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).
Heck, J. L. (2018). SCREENING FOR POSTPARTUM DEPRESSION IN AMERICAN INDIAN/ALASKA NATIVEWOMEN: A COMPARISON OF TWO INSTRUMENTS. American Indian and Alaska native mental health research (Online), 25(2), 74-102.
2. What was the definition of the construct?
The construct being measured postpartum depression measured by iEPDS and PHQ9 scales in American Indian and Alaskan Native populations. In this case, the paper does not explicitly define postpartum depression but instead highlights the many changes that occur physically and mentally and postpartum and notes, “these issues complicate assessing postpartum depressive symptoms; therefore, it is important to have reliable and valid PPD screening instruments”. The American College of Obstetrics and Gynecology (ACOG) specifically defines postpartum depression as, “intense feelings of sadness, anxiety, or despair that prevent them from being able to do their daily tasks… up to 1 year after having a baby, but it most commonly starts about 1–3 weeks after childbirth”.
3. How did the authors provide evidence on the validity of the measure? Could you think of additional approaches to validating the measure?
The Heck article is a metanalysis of 58 papers which looked at Edinburgh Postnatal Depression Scale (EPDS) and the Patient Health Questionnaire-9 (PHQ-9) paired with psychometric properties or validation in detecting postpartum depression in American Indian and Alaskan Native population. For both PHQ9 and EPDS the author concurrent validity of the two tests and predictive validity. In terms of concurrent validity, they found significant discordance between the EPDS and PHQ9 (17%), suggesting they categorize postpartum depression differently from one another. Both had similar sensitivity while PHQ9 had reduced specificity. The author also looked at predictive accuracy and found that EPDS has low discriminative validity for detecting PPD in samples of women drawn from non-Western cultures (positive predictive values in non-Western samples range from 21.1-90%). It would have been helpful to also look at discriminative validity for PHQ9 however it was not reported.
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 author used internal consistency of a marker of reliability as it speaks to items on a scale being interrelated and all measuring the same attribute. They measured internal consistency by computing Cronbach's alpha (a), and values of .70 and higher are desirable. Internal consistency values were above .70 for both for both EPDS and PHQ9 and were considered acceptable by the author. It would have been helpful to also include test-retest reliability; however, this was either absent from studies analyzed, or reported without values.
5. Describe the implications of a lack of measurement validity or reliability for future research applications.
The author was able to show that EPDS and PHQ9 have appropriate concurrent validity but low predictive accuracy in Native American and Alaskan Native population. The low predictive accuracy suggests that these may not be valid for PPD screening with AI/AN women. Efforts should be use to adapt these tools to this population and further studies should be performed to validate those tools.
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)
Vilchez, G., Dai, J., Kumar, K., Mundy, D., Kontopoulos, E., & Sokol, R. J. (2018). Racial/ethnic disparities in magnesium sulfate neuroprotection: a subgroup analysis of a multicenter randomized controlled trial. The Journal of Maternal-Fetal & Neonatal Medicine, 31(17), 2304-2311. PDF to large to upload, UC link: https://search-proquest-com.ucsf.idm.oclc.org/docview/2059594566?accountid=14525
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 construct being was measured was racial inequalities of maternal side effects and neonatal birth outcomes related to magnesium administration. None of the outcomes were initially found to vary by race or ethnicity. An unplanned subgroup analysis on side effects was performed, finding increased side effects from magnesium in Hispanic patients in the group that received magnesium. The authors then compared Hispanic patient side effects to those with lowest side effects (presuming Asians although this is not specifically said, only suggested).
3. What is the evidence for the validity and reliability of the measures?
A subgroup analysis of side effects in the group the received magnesium found that Hispanic patients had highest frequency of side effects. This is only found when removing the placebo group from data.
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?
Although not explicitly described, it appears that Asian patients on magnesium were used as the reference category as they only describe the significance when “compared to the group that presented the lowest rates of side effects” and Asians are listed as having lowest side effects. This analysis is perplexing as they did not find a difference between placebo and control in Hispanic group, this analysis was post hoc and the number of Asian participants was low (n=18). I think there is a possibility the results are due to chance and data mining.
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 outcomes were presented as relative risk with no statistically significant findings in initial analysis of maternal and fetal outcomes. The secondary analysis was presented as an adjusted odds ratio. Compared to the group that presented the lowest rates of side effect maternal side effects were “highest in Hispanics [OR (95%CI) = 6.6 (1.3–33.8)], followed by Caucasians [aOR (95%CI) = 5.1 (1.1–23.9)]”. These relative measurements are appropriate as the study was looking at disparities of outcomes between races.