## Info

The average value for the 25 phantom scans is 1.177 g/cm2. The SD is 0.006 g/cm2 and 1.5% of the average value is 0.018 g/cm2. Compare these average and SD values to those calculated for the 10 scans in Table 4-2.

The average value for the 25 phantom scans is 1.177 g/cm2. The SD is 0.006 g/cm2 and 1.5% of the average value is 0.018 g/cm2. Compare these average and SD values to those calculated for the 10 scans in Table 4-2.

The control table described earlier is simpler to create and maintain than the Shewhart chart but the ability to visually inspect the data for drifts or shifts is lost. The creation of a Shewhart control table or chart constitutes the minimum quality control program that should be in use in every facility performing densitometry.

The creation of an average baseline phantom value by scanning the phantom 10 times on the same day without repositioning may not reflect the day-to-day variability in machine values and the effects of repositioning that would be expected as the phantom is scanned over time. Several groups have consequently recommended that the baseline phantom value be established by scanning the phantom once a day for 15 to 25 consecutive days and then averaging these 15 to 25 scans. It is thought that this will more accurately reflect the day-to-day variability in machine values and result in fewer "false alarm failures." For example, the average BMD of the same Hologic spine phantom when scanned on 25 consecutive days as shown in Table 4-3 was 1.177 g/cm2 resulting in a range for the average ± 1.5% of 1.159 g/cm2 to 1.195 g/cm2. In both cases, 1.5% of the mean value was 0.018 g/cm2 but the range of acceptable values was different from that seen when the phantom was scanned 10 times on the same day without repositioning. Figure 4-2 is the graph of subsequent scans now plotted against the baseline phantom value obtained after scanning the phantom once on each of 25 consecutive days.

Notice in Fig. 4-2, when the mean was calculated using 25 scans performed on consecutive days, the same phantom values do not give any indication of a loss of random scatter. More sophisticated evaluations of this type of data can be done to determine if, in fact, there has been a shift in values. Nevertheless, this type of chart is the foundation of a good quality control program.

Fig. 4-1. A quality control chart with control limits of ±1.5%. The average BMD of the phantom was established by scanning the phantom 10 times on the same day. Arrow 1 indicates a point at which values appear to be drifting downward. Arrow 2 indicates a violation of the 1.5% rule.
Fig. 4-2. A Shewhart chart for quality control. The average BMD of the phantom was established by scanning the phantom once on 25 consecutive days.