Reader and case sample combinations needed for 0.80 power based on the data of:

Ohashi K, El-Khoury GY, Bennett DL, Restrepo JM, Berbaum KS. Orthopedic hardware complications diagnosed with multi-detector row CT. Radiology 2005;237:570-577.

Readers: 2 pediatric radiologists

Cases: 44 patients, 18 with clinically proven hardware complications, 26 with clinically proven absence of hardware complications

Rating: probability of abnormality on a 100-point subjective probability scale Conditions: (1) radiography, (2) multi-detector row CT

Results: Fixed Reader / Random Case Analysis

ROC Model = propROC, Parameter = AUC,

AUCs = 0.842, 0.999, respectively.

Readers = 2, Cases = 44,

MS( Treatment*Reader ) = 0.03649473,

MS( Treatment*Case ) = 0.20259783,

MS( Treatment*Reader *Case ) = 0.03236510.

AUC Difference = 0.03

Readers	Readers and Cases Random	Cases Random	Readers Random
3	.	1675	.
4	.	1628	.
5	.	1600	497
6	.	1581	260
7	.	1568	176
8	1997	1558	134
9	1918	1550	108
10	1861	1543	90
11	1818	1538	78
12	1784	1534	68
13	1756	1530	61
14	1733	1527	55
15	1714	1525	50

Note: Cells containing a period indicate that more than 2000 cases would be required.

AUC Difference = 0.05

Readers	Readers and Cases Random	Cases Random	Readers Random
3	842	605	1379
4	714	588	178
5	664	578	94
6	637	571	64
7	620	566	48
8	607	562	39
9	598	559	33
10	591	557	28
11	586	555	25
12	581	554	22
13	577	553	20
14	574	551	20
15	571	550	20

AUC Difference = 0.10

Readers	Readers and Cases Random	Cases Random	Readers Random
3	163	153	87
4	154	149	33
5	149	146	20
6	147	145	20
7	145	143	20
8	143	142	20
9	142	142	20
10	141	141	20
11	141	141	20
12	140	140	20
13	140	140	20
14	139	140	20
15	139	139	20

AUC Difference = 0.20

Readers	Readers and Cases Random	Cases Random	Readers Random
3	39	40	20
4	38	39	20
5	37	38	20
6	36	38	20
7	36	38	20
8	36	37	20
9	35	37	20
10	35	37	20
11	35	37	20
12	35	37	20
13	35	37	20
14	35	37	20
15	35	37	20