Missing data: Inclusion of sexual and gender identity items on a general population survey

Collection of representative data for LGBT populations is of critical importance in the effort to better understand and confront barriers to health care and other health disparities facing LGBT individuals and other sexual minorities. Although sexual identity and behavior are increasingly recognized as important social determinants of health, sexual identity and partner-gender questions remain uncommon on surveys not specifically targeting sexual behavior or risk. In addition, self-reported gender identity has rarely, if ever, been gathered on general population surveys, making this population artificially invisible. The Southeastern Pennsylvania Household Health Survey, fielded June to September 2006, is a random-digit-dial telephone survey of 10,100 households in Philadelphia and the four surrounding PA counties. Items about sexual identity, gender of sex partners in past year, gender identity, and sex at birth are presented, including demographics of this sample compared to heterosexuals. 208 survey respondents (2.1%) identified as homosexual, gay, or bisexual. Sexual identity was not always concordant with sexual behavior: for example, 23 of 91 (25.3%) men reporting exclusively male sex partners identified as heterosexual; 33 of 76 (43.4%) women reporting exclusively female sex partners identified as heterosexual. 35 respondents (0.4%) either identified as transgender, or had a current gender different than their sex at birth. Feasibility and limitations of using these types of measures and methodology will be discussed. General population surveys must include measures of sexual identity, partner-gender, and gender identity to produce a full picture of the health disparities and barriers to care for sexual minorities.