Trial Outcomes & Findings for The Provider Awareness and Cultural Dexterity Toolkit for Surgeons Trial (NCT NCT03576495)

Recruitment of general surgery residents and surgical patients at Johns Hopkins University, Brigham and Women's Hospital, Brown University, and Eastern Virginia Medical School, Massachusetts General Hospital, Beth Israel Deaconess Medical Center, Howard University, and Washington University in St. Louis. We randomized at the site level to two groups: Early Intervention and Delayed Intervention. We reported our aggregate results for the residents and surgical patients in each group.

Inclusion criteria (Residents): All general surgery residents at all participating sites Exclusion criteria (Residents): All non-surgical residents at all participating sites Inclusion criteria (Patients): Admitted to a surgical service under the care of a participating resident with the capacity to give informed consent Exclusion criteria (Patients): Mentally impaired, admitted to intensive care

Unit of analysis: Residency Sites

The number of participants analyzed differs from the number enrolled due to loss to follow up.

The effect of PACTS curriculum on surgical residents' questionnaire scores measuring knowledge about caring for culturally diverse patients at time Period 2 (18 months). At this time, the Early Intervention Group had already received the PACTS curriculum. The Delayed Intervention Group had not received the PACTS curriculum, serving as the control group at this time period. Resident knowledge: Percent score out of 100, with range 0-100%. Higher values represent a better outcome, with 100% as the highest score possible. This is the average score for the Early Intervention and Delayed Intervention groups at Period 2.

The effect of PACTS curriculum on surgical residents' cross cultural care survey about caring for culturally diverse patients at time Period 2 (18 months). At this time period, the Early Intervention group received the PACTS curriculum, while the Delayed Intervention Group had not received the PACTS curriculum, serving as the control group. The cross cultural care survey was evaluated using a modified Likert type scale (with scale ranging from lowest to highest: very unprepared, somewhat unprepared, somewhat prepared, very well prepared). Resident scores were dichotomized into two groups, those reporting "very unprepared" and "somewhat unprepared", and those reporting "somewhat prepared" and "very well prepared." Here reported values are representative of the percentage of participants who reported "somewhat prepared" and "very well prepared" at time Period 2.

The effect of PACTS curriculum on surgical residents' self-assessed skills for caring for culturally diverse patients at time Period 2 (18 months). Self-assessed skills ranged from levels 1 to levels 4, with level 1 indicated less skilled, and level 4 indicating skillful. For purposes of comparison, resident scores were dichotomized into two groups: less skilled (referring to skill levels 1 and 2), and skillful (levels 3 and 4). Here reported values are representative of the percentage of participants who reported skill levels 3 or 4, indicating skillful. Results here demonstrate the proportion of residents in the Early Intervention (Intervention) group and Delayed Intervention (control group) who self-evaluated their skills as skillful at Period 2.

The effect of PACTS curriculum on surgical residents' beliefs regarding caring for culturally diverse patients at time Period 2 (18 months). At this time, the Early Intervention Group had already received the PACTS curriculum. The Delayed Intervention Group had not received the PACTS curriculum, serving as the control group at this time period. The questionnaire was scored using a modified Likert type scale with a range from lowest to highest: strongly disagree, moderately disagree, mildly disagree, strongly agree, moderately agree, mildly agree. Answers were dichotomized into two groups: strongly disagree, moderately disagree, and mildly disagree; and strongly agree, moderately agree, and mildly agree. Here reported values are representative of the proportion of participants who answered "strongly agree, moderately agree, and mildly agree."

Standardized Patient observers evaluated surgical residents on multiple dimensions of cultural dexterity and communication skills using Likert-type scales. The scale range, from lowest to highest was: "Not at all; a little bit; somewhat; mostly; a great deal." Scores were put into two groups: 1) not at all, a little bit, and somewhat; 2) mostly and a great deal. The percentage of residents who received scores of "mostly" and "a great deal" in categories of trust, limited english proficiency, consent, and pain were reported.

We used an adapted version of the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) to assess patients' (1) satisfaction with pain management, (2) satisfaction with communication including specific measures for limited English proficiency (LEP), (3) trust, and (4) comprehension of informed consent. Patient satisfaction was captured using a modified Likert scale from lowest to highest: strongly disagree, disagree, neither agree nor disagree, agree, and strongly agree. The values reported below are the percentage of participants who reported "agree" or "strongly agree."

National Surgical Quality Improvement Program (NSQIP) metrics for each patient participant capturing hospital length of stay in days. Patients designated to Early Intervention Group were cared for by a resident enrolled in the Early Intervention group, where the PACTS curriculum was administered between period 1 (0 months) and period 2 (18 months). Patients designated to the Delayed Intervention group were cared for by a resident enrolled in the Delayed Intervention group, where the standard curriculum was administered between period 1 (0 months) and period 2 (18 months). We are comparing median length of stay at period 2.