Journal Entry #3: Current database

                                                                                                            Alanna M. Schauer

                                                                                                            Module 3: Standard Databases Used 

                                                                                                                               in Healthcare Delivery

 

Van Le, H., Beach, K., Powell, G., Pattishall, E., Ryan, P., & Mera, R. (2013). Performance of a semi-

                automated approach for risk estimation using a common data model for longitudinal

 

                healthcare  databases. Statistical Methods In Medical Research, 22(1). 

 

                doi:10.1177/0962280211403599

               Retrieved from http://resources.njstatelib.org/login?
   
                 url=http://search.ebscohost.com/login.aspx?     
                 direct=true&db=aph&AN=85740190&site=ehost-live

                This article is interesting because it demonstrates the adaptability of databases to meet the needs of many healthcare and research entities. Drug safety signals affect patient care and determining patterns in drug interactions or adverse reactions. "The Council for International Organizations of Medical Sciences (CIOMS), which defines a safety signal as information that arises from one or multiple sources which suggests a new, potentially causal association, or a new aspect of a known association between an intervention [e.g., administration of a medicine] and an event or set of related events, either adverse or beneficial, that is judged to be of sufficient likelihood to justify verificatory action” (What is a, 2011, pg. 1).  In the past the United States drug safety surveillance encompassed a process that they had to wait for voluntary reporting of adverse drug reactions mandated by the pharmaceutical industry.  This database creation took a passive data collection system and converted it to current active database with real time complications. This was a massive undertaking because they had to retrospectively review data from the electronic health record to create a semi automated common data model (CDM).

                This article was a research study utilizing health care claims database to determine established drug safety associations, publications, and review of cohort and case study to obtain the twenty-seven drug safety signals. The population and inclusion/exclusion criteria were obtained and then the CDM produced consistent risk estimates for the selected drug outcomes.

                The need that was addressed was the development of an active surveillance system that can identify "hundreds of potential drug signals," (Van Le, Beach, Powell, Pattishall, Ryan, & Mera, 2013, pg. 98). The CDM can also estimate risk potential of these drug signals with evaluation in a timely manner. They took information from the PharMetrics (PM) which is a US administrative health claims database with thirty-one millions patient's information and General Electric Centricity (GE) which is a US electronic health record providing medication and prescription history on almost nine million patients and populated the CDM. They used the CDM to extract ICD-9 codes, medications, adverse reactions, patient information, medical conditions, and determine risk estimation. They compared the data from literature reviews to the healthcare database to determine the confidence interval (CI).  Tables for the database were created for each drug-condition association to include the data source and the outcomes evaluated.  The tables also included the risk estimation for each drug signal. These were compared to the literature review against the healthcare data base.  

                Technology was used to ensure that drug safety signals are identified quickly and accurately so they can be dealt with without haste. The benefits of the CDM creation was to standardize the  longitudinal healthcare databases by including observational data like drug interaction or reactions, standardize the vocabulary used and gather information for the potential need of public health crisis.

                The implications to healthcare delivery include the existence of patterns forming at an early juncture. With knowledge of a drug safety signal occurring, actions can be more pre-emptive rather than reactive.  The drug signals will be collected in an active manner instead of waiting for the stakeholders involved with pharmaceuticals to come forward and voluntarily report a problem.

                I have learned that going forward many existing processes will be analyzed to determine if they can be placed in a database for their efficiency and accuracy. With the regulatory and accrediting agencies requirements of patient safety and maintaining quality outcomes, databases are an excellent tool for actively capturing patterns and trends as they emerge. Pharmaceuticals are a huge part of healthcare management and identifying the drug safety signals and mitigating them when necessary will be crucial.      

                                                                       Reference

What is a safety signal?. (2011, October). Retrieved from www.pfizer.com/files/health/..safety/2-

                4_What_is_a_Safety_Signal