RESUMEN
We randomly allocated family physicians meeting explicit entry criteria to experimental and control groups to determine whether CME affects the quality of patient care. The experimental group received educational packages, the control group did not. These educational packages were closely matched with explicit criteria used in the indicator condition-our method of measuring the quality of care. The indicator conditions were divided into elective, mandatory and hidden categories. We compared over 4,500 patient encounters, before and after the educational maneuver, with explicit clinical criteria in the indicator condition. These episodes of care were then classified according to quality of care. Though knowledge increased from the educational packages, overall quality of care improved very little. If the topics were elective, quality of care improved equally in both the experimental and control groups. When the topics were mandatory, quality of care provided by the experimental group improved (P < 0.05). Topics covered by the hidden indicator conditions showed no spillover effect.