Now let's talk about strategies for reducing medication errors. I want to thank Amy Rosenberg from the University of Florida Shands Pharmacy, who is working on the front lines of our hospital to prevent medication errors. She has provided much of the content for this session. Let's start with an all too typical case. The medical literature has reported many cases like this hypothetical example. A 25 year old man comes to the ER having severely injured his knee in a bicycle accident. He is in extreme pain, and the new resident orders the narcotic hydromorphone called Dilaudid. He mistakenly this is this pain medication is equivalent to morphine. He orders 5 mg to be given by IV push. Ten minutes later the patient has a respiratory arrest. Fortunately, the nurse is at the bedside when he stops breathing, and he is immediately resuscitated and given Narcan to reverse his narcotic overdose. The problem is that hydromorphone, Dilaudid, is seven times more potent than morphine. 1 mg of Dilaudid is equal to 7 mg of morphine. Therefore, giving this young man 5 mg of Dilaudid was the same as giving him 35 mg of morphine, a potentially lethal dose. Too often, caregivers assume Dilaudid equals morphine. The result, respiratory arrest and deaths. How can we prevent this common error? Amy Rosenberg led the effort to solve this problem. She worked with the support staff for our electronic medical record epic to change the pull-down menu for ordering Dilaudid. The pull down menu for dosing was lowered to a maximum dose of .5 milligrams of IV. Previously physicians were able to pick doses of 1 and 2 milligrams and in the narcotic naive patients it is well known that these higher doses can produce dangerous respiratory depression. Also a physician tried to hand enter a higher dose, a warning now pops up, message, caution Dilaudid 1 milligram IV equals morphine 7 milligrams IV. Amy and her team used the second most effective approach for preventing errors, automation, and computerization. Although the risk of overdose is lowered, it is not completely eliminated. Let's look at another case. This case has been modified to protect confidentiality. A 30 year old female recently underwent a kidney transplant and is receiving agents that depress her immunity to prevent rejection of her kidney. Unfortunately, she breathes in contaminated air and develops pneumonia caused by the fungus Aspergillosis. The senior physician attending instructs the intern to order the anti-fungal agent Amphotericin B, meaning, the lipid form, which is a dose of 5 milligram per kilogram per day. However, the intern orders 5 mg/kg/day of the conventional Amphotericin B preparation. The dose for conventional form should be 1 mg/kg/day. So the intern inadvertently orders five times the recommended dose. The patient develops severe kidney failure and subsequently dies. The lipid based amphotericin dose is generally 5 times higher than the conventional amphotericin dose. Amphotericin is used to treat serious invasive fungal infections, and the lipid form is far less toxic. Conventional Amphotericin is a nephrotixic that is toxic to the kidney medication even in the correct doses. With the overdose this toxicity is magnified. Mix-ups can occur during prescribing or dispensing. What were the risk reduction strategies pharmacies used for this dangerous condition? Force the right choice by limiting the formulary options to one product. Remove the conventional product within the pharmacy. Alternatively, the pharmacy could store the liposomal form on the pharmacy shelf and sequester conventional Amphtericin B in another area. If this second strategy is chosen, a second step needs to be added. You must create a fail safe way to alert the dispenser using an automated dispensing cabinet within the pharmacy. Alerts can be implemented to ensure that those removing the product are aware of the safety concerns with product mix ups. The second two step strategy is not foolproof and the possibility of harm is not eliminated. Forcing functions make it impossible to do the wrong thing. When conditions are life threatening, this is the best approach. What can go wrong during anesthesia? A 52 year old woman is undergoing anesthesia for a knee repair. 15 minutes into surgery, she has a cardiac arrest and dies. After death they discover that the nitrous oxide tank had been hooked up to the oxygen line and the woman had received no oxygen. Specific colors for hoses can reduce the chances of this fatal error, but do not eliminate the risk. Oxygen equals green, nitrous oxide equals blue, compressed air equals yellow. But color codes are not enough. You need connecting valves with specific pins, make it impossible to hook nitrogen oxide to the oxygen dispenser. Prior to the design of these lock and key valves, cases like this 52 year old woman's death from hypoxia were quite common. However, since this forcing function was created this mix up never happens. Forcing functions saves lives. The second most effective strategy for reducing errors is automation, and when it comes administering medications, a barcode reader provides the best approach for reducing medication administration errors. There is a bar code on every medication and a bar code on the wristband of every patient. The right medication to the right patient equals green light. The wrong patient or the wrong medication equals red light. Reliability is 1.43 per 10,000 or 143 errors per million, not quite six sigma, but getting there. What are some other ways that we can prevent errors? Let's return to the error prevention pyramid. Standardization is of intermediate effectiveness, but it certainly has its place. Care bundles are a form of standardization. They represent a collection of evidence based processes that, when performed together, improve outcome. For example, central venous lines, CVL infections, in our ICUs. This is a serious problem, 80,000 infections per year resulting in 28,000 deaths per year. In other words, one in three of these infections is fatal. Peter Pronovost of Johns Hopkins began promoting the catheter insertion bundle, and this bundle was applied to over 100 Michigan hospitals. CVL infections dropped from a mean rate of 7.7 to 1.4 per thousand catheterizations over a 16 to 18 month period. The bundle consisted of hand-washing, use of full barrier precautions with central line placement, that means drapes, gowns, masks and gloves. Cleaning the insertion site with chlorhexidine, a very strong disinfectant. Avoiding the use of femoral catheters whenever possible. And removing all unnecessary catheters as soon as possible. As this bundle illustrates, seemingly simple evidence based interventions can profoundly reduce patient harm. Another standardization tool is the anti-coagulation protocol for treating thrombophlebitis. If this had been used during Mary's hospitalization, she would not have suffered a pulmonary embolism and probably would not have ended up in the ICU. Next on the pyramid of prevention measures are reminders, checklists, and double checks. Checklists and reminders are used for airline takeoffs, and are particularly useful prior to a surgical procedure. They serve as a cognitive safety net tha ensures that obvious steps are not overlooked. This method needs to be used sparingly because creating too many checklists can lead to checklist fatigue. Also because the working memory can only manage seven facts at a time, checklists should include seven or less items. Finally, use the do-confirm method. In other words, complete the tasks and then use the checklist to assure that you didn't miss any steps. This is the quickest approach and has been found to be the most effective way to use checklists. At the bottom of the pyramid are rules, policies, and mandates. In the past, administrators thought they could just simply mandate quality improvements, but it turns out human beings, particularly us freedom loving Americans, are not very responsive to this approach. Too often, front line caregivers simply ignore these mandates and rules. And finally at the very bottom is eduction. Education at best results in a 5% change in behavior. This is a disappointing fact. Now you realize that educational load will not fix healthcare delivered. How we can all help by creating forcing functions, helping to simplify and standardize and by encouraging check lists as well as reminders. By practicing what you have learned in this class, you can continue to learn by doing. Together we can save lives. Thank you.