ADDRESS Development and Results
Development of the ADDRESS system has involved a great deal of effort and many thousands of man-hours of work. The system is not a simple listing of all possible side effects for any given drug. We determined very early on that such an approach was hopelessly cumbersome and unlikely to prove of any real use in day to day practice in the ambulatory setting. Practitioners specializing in specific disease areas tend to be quite familiar with the ADRs of the relatively limited tool-box of therapeutic agents they use regularly. The intended purpose of ADDRESS is to help manage the pharmacotherapy of the general population, using commonly prescribed drugs for mostly chronic diseases, who are not institutionalized or receiving intense levels of clinical attention. In other words, the great majority of general patients. We KNOW patients will experience certain ADRs at relatively well-established incidence rates. What was necessary was to develop a system that utilized the optimum level of psychometric efficiency - that is, that produced the highest number of “side effect” hits per question asked.
While the exact technique is proprietary, ADDRESS has been designed from the ground up to achieve our goal of optimal psychometric efficiency. Ultimately, we turned the traditional system of drug categorization (based on therapeutic type or class, followed by side effects) on its head, and worked back from the side effects up to the drug. Some early posters are listed below.
- Prevalence of Adverse Drugs Effects and Adverse Drug Reactions in the 200 Most Commonly Prescribed Drugs
- Prevalence of Adverse Drugs Effects and Adverse Drug Reactions in the 200 Most Commonly Prescribed Drugs Corrected for Prescription Volume
- Initial Testing of Instruments to Discern Adverse Drug Reactions and Adverse Drug Experiences in an Ambulatory Patient Population: Preliminary Results
We published a basic science piece in Pharmacy Innovations that lays out the approach to categorization of drug information we had to develop to make the ADDRESS system work. It is available online for you to read. We published in the Journal of the American Pharmacist’s Association an article describing our initial “proof of concept” results using paper and pencil versions of the questions used by our automated side effect screening system. It is available online for you to read. At this time, we have some very initial data about the performance of the automated application. There is a slide set describing how much time pharmacists spend, on average, programming the application for each drug, and the time patients spend responding. View the slideshow on performance. The resulting manuscript is in preparation and final write-up.
We have also gathered additional data sets that should prove interesting. One consists of screening of four groups- The Purdue University Retirement Association, The Wabash Alliance for Life-Long Learning, The “Roads-Scholar’s Program” (a 2 week series of seminars for older folks interested in learning), and the Arch Diocese of Indianapolis’ Senior Companions Program, which consists of people over 65 and under the federal poverty line who are hired to provide care and companionship to the frail elderly, and are almost entirely African-American. This will present a chance to examine the possibility of the existence of differences in ADR burden in different racial/socio-economic groups. The first three groups are almost entirely white middle or upper middle class, and of retirement age or older. The final group is of similar age, but poverty level and African American.
The next data set is small (56 patients) collected randomly from an independent community Pharmacy in North-Central Indiana (Webb’s Pharmacy). The interesting aspect of this data set is that the store’s dispensing system generates adherence metrics for all of a patient’s drugs, so we may be able to take a first look at the possibility of an association between ADR burden and adherence.
A final note- as of June 2017, we have initiated a program here at Purdue where student pharmacists on their Community APPE rotations are provided with the technology and required to screen at least two patients each day of their rotations. We are assessing communications reluctance and knowledge of the ADRs for the top 15 drugs (by Rx volume) to assess the impact of the use of technology on enhancement of the student‘s rotational experience. It will also, over time, provide a massive sample of patients in the ambulatory population- approximately 7,000 subjects each year. The graphs on the opening page represent a real-time representation of this data, as it comes in. (We hope the students will be enthused by seeing their work being reflected in this manner).