OBJECTIVES: To assess the geographical inequities of per capita consumption for 4 product categories, ATC groups: B01AE direct thrombin inhibitors & B01AF direct factor Xa inhibitors (anticoagulants); G03G gonadotropins and other ovulation stimulants (fertility); OTC groups: 86H2 hair loss (alopecia); and 13A sleep without prescription (sleep).

METHODS: An ecological study was conducted using aggregated information on product sales, mean income, age groups and gender. Information was retrieved from the IQVIA sell-in database, which is provided with a wholesalers’ panel that cover 99% of the market and regularly provides the sales units distributed to retail pharmacies. This database was used as it has the best coverage rate, with information provided for each of the 2,568 sanibricks, which are geographic territories that comprise the area of influence of the Primary Care centers of Spain.

To assess the inequity in consumption an ecological correlation was estimated to assess the linear relationship between per capita consumption and socioeconomic characteristics in each geographical territory. The Pearson correlation coefficients were used to estimate the linear relationship between two continuous variables. A sensitivity analysis was done using the Spearman correlation to evaluate non-linear, monotonic relationship between each of the pair of variables.

RESULTS: During a 3-year period the mean (standard deviation, minimum-maximum) units of consumption per capita for each product category was: anticoagulants 0.56 (0.14, 0.08-1.92), fertility 0.02 (0.03, 0.00-0.57), alopecia 0.10 (0.07, 0.01-1.58) and sleep 0.41 (0.24, 0.07-2.58). A moderate positive correlation (0.669, p<0.001) between anticoagulants and proportion of ≥65 years old, and low correlation (0.42, p<0.001) between fertility and mean income were observed. Interestingly a high positive correlation (0.74, p<0.001) between alopecia and sleep consumption was also observed.

CONCLUSIONS: Low positive correlations of consumption with income level were observed for the 4 selected categories, and a moderate correlation with age for one of the categories.