Study design and setting
Eight different dental chairs were used in this study. #1 (EOM-PLUS SS®; GC, Tokyo, Japan), #2 (EOM ∑®; GC, Tokyo, Japan), #3 (EOMαⅡ®; GC, Tokyo, Japan), #4 (Celeb BM Type Clair®; TAKARA, Tokyo, Japan), #5 (SPACELINE EMCIA Type II®; MORITA, Tokyo, Japan), #6 (SPACELINE EMCIA Type III UP®; MORITA, Tokyo, Japan), #7 (NOVA SERIO®; YOSHIDA, Tokyo, Japan), #8 (STAGE II®; YOSHIDA, Tokyo, Japan). Each dental chair was installed in four private dental offices. Three health care providers, who completed AHA-certified Basic Life Support course, participated in this study; A: 47 years-old man,175cm, 93kg. B: 44 years-old man, 177cm, 60kg. C: 44 years-old woman, 157cm, 50kg.
The CPR manikin (Resusci Anne Torso Basic version 2011; Laerdal Medical AS, Stavanger, Norway) was laid on the horizontal backrest of the dental chair. The upper end of the torso of the manikin was aligned with the top edge of the backrest (Figure. 1A, Red line). The surface of the backrest under the lower half of the sternum of the manikin was levelled using a levelling instrument (Z-340; Hozan Co., Osaka, Japan).
The hand position for MCC was the center of the chest (the lower half of the sternum, Figure 1B) as recommended in the European Resuscitation Council Guidelines for Resuscitation [4] and the American Heart Association (AHA) Guidelines [5]. Three health care providers performed MCC on the resuscitation manikin in eight different dental chairs. The displacement of the point P (Figure. 2A) on the lower surface of the backrest (vertically under the area for MCC) was fixed (Figure. 2B). The metal indicator (point P) was attached the instrument by using a level gauge (Z-340; Hozan Co., Osaka, Japan) horizontally to the ground. The point P was measured at the same time as MCC-induced vertical movements of the backrest. The depth of MCC was kept between 5.1 to 6.0 cm with and without a stool. The actual depth of MCC was evaluated by the skill-reporter® system equipped with the manikin. The green light of the skill-reporter® indicates 3.8 to 5.0 cm of MCC depth, and red light indicates 5.1 to 6.0 cm of MCC depth ((Figure. 2C). When the compression depth in the chest of manikin by MCC was 5.1 to 6.0 cm, the vertical displacements of the backrest from its basal position (the width of a starting point to an ending point) were recorded by the camcorder (HC-W580M; Panasonic, Osaka, Japan). Video data were transferred to a computer (Dell; Windows 7, intel: Core i3, Cupertino CA, USA) using a camcorder’s dedicated software (HD Writer 3.1; Panasonic, Osaka, Japan). the vertical displacements (degree of instability) of the backrest were measured using the simultaneously captured ruler as a reference.
To compare the efficacy of a stool as a stabilizer on MCC in eight types of dental chair, a round stool with a hard seating surface (45cm in diameter, 46cm in height; FB-01ALLBK, Fuji Boeki Co., Ltd. Fukuoka, Japan) was placed under the backrest of the dental chair. The edge of the seating surface of the round stool was set to (vertically) touch the backrest under the area for ECC (Figure. 1A, Green line). MCC was performed with or without the round stool as a stabilizer. The manikin was on a fully reclined chair.
Protocol
Three health care providers individually performed ten rounds of continuous MCC for 20 times each at a pace of 100 compressions per minute by synchronizing to a metronome. Chest compressions of 5.1 to 6.0 cm were performed with and without a stool. The health care providers and the research team were blinded to the information during MCC. Therefore, for each participant, 200 records of chest compressions were gotten for each dental chair.
Statistical analysis
The programming language R (version 3.4.3; The Comprehensive R Archive Network, USA) was used for statistical analysis.
Each combined measurement data set of the chair’s reference point displacement during ECC treatment by 3 practitioners were applied to the Shapiro-Wilk test (with the function shapiro.test) to see whether they were sampled from a population with the normal distribution. The data sets were analyzed using the non-parametric Wilcoxon rank sum test (wilcox.exact: exactRankTests package).