Spanish instructional video did not improve reported knowledge scores vs no video (3.5 vs 4, P=.908) in Spanish-speaking surgical patients

Description #

In this prospective, randomized, nonblinded pilot at Massachusetts General Hospital, adding a 4-minute Spanish-language anesthesia video before the preanesthesia interview produced no detectable difference in self-reported knowledge-improvement score between the video and no-video groups (3.5 vs 4; P = 0.908). Self-reported knowledge about general anesthesia rose across all surveys, but the between-group change was not statistically significant, and the authors note they did not specifically test for objective knowledge acquisition.

"There was no difference detected in reported knowledge improvement score between the two groups (3.5 vs 4; P = 0.908)." (West, 2014, p. 328)

"As we did not specifically test for knowledge acquisition, our study did not show a significant change in knowledge base. Interestingly, patients' self-reported knowledge about general anesthesia increased in all of our surveys, although the change was not statistically significant." (West, 2014, p. 328)

Methods Context #

What? #ⓘ

The observable: the self-reported knowledge-improvement score about general anesthesia procedures recorded on a 10-inch visual analog scale (VAS) anchored from "no knowledge" to "maximal knowledge."

"On confirmation of patient's participation, an assessment of anxiety, knowledge, and satisfaction was obtained using a visual analog scale (VAS). The VAS was based on a 10-inch (25.4 cm) line drawn on paper. The extreme left side indicated no anxiety, no knowledge, or dissatisfaction; the extreme right denoted maximal anxiety, knowledge, or satisfaction." (West, 2014, p. 326)

How? #ⓘ

Prospective, randomized, nonblinded pilot: patients were randomized by a computer-generated algorithm to view A-0003ArtifactA-0003Initial AI draftPreoperative Spanish-language anesthesia instructional videoTo reduce preoperative anxiety and improve knowledge and satisfaction for Spanish-speaking patients with limited English proficiency (LEP) by delivering standardized anesthesia education in the patient's native language,… (Group V) or not (Group NV) before the preanesthesia interview, and pre-to-post VAS changes were compared between arms with a two-sample t-test or Wilcoxon rank sum test as appropriate.

"Following a computer-generated algorithm, patients were randomized to Group V (video) or Group NV (no video)." (West, 2014, p. 327)

"Continuous variables (summarized by means with SD or medians with interquartile ranges) were compared using the two-sample t-test or Wilcoxon rank sum test for normally and nonnormally distributed variables, respectively." (West, 2014, p. 327)

Who? #ⓘ

20 adult, ASA physical status 1–3 Spanish-speaking patients requiring a Spanish interpreter, scheduled for elective gynecological, orthopedic, and intra-abdominal surgery under general anesthesia at Massachusetts General Hospital.

"Twenty adult, ASA physical status 1, 2, and 3 patients, scheduled for elective surgery (gynecological, orthopedic, and intrabdominal surgery) during general anesthesia were studied." (West, 2014, p. 325)

Other Notes #

The authors attribute the null in part to measuring self-reported knowledge improvement rather than objectively testing knowledge acquisition, and suggest a future study administer a short before/after knowledge questionnaire (West, 2014, pp. 328–329).

Caveats #

• Tiny (n=20), single-center, nonblinded pilot with self-report VAS outcomes The findings come from a very small (n = 20), single-center, prospective randomized but nonblinded pilot — an interim analysis of the first 20 patients — with outcomes captured only by self-report on a visual analog scale rather than objective measurement. The tiny, single-institution Spanish-speaking sample constrained extrapolation to LEP patients of other ethnic/cultural backgrounds, the study was not powered to assess modifiers such as prior anesthetic experience, PATA visit, gender, or age, and the knowledge outcome in particular reflected self-reported improvement rather than objectively tested knowledge acquisition. Together these features (small n, no blinding, single site, subjective self-report) make the effect estimates fragile and imprecise, and warrant treating the anxiety, satisfaction, and null-knowledge results as hypothesis-generating.