My trust/network has very incomplete data on some risk-factors. Is this likely to affect our results?

Where patients are missing any risk factors, values are imputed based on their other risk factors. This only leads to unbiased estimates of the outcome if there is no systematic difference between patients with a risk factor missing and those with it recorded, after taking into account all of the data items used in the multiple imputation.