The Determination of LSC

The situation then, of one measurement at baseline and one measurement at follow-up effectively changes equation 10 to equation 11, used for the calculation of the 1x1LSC:

If two measurements are done at baseline and again at follow-up, the sum under the square root sign in equation 9 becomes 1 as shown in equation 12 for the calculation of the 2x2lsc.

This effectively changes the equation for the calculation of the ^LSC to:

Thus, for any level of statistical confidence, the magnitude of the LSC is reduced by performing duplicate measurements at baseline and follow-up rather than single measurements because the product of the Z' and precision values is being multiplied by only 1 instead of 1.414. The LSC is effectively reduced by approximately 30%.

If the Z' values shown in Table 11-5 for 95% and 80% are substituted in the formulas for the 1x1LSC and the 2x2LSC the formulas become:

For example, if the RMS-SD precision of PA lumbar spine DXA studies at a facility was determined to be 0.015 g/cm2, this value would be substituted in Equation 14 if 95% confidence was desired and one measurement was performed both at baseline and follow-up. In that case, the changes in Equation 14 are reflected in Equations 18 and 19.

For 80% confidence, the precision value of 0.015 g/cm2 is substituted in equation 15, resulting in an LSC of0.027 g/cm2. The RMS-%CV can be substituted in a similar fashion for the precision value in equations 14 through 17 to give the LSC as a percent change from baseline for the various levels of confidence and numbers of measurements.

The timing of the repeat measurement is a direct consequence of the LSC. The follow-up measurement(s) should be done when enough time has passed for the LSC to be achieved. Therefore, once the magnitude of the LSC has been determined, the time required between measurements is:

Expected rate of change per year

Table 11-6

The Interval Between BMD Measurements Required to Obtain the 1x1LSC95 for Various Levels of Precision and Expected Rates of Change

Interval Between BMD Measurements

Table 11-6

The Interval Between BMD Measurements Required to Obtain the 1x1LSC95 for Various Levels of Precision and Expected Rates of Change

Interval Between BMD Measurements

Precision as % CV

% Change/Year

Months

Years

Was this article helpful?

0 0

Post a comment