Photo Credit: LuckyTD
Various sessions at the 2024 ASCO Genitourinary (GU) Cancers Symposium reviewed discrimination and disparities in healthcare among patients with prostate cancer.
Differences in Perceived Healthcare Discrimination
In one study, a team of researchers investigated the association between race/ethnicity and perceived healthcare discrimination among individuals diagnosed with prostate cancer, aiming to understand how this association impacts health outcomes. They focused on individuals with prostate cancer in the “All of Us” database and analyzed variables related to perceptions of respect, courtesy, and service quality from healthcare providers, categorized by race. The study hypothesized that racial/ethnic minorities with prostate cancer would be more likely to perceive healthcare discrimination and that such perceived discrimination would be associated with self-reported poor health.1
The study cohort consisted of 3,129 patients with prostate cancer, including patients who were of the following race/ethnicity:
- White (n=2,665);
- Black or African American (n=150);
- Hispanic/Latino (n=98);
- Other (n=171);
- Asian (n=27); and
- Multirace (n=18).1
A significant number of patients reported experiencing discrimination in healthcare settings, with Black or African American patients being significantly more likely to perceive discrimination compared with other racial/ethnic groups, according to the study.1
The analysis revealed that Black or African American patients had the highest OR for perceiving discrimination (OR=5.20; P<0.001), followed by Asian patients (OR=2.11; P=0.17), individuals of multiple races (OR=2.43; P=0.16), and Hispanic/Latino patients (OR=1.53; P=0.19). Furthermore, patients who felt respected by their healthcare provider were less likely to report worse health outcomes, while those who perceived a lack of respect had higher odds of poor health.1
Overall, the study underscores the persistent racial differences in perceived healthcare discrimination among patients with prostate cancer, highlighting the urgent need for systemic interventions in healthcare settings. The negative experiences reported by patients are correlated with worsened health outcomes, according to the study, emphasizing the importance of addressing these disparities to enhance care quality and equity.1
Using Machine Learning to Identify Racial Disparities
Another study presented at ASCO GU employed machine learning models to examine outcomes following prostate cancer surgery in a large patient cohort.
According to the study authors, machine learning offers promising avenues for predicting post-operative outcomes, which helps facilitate enhanced patient selection and risk assessment. However, persistent healthcare disparities challenge equitable health outcomes across patient demographics. This study leveraged AI and machine learning methodologies to forecast adverse outcomes following prostate cancer surgery, aiming for early intervention to optimize patient care.2
Using data from the 2016–2021 ACS National Surgical Quality Improvement Program, the study team analyzed 84,064 patients undergoing prostatectomy or related surgeries. Utilizing gradient-boosted trees, random forests, and deep learning models, the researchers scrutinized variables that included demographics, clinical data, preoperative labs, medical history, and discharge dispositions.2
Based on study criteria, the median age of the sample was 65 years, with a median hospital stay of 1 day and an operation time of 200 minutes. White non-Hispanic patients comprised 62% of the sample, whereas Black non-Hispanic patients constituted 11%. Smoking prevalence and hypertension medication requirement were higher among Black non-Hispanic patients.2
According to the findings, 30-day readmission rates varied marginally among racial groups, with minimal disparities in mortality rates. However, non-home discharge rates were notably disparate, with Black non-Hispanic patients exhibiting higher rates compared with White non-Hispanic and White Hispanic counterparts. The deep learning model yielded the highest accuracy (80.48%) in predicting discharge dispositions, with race emerging as a significant factor influencing outcomes.2
Based on the findings, this study underscores the efficacy of machine learning in identifying patient populations affected by healthcare disparities. Despite adjustments for covariates, racial disparities persist in adverse discharge dispositions post-prostate cancer surgery. The authors emphasized the necessity for targeted interventions to ameliorate disparities and improve QOL for affected patient subgroups.2
