Computer-Based Diagnostic Expert Systems in Rheumatology: Where Do We Stand in 2014?

Table 2

Validation of the identified expert systems.

Name of ES^a or first author	Number of cases used for validation	Percentage of diagnoses correct	Sensitivity	Specificity	Reference
Romano	32				[23]
Watt	200	100%			[24]
Provenzano	511	22.9–69.7%^b			[25]
Binder	325		82.6%; CI^c: 68.0–91.7	93.2%; CI^c: 89.4–95.7	[26]
Liu	90	95%	100%	88%	[27]
Lim	No validation				[28]
CADIAG^d	54	48%^e			[29]
RENOIR^d	32	75%			[36]
RHEUMexpert	252		32–77%^f	70–73%^f	[37]
Zupan	462	46.8%; SD^g: 3.9			[38]
AI/RHEUM^d	94	80%			[42]
Dzeroski	462	47.2–50.9%^b			[44]
Heller	12000 computer simulated cases	84.15–99.9%^f			[45]
Astion	807		94.4%	91.9%	[46]
Barreto	No validation				[47]
MESICAR	No validation				[48]
RHEUMA	51	89%^e			[49]
Bernelot Moens^d	570	76%/80%^b; SE^h: 10.2/9.5	62%	98%	[51]
Sereni	341				[53]
Rigby	No validation				[54]
Schewe	358	74.4%			[55]
Prust	No validation				[56]
Gini	No validation				[57]
Dostál	553	80%			[58]
Fries	190	76%			[59]

aExpert system, ^bmultiple formulas were applied, ^cCI: 95% confidence interval, ^dmore than one evaluation, ^eevaluated in other clinic than developed, ^fresults depending on disease, ^gSD: standard deviation, ^hSE: standard error.