Accepted Manuscript Measuring patient respect in the intensive care unit: Validation of the ICU-RESPECT instrument
Gail Geller, Hildy Schell-Chaple, Kathleen Turner, Wendy G. Anderson, Mary Catherine Beach PII: DOI: Reference:
S0883-9441(18)30006-6 doi:10.1016/j.jcrc.2018.03.026 YJCRC 52895
To appear in: Please cite this article as: Gail Geller, Hildy Schell-Chaple, Kathleen Turner, Wendy G. Anderson, Mary Catherine Beach , Measuring patient respect in the intensive care unit: Validation of the ICU-RESPECT instrument. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Yjcrc(2017), doi:10.1016/j.jcrc.2018.03.026
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Measuring Patient Respect in the Intensive Care Unit: Validation of the ICU-RESPECT Instrument Authors:
Gail Geller, ScD, MHS1, 2, 3 Hildy Schell-Chaple, RN, PhD, CCNS, FAAN 4 (
[email protected]) Kathleen Turner, RN, CHPN, CCRN-CMC4 (
[email protected]) Wendy G. Anderson, MD, MS5
CR
IP
Mary Catherine Beach, MD, MPH1, 2, 3, 6 (
[email protected])
T
(
[email protected])
Affiliations:
Berman Institute of Bioethics, Johns Hopkins University, 1809 Ashland Ave, Baltimore, MD 21205, USA
2
Department of Medicine, Johns Hopkins School of Medicine, 733 N Broadway, Baltimore, MD 21205, USA
3
Bloomberg School of Public Health, Johns Hopkins University, 615 N Wolfe St, Baltimore, MD 21205, USA
4
University of California San Francisco Medical Center, Long Hospital L-976, 505 Parnassus Ave, San
M
AN
US
1
5
ED
Francisco, CA 94143
Division of Hospital Medicine and Palliative Care Program, University of California San Francisco, 533
Welch Center for Prevention, Epidemiology and Clinical Research, 2024 E. Monument Street, Suite 2-600,
CE
6
PT
Parnassus Avenue, San Francisco, CA 94143
Baltimore, MD 21205, USA
AC
Corresponding Author:
Gail Geller, Sc.D., M.H.S. Johns Hopkins University Berman Institute of Bioethics 1809 Ashland Ave, Room 202 Baltimore, MD 21205 (410) 614-5556 e-mail:
[email protected] fax: 410-614-5360
1
ACCEPTED MANUSCRIPT ABSTRACT (200 words including subheadings) Purpose: To validate a brief index of patient and family experiences of respect in the intensive care unit. Material and Methods: A survey including the 10- item ICU-RESPECT scale was administered to patients and family members in one ICU at a large west coast academic medical center. Confirmatory psychometric
T
analyses were conducted.
IP
Results: Based on 142 completed surveys, factor analysis confirmed a unidimensional scale with an alpha of
CR
0.90, an Eigen value of 4.9, and factor loadings from .50 to .86. The mean total score was 7.59 (SD=3.06) out of a maximum of 10. Among the 106 surveys that included demographics, overall scores did not differ
US
by type of respondent (patient or family) or by gender. There were modest differences in overall scores by patient race. Two individual items differed by type of respondent.
AN
Conclusions: The ICU-RESPECT index demonstrates reliability and concurrent validity in a different ICU
M
setting from the one where the index was developed. Future research should assess the scale’s predictive validity, and factors associated with variation in scores. As hospitals address patient experience more
ED
broadly in response to national metrics, the index could identify particular behaviors or ICUs that would
PT
benefit from interventions to enhance respectful treatment.
CE
Key Words: Ethics; Intensive Care Units; Psychometrics; Patient Experience; Human Dignity; Professional-
AC
Patient Relationship, Respect
2
ACCEPTED MANUSCRIPT INTRODUCTION (1645 WORDS) The importance of respect for persons as a foundational ethical principle in patient care is well described [1-4]. Recently, attention has focused on how treatment with respect, or lack thereof, manifests itself in the intensive care (ICU) setting, and how such treatment can be measured. [5-14]. Based on a
T
robust, multi-method approach to describing specific attitudes and behaviors that characterize respectful
IP
treatment, we developed a 10 item ICU-RESPECT index at one large, academic medical institution on the
CR
east coast [15]. The primary aim of the current study was to validate the ICU-RESPECT instrument at another academic medical center that has a different culture and patient population. A secondary aim was
US
to determine whether there were differences between patient and family responses. MATERIAL AND METHODS
AN
Data Collection
M
Data were collected between February and October 2016 from patients and family members in one 32 bed medical-surgical ICU in a large, west coast academic medical center. To ensure that we sampled
ED
patients and families who had had opportunities to interact with ICU clinicians, we excluded those who
PT
were in the ICU less than 3 days. Patients were eligible if they had been in the ICU for at least 3 days, and could provide informed consent and complete the survey. Eligible family members were defined as family
CE
or friends who were involved in the patient’s care and were at the patient’s bedside when the research
AC
assistant (RA) approached. RAs screened patients and family members and provided information about the survey to those who were eligible. Patients and families completed signed informed consent and then completed the survey immediately after enrollment. Surveys were completed on a tablet device using the Research Electronic Data Capture (REDCap) application (version 5.12.2). Participants completed the survey themselves, with RAs providing assistance only if they were unable to manually select responses on the tablet. Family members were only asked to complete the survey if the patient could not do so. This study was reviewed and approved by the Institutional Review Boards at the University of California, San Francisco and the Johns Hopkins University School of Medicine. The development of the 3
ACCEPTED MANUSCRIPT ICU-RESPECT index is described elsewhere (15). In brief, each item is ranked as 1 to 4 based on frequency of respectful treatment. Raw total scores ranged from 10 to 40 but, because of skewness, we recoded the responses as described below so the final scores ranged from 1 to 10. Statistical analysis and confirmatory psychometric testing
T
Data were entered into a REDCap database [16] and exported to Stata 13.1 [17] for analysis.
IP
Frequency distributions revealed that the responses to each of the 10 respect items were skewed, with a
CR
majority of respondents selecting “all of the time” for positively-worded items or “never” for negativelyworded items. Because of the skewness, we dichotomized all positively-worded items into “all of the time”
US
versus “most of the time”, “occasionally”, or “rarely/never”. Responses of “not applicable” were treated as missing values. The initial set of dichotomized items was subjected to factor analysis using the underlying
AN
variable approach where observed dichotomous variables are considered to be realizations of whether
M
underlying continuous variables pass a hypothetical threshold [18]. Following standard psychometric methods [19], a matrix of the correlations among the
ED
dichotomous items was produced, and factor analysis was conducted on the correlation matrix. Eigenvalues
PT
were calculated to confirm the unidimensional nature of the draft scale, and oblique (promax) rotation was used to evaluate the loading of the items on the factor. An overall Cronbach’s alpha was calculated, as well
CE
as item-to-total correlations, item-to-rest correlations, and alpha values if an item is removed.
AC
RESULTS
We collected surveys from patients themselves (n=58) and/or their family members (n=84) for a total of 142 surveys, representing the care of 106 discrete patients. We collected surveys from more than one family member if they were present and chose/consented to participate. Because demographic information on family members was unreliable (we did not know whether they were answering the demographic questions about themselves or the patient), we only present demographic data on the 106 patients in Table 1.
4
ACCEPTED MANUSCRIPT Table 2 presents the distributions of the most positive responses vs. less positive responses for each of the 10 respect and dignity items. “Care team made an effort to understand what matters” and “Care team made an effort to know the patient as a unique individual” were the only items with a significant minority of respondents endorsing the less positive response (32% and 42%, respectively). These
T
two items also reflected a statistically significant difference between patients and families, with families
IP
reporting more respectful treatment than patients. For all other items, responses were comparable among
CR
patients and families with the most positive response endorsed by the majority of respondents (73-86%). Psychometric analyses on all 142 surveys confirmed that the ICU-RESPECT index is unidimensional,
US
i.e., it can be measured based on a single score. Item-item correlations ranged from .45 to .93. In factor analysis using all 10 items, the eigenvalues decreased markedly after the first factor, from 4.9 with the first
AN
factor to .39 with the second factor. The factor loadings ranged from .54 (“care team introduced
M
themselves when first met patient”) to .86 (“felt that care team was attentive to my requests”). The overall alpha was .90.
ED
The mean score of the final index was 7.59 with a standard deviation of 3.06. Although scores
PT
ranged from 1 to 10, the distribution was skewed, with 49% of observations having total scores of 9 or 10. As shown in Table 3, there was no significant variation in scores between the type of respondent. Overall
CE
scores did not differ by gender but there were modest differences by race. Respondents who selected
DISCUSSION
AC
“Other” as their race scored lower than whites, blacks, Asians and Hispanics.
This study confirms the psychometric properties of the 10-item self-reported “ICU-RESPECT” index [15], a scale that was designed to assess patient perceptions of respect in the ICU setting but tested in only one east coast academic medical center. The current study was performed in a west coast academic medical center that has a different culture and patient population. The concurrent validity of any scale is demonstrated by its ability to distinguish between groups. In the original study, ICU-RESPECT scores
5
ACCEPTED MANUSCRIPT differed between different types of ICUs [15]. In the current study, we were not able to examine differences between ICUs because we only collected data from one ICU. However, we did identify modest differences among racial/ethnic groups. In the current study patients who designated “other” as their racial/ethnic group reported lower scores than white, blacks,
T
Asians and Hispanics. Since we do not know what the “other” category is comprised of, and there were so
IP
few in that category, we cannot interpret this finding. However, barriers to effective communication in
CR
ICUs have been well-described, in part based on racial and socioeconomic differences [20]. Future research to verify and understand patterns of racial and ethnic variation in experiences of respect among ICU
US
patients will be facilitated by the ability of the ICU-RESPECT index to identify such differences. In addition to detecting differences in overall ICU-RESPECT scores, we observed some important
AN
patterns in individual items. The good news is that, for each item, the majority of respondents endorsed
M
the most positive response. However, there is still room for improvement. If our goal is for 100% of patients and families to experience these respectful behaviors “all of the time”, our findings indicate a
ED
shortfall in every category, particularly the “Care team made an effort to understand what matters” and the
PT
“Care team made an effort to know the patient as a unique individual”. Moreover, that fewer patients than families report experiencing these behaviors “all of the time” deserves further scrutiny.
CE
Our study has several limitations. First, our sample size is small, although the ratio of respondents
AC
to items in the scale is reasonable. Second, the response rate was not as high as we had hoped. We do not know if respondents differed from non-respondents in meaningful ways. We were unable to conduct a non-response analysis because we did not collect any data on non-respondents. There likely is selection bias operating in this sample, highlighting the views of ICU patients who are well enough, able, and willing to complete the questionnaire. Although the small size and absence of information on non-respondents threatens the external validity of our study, our experience with a moderate participation rate may not be unique to our study and may explain why very few studies try to recruit ICU patients and families while they are acutely ill. A related limitation is that survey responses may be confounded by patient health status. 6
ACCEPTED MANUSCRIPT ICU patients have recently survived a critical medical event which may eclipse their concerns about other aspects of care, particularly whether or not they have been treated with respect. Finally, the validation of the ICU-RESPECT scale was limited because this study was conducted at only one adult ICU that is part of a single academic health system. Findings may vary across ICUs in other geographic locations. Despite these
T
limitations, our findings demonstrate that the “ICU-RESPECT” index is reliable in two different ICU settings
IP
and is useful in describing and quantifying the nature of respectful treatment in intensive care units.
CR
CONCLUSION
Our study contributes to a small literature on measurement of the extent to which patients and
US
families are treated with respect in the critical care setting. The valid and reliable “ICU-RESPECT” index could enable the development of a richer understanding of how ICU patients’ and their families’
AN
experiences of respect are affected by clinician behaviors, by patient characteristics, by interpersonal
M
dynamics among clinicians, patients, and their families, and by characteristics of the care unit. The use of “ICU-RESPECT” can inform the development and evaluation of educational interventions designed to
ED
counteract particular types of disrespectful treatment of patients and their families in the ICU. A careful
PT
review of respondents’ reports of respectful treatment might help target interventions toward particular
CE
ICUs or clinician behaviors to improve care. Future research should validate this index among a large and diverse population of ICU patients,
AC
perhaps oversampling different minority groups, and assess its ability to predict or measure specific behaviors associated with respect. As hospitals address patient experience more broadly in responding to national metrics, this short and validated ICU-RESPECT index could be useful in ICU quality improvement efforts.
7
ACCEPTED MANUSCRIPT Acknowledgements This research was part of the Emerge Collaboration funded by the Gordon and Betty Moore Foundation. The authors wish to thank the patients and families who participated in this study. This work would not have been possible without the support of the ICU staff and research assistants, Megan Rathfon, Jenica
IP
T
Cimino, Trese Biagini, and Cindy Beavon.
CR
Study data were collected and managed using REDCap electronic data capture tools hosted at Johns Hopkins University (Harris et al., 2009). REDCap (Research Electronic Data Capture) is a secure, web-based
US
application designed to support data capture for research studies, providing 1) an intuitive interface for validated data entry; 2) audit trails for tracking data manipulation and export procedures; 3) automated
AN
export procedures for seamless data downloads to common statistical packages; and 4) procedures for
AC
CE
PT
ED
M
importing data from external sources.
8
ACCEPTED MANUSCRIPT REFERENCES
1. Beauchamp TL, Childress JF. Principles of Biomedical Ethics. 5th ed. New York: Oxford University Press; 2001.
IP
health professionals to respect patients. J Gen Intern Med 2007; 22:692-695.
T
2. Beach MC, Duggan PS, Cassel CK, Geller G. What does 'respect' mean? Exploring the moral obligation of
CR
3. Beach MC, Roter DL, Wang NY, Duggan PS, Cooper LA. Are physicians' attitudes of respect accurately perceived by patients and associated with more positive communication behaviors? Patient Educ Couns
US
2006; 62:347-354.
4. Dickert NW, Kass NE. Understanding respect: Learning from patients. J Med Ethics 2009;35: 419-423.
M
harm. BMJ quality & safety 2015;24: 550-553.
AN
5. Sokol-Hessner L, Folcarelli PH, Sands KE. Emotional harm from disrespect: the neglected preventable
6. Henry LM, Rushton C, Beach MC, Faden R. Respect and dignity: a conceptual model for patients in the
ED
intensive care unit. Narrat Inq Bioeth 2015;5(1A):5A-14A. doi: 10.1353/nib.2015.0007.
PT
7. Gazarian PK, Morrison CR, Lehmann LS, Tamir O, Bates DW, Rozenblum R. Patients' and care partners' perspectives on dignity and respect during acute care hospitalization [published online February 22 2017]. J
CE
Patient Saf. 2017. doi: 10.1097/PTS.0000000000000353.
AC
8. Chochinov HM, Hassard T, McClement S, Hack T, Kristjanson LJ, Harlos M, et al. The patient dignity inventory: a novel way of measuring dignity-related distress in palliative care. J Pain Symptom Manage. 2008;36:559-571.
9. Aboumatar HA, Forbes L, Branyon E, Carrese J, Geller G, Beach MC, et al. Understanding treatment with respect and dignity in the intensive care unit. Narrative Inquiry in Bioethics 2015; 5(1A): 55–67. 10. Aboumatar HA, Beach MC, Yang T, Branyon E, Forbes L, Sugarman J. Measuring patients’ experiences of respect and dignity in the intensive care unit: A pilot study. Narrative Inquiry in Bioethics 2015; 5(1A): 6984. 9
ACCEPTED MANUSCRIPT 11. Beach MC, Forbes L, Branyon E, Aboumatar H, Carrese J, Sugarman J, et al. Patient and family perspectives on respect and dignity in the intensive care unit setting. Narrative Inquiry in Bioethics 2015; 5(1A): 15–25. 12. Carrese J, Forbes L, Branyon E, Aboumatar H, Geller G, Beach MC, et al. Observations of Respect and
T
Dignity in the Intensive Care Unit. Narrative Inquiry in Bioethics 2015; 5(1A): 45-53.
IP
13. Geller G, Branyon E, Forbes L, Rushton CH, Beach MC, Carrese J, et al. Health Care Professionals’
CR
Perceptions and Experiences of Respect and Dignity in the Intensive Care Unit. Narrative Inquiry in Bioethics 2015; 5(1A): 27-42.
US
14. Carrese JA, Geller G, Branyon ED, Forbes LK, Topazian RJ, Weir BW, Khatib O, Sugarman J. A Direct Observation Checklist to Measure Respect and Dignity in the Intensive Care Unit. Crit Care Med
AN
2017;45:263-270. doi: 10.1097/CCM.0000000000002072.
M
15. Geller G, Branyon ED, Forbes LK, Topazian RJ, Weir BW, Carrese JA, Beach MC, Sugarman J. ICURESPECT: An Index to Assess Patient and Family Experiences of Respect in the Intensive Care Unit. J Crit
ED
Care 2016;36:54-59. doi: 10.1016/j.jcrc.2016.06.018.
PT
16. REDCap (2014). Computer Software (Version 5.12.2). Vanderbilt University. SAS Institute Inc. 2012. SAS 9.4. Cary, NC, USA.
CE
17. StataCorp. 2013. Stata Statistical Software: Release 13. College Station, TX: StataCorp LP.
press, 2008.
AC
18. Bartholomew, DJ, Steele F, Galbraith J, Moustaki I. Analysis of multivariate social science data. CRC
19. Nunnally JC, Bernstein, IH. Psychometric theory. 3rd ed. New York: McGraw–Hill, 1994. 20. Muni S, Engelberg RA, Treece PD, Dotolo D, Curtis JR. The influence of race/ethnicity and socioeconomic status on end-of-life care in the ICU. Chest 2011;139:1025-1033.
10
ACCEPTED MANUSCRIPT
(%)
Age 19 to 30 31 to 50 51 to 60 61 to 70 71 to 90
7 21 28 31 17
(7) (20) (27) (30) (16)
Gender Male Female
59 47
60 11 13 17 5
(56) (44)
(57) (10) (12) (16) (5)
AC
CE
PT
ED
M
AN
Race White Black/African American Asian Latino/Hispanic Other
US
CR
IP
n
T
Table 1. Patient Characteristics (n = 106)
11
ACCEPTED MANUSCRIPT Table 2: Item frequencies for patient and family responses to ICU-RESPECT index
Patients N=58 16 (28%) 42 (72%)
19 (23%) 65 (77%)
35 (25%) 107 (75%)
11 (19%) 47 (81%)
17 (20%) 67 (80%)
28 (20%) 114 (80%)
23 (28%) 60 (72%)
45 (32%) 96 (68%)
17 (20%) 67 (80%)
32 (22%) 110 (78%)
18 (31%) 40 (69%)
21 (25%) 63 (75%)
39 (27%) 103 (73%)
28 (50%) 28 (50%)
31 (37%) 53 (63%)
59 (42%) 81 (58%)
8 (14%) 50 (86%)
19 (23%) 65 (77%)
27 (19%) 115 (81%)
12 (21%) 46 (79%)
15 (18%) 68 (82%)
27 (19%) 114 (81%)
8 (14%) 49 (86%)
12 (14%) 72 (86%)
20 (14%) 121 (86%)
12 (21%) 46 (79%)
19 (23%) 65 (77%)
31 (22%) 111 (78%)
IP
22 (38%) 36 (62%) 15 (26%) 43 (74%)
US
AN
M
ED
PT
CE
Combined N=142
CR
Care team introduced themselves when first met patient Never to most of the time All of the time Treated with courtesy by care team Never to most of the time All of the time Care team made effort to understand what matters* Never to most of the time All of the time Care team attentive to requests Never to most of the time All of the time Felt that care team really listened Never to most of the time All of the time Care team made effort to know patient as unique individual* Never to most of the time All of the time Care team kept body covered Never to most of the time All of the time Care team treated patient as equal Never to most of the time All of the time Care team managed pain Never to most of the time All of the time Care team treated patient as they would like to be treated Never to most of the time All of the time
Family N=84
T
Item
AC
ICU-Respect responses dichotomized If we aspire to 100% “all of the time” responses, we fall short on every item. *The 2 items that deserve the most immediate attention both overall and by subgroup.
12
ACCEPTED MANUSCRIPT Table 3. Relationship of ICU-RESPECT scores with respondent characteristics (N=106)
Mean
SD
Patients (N=58)
7.4
2.99
Family/friends (N=48)
7.5
3.17
Gender: Female (N=47)
7.9
31-50 (N=21) 51-60 (N=28)
M
61-70 (N=31)
ED
71-95 (N=17) Race
CE
PT
White (N=60)
Hispanic (N=17)
.197
7.1 9.0
1.41
.140
3.14
7.0
3.33
7.8
2.83
7.3
2.89
7.0
3.87
7.7
2.74
8.0
3.29
7.2
3.29
7.7
3.26
3.6
3.58
.065
AC
Other (N=5)
2.94
AN
18-30 (N=7)
Asian (N=13)
IP
Age
Black (N=11)
.809
US
Male (N=59)
p
CR
Type of Respondent:
T
ICU-RESPECT Scores
13
ACCEPTED MANUSCRIPT Highlights - Journal of Critical Care ICU-RESPECT Manuscript
The psychometric properties of the 10-item self-reported “ICU-RESPECT” index were confirmed in a different ICU setting from the one where the index was developed.
The ICU-RESPECT index is unidimensional with an alpha of 0.90, an Eigen value of 4.9, factor loadings from .50 to .86, and a mean total score of 7.59 out of 10.
The “ICU-RESPECT” index can be used to inform the development and evaluation of educational interventions designed to counteract particular types of disrespectful treatment of patients and their families in the ICU.
AC
CE
PT
ED
M
AN
US
CR
IP
T
14