Use and scoring of the Iowa Satisfaction with Anesthesia Scale

To the Editor

I thoroughly enjoyed Said et al.’s paper (Said et al. 2016), but was concerned with the use of the Iowa Satisfaction with Anesthesia Scale (ISAS) (Dexter et al. 1997; Dexter and Candiotti 2011).

The ISAS is scored from − 3 to + 3 and the mean taken (Dexter et al. 1997; Dexter and Candiotti 2011). The authors report (page 269, column 1) using 1 to 6 (Said et al. 2016). The maximum mean of the 11 items would then be a score of 6. However, the authors’ Figure 3 on page 271 shows scores with a median of 9 in each of the two groups.

Quoting from the titles of both References Dexter et al. 1997 and Dexter and Candiotti 2011, the ISAS measures “patient satisfaction with monitored anesthesia care.” From https://www.franklindexter.net/FAQ/FAQ_a32.pdf (accessed May 17, 2016), “the instrument” (Dexter et al. 1997; Dexter and Candiotti 2011) “should not be used for … regional anesthesia”. “For measuring patient satisfaction with regional anesthesia, click here. For satisfaction with the perioperative period including regional anesthesia, click here.” The “click here” PubMed hyperlinks are for References Maurice-Szamburski et al. 2013 and Mui et al. 2011, respectively. The ISAS likely is invalid for measuring patient satisfaction for the authors’ (Said et al. 2016) epidural anesthesia group.

The authors’ text on page 270, column 1, refers to the mean differing (“the mean ISAS score was significantly higher (P = 0.018) in group [epidural] compared with group [local anesthesia]”) (Said et al. 2016). An unpaired t test for comparing means would be an appropriate analysis for the ISAS (Dexter and Candiotti 2011). Yet the authors’ methods (page 269, column 1) refer to “Wilcoxon ranked test,” which would not compare means.

The authors’ sample size (N = 31 and 32 in each group) was approximately half the N = 70 we calculated for each group to achieve an 80% statistical power to detect a moderate effect size (Dexter et al. 1997).

The ISAS is copyrighted by Franklin Dexter and the University of Iowa Research Foundation. This is stated in each table of the original paper and throughout the second paper (Dexter et al. 1997; Dexter and Candiotti 2011). Web search with Google or Bing (both performed May 17, 2016) using “Iowa Satisfaction with Anesthesia Scale” (with quotation marks) returned the authorization form for use as their first and second entries, respectively; Bing provides the original paper (Dexter et al. 1997) as the first entry. This is in part to assure the best possible use of the scale and help investigators with the analysis.