To the Editor.
—The large study by Greenberg et al1 on the completeness of reporting acquired immunodeficiency syndrome (AIDS) in New York City provides important information on the adequacy of AIDS surveillance, but may be misleading regarding reporting by physicians. Using data from five projects, Greenberg et al estimated that overall reporting was 84% complete. However, the completeness of reporting of hospitalized cases by four hospitals (87%) was nearly identical to reporting by 14 private physicians' offices (88%). We suspect that biases in the study resulted in both an underestimate of hospital reporting and an overestimate of physician reporting.Evidence that hospital reporting was underestimated comes from a study that we conducted in Upstate New York (New York State excluding New York City), where surveillance is conducted in a fashion nearly identical to that in New York City. We assessed the completeness of AIDS reporting in a population-based sample