Medical Blog - Part 168

Instrument Development

In developing the American Medical Student Association Diversity Survey (AMSA-DS), we first conducted a review of the relevant literature using a Medline search for combinations of the following key words: “underrepresented minorities,” “medical school admissions,” “diversity,” “recruitment,” “retention” and “representation.” Included were all articles, both research studies and commentary, published between 1985 and 2000 and having any discussion of minority representation in medical schools. A summary of the findings in these articles with an annotated bibliography were presented to AMSA’s Diversity Coalition, a group of eight membership-based medical associations with a stated commitment to diversity in their mission statement (Table 1). Representatives of these organizations drafted the AMSA-DS. In the fall of 2001, the instrument was pilot-tested at four medical schools. The deans of students at each of these schools were asked to comment on content, length and clarity of the survey and survey items. Based on their feedback, the instrument was revised by modifying some questions, adding certain items and excluding other questions. The final survey contained 100 items and took approximately 20 minutes to complete [instrument available from Carrasquillo (author)]. Survey items included:

Progress and Pitfalls in Underrepresented Minority Recruitment


Despite nearly a quarter century of diversity initiatives by government agencies, medical schools and other organizations, the percentage of medical students who belong to historically underrepresented racial and ethnic minority (URM) groups has remained fairly uniform, fluctuating from 8.0% to 12.5%/ while representation of these same minority groups—African Americans, Hispanic Americans and Native Americans—in the U.S. population has reached 26% and continues to grow.

Racial and Ethnic Disparities DISCUSSION

This study used the PPOR model to categorize fetal and infant deaths in New York City, describe racial/ethnic disparities and guide interventions. During 1996-2000, the average feto-infant mortality rate was 11.5/1,000 live births plus fetal deaths for the city. Statistically significant racial/ethnic differences in feto-infant mortality were found with black non-Hispanic women experiencing a much higher rate of feto-infant mortality compared to other groups.

In New York City from 1996-2000, there were 3,443 fetal deaths (>20 weeks’ gestation), 574,476 live births and 3,188 deceased infants >500 g and were born to city residents. The feto-infant mortality rate among this population was 11.5/1,000 live births plus fetal deaths (Figure 1). Maternal health and prematurity was the largest contributing area to feto-infant mortality (5.9/1,000) followed by maternal care (3.0/1,000), infant health (1.4/1,000) and newborn care (1.2/1,000).

The PPOR model was applied to fetal and infant mortality data collected in New York City from 1996 to 2000. In accordance with the model, birthweight was imputed from gestational age where weight was missing, as described in the cited reference. Imputation was needed for 7.7% of spontaneous abortion records, 1.5% of infant death records and <1% of birth records.

Racial and Ethnic Disparities in Perinatal Mortality

Consistent with the decline seen in the early 1990s, the infant mortality rate in New York City continued to decrease over the past five years from 7.8/1,000 live births in 1996 to 6.7/1,000 in 2000, exceeding the Healthy People 2000 objective of 7.0 but still higher than the Healthy People 2010 objective of 4.5/1,000. The fetal death rate also declined during this period from 12.6/1,000 live births plus fetal deaths >20 weeks’ gestation in 1996 to 10.8/ 1,000 in 2000. However, despite the declines in infant mortality, racial/ethnic disparities persist, with the infant mortality rate and fetal death rate for black non-Hispanic women much higher than the rate for white non-Hispanic women.

Bisexual Black Men

This literature review adds key points to the dialogue about men on the down-low. First, black MSM are more likely than MSM of other races and ethnicities to identify as bisexual and to be bisexually active. Second, heterosexual identity and corresponding sexual behavior among black men are sometimes incon-gruent, but this discordance is not exclusive to nor greatest among black men. Third, black MSM are less likely than other MSM to disclose their homosexual behavior or identity, but nondisclosing black MSM may engage in fewer sexual risks with male sex partners than disclosing black MSM. Last, a large multi-site study found that more gay- or bisexually identified HIV-positive black MSM reported sex with women than heterosexually identified HIV-positive black MSM.