Background: Studies suggest that complementary and alternative medicine (CAM) is widely used in the European Union (EU). We systematically reviewed data, reporting research quality and the prevalence of CAM use by citizens in Europe; what it is used for, and why. Methods: We searched for general population surveys of CAM use by using Ovid MEDLINE (1948 to September 2010), Cochrane Library (1989 to September 2010), CINAHL (1989 to September 2010), EMBASE (1980 to September 2010), PsychINFO including PsychARTICLES (1989 to September 2010), Web of Science (1989 to September 2010), AMED (1985 to September 2010), and CISCOM (1989 to September 2010). Additional studies were identified through experts and grey literature. Cross-sectional, population-based or cohort studies reporting CAM use in any EU language were included. Data were extracted and reviewed by 2 authors using a pre-designed extraction protocol with quality assessment instrument. Results: 87 studies were included. Inter-rater reliability was good (kappa = 0.8). Study methodology and quality of reporting were poor. The prevalence of CAM use varied widely within and across EU countries (0.3–86%). Prevalence data demonstrated substantial heterogeneity unrelated to report quality; therefore, we were unable to pool data for meta-analysis; our report is narrative and based on descriptive statistics. Herbal medicine was most commonly reported. CAM users were mainly women. The most common reason for use was dissatisfaction with conventional care; CAM was widely used for musculoskeletal problems. Conclusion: CAM prevalence across the EU is problematic to estimate because studies are generally poor and heterogeneous. A consistent definition of CAM, a core set of CAMs with country-specific variations and a standardised reporting strategy to enhance the accuracy of data pooling would improve reporting quality.

Background: Studies suggest that complementary and alternative medicine (CAM) is widely used in the European Union (EU). We systematically reviewed data, reporting research quality and the prevalence of CAM use by citizens in Europe; what it is used for, and why. Methods: We searched for general population surveys of CAM use by using Ovid MEDLINE (1948 to September 2010), Cochrane Library (1989 to September 2010), CINAHL (1989 to September 2010), EMBASE (1980 to September 2010), PsychINFO including PsychARTICLES (1989 to September 2010), Web of Science (1989 to September 2010), AMED (1985 to September 2010), and CISCOM (1989 to September 2010). Additional studies were identified through experts and grey literature. Cross-sectional, population-based or cohort studies reporting CAM use in any EU language were included. Data were extracted and reviewed by 2 authors using a pre-designed extraction protocol with quality assessment instrument. Results: 87 studies were included. Inter-rater reliability was good (kappa = 0.8). Study methodology and quality of reporting were poor. The prevalence of CAM use varied widely within and across EU countries (0.3–86%). Prevalence data demonstrated substantial heterogeneity unrelated to report quality; therefore, we were unable to pool data for meta-analysis; our report is narrative and based on descriptive statistics. Herbal medicine was most commonly reported. CAM users were mainly women. The most common reason for use was dissatisfaction with conventional care; CAM was widely used for musculoskeletal problems. Conclusion: CAM prevalence across the EU is problematic to estimate because studies are generally poor and heterogeneous. A consistent definition of CAM, a core set of CAMs with country-specific variations and a standardised reporting strategy to enhance the accuracy of data pooling would improve reporting quality.