Clinical medicine faces many challenges, e.g. applying personalized medicine and genomics in daily practice; utilizing highly specialized diagnostic technologies; prescribing costly therapeutics. Today's population is aging and patients are diagnosed with more co-morbid conditions than in the past. Co-morbidity makes management of the elderly difficult also in terms of pharmacotherapy. The high prevalence of hypertension and diabetes as co-morbidities is indicative of the complexities that can impact accuracy in diagnosis and treatment, with poly-pharmacy being a significant component. It is essential to apply analytic methods to evaluate retrospective data to understand real world patients and medical practice. This study applies social network analysis, a novel method, to administrative data to evaluate the scope and impact of poly-pharmacy and reveal potential problems in management of elderly patients with diabetes and hypertension. Social Network Analysis (SNA) enables the examination of large patient data sets to identify complex relationships that may exist and go undetected either because of infrequent observation or complexity of the interactions. The application of SNA identifies critical aspects derived from over-connected portions of the network. These criticalities mainly involve the high rate of poly-pharmacy that results from the observation of additional co-morbid conditions in the study population. The analysis identifies crucial factors for consideration in developing clinical guidelines to deal with real-world patient observations. The analysis of routine health data, as analyzed using SNA, can be further compared with the inclusion/exclusion criteria presented in the current guidelines and can additionally provide the basis for further enhancement of such criteria.