The ROC curve shows you the sensitivity and specificity of the lab results you entered. It does this by comparing the results in a group of patients with a group of controls. The diagnosis of patient or control must be made independently, not as a result of the lab test you are assessing.
Prism computes the ROC curve from raw data. Don't enter sensitivity and specificity directly and then run the ROC analysis.
If some people are in the wrong group, the ROC curve won't be accurate. The method used to discriminate between patient and control must truly be a gold standard.