DBCG hypo trial validation of radiotherapy parameters from a national data bank versus manual reporting

Figures & data

Table 1. Contingency tables between manual reporting and data submitted to the data bank.

	Manual
A: Trial arm	40 Gy	50 Gy	Missing	Sum
Data bank
40 Gy	745 (48.9%)	3 (0.2%)	7 (0.5%)	755 (49.6%)
50 Gy	3 (0.2%)	753 (49.5%)	11 (0.7%)	767 (50.4%)
Sum	748 (49.1%)	756 (49.7%)	18 (1.2%)	1522 (100.0%)
	Manual
B: Laterality	Right	Left	Missing	Sum
Data bank
Right	730 (48.0%)	4 (0.3%)	1 (0.1%)	735 (48.3%)
Left	11 (0.7%)	772 (50.7%)	4 (0.3%)	787 (51.7%)
Sum	741 (48.7%)	776 (51.0%)	5 (0.3%)	1522 (100.0%)
	Manual
C: Boost	None	10 Gy	16 Gy	Missing	Sum
Data bank
None	1297 (85.5%)	2 (0.1%)	0 (0.0%)	2 (0.1%)	1301 (85.8%)
10 Gy	1 (0.1%)	169 (11.1%)	0 (0.0%)	0 (0.0%)	170 (11.2%)
16 Gy	1 (0.1%)	0 (0.0%)	45 (3.0%)	0 (0.0%)	46 (3.0%)
Sum	1299 (85.6%)	171 (11.3%)	45 (3.0%)	2 (0.1%)	1517 (100.0%)

Tables are shown for A: trial arm, B: laterality of treatment, and C: type of boost. For boost type five patients were excluded since they were detected in the automatic system, prior to data comparison, to be boost treatments. The treatment plans related to these patients were not available in the data bank since they were electron boost based on a manual dose calculation. Since analysis was performed after collection of all data in the data bank there are no missing values in the data bank.

Figure 1. Comparison between values of manual and data bank for: date of RT, CTV volume, minimum dose to CTV, maximum dose to LADCA, CTV maximum, and 1-V95 values (Vx–x% of volume covered by the prescribed dose). IQ is the inter quartile range of the differences between the data bank and manual value. The percentage of outliers and extreme outliers (definition see text) is shown as well as the number of missing data on the x- and y-axis. Red points represent extreme outliers. Two points in subsection c with manual CTV values above 700% are outside the displayed area. Misunderstanding of the interpretation of the specific values in the manual reporting system is seen in subsection e–f, but most pronounced in subsection f.