Article Text

Download PDFPDF

13 How to diagnose, treat and eliminate racial and ethnic disparities in our healthcare systems
  1. Daniel Low
  1. University of Washington


Background Racial and ethnic disparities exist.

We can stratify processes of care and outcomes by race & ethnicity and language to identify disparities.

Objectives Identify and eliminate disparities in patients having tonsillectomy surgery.

Methods We analyzed three measures, stratified by race & ethnicity and language.

  1. Rate of prophylaxis for post-operative nausea and vomiting – PONV

  2. Rate of PONV

  3. Rate of return to the operating room with 30 days

We used IHI’s improvement methodology – PDSA cycles, SPC charts to close disparity gaps.


  1. Black patients had a lower rate of prophylaxis at our main hospital (77%) compared to Non-Hispanic White patients (90%). At our ambulatory surgery center no disparity existed, the rate of prophylaxis 98-99% across all patients. (figure 1A)

  2. PONV rates were much higher in Black patients (7.3%) at our main hospital compared to Non-Hispanic White patients (3.1%). No disparity existed at our ambulatory surgery center where the rates were 1-2% for all patients. (figure 2A)

  3. The 30-day return to OR rate across our system was 1.4% (figure 3A), the funnel plots revealed disparities for Vietnamese (3.5%), Somali (2.4%) and Spanish speaking (1.8%) patients (figure 3B). English speaking patients had a rate of 1.3%. When stratified by race and ethnicity. We found disparities in outcomes for Asian patients (2.7%), Black patient (2.4%) and Hispanic patients (1.6%). Non-Hispanic white had a rate of 0.95% (figure 3C).

ResultsA rapid series of PDSA cycles followed – annotated in (figure 3A)

  1. 98-100% of all patients now receive PONV prophylaxis (figure 1B).

  2. 3x reduction PONV rates for Black patients (3.6% to 1%), 6x reduction for Hispanic patients (2.4% to 0.3%). (figure 2B)

  3. Reduction in re-operation rate for Asian patients (2.7% to 0%). Rates for black patients and Hispanic patients are unchanged. (figure 3B and 3C)

Conclusions and Implications We can identify and eliminate disparities in treatment process and outcomes.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.