Comparison of nursing diagnostic accuracy when aided by Knowledge-Based Clinical Decision Support Systems with Clinical Diagnostic Validity and Bayesian Decision Models for psychiatric care plan formulation among nursing students: a quasi-experimental study

Table 4 Pretest and posttest results on nursing students’ decision-making competency

Outcomes	Group	Pretest		Posttest		P value
Outcomes	Group	n (%)	P value (x²)	n (%)	P value (χ²)	P value
Compliance with NANDA-I suggestions^d	Control^a	144 (69.90)	.53(1.29)	155 (75.24)	< .001 (29.57) b > c > a	.15^j
	CDV^b	152 (74.88)		191 (94.09)		< .001^j
	BADE^c	142 (71.72)		172 (86.87)		< .001^j
False positives^e	Control^a	44 (21.36)	.55(1.22)	37 (17.96)	0.01 (8.53) a > b	.12^j
	CDV^b	47 (23.15)		17 (8.37)		< .001^j
	BADE^c	37 (18.69)		24 (12.12)		.03^j
False negatives^f	Control^a	104 (50.49)	.40(1.83)	106 (51.46)	< .001 (20.72) a > b; a > c	.87^j
	CDV^b	95 (46.80)		60 (29.56)		< .001^j
	BADE^c	106 (53.54)		76 (38.38)		< .001^j
True positives^g	Control^a	58 (28.16)	.86(.29)	63 (30.58)	< .001 (41.24) b > c > a	.51^j
	CDV^b	61 (30.05)		126 (62.07)		< .001^j
	BADE^c	55 (27.78)		98 (49.49)		< .001^j
Positive predictive value^h	Control^a	0.57		0.63		< .001 (F = 9.78)^k b > c > a
	CDV^b	0.56		0.88
	BADE^c	0.60		0.80
Sensitivityⁱ	Control^a	0.36		0.37		< .001 (F = 23.26)^k b > c > a
	CDV^b	0.39		0.68
	BADE^c	0.34		0.56

^aControl, control group using the psychiatric care planning system
^bCDV, group using the knowledge-based clinical decision support system (KBCDSS) with the clinical diagnostic validity inference engine
^cBADE, group using the KBCDSS with the Bayesian inference engine
^dCompliance with NANDA-I suggestions, the frequency with which defining characteristics or risk factors were identified by participants in accordance with the suggestions of NANDA-I nursing diagnoses
^eFalse positives, higher frequency of participants identifying defining characteristics than that of the researcher
^fFalse negatives, lower frequency of participants identifying defining characteristics than that of the researcher
^gTrue positives, equal frequency of participants and the researcher of identifying defining characteristics
^hPositive predictive value = true positives/(true positives + false positives)
ⁱSensitivity = true positives/(true positives + false negatives)
^jMcNemer’s test for within-group differences in proportional variables
^kAnalysis of covariance—P values are followed by F values in parentheses

ISSN: 1472-6955