114 CHAPTER 2.3 Some criteria from STOPP/START are rather non-specific and ambiguous. Consequently, undesirable variations in interpretation and application could emerge. In order to develop software applications based on STOPP/START version 2, these criteria need further specification. Consensus is required to define STOPP/ START version 2 more clearly [15]. The aim of the current study was to develop and provide logically unambiguous algorithms of STOPP/START criteria version 2, encoded with international disease and medication classification codes, to facilitate the development of software applications for multiple purposes. Methods The current study involved a multidisciplinary consensus and validation procedure in order to develop a specification of STOPP/START criteria version 2, encoded with international disease and medication classification codes, ultimately providing implementable coded algorithms for software applications. STOPP/START criteria For this study we used the original Irish version 2 of STOPP/START as published by O’Mahony et al. consisting of 80 STOPP and 34 START criteria [13]. Classification databases To facilitate extractions both in hospital and general practices, two widely used classification systems for coding diseases were selected: the International Classification of Disease (ICD) version 9 and 10 and the International Classification of Primary Care (ICPC) version 1 and 2 [21–23]. Medication was specified according to the Anatomic Therapeutic Chemical (ATC) classification system formulated by the World Health Organization Collaborating Center for drug statistics methodology. They were defined as either medication classes (ATC 3 and 4 level) or singular drug compounds (ATC 5 level) [24]. The Logical Observation Identifiers Names and Codes (LOINC) database was used to code laboratory values and measurements [25]. All these databases are freely accessible. Consensus procedure The multidisciplinary consensus procedure consisted of four rounds. A flowchart illustrating the procedure is shown in Figure 1.
RkJQdWJsaXNoZXIy MTk4NDMw