The impact of patient ethnicity on cancer incidence following platelet count and C-reactive protein tests in English primary care: a cohort study of 5 million patients

This large cohort study of 5 million patients in England found that while overall cancer incidence following abnormal platelet or CRP tests was highest among White patients and similar diagnostic odds ratios were observed across ethnic groups, Black patients exhibited significantly higher diagnostic odds ratios for colorectal cancer specifically, highlighting the critical need for ethnically diverse cohorts to prevent widening healthcare inequalities.

The Gist

Imagine you are a detective trying to solve a mystery: Is a patient hiding a secret cancer?

In the UK, doctors often use two simple "clues" (blood tests) to help them find the answer:

1. Platelet Count: Think of these as the body's "construction crew." When there's a wound or inflammation, they build more of themselves. Sometimes, a hidden cancer tricks them into building too many.
2. C-Reactive Protein (CRP): Think of this as the body's "smoke alarm." When there is inflammation (like an infection or a tumor), the alarm goes off, and the levels rise.

For a long time, doctors have used these clues to decide who needs urgent checks for cancer. But a big question remained: Do these clues work the same way for everyone, regardless of their ethnicity? Or do the "rules" change depending on whether a patient is White, Asian, Black, or from another background?

This massive study, involving nearly 5 million people in England, set out to answer that question. Here is what they found, explained simply:

1. The "Smoking Gun" is Usually the Same

The researchers looked at whether high platelet counts or high CRP levels were good at predicting cancer in the year following the test.

The Good News: For the vast majority of cancers, the "clues" work just as well for Black, Asian, and White patients. If a White patient has a high platelet count, their risk of having cancer is about the same as a Black or Asian patient with the same high count. The "smoke alarm" rings with the same urgency for everyone.

2. The "Special Case" Exception (Colorectal Cancer)

However, the study found one interesting twist, like a specific type of lock that only one key fits perfectly.

When looking specifically at colorectal cancer (cancer of the colon or rectum), the "clues" were even more powerful for Black patients than for White patients.

• The Analogy: Imagine you are looking for a lost coin in a dark room. For White patients, a high platelet count is like finding a faint glimmer of light. For Black patients, that same high platelet count is like a bright flashlight. It signals a much stronger connection to colorectal cancer in this specific group.
• The study found that Black patients with high platelets were significantly more likely to have colorectal cancer than White patients with the same high platelets. The same was true for high CRP levels.

3. Why This Matters (The "One-Size-Fits-All" Trap)

The study warns us against assuming that medical rules are exactly the same for every single person, but also warns against over-complicating things.

• The Danger of Ignoring Differences: If we assume the tests work exactly the same for everyone, we might miss the fact that for Black patients, a high platelet count is a stronger warning sign for colon cancer. We might not take them seriously enough.
• The Danger of Over-Correction: On the flip side, the study found that for most cancers, the tests work the same. We shouldn't invent new, complicated rules for every single group if the basic science is the same.

4. The Big Picture: A Diverse Team is Needed

The authors use a great metaphor for the future of medicine: You can't test a new car only on a racetrack in one country and assume it will drive perfectly on every road in the world.

To make sure our "cancer detectors" (blood tests) work for everyone, we need to test them on a team that looks like the real world. This study included millions of people from diverse backgrounds, which is a huge step forward.

The Takeaway

• For most cancers: The blood tests (platelets and CRP) are reliable "smoke alarms" for everyone, regardless of ethnicity.
• For colon cancer: The alarm might be even louder for Black patients. Doctors should be extra vigilant when they see high platelets or CRP in Black patients, as it could be a very strong signal of trouble.
• The Goal: We need to keep studying diverse groups to make sure we aren't leaving anyone behind or missing clues that could save lives.

In short: The rules are mostly the same, but knowing the specific nuances for different groups helps us catch the bad guys (cancer) faster and fairer.

Technical Summary

1. Problem Statement

Despite the National Cancer Plan for England prioritizing earlier diagnosis and reducing inequalities, significant disparities persist. Patients from Asian and Black ethnic groups are frequently diagnosed at advanced stages and experience longer diagnostic intervals compared to White patients, despite generally having lower overall cancer incidence rates.

Platelet count (specifically thrombocytosis, $\ge$400$\times$10$^9$/L) and C-reactive protein (CRP, $\ge$7mg/L) are routinely used inflammatory markers in primary care to flag undiagnosed cancers. However, it is unknown whether the predictive value of these tests varies by ethnicity. Previous literature suggests baseline differences in these biomarkers across ethnic groups (e.g., higher levels in Black populations in the US), but UK-specific data on how ethnicity influences the association between abnormal test results and subsequent cancer diagnosis is lacking.

2. Methodology

Study Design:
A large-scale retrospective cohort study utilizing routinely collected primary and secondary care data in England, linked to national cancer registry data.

Data Sources:

• CPRD Aurum: Electronic primary care records.
• HES APC: Hospital Episode Statistics (Admitted Patient Care).
• NCRAS: National Cancer Registration and Analysis Service.

Study Population:

• Inclusion Criteria: Patients aged $\ge$40 years with an ethnicity record, no prior malignancy (except non-melanoma skin cancer), and a recorded platelet count or CRP test between January 1, 2010, and December 31, 2017.
• Cohort Size: 4,948,342 patients with platelet records; 811,559 patients with CRP records.
• Ethnicity Categorization: White, Asian, Black, Other, and Mixed (aligned with 2021 UK census and previous literature).

Variables:

• Exposure: First recorded blood test result.
  • Thrombocytosis: Platelet count $\ge$400$\times$10$^9$/L.
  • High CRP: CRP level $\ge$7mg/L.
• Outcome: Cancer diagnosis within one year of the test (to capture cancers present at the time of testing).
• Secondary Outcomes: Cancer stage at diagnosis (Advanced: T3/T4 or M1) and site-specific incidence (Colorectal, Lung, Oesophageal-gastric, Uterine).

Statistical Analysis:

• Model: Multi-level logistic regression models clustering patients within GP practices.
• Adjustments: Age (10-year bands), sex, smoking status, Cambridge Multimorbidity Score (CMS), Body Mass Index (BMI), Index of Multiple Deprivation (IMD), and year of test.
• Interaction Analysis: Interaction terms were included to test if the association between abnormal test results and cancer incidence differed significantly by ethnicity.
• Metrics: Adjusted cancer incidence rates and Diagnostic Odds Ratios (dOR).

3. Key Contributions

• Scale: This is the largest study of its kind, analyzing nearly 5 million patients, providing unprecedented statistical power to detect subtle ethnic variations.
• Ethnic Stratification: It moves beyond aggregate data to specifically evaluate how diagnostic performance varies across White, Asian, and Black populations in the UK context.
• Differentiation of Baseline vs. Predictive Value: The study distinguishes between baseline differences in biomarker distribution (e.g., Black patients having higher median CRP) and the predictive value of those markers for cancer detection.
• PPIE Integration: The study incorporated Patient and Public Involvement and Engagement (PPIE) representatives from diverse ethnic backgrounds to aid in result interpretation.

4. Key Results

General Cancer Incidence:

• Baseline Incidence: One-year cancer incidence was highest among White patients and lowest among Asian patients.
• Predictive Consistency: For overall cancer incidence, there were no significant interactions between ethnicity and blood test results. The diagnostic odds ratios (dOR) for thrombocytosis and high CRP were similar across all ethnic groups.
  • Example (Thrombocytosis): dOR was consistent across White, Asian, and Black groups, suggesting the test performs equally well for general cancer detection regardless of ethnicity.

Site-Specific Findings (Colorectal Cancer):

• Significant Interaction: A notable exception was found for colorectal cancer.
  • Black Patients: Showed significantly higher diagnostic odds ratios for colorectal cancer following abnormal results compared to White patients.
    • Thrombocytosis: dOR 11.1 (95% CI 7.8–15.6) for Black vs. 5.7 (5.4–6.0) for White ($p < 0.001$).
    • High CRP: dOR 4.1 (2.6–6.6) for Black vs. 2.5 (2.3–2.7) for White ($p = 0.043$).
• Other Sites:
  • Kidney and Leukaemia: Similar patterns of higher dORs for Black patients with thrombocytosis were observed, though these require cautious interpretation due to smaller sample sizes and multiple comparisons.
  • Mixed/Other Groups: Some elevated odds were seen for oesophageal-gastric (Mixed) and uterine (Other) cancers, but these were based on very small numbers and interpreted with caution.

Cancer Stage:

• There was no evidence of ethnic differences in the proportion of cancers diagnosed at an advanced stage following abnormal test results. High test results were generally indicative of advanced stage across all groups.

Biomarker Distribution:

• Platelets: Black patients aged 50+ had lower proportions of thrombocytosis compared to White and Asian patients, despite higher baseline inflammation markers in other contexts.
• CRP: Black patients showed a more gradual increase in CRP with age compared to the pronounced rise seen in White and Asian patients.

5. Significance and Implications

• Diagnostic Equity: The study suggests that for general cancer detection, current thresholds for platelet count and CRP are likely applicable across ethnic groups in the UK. This supports the use of these markers as universal tools in primary care.
• Nuanced Risk: The finding that Black patients have a higher likelihood of colorectal cancer following abnormal results suggests that while the test works, the risk profile may differ. Clinicians should maintain high vigilance for colorectal cancer in Black patients with elevated markers.
• Methodological Caution: The authors warn against simply adjusting diagnostic thresholds based on ethnicity. Such adjustments could oversimplify complex socio-economic and structural determinants of health and potentially widen inequalities.
• Future Research: The study emphasizes the need for ethnically diverse cohorts in all diagnostic validation studies. It also highlights the need to investigate why Black patients show higher odds for specific cancers (e.g., colorectal) despite similar test performance, pointing toward biological or systemic factors beyond the test itself.
• Limitations: The study acknowledges potential heterogeneity within broad ethnic categories (e.g., African vs. Caribbean) and the inability to account for the clinical rationale behind test ordering.

Conclusion:
The study concludes that while ethnic groups differ in baseline biomarker distributions and specific cancer risks, the predictive performance of platelet count and CRP for cancer detection is largely consistent across ethnicities in English primary care. However, specific site-based risks (notably colorectal cancer in Black patients) warrant targeted clinical attention to ensure equitable early diagnosis.