Health Disparities Forum

Part 1:

1. Choose a paper describing the development or validation of a measure of relevance in health disparities research (please give the full citation and/or upload the paper if that's possible).

Shariff-Marco S, Breen N, Landrine H, et al. MEASURING EVERYDAY RACIAL/ETHNIC DISCRIMINATION IN HEALTH SURVEYS: How Best to Ask the Questions, in One or Two Stages, Across Multiple Racial/Ethnic Groups?. Du Bois Rev. 2011;8(1):159-177.

2. What was the definition of the construct?

The construct in this study was self-reported discrimination as measured using the Discrimination Module. The goal was to validate the construct across multiple racial/ethnic groups and to assess how to ask about discrimination experiences related to race/ethnicity.

3. How did the authors provide evidence on the validity of the measure? Could you think of additional approaches to validating the measure?

The authors used Differential Item Functioning (DIF) analysis to determine if the individual questions in the Discrimination Module performed differently across racial/ethnic groups. They demonstrated that the DIF for each individual question was small, supporting validity, and that the differences in a particular DIF were cancelled out when the questions were considered in aggregate. They used this conclude that the construct has validity and that the same measure can be used across a diverse population as long as it is inclusive of a range of discrimination experienced by the populations studied. While this approach seems valid, the study’s sample size was limited and therefore could not control for other important factors that may impact how patients respond to the module including gender and education.

4. How did the authors provide evidence on the reliability of the measure? Could you think of additional approaches to evaluating the reliability of the measure?

The authors measured inter-item reliability using Cronbach’s alpha, a measure of internal consistency for one-stage and two stage approaches, and reported 0.88 and 0.81 respectively. Both approaches were considered acceptable for a group level measurement (>0.70) but did not meet the acceptable standard for individual level assessment (>0.90).

5. Describe the implications of a lack of measurement validity or reliability for future research applications.

If there were a lack of validity or reliability in this study than it would not be wise for future use of the Discrimination Module (developed for African American subjects) to be used in diverse racial/ethnic population as we would not be able to successfully measure our predictor variable. While instrument has demonstrated and validity in a diverse population, the study was not large enough to control for important confounders/mediators (including gender and education) and therefore should be validated in a larger population.

Part 2:

1. Find a paper describing a health disparity (please give the full citation or, even better, upload the paper so everyone else can look at it too)

Le Cook B, Manning W, Alegria M. Measuring disparities across the distribution of mental health care expenditures. J Ment Health Policy Econ. 2013;16(1):3-12.

2. Summarize the construct and measurement of the dimension of disparity (e.g., racial inequalities?, SES inequalities?) and the outcome measured (e.g., self-rated health).

The authors used pooled five-year cross sectional data from the the Medical Expenditure Panel Surveys 2004-2008. They assess the association between race/ethnicity on mental health care expenditures, as defined as the sum of all direct payments in the year for mental health-related prescription drugs, inpatient or outpatient care and emergency room visits.

3. What is the evidence for the validity and reliability of the measures?

The authors do not explicitly address validity or reliability of the measures. In reviewing other sources, the AHRQ published a working paper on the validity of Medical Expenditure Panel Survey reported Medicare-covered HHA use and spending, which found no statistically significant differences in mean reported utilization or spending measures, and moderate concordance at the individual level. A separate paper compared reports from participants in MEPS with Medicare coverage to their Medical enrollment and claims data and found that household respondents accurately report inpatient hospitalizations but underreport ED and office visits, and this did appear to differ by race/ethnicity. This is a limitation to the above study in that the majority of mental health services occur in the outpatient setting and reporting of services may differ by race/ethnicity.

4. What is the reference category used for the disparity measure (ie, who is the comparison group)?  Why does this reference category make sense (or not) for this research question?

The outcome was overall expenditure and the reference category was individuals who self-identified as white. I am not sure how to define the appropriate reference category without being able to disentangle utilization. In looking at the racial disparities in health expenditures it is unclear whether this is driven by underuse or overuse of services.

5. How is the disparity quantified or measured?  Is this an absolute or relative measure or are both provided?  Describe which type of measure you would prefer for this research area, or, if both, why.

The disparity was quantified - health expenditures at each quantile of use with focus on the 95th, 97.5th and 99th quantiles (individuals with high mental health expenditures). However to fully understand this issue we also need to better understand how both health services use and expenditures varied to inform changes in allocation of services or efforts to address access.

Part 3:

1. Read someone else's response to part 1 above (identifying a construct) and comment, specifically noting whether you can see any additional implications of measurement quality for future research or whether you agree with those noted by your classmate.