"We need better data; better dissemination of that data; and better communication of that data, in ways that help people make decisions which reflect their wis

"We need better data; better dissemination of that data; and better communication of that data, in ways that help people make decisions which reflect their wishes." This is where I think good EMR software can make a difference - in gathering and organizing data. The quantitative data like lab values are probably stored electronically, but the clinical assessments are still scribbles on paper. Most of the physician notes are in SOAP format (Subjective Objective Assessment Plan), and many require educated guesses to decipher (I'm a pharmacist so I have to make these guesses daily). This alone makes data very difficult to gather. You can't do a SELECT * query on this stuff. Every time I call a doctor's office, they'll put me on hold for some time while they flip through a patient's chart. Combine that with the fact that there are multiple clinicians working on the same patient, each with their own set of medical records, the cardiologist might not know what the endocrinologist is doing. Also almost every medical institution uses fax machines. It's 2014. Fax machines should not exist. So the first step to better data is better data-gathering. Prescribing can also be a mess. Evidence-based medicine means there are standards for diagnosing and treating. For example, if a patient was recently diagnosed with Type 2 Diabetes, there is a certain protocol on what to prescribe, with considerations based on a patient's status (say, kidney function). I think this is medicine equivalent of idiomatic programming constructs - just as "cowboy coding" is frowned upon, medicine shouldn't be like House MD. But I frequently see patients being started on non-standard regimens. So not only are standards difficult to come by, sometimes they are not even being followed - and this too compounds the difficulty in gathering good data. Relating to dissemination of data: My aunt was recently diagnosed with non small-cell lung cancer. The first step was surgery. Her next option was infusions of chemotherapy. She didn't consult me on this matter, but I dug up some studies, and it appears that if she chose chemo it'd increase her 5-year survival rate by 5%. Currently based on her staging it's 50%. Granted, aside from profound hair loss, one seizure, and likely adverse hematologic effects, I don't observe any other damage, but I wonder if she would have made the choice to do chemo if she knew about that 5%. Side note: I wish I hadn't been so lazy in pharmacy school and really put in the effort to learn how to code. I remember opening up K&R a few times half-heartedly and not getting past the first three pages. Learning this stuff by myself is an uphill battle, even given the wealth of documentation out there. Most of pharmacy school was just "memorize and regurgitate" - once you got past Organic Chem it was all smooth sailing. Progress is super-slow, because I give up frequently, only to pick myself up the next day to start again. It's only my vision of better medical software that keeps me going.