It is important to understand the clinical properties of instruments used to measure patient safety before they are used in the setting of an intensive care unit (ICU).
The Safety Climate Survey (SCSu), an instrument endorsed by the Institute for Healthcare Improvement, the Safety Culture Scale (SCSc), and the Safety Climate Mean (SCM), a subset of seven items from the SCSu, were administered in four Canadian university affiliated ICUs. All staff including nurses, allied healthcare professionals, non-clinical staff, intensivists, and managers were invited to participate in the cross sectional survey.
The response rate was 74% (313/426). The internal consistency of the SCSu and SCSc was 0.86 and 0.80, respectively, while the SCM performed poorly at 0.51. Because of poor internal consistency, no further analysis of the SCM was therefore performed. Test-retest reliability of the SCSu and SCSc was 0.92. Out of a maximum score of 5, the mean (SD) scores of the SCSu and SCSc were 3.4 (0.6) and 3.4 (0.7), respectively. No differences were noted between the three medical-surgical and one cardiovascular ICU. Managers perceived a significantly more positive safety climate than other staff, as measured by the SCSu and SCSc. These results need to be interpreted cautiously because of the small number of management participants.
Of the three instruments, the SCSu and SCSc appear to be measuring one construct and are sufficiently reliable. Future research should examine the properties of patient safety instruments in other ICUs, including responsiveness to change, to ensure that they are valid outcome measures for patient safety initiatives.
Cites: BMJ. 2000 Mar 18;320(7237):745-910720356
Cites: Crit Care Med. 2003 Mar;31(3):956-912627011
Cites: Qual Saf Health Care. 2003 Dec;12(6):405-1014645754
Cites: CMAJ. 2004 May 25;170(11):1678-8615159366
Cites: Stat Med. 1998 Jan 15;17(1):101-109463853
Cites: Biometrics. 1977 Mar;33(1):159-74843571
Comment In: Qual Saf Health Care. 2005 Aug;14(4):231-316076784
Comment In: Qual Saf Health Care. 2006 Jun;15(3):154-516751461