Anaesthesia, 2009, 64, pages 10771080 doi:10.1111/j.1365-2044.2009.06036.x .....................................................................................................................................................................................................................

Pre-operative forced-air warming as a method of anxiolysis

R. J. Wen,1 K. Leslie2,3 and P. Rajendra4
1 Medical Student, 2 Head of Research and 4 Provisional Fellow, Department of Anaesthesia and Pain Management, Royal Melbourne Hospital, Melbourne, Australia 3 Associate Professor, Department of Pharmacology, University of Melbourne, Melbourne, Australia Summary

We tested the hypothesis that pre-operative forced-air warming is as effective for anxiolysis as intravenous midazolam, using a blinded, placebo controlled factorial design. One hundred and twenty patients were randomly assigned to cotton blanket and saline injection (n = 30), forced-air warmer and saline injection (n = 30), midazolam 30 lg.kg)1 and cotton blanket (n = 30), and forced-air warmer and midazolam 30 lg.kg)1 (n = 30). Patients completed visual analogue scales for anxiety and thermal comfort, and the State-Trait Anxiety Inventory, at baseline and after 20 min. The estimated effect of midazolam on visual analogue scores for anxiety was )10 (95% CI )3 to )18; p = 0.007) and on state anxiety was )5 (95% CI )7 to )4; p = 0.03). Warming had no inuence on visual analogue scores for anxiety (p = 0.50) or state anxiety (p = 0.33), but its estimated effect on thermal comfort was +23 (95% CI 1927; p < 0.0001). There was no interaction between midazolam and warming. Pre-operative warming was not equivalent to midazolam for anxiolysis and cannot be recommended solely for this purpose.
. ......................................................................................................

Correspondence to: Associate Professor Kate Leslie E-mail: kate.leslie@mh.org.au Accepted: 11 June 2009

Anxiety is a normal reaction to a major life event such as surgery but is a state that many patients wish to avoid or have treated [1]. Midazolam is commonly administered for pre-operative anxiety, but may cause excessive sedation and arterial desaturation [2]. An effective, nonpharmacological anxiolytic would be a useful alternative. Pre-operative warming is a possible method of anxiolysis that is quick to implement, minimally disruptive and has additional peri-operative benets [3, 4]. However, two recent studies have reported conicting results [3, 4]. Wagner et al. [3] demonstrated that patients prewarmed with a warming suit were less anxious than those who received routine care. However, the patients in that study were volunteers who responded to a yer in the pre-admission clinic and this may have introduced a selection bias. Kimberger et al. [4] reported no difference in anxiety between forced-air warmer treatment and routine care. These patients were presenting for neurosurgery and had an indwelling urinary catheter inserted before the study commenced. This would almost certainly have increased anxiety. In both studies, patients were not blind to the assigned treatment and an objective measure of hypnosis was not used.
2009 The Authors Journal compilation 2009 The Association of Anaesthetists of Great Britain and Ireland

We therefore conducted a blinded, placebo controlled randomised trial in an elective general surgical population, using bispectral index (BIS) monitoring as an objective measure of hypnosis [5, 6]. We tested the hypothesis that pre-operative forced-air warming is as effective for anxiolysis as intravenous midazolam 30 lg.kg)1.
Methods

Prospective approval was obtained from the Melbourne Health Human Research and Ethics Committee. Eligible patients were scheduled for elective surgery under general anaesthesia, were English-speaking, 1870 years old and of ASA physical status 13. Patients were excluded if they were scheduled for cardiothoracic or intracranial surgery, had a pre-existing diagnosis of anxiety, were taking psychotropic drugs, illicit drugs or excessive alcohol, or had been previously diagnosed with obstructive sleep apnoea (where sedatives are a relative contraindication). Written informed consent and baseline data (including visual analogue scale (VAS) scores for anxiety and thermal comfort, and the State-Trait Anxiety Inventory (STAI) were obtained before patients were brought to the
1077

R. J. Wen et al. Pre-operative warming and anxiolysis Anaesthesia, 2009, 64, pages 10771080 . ....................................................................................................................................................................................................................

operating suite. Patients arrived wearing a cotton gown and covered by a cotton blanket. In the anaesthetic room, oxygen was applied (6 l.min)1 via a clear plastic mask), pulse oximetry and blood pressure monitoring were applied and BIS monitoring (Aspect Medical Systems Inc., Newton, MA, USA) was commenced. Curtains were drawn around the patients to minimise interruptions and the lights were dimmed. Computer generated randomisation results (block randomised for sex) were concealed in opaque envelopes until preparations were complete. Patients were randomised to four groups as follows: 1 Control: single cotton blanket; saline injection. 2 Warming: full-body Bair HuggerTM blanket (Arizant Healthcare Inc., Eden Prairie, MN, USA) and Bair Hugger forced-air warmer (Model 505; Augustine Medical Inc., Eden Prairie, MN, USA) set to high (43C); saline injection. 3 Midazolam: single cotton blanket; intravenous midazolam 30 lg.kg)1. 4 Combined: forced-air warming as above; intravenous midazolam 30 lg.kg)1. The patients anaesthetist administered the treatments. Saline was given in an equivalent volume to that of midazolam 30 lg.kg)1. Injections were made into a port out of sight of the patient. The observer was blinded to the intravenous solution given, but could not be blinded to forced-air warming or cotton blanket. With the approval of the Ethics Committee, patients were not informed of the specic interventions being tested, only that the study involved looking at the efcacy of various treatments for anxiety. Since attention was not drawn to specic interventions as they were applied, we deemed the patients to be blinded. We did not measure ambient, core or skin temperature. The study was 20-min in duration, with monitored variables recorded at 0, 5, 10, 15, and 20 min. At the end of the study period, VAS scores for anxiety and thermal comfort were recorded and the state portion of the STAI administered. The anaesthetic then proceeded at the anaesthetists discretion. The anxiety VAS was scored on a 100-mm line, with one end marked not at all anxious (0 mm) and the other end marked extremely anxious (100 mm). The VAS for anxiety has been validated [7]. The thermal comfort VAS was scored on a 100-mm line, marked with worst imaginable cold (0 mm), neutral (50 mm), and worst imaginable heat (100 mm). The STAI consists of two 20-statement questionnaires, measuring a patients trait anxiety (anxiety proneness) and state anxiety (situational anxiety) [8]. Each statement is scored from 1 to 4, giving a range of scores from 20 to 80 for each part of the questionnaire.
1078

This was a trial of equivalence. Taylor et al. [9] reported that midazolam reduces the mean (SD) VAS for anxiety from 60 mm (20 mm) to 20 (20 mm). A difference in VAS of 20 mm was thought to be clinically meaningful. We therefore considered a reduction in anxiety score from 60 mm to 40 mm in the forced-air warming group to be the outer boundary of acceptable equivalence (d = 20 mm). Based on a = 0.05 and b = 0.1, a sample size of 17 patients per group was estimated. One hundred and twenty patients were studied as the effect of treatment allocation on the secondary endpoints was also of interest. Continuous data were tested for normality. Categorical data were compared using the chi-squared or Fishers exact test. One-way ANOVA was used to determine inter-group differences in VAS anxiety, VAS thermal comfort and STAI-state. Factorial analysis was undertaken using two-way analysis of variance, including an interaction term for midazolam and forced-air warming. A generalised linear model was used to determine predictors of a decrease in VAS anxiety. Repeated measures ANOVA was used to determine changes in BIS over time and differences in BIS between groups. Stata 10.1 (Stata Corporation, USA) was used for all analyses; p < 0.05 was considered statistically signicant.
Results

Twenty-two patients refused to participate, three patients were excluded before randomisation because of previously unidentied antidepressant use and ve patients were replaced after randomisation due to a change in the anaesthetic plan. One hundred and twenty patients completed the study. No patient required rescue treatment for anxiety or became unresponsive to command. Baseline characteristics were similar in the two groups (Table 1). Women were more anxious than men at baseline: median (IQR [range]) VAS for anxiety was 33 (1755 [096]) for women vs 26 (1043 [095]) for men (p = 0.03) and STAI-state was 40 (3144 [2072]) for women vs 34 (2443 [2064]) for men (p = 0.03). However, trait anxiety was similar: 33 (2739 [2261]) for women vs 32 (2739 [2067]) for men (p = 0.87). Signicant differences were observed between the randomised groups in terms of post-treatment anxiety and thermal comfort (Table 2). Midazolam reduced anxiety, with an estimated effect on VAS for anxiety of )10 (95% CI )3 to )18; p = 0.007) and on STAI-state of )5 (95% CI )7 to )4; p = 0.03). Midazolam had no effect on thermal comfort (p = 0.88). Forced-air warming had no inuence on VAS for anxiety (p = 0.50) or STAI-state (p = 0.33), but its estimated effect on thermal comfort was +23 (95% CI +19 to +27; p < 0.0001).
2009 The Authors Journal compilation 2009 The Association of Anaesthetists of Great Britain and Ireland

Anaesthesia, 2009, 64, pages 10771080 R. J. Wen et al. Pre-operative warming and anxiolysis . ....................................................................................................................................................................................................................

Table 1 Baseline characteristics of patients. Control: cotton blanket and placebo injection; Warming: full body forced-air warming and placebo injection; Midazolam: cotton blanket and 30 lg.kg)1 midazolam; Combined: full body forced-air warming and 30 lg.kg)1 midazolam. Data are mean (SD) or number (%).

Characteristic Age; years Male:female Weight; kg ASA physical status 1 2 3 Type of surgery Orthopaedic General Other Prior experience of anaesthesia

Control (n = 30) 47 (15) 15:15 76 (20) 13 (43%) 16 (53%) 1 (3%) 4 (13%) 13 (43%) 13 (43%) 27 (90%)

Warming (n = 30) 45 (15) 15:15 76 (19) 17 (57%) 9 (30%) 4 (13%) 5 (17%) 13 (43%) 12 (40%) 27 (90%)

Midazolam (n = 30) 38 (13) 15:15 74 (16) 15 (50%) 14 (47%) 1 (3%) 7 (23%) 12 (40%) 11 (37%) 24 (80%)

Combined (n = 30) 39 (13) 15:15 74 (16) 20 (67%) 10 (33%) 0 (0%) 6 (20%) 10 (33%) 14 (47%) 24 (80%)

Table 2 Anxiety and thermal comfort scores. Control: cotton blanket and placebo injection; Warming: full body forced-air warming and placebo injection; Midazolam: cotton blanket and 30 lg.kg)1 midazolam; Combined: full body forced-air warming and 30 lg.kg)1 midazolam. Data are median (IQR [range]).
Characteristic Baseline VAS anxiety DVAS anxiety Baseline VAS thermal comfort DVAS thermal comfort Baseline STAI-trait Baseline STAI-state DSTAI-state Control 27 (849 [082]) )3 ()13 to +5 [)48 to +30]) 50 (4850 [2975]) 0 ()8 to +1 [)25 to +15]) 32 (2638 [2161]) 36 (2544 [2072]) )1 ()6 to +1 [)33 to +5]) Warming 40 (2351 [096]) )5 ()15 to +12 [)90 to +59]) 50 (5050 [960]) 12 (+4 to +26 [)11 to +46]) 36 (3343 [2050]) 37 (2750 [2061]) )1 ()9 to +2 [)34 to +14]) Midazolam 24 (1235 [078]) )8 ()23 to +1 [)57 to +29]) 50 (5050 [3174]) )4 ()16 to 0 [)27 to +12]) 31 (2734 [2245]) 33 (2841 [2256]) )5 ()11 to )1 [)29 to +9]) Combined 21 (1756 [091]) )18 ()32 to )5 [)50 to +7]) 50 (5050 [2551]) 18 (+8 to + 28 [0 to +35]) 32 (2643 [2367]) 41 (3145 [2064]) )8 ()14 to +5 [)24 to +5]) p value* 0.33 0.02 0.12 0.0001 0.11 0.48 0.11

*From one-way ANOVA.

There was no interaction between midazolam and forcedair warming for VAS for anxiety (p = 0.12), STAI-state (p = 0.90) or thermal comfort (p = 0.08). Administration of midazolam (p = 0.006) and female sex (p = 0.04) were signicant independent predictors of a reduction in VAS-anxiety in the generalised linear model. Signicant between-group (p < 0.0001) and within-group (p < 0.0001) effects were observed in BIS values. The lowest BIS values occurred at 5 min in the midazolam group (median (IQR [range]) 84 (8190 [6197])) and combined group (83 (8191 [7198])), and at 20 min in the control group (90 (8693 [6798])) and warming group (92 (8897 [8398])). There were no signicant differences in heart rate, systolic blood pressure or arterial oxygen saturation between the randomised groups (results not shown).
Discussion

Pre-operative forced-air warming improved thermal comfort, but did not reduce pre-operative anxiety. In
2009 The Authors Journal compilation 2009 The Association of Anaesthetists of Great Britain and Ireland

addition, there was no evidence for a benecial effect of warming in combination with midazolam. Our results are similar to those of Kimberger et al. [4] who also investigated forced-air warming, but contrast with those of Wagner et al. [3] who reported less anxiety in patients warmed with a warming suit than in patients who received routine care. Whilst 30 lg.kg)1 midazolam produced anxiolysis in our patients, we were concerned that it might also cause excessive sedation. We monitored BIS because we wanted to measure sedation objectively without stimulating the patients during the treatment period [5, 6]. Midazolam produced signicant reductions in BIS values which is consistent with previous reports [5, 6]. However, no loss of responsiveness nor arterial desaturation was observed. Forced-air warming produced signicant improvements in thermal comfort [10], so we can exclude lack of warming efcacy as a reason for our result. However, because we did not conduct a head-to-head comparison of forced-air warming and a warming suit (like that used by Wagner et al. [3]), we cannot comment
1079

R. J. Wen et al. Pre-operative warming and anxiolysis Anaesthesia, 2009, 64, pages 10771080 . ....................................................................................................................................................................................................................

about the superiority of one type of warming over another. Female sex was an independent predictor of a decrease in anxiety. Previous work has reported a greater placebo response in patients with higher levels of anxiety [11]. Given that women have a higher baseline level of anxiety [2], they may be predisposed to a greater placebo response and this may explain our results. The role of gender in the placebo effect, which has rarely been investigated in a clinical trial setting, warrants further study [11]. Our patients had lower baseline STAI scores than those collected by Spielberger et al. [8] in surgical patients during development of the score in the 1980s, and by Kimberger et al. in their neurosurgical population [3]. Perhaps modern-day patients having minor surgery are less anxious pre-operatively due to various improvements in safety and quality of care. In addition, the effect of midazolam was not particularly strong (only a 10-mm reduction in VAS) although this also may have been a function of low general anxiety. In conclusion, whilst forced-air warming increased patients thermal comfort, it was not equivalent to midazolam as an anxiolytic. Pre-operative warming of patients does not cause a clinically signicant reduction in peri-operative patient anxiety and as such cannot be recommended solely for this purpose.
References 1 Ozalp G, Sarioglu R, Tungel G, Aslan K, Kadiogullari N. Preoperative emotional states in patients with breast cancer

8 9

10

11

and postoperative pain. Acta Anaesthesiologica Scandinavica 2003; 47: 269. Fredman B, Lahav M, Zohar E, Golod M, Paruta I, Jedeikin R. The effect of midazolam premedication on mental and psychomotor recovery in geriatric patients undergoing brief surgical procedures. Anesthesia & Analgesia 1999; 89: 11616. Wagner D, Byrne M, Kolcaba K. Effects of comfort warming on preoperative patients. AORN Journal 2006; 84: 42730. Kimberger O, Illievich U, Lenhardt R. The effect of skin surface warming on pre-operative anxiety in neurosurgery patients. Anaesthesia 2007; 62: 1405. Glass PS, Bloom M, Kearse L, Rosow C, Sebel PS, Manberg P. Bispectral analysis measures sedation and memory effects of propofol, midazolam, isoflurane and alfentanil in healthy volunteers. Anesthesiology 1997; 86: 83647. Liu J, Singh H, White PF. Electroencephalographic bispectral analysis predicts the depth of midazolam-induced sedation. Anesthesiology 1996; 84: 649. Kindler C, Harms C, Amsler F, Ihde-Scholl T, Scheidegger D. The visual analog scalre allows effective measurement of preoperative anxiety and detection of patients anesthetic concerns. Anesthesia & Analgesia 2000; 90: 70612. Spielberger C. Manual for the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press, 1983. Taylor E, Ghouri A, White P. Midazolam in combination with propofol for sedation during local anesthesia. Journal of Clinical Anesthesia 1992; 4: 2136. Sessler D, Schroeder M, Merrifield B, Matsukawa T, Cheng C. Optimal duration and temperature of pre-warming. Anesthesiology 1995; 82: 67481. Enck P, Benedetti F, Schedlowski M. New insights into the placebo and nocebo responses. Neuron 2008; 59: 195206.

1080

2009 The Authors Journal compilation 2009 The Association of Anaesthetists of Great Britain and Ireland