Table 1 Simulated clinical record review (CRR) parameters and parameter values and their effect on the precision and power of harm rate estimates
Parameters | Selected parameter values | Parameter affects precision | Parameter affects power |
|---|---|---|---|
Parameters with variable values | |||
Number of general practices conducting CRR | 1, 10, 20, 50, 100, 150, 200, 250, 300 | √ | √ |
Number of unique patient records reviewed by each practice at a given point in time | 20, 25, 50, 100, 150, 200 | √ | √ |
The real harm rate in the sampled patient population, expressed as incidents per 100 patients per year | 2, 5, 10, 20 | √ | √ |
The actual reduction in the real harm rate over 12 months | 20%, 50% | √ | |
Inter-patient variation in harm susceptibility, expressed as a median rate ratio (MRR) | 1.2, 2 | √ | √ |
Parameters with fixed values | |||
Inter-practice variation in harm susceptibility, expressed as a MRR | 1.2 | √ | √ |
Period of time reviewer reviewed in each record (calendar months) | 3 | √ | √ |
Period of time over which changes in harm rates are examined (months) | 12 | √ | |
Number of reviews during the simulated 12-month period | 2 | √ | |
Reviews at different time points are conducted on the same or different samples of patient records | Same | √ |