OBJECTIVE: The International Outcomes Survey in Dementia (IOSID) was initiated to observe the effects of current standard of care for Alzheimer's disease (AD) on patient outcomes and caregiver burden in a real-life setting.RESEARCH DESIGN AND METHODS: This 2-year, international, prospective, longitudinal and observational cohort survey involved patients with mild-to-moderate AD (Mini-Mental State Examination [MMSE] scores of 10-26 points) living in an ordinary household at baseline. There was no intervention with regard to patient management. Primary informal caregivers were also included in the survey.MAIN OUTCOME MEASURES: Patient parameters assessed included the MMSE, Disability Assessment for Dementia (DAD), Neuropsychiatric Inventory (NPI), and Clinical Global Impression (CGI). Caregiver burden was evaluated with the Zarit Burden Interview and caregiver distress was assessed as part of the NPI. Adverse events (AEs) were monitored.RESULTS: Of 2288 patients recruited, 1382 (60.4%) completed the survey. At baseline, the majority (79.3%) of patients were receiving treatment with acetylcholinesterase inhibitors (AChEIs) or/and memantine. MMSE, DAD, NPI and CGI scores all showed that patients experienced deterioration of AD symptoms during the survey. MMSE scores declined less steeply than might have been expected based on historical data. Scores on the four outcome scales were significantly correlated at all time points. Mean caregivers' feeling of strain and caregiver distress increased during the survey. AEs occurring in more than 2% of patients were nausea (3.0%), injury (2.6%), fall (2.4%), depression (2.2%) and urinary tract infection (2.2%).CONCLUSIONS: Community patients with AD experience progressive and interconnected decline in cognition, behaviour and functioning over time, placing increased burden on caregivers. However, improved care in recent years, including AChEI use, might be reflected in slower rates of patient decline than were evident in the past. Overall, relatively low rates of AEs were apparent during the survey. Limitations of this survey included a smaller than anticipated number of recruited patients confounding the possibility of performing comprehensive subgroup analyses, and the lack of randomisation inherent in the survey methodology.