Validity and reliability of a self-administered health history questionnaire.

A self-administered, health history questionnaire devised for routine use in a general medical clinic is completed without the assistance of clinic personnel and used, unedited, by the providers. The reliability and validity of the responses of 23 patients to this questionnaire were tested statistically. In our setting, more than 90 percent of the patients referred for care are capable of completing the questionnaire. The 23 patients averaged 32 minutes to complete the questionnaire. An average of 34 minutes of encounter time is required to obtain the same historical data by interview. Test-retest reliability of patients' responses to the questionnaire was 90 percent. More than 92 percent of the patients' written responses to health history items agreed with the data obtained in a blinded fashion by internists in the traditional interview. The questionnaire accurately obtains items of history frequently missing from the recorded ambulabory care data base, and in some instances obtains items of history more effectively than the interviewing physician. The study results showed a low incidence of false positive (1.8 percent) and false negative (2.8 percent) responses to questionnaire items.