. 2018 Sep 14;18:153. doi: 10.1186/s12890-018-0696-3

Table 2.

Comparison of number of exams with insufficient sweat weight (below 75 mg) between continuous constant current and sinusoidal and triangular continuous pulsed current

		Continuous constant current
	Sweat weight	Sufficient	Insufficient	p-value
Sinusoidal pulsed current	Sufficient	49 (89.1%)	3 (5.5%)	0.655^a
Sinusoidal pulsed current	Insufficient	2 (3.6%)	1 (1.8%)	0.655^a
Triangular pulsed current	Sufficient	172 (83.9%)	9 (4.4%)	0.523^b
Triangular pulsed current	Insufficient	13 (6.3%)	11 (5.4%)	0.523^b
Continuous pulsed current	Sufficient	221 (85%)	12 (4.6%)	0.333^c
Continuous pulsed current	Insufficient	15 (5.8%)	12 (4.6%)	0.333^c

Statistical analysis conducted through McNemar’s tests. Alpha = 0.05

^aNumber of observed agreements = 50 (90.91%); Number of agreements expected by chance = 48.4 (88.07%); Kappa = 0.238; SE of kappa = 0.232; 95% confidence interval (95%CI) = − 0.217 to 0.693; strength of agreement = fair

^bNumber of observed agreements = 183 (93.85%); number of agreements expected by chance = 163.9 (84.04%); Kappa = 0.614; SE of kappa = 0.101; 95%CI = 0.416 to 0.813; strength of agreement = good

^cNumber of observed agreements = 233 (89.62%); number of agreements expected by chance = 214 (82.3%); Kappa = 0.413; SE of kappa = 0.093; 95%CI = 0.232 to 0.595; strength of agreement = moderate. As shown in the table, the minor prevalence of insufficient sweat weight occurred for sinusoidal pulsed current (1.8%)