Table 3:
Simulation Scenario | Simulated levels of outcome misclassification | Mean overall outcome PPV PPV1, PPV0 | Mean overall outcome NPV NPV1, NPV0 | Median RR(95% SI) |
||
---|---|---|---|---|---|---|
From analysis of electronic data with outcome misclassification present | Following QBA using PPV1, PPV0, NPV1, NPV0 | Following QBA using PPV1, PPV0, and assuming nondifferential sensitivity | ||||
Formula 1 | Formula 2 | Formula 3 | ||||
Simulated RR = 2.00 | ||||||
Nondifferential misclassification | ||||||
1 | SN = 95%, SP = 99% | 84.6% 91.8%, 84.1% | 99.7% 99.4%, 99.7% | 1.83 (1.70 - 1.98) | 2.00 (1.86 - 2.16) | 2.00 (1.85 - 2.17) |
2 | SN = 90%, SP = 98% | 72.2% 84.2%, 71.4% | 99.4% 98.8%, 99.4% | 1.70 (1.57-1.83) | 2.00 (1.86 - 2.16) | 2.00 (1.84-2.16) |
Differential specificity | ||||||
3 | SN1 = 95%; SN0 = 95% | 80.6% | 99.7% | 3.19 | 2.00 | 2.00 |
SP1 = 90%; SP0 = 99% | 52.8%, 84.1% | 99.3%, 99.7% | (3.01-3.36) | (1.86-2.16) | (1.85-2.16) | |
4 | SN1 = 95%; SN0 = 95% | 73.7% | 99.7% | 1.58 | 2.00 | 2.00 |
SP1 = 99%; SP0 = 98% | 91.8%, 72.6% | 99.4%, 99.7% | (1.46-1.70) | (1.86-2.16) | (1.85-2.16) | |
Differential sensitivity | ||||||
5 | SN1 = 70%; SN0 = 95% | 84.4% | 99.6% | 1.39 | 2.00 | 1.47 |
SP1 = 99%; SP0 = 99% | 89.2%, 84.1% | 96.6%, 99.7% | (1.27-1.51) | (1.86-2.16) | (1.34-1.61) | |
6 | SN1 = 50%; SN0 = 95% | 84.2% | 99.5% | 1.04 | 2.00 | 1.06 |
SP1 = 99%; SP0 = 99% | 85.5%, 84.1% | 94.4%, 99.7% | (0.94-1.14) | (1.86-2.16) | (0.95-1.17) | |
Differential specificity and sensitivity | ||||||
7 | SN1 = 70%; SN0 = 95% | 73.3% | 99.6% | 1.20 | 2.00 | 1.47 |
SP1 = 99%; SP0 = 98% | 89.1%, 72.6% | 96.5%, 99.7% | (1.10-1.30) | (1.86-2.16) | (1.34-1.61) | |
8 | SN1 = 90%; SN0 = 95% | 80.6% | 99.7% | 3.10 | 2.00 | 1.89 |
SP1 = 90%; SP0 = 99% | 51.5%, 84.1% | 98.7%, 99.7% | (2.93-3.27) | (1.86-2.16) | (1.74-2.05) | |
Simulated RR = 0.50 | ||||||
Nondifferential misclassification | ||||||
1 | SN = 95%, SP = 99% | 83.8% 72.1%, 84.1% | 99.7% 99.9%, 99.7% | 0.58 (0.51-0.66) | 0.50 (0.43-0.58) | 0.50 (0.43-0.58) |
2 | SN = 90%, SP = 98% | 71.1% 55.0%, 71.5% | 99.5% 99.7%, 99.4% | 0.65 (0.58-0.73) | 0.50 (0.43-0.58) | 0.50 (0.42-0.58) |
Differential specificity | ||||||
3 | SN1 = 95%; SN0 = 95% | 79.4% | 99.7% | 2.05 | 0.50 | 0.50 |
SP1 = 90%; SP0 = 99% | 20.5%, 84.1% | 99.9%, 99.7% | (1.92-2.20) | (0.43-0.58) | (0.42-0.58) | |
4 | SN1 = 95%; SN0 = 95% | 72.5% | 99.7% | 0.50 | 0.50 | 0.50 |
SP1 = 99%; SP0 = 98% | 72.1%, 72.5% | 99.9%, 99.7% | (0.44-0.58) | (0.43-0.58) | (0.42-0.58) | |
Differential sensitivity | ||||||
5 | SN1 = 70%; SN0 = 95% | 83.8% | 99.7% | 0.47 | 0.50 | 0.37 |
SP1 = 99%; SP0 = 99% | 65.5%, 84.1% | 99.2%, 99.7% | (0.40-0.55) | (0.43-0.58) | (0.30-0.44) | |
6 | SN1 = 50%; SN0 = 95% | 83.7% | 99.7% | 0.39 | 0.50 | 0.26 |
SP1 = 99%; SP0 = 99% | 57.5%, 84.1% | 98.7%, 99.7% | (0.32-0.45) | (0.43-0.58) | (0.21-0.32) | |
Differential specificity and sensitivity | ||||||
7 | SN1 = 70%; SN0 = 95% | 72.4% | 99.7% | 0.41 | 0.50 | 0.37 |
SP1 = 99%; SP0 = 98% | 65.5%, 72.5% | 99.2%, 99.7% | (0.35-0.47) | (0.43-0.58) | (0.31-0.44) | |
8 | SN1 = 90%; SN0 = 95% | 79.3% | 99.7% | 2.03 | 0.50 | 0.47 |
SP1 = 90%; SP0 = 99% | 19.7%, 84.1% | 99.7%, 99.7% | (1.90-2.17) | (0.43-0.58) | (0.40-0.55) |
The 95% SI represents the 2.5th and 97.5th percentiles of the RR from across 1000 replications for each simulated scenario. When the simulated relative risk was 2.00, the 95% simulation interval before applying misclassification was 1.86-2.16. When the simulated relative risk was 0.50, the 95% SI before applying misclassification was 0.43-0.58.
PPV0: positive predictive value among unexposed; PPV1: positive predictive value among exposed; QBA: quantitative bias analysis; SI: simulation interval; SN: overall outcome sensitivity; SN0: outcome sensitivity among unexposed; SN1: outcome sensitivity among exposed; SP: overall outcome specificity; SP0: outcome specificity among unexposed; SP1: outcome specificity among exposed.