Jurisdictions in Canada, the US, the EU and Australia are struggling with regulation of ever evolving software in medicine. Recently this discussion has had a focus on electronic medical records (EMRs). There is a mountain of evidence that EMRs have actualized potential to lead to the injury of patients through the information they offer to facilitate care. We are undertaking a systematic review of relevant literature in the field to uncover some of the latent hazards. We hypothesize that this exploration, using a variation on Leveson's system theoretic accidents models and processes (STAMP) model as a classification tool, will provide two benefits. First, the model will be sufficient to capture the complexity of the domain and its hazards, thus providing a holistic perspective on the problem. Second, the classification process will provide insight as to what steps might be taken to mitigate the risk that medical errors associated with these software tools will arise in health care systems which employ them. In this continuation of our study we still have not been able to produce evidence which contradicts either hypothesis.