Skip to main content

Advertisement

Table 6 BOLD Power Estimates (registry size = 5,000 patients)*

From: Study protocol: The back pain outcomes using longitudinal data (BOLD) registry

  Index Case Prevalence
  2% 3% 5% 10%
Number of Cases 100 150 250 500
Effect Size (D)
RDQ†
1.5 - - 61% 88%
2.0 - 63% 84% 98%
2.5 65% 82% 96% 99%
Pain NRS†
0.7 - 52% 74% 96%
1.0 65% 82% 96% 99%
1.3 87% 96% 99% 99%
Incremental Cost per QALY Gain < $50,000††
$500 / 0.03 D QALY - - - 38%
$500 / 0.06 D QALY 45% 61% 82% 98%
$800 / 0.03 D QALY - - - 21%
$800 / 0.06 D QALY 36% 51% 72% 95%
Rare Events (Surgery, Complication, or Binary Event Rate)**
2% Incidence Rate
RR = 1.5 - - - -
RR = 2.0 - - - 62%
RR = 2.5 - - 60% 88%
5% Incidence Rate
RR = 1.5 - - - 51%
RR = 2.0 - 51% 73% 95%
RR = 2.5 - 79% 95% 99%
  1. *Comparisons where power estimates are below 50% are not provided
  2. † Conservatively assumes 1:1 case:control ratio, SD estimates of 7.5 and 3.0 for RDQ and Pain NRS respectively.
  3. †† Conservatively assumes 1:1 case to control ratio, calculated for one-year time period, cost standard deviations of $4,000 per group, QALY standard deviations of 0.2 per group based on EQ-5D and 6-month effect improvement, and a positive correlation of 0.3 between costs and changes in EQ-5D [47]. Assuming a negative correlation did not affect power substantially.
  4. **Assumes 3:1 matching of controls to index cases.