Development and Validation of CSAT Instrument

Data from two different participant samples were used for the development and testing of the CSAT instrument. First, a pilot sample (N = 120) was used for the initial development of the instrument, which included item selection, preliminary psychometrics testing, and usability testing. Second, a subsequent set of early CSAT users administered the final 35-item version of the CSAT. These data were used for more in-depth psychometric testing, including subdomain analyses and structural invariance testing (see below).
The first CSAT pilot sample of participants was selected from different clinical work environments as well as from different healthcare professions. Recruitment efforts used a snowball sampling approach and included identifying and contacting stakeholders that could potentially benefit from using the tool, as well as promoting the CSAT at local and national dissemination and implementation conferences. Respondents had the option to forward the link to peers or nominate individuals to complete the CSAT. To incentivize participation, all respondents were offered an optional tailored sustainability results report and the opportunity to enter a drawing for one of five $50 gift cards. The final pilot sample size was 120 participants (Table 1).
Table 1

CSAT development participant demographic characteristics

	Pilot (N=120)		Early users (N=166)
Characteristic	N	%	N	%
Profession
Nurse	18	16	42	27
Pharmacist	37	33	11	7
Physician	29	26	34	22
Admin/research	13	12	29	19
Ancillary	7	6	18	11
Other	8	7	22	14
Position/role
Bedside provider	44	40	65	42
Unit level management	7	6	–	–
System leadership	6	5	60	38
Program leader	27	24	4	3
Other	27	24	27	17
Environment
Academic medical center	67	60	65	42
Private practice	6	5	16	10
Community hospital	21	19	19	12
Community health center	6	5	48	31
Other	12	11	8	5
Setting
Inpatient	56	55	35	25
Outpatient	26	26	79	56
Both	19	19	27	19
Patient
Pediatric	54	53	49	32
Adult	47	47	94	63
Both	–	–	6	4

Note: Frequencies add up to less than sample totals because of missing responses

The second set of early user program participants (N = 166) came from two separate research studies. The first study recruited clinical staff working on the cancer control continuum in Missouri. This included primary care environments, screening programs, and cancer care centers that are focused on the diagnosis and treatment of those with cancer. The second study recruited participants in antimicrobial stewardship teams working to implement surgical prescribing guidelines. The contact at each site forwarded the CSAT to stakeholders they identified to participate. The early user sample size was 166 participants who represented a mix of professions and roles. The early users differed from the pilot sample with more early users representing adult care and outpatient settings (Table 1).