Variability in the prescription of drugs with uncertain effectiveness. The case of SYSADOA in the Basque Country

Background: The majority of clinical practice guidelines do not recommend the use of SYSADOA (Symptomatic Slow Action Drugs for Osteoarthritis) for the treatment of osteoarthritis because of the lack of evidence or uncertainty around their efficacy. Nevertheless, the Spanish Public Health Service continues funding these drugs. Aim: The aim of this study is to describe the prescription status of SYSADOA in the primary care units of the Basque Country during 2011; to determine if variability exists among them; and to examine if the variability could be explained by the health care region each PC unit belongs to. Methods: Prescription data for SYSADOA during 2011 was obtained from the Basque Ministry for Health. In the Basque Country, primary care is divided into seven regions, each region consisting of several primary care units, which were used as the unit of analysis. Defined daily doses (DDD) per 1,000 inhabitant-days (DHD) were calculated. Data were standardized by sex and age using the total population of the Basque Country as the reference population. Small area statistics were calculated (extremal quotient, coefficient of variation and systematic component of variation). The influence of the region to which primary care units belonged was also analysed. R software (version R-2.15.0) was used for the analysis. Results: SYSADOA prescription during 2011 accounted for an expense of 4.5 million euros for the Basque Health Service. The crude rate of consumption of SYSADOA was 7.81 DDD per 1,000 inhabitant-days. The obtained external quotient was 13.67. The prescription of SYSADOA of the primary care units located in the 95th percentile was six times higher than the ones located in the 5th percentile. The region to which units belonged accounted for 57% of the observed variability. Discussion: The uncertainty around these drugs could be reflected in the existing variability of their prescription level. The analysis of the variability in the prescription of drugs with no demonstrated efficacy could help in allocating resources into other services or health technologies supported by evidence, thereby contributing to the improvement of health outcomes.