Adapting Causal Inference Methods to Improve Identification of - PowerPoint PPT Presentation

Adapting Causal Inference Methods to Improve Identification of Healthcare Disparities Benjamin Cook, PhD MPH Director, Health Equity Research Lab Cambridge Health Alliance/Harvard Medical School healthequityresearch.org @cmmhr June 26, 2017

Identifying Health Disparities and Pathways Amenable for Interventions to Reduce Disparities 2

3

4

Quantifying Disparities and How They Arise 5 Jones CP et al. J Health Care Poor Underserved 2009

Overview Identifying healthcare disparities: applying concepts from a causal inference framework • Brief background on race and causal inference • Measuring disparity using the notion of the “counterfactual” to measure healthcare disparities. 6

Overview Identifying healthcare disparities: applying concepts from a causal inference framework • Brief background on race and causal inference • Measuring disparity using the notion of the “counterfactual” to measure healthcare disparities. 7

Adapting counterfactual methods in disparities studies The causal effect α is a difference in α = E(Y | X=x 1 ) - E(Y | X=x 2 ) outcome Y between treatment (X=x1)and control (X=x2) Difference between an individual receiving treatment and the same individual not receiving treatment. Because the individual can only take one of these values, one of these is a counterfactual. Dabady, M., Blank, R. M., & Citro, C. F. (Eds.). (2004). Measuring racial discrimination . National Academies Press. 8 Holland 1986; 2003; Rubin 1974, 1977, 1978; Pearl 2000.

Adapting counterfactual methods to disparities studies The causal effect α is a difference in α = E(Y | X=x 1 ) - E(Y | X=x 2 ) outcome Y between treatment (X=x1)and control (X=x2) Z X Y Randomized experiments and quasi-experiments at the U population level allow us to calculate average treatment Z effects that estimate this causal effect. X Randomization Y U Dabady, M., Blank, R. M., & Citro, C. F. (Eds.). (2004). Measuring racial discrimination . National Academies Press. 9 Holland 1986; 2003; Rubin 1974, 1977, 1978; Pearl 2000.

Adapting counterfactual methods to disparities studies The causal effect α is a difference in α = E(Y | X=x 1 ) - E(Y | X=x 2 ) outcome Y between treatment (X=x1)and control (X=x2) Z Randomization breaks the link between X and all other observables (Z) and unobserved variables (U) X Y except the outcome (Y) U By randomizing at the population level, we are able to infer the Z difference between the outcome if an individual received the treatment and the outcome if the X Randomization Y same individual did not receive the treatment. Remember that one of these is a counterfactual. U Dabady, M., Blank, R. M., & Citro, C. F. (Eds.). (2004). Measuring racial discrimination . National Academies Press. 10 Holland 1986; 2003; Rubin 1974, 1977, 1978; Pearl 2000.

Adapting counterfactual methods to disparities studies  For causation to occur, manipulability of the potential causal variable is required (Holland 2003)  Is race manipulable?  “Racial categories, differential perceptions and treatment of racial groups, and associations between race and health outcomes are modifiable.” Z Race??? Randomization Y U VanderWeele, T. J., & Robinson, W. R. (2014). On the causal interpretation of race in regressions adjusting for 11 confounding and mediating variables. Epidemiology , 25 (4), 473-484. see Krieger letter to editor and response.

Adapting counterfactual methods to improve identification of healthcare disparities  In disparities studies, minority race is the “treatment” of interest.  Ideally, the counterfactual group is a group identical in all aspects to the minority group except for minority race status.  “Balancing” can be achieved (i.e., videos with actors (Schulman 1999) , job applications given names typical of blacks and whites ( Bertrand and Mullainathan 2004 )).  Implementing the IOM definition of healthcare disparities requires a hypothetical group with counterfactual distributions of health status variables (Cook et al. 2009) … 12

Overview Identifying healthcare disparities: applying concepts from a causal inference framework • Brief background on race and causal inference • A framework that uses the notion of the “counterfactual” to measure healthcare disparities. 13

Unequal Treatment 1) Racial & ethnic disparities in care associated with worse outcomes, thus unacceptable 2) Disparities reflect broader inequality & discrimination in American society 3) Health systems, providers, managers & patients contribute to disparities 4) Provider uncertainty, stereotyping, & bias contribute to disparities 5) Small differences in refusal rates Institute of Medicine, do not explain disparities 2003

Defining Racial/ethnic healthcare disparities  Unpacking healthcare “disparity” to make it more relevant to practice / policy 15

Health care differences are due to many factors: • African-Americans and Latinos have lower rates of education and income.  more likely to be uninsured.  • Asians have lower rates of illicit drug and alcohol use than whites • Latinos are on average younger than whites and more likely to be in age groups that have higher prevalence of mental illness • Providers have biases that may lead to discrimination. • Hospitals and community health centers have had a legacy of racist policies Simkins v Moses H. Cone Memorial Hospital (1963), challenged the federal  government’s use of public funds to expand and maintain segregated hospital care differential harm from research, detention, involuntary commitment  16

Should differences due to all of these factors be considered a disparity? Differences due to:  Are these allowable or Income justified differences? Education Rates of Substance Use  Should the health care Age system be held accountable Geography for these differences in care? Discrimination Racism  To track progress in a way Insurance that is useful for policy, do we count all these Employment differences? Comorbidities 17

Defining Healthcare Disparity: Differences, Discrimination, and Disparity Difference The difference is due to: Clinical Need & Appropriateness & Whites Patient Preferences Healthcare Systems & Legal / Regulatory Quality of care Blacks Systems Disparity Discrimination : Bias, Stereotyping, and Uncertainty IOM Unequal Treatment 2002 18

In Unequal Treatment , the IOM made a distinction between allowable and unallowable differences Allowable / Justified Unallowable / Unfair Need for Care Discrimination (Substance abuse rates) Income Prevalence of MI Education Employment Preferences for Care Insurance The IOM Definition The IOM Definition of Healthcare Disparities of Healthcare Disparities ? Clinical Need & Clinical Need & Clinical Need & Appropriateness, Appropriateness, Appropriateness, Comorbidities Patient Preferences Patient Preferences Patient Preferences Quality of Care Quality of Care Difference Difference Healthcare Systems & Healthcare Systems & Healthcare Systems & y Legal / Regulatory Legal / Regulatory Legal / Regulatory Whites t i Geography r o Systems Systems Systems n y Blacks i t Disparity Disparity M i r o - n n Discrimination: Discrimination: Discrimination: o i M N Bias, Stereotyping, Bias, Stereotyping, Bias, Stereotyping, Legacy of racist care & Uncertainty & Uncertainty & Uncertainty IOM, 2002 19

Definition of Racial Disparities: IOM  Disparities do not include differences related to health status (clinical appropriateness and need), and patient preferences  Disparities do include differences due to SES (differential impact of healthcare systems and the legal/ regulatory climate), and discrimination.  Different than HHS definition 1,2 : “All differences among populations in measures of health and health care.” 1 Healthy People 2010 2 National Healthcare Disparities Report, 2003

Definition of Racial Disparities: IOM  Example 1: Difference overestimates disparity • Hispanics are on average younger and therefore use less medical care. This is not an “unfair” difference.  Example 2: Difference underestimates disparity • African-Americans are on average less healthy than Whites but may have very similar rates of utilization. • If Blacks were made to be as healthy as Whites, we would see much less use for Blacks compared to Whites - an “unfair” difference.

Commonly Used Disparities Methods  Typical method of measuring disparities using a regression framework from previous studies 1) y=  0 +  R RACE i +  A Age i +  G Gender i +ε 2) y=  0 +  R RACE i +  A Age i +  G Gender i +  H Health i +ε 3) y=  0 +  R RACE i +  A Age i +  G Gender i +  H Health i +  I Income i +ε   R represents a “residual direct effect”  Omitted variable bias -  R difficult to interpret  Difficult to track this coefficient (or change in coefficient) over time and across studies