor consisted of five items (with factor loadings from 0.64 to 0.79) relating to experiences of being safe in the environment, and the factor was labelled Ôa climate of safetyõ. The second factor also consisted of five items ( ), and was labelled Ôa climate of everydaynessõ as item content related to the environment as having an everyday and neat character. The third factor contained four items ( ) and involved possibilities to keep previous and establish new social contacts in the environment, and was labelled Ôa climate of communityõ. The item order in the final scale is based on the factor structure, with items placed in descending order of importance. The three-factor model obtained by the explorative factor analyses was also Table 2 Scale content, factor loadings, total variance explained and CronbachÕs alpha for the final varimax rotated three-factor solution (n = 600); loadings >0.4 included Item number Item content Factor 1 a climate of safety Factor 2 a climate of everydayness Factor 3 a climate of community 1 A place where I feel 0.79 welcome 2 A place where I feel 0.78 acknowledged as a person 3 A place where I feel 0.77 I can be myself 4 A place where the patients 0.70 are in safe hands 5 A place where the staff use a language that the patients can understand A place which feels homely 0.78 even though it is in an institution 7 A place where there is 0.77 something nice to look at 8 A place where it is quiet and 0.68 peaceful 9 A place where it is possible 0.67 to get unpleasant thoughts out of your head 10 A place which is neat and clean A place where it is easy for 0.82 the patients to keep in contact with their loved ones 12 A place where it is easy for 0.80 the patients to receive visitors 13 A place where it is easy for 0.60 the patients to talk to the staff 14 A place where the patients have someone to talk to if they so wish 0.58 Total variance explained (%) CronbachÕs Alpha The probing question used is: ÔI experience my workplace asõ, and the response options used are: No, I disagree completely, No, I disagree, No, I partly disagree, Yes, I partly agree, Yes, I agree, Yes, I agree completely. ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd, Journal of Nursing Management, 17,

5 D. Edvardsson et al. evaluated using confirmative factor analysis, and goodness of fit was estimated using the Root Mean Square Error of Approximation (RMSEA), the Normed Fit Index (NFI) and the Comparative Fit Idex (CFI). The RMSEA was 0.08, the NFI was 0.88 and the CFI was Thus, the confirmatory factor analyses supported the exploratory findings, and the three-factor model was interpreted to offer a good fit for the questionnaire. As shown in Table 3, the PCQ-S was evaluated for how well it was able to detect systematic differences between groups. The total scores ranged from 3.99 to 5.26, and sub-scale scores ranged from 2.94 and Thus, it is useful for measuring group differences between perceptions of person centredness. The final 14-item questionnaire had an internal consistency coefficient (CronbachÕs alpha) of 0.88 in the sample, indicating a reliable scale that seems to measure the same underlying construct. Furthermore, as factor analyses of the scale resulted in a three-factor solution these three subscales were independently evaluated for internal consistency. They all showed satisfactory to acceptable internal consistency when evaluated separately: safety 0.84, everydayness 0.80 and community A sub-sample of 37 health care staff completed the questionnaire twice with 1 week in between. The intra-class correlation coefficient to evaluate the test retest reliability of the questionnaire showed an average measure intra-class correlation coefficient of 0.51 with a 95% confidence interval ranging from 0.47 to Thus, the questionnaire was estimated as being valid and reliable. From these analyses, the preliminary conclusion is that the Swedish PCQ-S is a valid and reliable selfreport instrument for evaluating to what extent the climate of health care settings are perceived as being person centred, and that it consists of three sub-scales Table 3 Distribution of scores in the Person-centred Climate Questionnaire staff version (PCQ-S) among participating units with n > 20 Unit n Total score Safety Everydayness Community (safety, everydayness and community). It is suggested that simple sum-scores are used for the scale and subscales respectively, and thus the total scores can range from 14 (a climate minimally person centred) to 84 (a climate maximally person centred). Discussion The items comprising the PCQ relate to the dimensions of person-centred care as described by other researchers (Kitwood 1997, Brooker 2004, Kjeldman et al. 2006, McCormack & McCance 2006, Slater 2006, OÕDonovan 2007, Penrod et al. 2007). In comparison with the frameworks of person-centred nursing as described by McCormack (2004) and McCormack and McCance (2006), the first factor comprising perceptions of safety relate to the competency and interpersonal skills of staff. Also, the second factor comprising perceptions of the environment as having an everyday character relate to supportive organizational systems, and the third factor embracing a climate of community relate to facilitating patient participation (McCormack & McCance 2006). The content of the questionnaire also reflect the fundamental constituents of person-centred care as described by Brooker (2007), namely perceptions of being valued as a member of staff, prioritizing needs of patients, using the perspective of the person and creating a positive social environment to promote well-being. By presenting a valid and reliable tool for the measurement of person-centred climates, this article contributes to the current knowledge base of personcentred care. The overall results from this study provided support for the psychometric properties of the scale when used in a Swedish sample of hospital staff from heterogeneous wards. This means that the extent to which staff perceives the climate of health care settings to be person centred can now begin to be explored and perhaps also linked to patient and staff outcomes. No such self-report instrument was previously identified in the literature. However, the unique focus of this instrument is on the climate of the setting, and as such it relates strongest to previous descriptions of the influence of place on patient experiences and person centredness (McCormack 2004). Even although further evaluation of psychometric properties of the scale when used in other samples is necessary, the scale is a tentative contribution to the current knowledge base by providing a means to further explore how the person centredness of care environments might influence well-being of staff. By making it possible to correlate staff ratings of the climate with ratings of work satisfaction, burnout and health for 794 ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd, Journal of Nursing Management, 17,

6 Evaluation of the PCQ-S example, staff well-being can be explored in relation to the environment of the setting. Furthermore, as a patient version PCQ exists there is a possibility to perform studies exploring to what extent staff and patient perceptions of the person-centredness of the climate correspond. There are limitations to this study that needs consideration. One limitation is the response rate of 57%. As the study employed anonymous return and implied consent, it was not possible to identify staff who did not participate to explore reasons for not participating. Thus, analyses of the characteristics of non-respondents are not possible to make as they remain unknown. Another limitation of the study is that analyses of criterion related validity in relation to an established scale were not possible to perform as no gold standard instrument was identified in the literature. Validity was evaluated in terms of construct and content validity instead (Nunnally & Bernstein 1994). Construct validity was deemed to be satisfactory as analyses showed a stable three-factor solution with items loading high and exclusively in each factor with satisfactory goodness-of-fit. Content validity was estimated satisfactory by a consulting expert group, and was also strengthened by the use of a theoretical foundation for item generation. However, further evaluation of validity and reliability of the questionnaire is needed and invited. For example, the replicability of factor structure and internal consistency of the staff PCQ needs to be tested in other samples as well as further examination of the sensitivity of the instrument as measured by other sociodemographic factors such as age, profession, education etc. Nevertheless, the sample size of the study is considered substantial enough to provide a trustworthy psychometric evaluation (Nunnally & Bernstein 1994). In conclusion, the psychometric evaluation provides tentative support that the Swedish language PCQ-S offers a valid and reliable tool for evaluation of to what extent the climate of health care settings is experienced as being person centred by health care staff. The questionnaire can be suggested for use in describing person centredness in different care environments, exploring correlates between staff perceptions of personcenteredness and measures of job strain or measures of well-being in patients, and/or in evaluating the effects of interventions. An English version staff scale is currently being evaluated and will be presented in forthcoming publications. Acknowledgement The study was supported by grants from Swedish Brain Power, Umeå University, La Trobe University, Swedish Council for Working Life and Social Research, and King Gustaf V and Queen VictoriaÕs Foundation. References Brooker D. (2004) What is person-centred care in dementia? Reviews in Clinical Gerontology 13, Brooker D. (2007) Person Centred Dementia Care: Making Services Better. Jessica Kingsley Publishers, London. Edvardsson D. (2008) Therapeutic environments for older people: a descriptive and interpretative study. Journal of Gerontological Nursing 34, Edvardsson D. & Street A. (2007) Sense or no-sense: the nurse as embodied ethnographer. International Journal of Nursing Practice 13, Edvardsson D., Sandman P.O. & Rasmussen B. (2005) Sensing an atmosphere of ease a tentative theory of supportive care settings. Scandinavian Journal of Caring Sciences 19, Edvardsson D., Sandman P.O. & Rasmussen B. (2008) Swedish language Person-centred climate questionnaire patient version: construction and psychometric evaluation. Journal of Advanced Nursing 63, Kitwood T. (1997) Dementia Reconsidered: The Person Comes First. Open University Press, Buckingham. Kjeldman D., Holmstrom I. & Rosenqvist U. (2006) How personcentred am I? A new method to measure physiciansõ personcentredness. Patient Education and Counseling 62, McCormack B. (2004) Person-centredness in gerontological nursing: an overview of the literature. International Journal of Older People Nursing 13, McCormack B. & McCance T.V. (2006) Development of a framework for person-centred nursing. Journal of Advanced Nursing 56, Nolan M.R., Davies S., Brown B., Keady J. & Nolan J. (2004) Beyond Ôperson-centredÕ care: a new vision for gerontological nursing. International Journal of Older People Nursing 13, Nunnally J.C. & Bernstein I. (1994) Psychometric Theory. McGraw-Hill, New York. OÕDonovan A. (2007) Patient-centred care in acute psychiatric admission units: reality or rhetoric? Journal of Psychiatric and Mental Health Nursing 14, Packer T. (2000) Does person-centred care exist? Journal of Dementia Care 8, Penrod J., Yu F., Kolanowski A., Fick D.M., Loeb S.J. & Hupcey J.E. (2007) Reframing person-centred nursing care for persons with dementia. Research and Theory for Nursing Practice 21, Pett M.A., Lackey N.R. & Sullivan J.J. (2003) Making Sense of Factor Analysis: The Use of Factor Analysis for Instrument Development in Health Care Research. Sage, London. Slater L. (2006) Person-centredness: a concept analysis. Contemporary Nurse 23, ª 2009 The Authors. Journal compilation ª 2009 Blackwell Publishing Ltd, Journal of Nursing Management, 17,