Spatial accessibility to health care plays a vital role in the evaluation of medical resource equality. A widely used method of spatial accessibility is the Two-Step Floating Catchment Area (2SFCA) method. However, the 2SFCA model (and its later variants) implicitly assumes that each doctor has the same attraction (unlimited resources) to care seekers and each care seeker's need is the same; it does not consider insurance that doctors accept or patients' different needs by age and gender. In fact, patients usually choose doctors within their insurance network and seniors and females usually have higher health care needs/demands than others. Here we present an improvement to the 2SFCA method to address these shortcomings. On the supply side, we allocate each doctor's resource equally to the insurance plans that he/she accepts. On the demand side, we adjusted the population based on their health care needs by age and gender and estimated the population holding each insurance based on the insurance's market share (assuming each insurance's market share is a reasonable representation of the population using that insurance). Next we calculate the accessibility score of each insurance plan following the 2SFCA approach and sum them at each population location as the accessibility at that location. We call the new improved approach Supply-Demand Adjusted 2SFCA. The results indicate that the SDA-2SFCA model could better reflect the actual supply and demand situation of health care and thus provide a better measure of spatial accessibility to health care. The SDA-2SFCA model can help researchers and government agencies better allocate limited resources to the neediest areas.