The effect of user fees on prescribing quality in rural Nepal: two controlled pre-post studies to compare a fee per drug unit vs. a fee per drug item.

OBJECTIVE To compare prescribing quality with a fee per drug unit vs. a fee per drug item. METHODS Prescribing data were collected prospectively over 10 years from 21 health facilities in two districts of rural eastern Nepal. In 1995, both districts charged a fee per drug item. By 2000, one district was charging a fee per drug unit, and the second district continued to charge a fee per drug item (control group). By 2002, the second district was also charging a fee per drug unit. These fee changes allowed two pre-post 'cohort' with control analyses to compare INRUD/WHO drug use indicators for a fee per drug unit vs. a fee per drug item. RESULTS Charging a fee per drug unit increased the percentage of antibiotics prescribed in under-dosage by 11-12% (P = 0.02 and 0.02), decreased the percentage of patients prescribed injections by 4-6% (P = 0.002 and 0.02), reduced the units per drug item prescribed by 1.7 (P = 0.02 and 0.03), and decreased compliance with standard treatment guidelines by 11-15% (P = 0.02 and 0.06). CONCLUSION A fee per unit was associated with prescription of fewer units of drugs and fewer expensive drugs (such as injections), resulting in significantly poorer compliance with standard treatment guidelines. This finding is of great concern for public health in countries where patients are charged a fee per unit of drug.