Comparative Study on Prevalence of Anaemia Using Hemoglobin Meters and Fully Automated Method

A comparative study conducted at Aniniwaa Medical Centre in Kumasi found that while portable hemoglobin meters offer 100% sensitivity for anemia screening, they significantly underestimate hemoglobin levels and overestimate prevalence compared to the more accurate and reliable fully automated analyser.

The Gist

Imagine your body is a bustling city, and hemoglobin is the fleet of delivery trucks that carry oxygen to every neighborhood. When there aren't enough trucks, the city gets tired and sluggish—that's anemia.

To keep the city running, doctors need to count these trucks. But here's the problem: there are two very different ways to do the count.

1. The "Gold Standard" Lab: This is like a high-tech, automated factory with giant robots that count every single truck with perfect precision. It's accurate, but it's expensive, needs electricity, and can't be moved easily.
2. The "Pocket Counters" (Hemoglobin Meters): These are small, handheld devices, like a digital thermometer. They are cheap, portable, and can be used in a village with no electricity. But are they as accurate as the giant factory robots?

This study set out to find the answer. The researchers in Kumasi, Ghana, took blood from 100 people and ran it through both the high-tech factory robot (a fully automated analyzer) and two different types of pocket counters (called Urit and Mission).

Here is what they found, broken down simply:

1. The "False Alarm" Problem

The most surprising result was how differently the machines saw the world.

• The Factory Robot said: "Only 28% of these people are low on trucks (anemic)."
• The Pocket Counters said: "Whoa! 60% to 64% of them are low on trucks!"

The pocket counters were essentially crying wolf. They were much more likely to say someone was sick when they were actually healthy. In the study, the pocket counters had 100% sensitivity (they never missed a sick person), but they had very low specificity (they guessed that healthy people were sick).

The Analogy: Imagine a metal detector at an airport.

• The Factory Robot is a security guard who checks every bag carefully. They only flag the bags that actually have weapons.
• The Pocket Counters are like a super-sensitive metal detector that beeps at everything—keys, coins, belt buckles, and even a soda can. They catch every real weapon (great!), but they also make you take off your shoes for a soda can (annoying and wasteful).

2. The "Under-Reading" Glitch

When the researchers compared the numbers directly, they found that the pocket counters consistently underestimated the hemoglobin levels.

• If the Factory Robot said a person had 12.0 g/dL (a healthy level), the pocket counters might say 10.5 g/dL (anemic).
• This is like a scale that is slightly broken and always says you weigh 5 pounds less than you actually do. Because the numbers were lower, the machines thought more people were "underweight" (anemic) than they really were.

3. Who Was Most Affected?

The study confirmed what we already know: Women are much more likely to be anemic than men, likely due to monthly blood loss and pregnancy needs.

• The Factory Robot found that about 37% of women were anemic.
• The Pocket Counters found that 73% to 76% of women were anemic.

While the trend was the same (women are more at risk), the pocket counters made the problem look almost twice as big as it actually was.

The Big Takeaway

So, are the pocket counters useless? No!

Think of them like a smoke alarm.

• A smoke alarm is designed to be super sensitive. It's better to have a false alarm (it beeps because you burned toast) than to miss a real fire.
• Similarly, in remote villages or during mass screening campaigns, the Pocket Counters are excellent. They ensure that no one who is truly sick gets missed.

However, they are not good for making the final diagnosis. If the pocket counter says you are anemic, you shouldn't just start taking iron pills immediately. You need to go to the "Factory" (the lab) to get the Gold Standard test to confirm it. Otherwise, we might waste money and resources treating healthy people who aren't actually sick.

In short: The portable meters are great for casting a wide net to catch everyone who might be sick, but the big lab machines are needed to figure out who is actually sick.

Technical Summary

1. Problem Statement

Anaemia is a critical global public health issue, particularly affecting women of reproductive age and children. Accurate hemoglobin (Hb) estimation is essential for diagnosis and monitoring. While fully automated hematology analysers are considered the "gold standard" in clinical laboratories, they are often unavailable in resource-limited settings (such as rural Ghana) due to high costs, maintenance requirements, and the need for stable electricity. Consequently, point-of-care (POC) portable hemoglobin meters (e.g., Urit and Mission) are widely used for rapid screening.

However, there is ongoing concern regarding the accuracy and reliability of these portable devices compared to automated systems. Specifically, there is a need to determine if POC meters provide valid prevalence data or if they systematically over/underestimate anaemia, which could lead to misallocation of public health resources or incorrect clinical diagnoses.

2. Methodology

• Study Design: A comparative cross-sectional study.
• Setting: Aniniwaa Medical Centre, Kumasi, Ghana.
• Participants: 100 individuals (71 females, 29 males) visiting the laboratory for a Complete Blood Count (CBC).
• Sample Collection: Venous blood samples (approx. 2 mL) were collected into EDTA tubes. Samples were mixed and analyzed within 30 minutes.
• Methods Compared:
  1. Reference Method: Fully automated Mindray haematology analyser.
  2. Test Methods: Two portable Hb meters: URIT-12 and Mission Hb Meter.
• Statistical Analysis:
  • Descriptive statistics for demographics.
  • Chi-square tests for prevalence differences and gender/age associations.
  • Bland–Altman plots to assess agreement and bias between methods.
  • Sensitivity and Specificity calculations using the automated analyser as the reference standard.
• Anaemia Definition: Hb < 11.0 g/dL (consistent with WHO criteria for the study context).

3. Key Contributions

• Direct Comparison in Local Context: The study provides empirical data comparing two specific brands of portable Hb meters (Urit and Mission) against a fully automated analyser using venous blood (which is more reliable than capillary blood often used in field settings) in a Ghanaian healthcare setting.
• Quantification of Diagnostic Error: It moves beyond simple correlation to calculate specific diagnostic metrics (Sensitivity, Specificity, False Positives) and agreement limits (Bland–Altman), offering a clear picture of the trade-offs between portability and accuracy.
• Prevalence Discrepancy Analysis: The study highlights the magnitude of overestimation of anaemia prevalence when using POC devices compared to laboratory standards.

4. Key Results

A. Prevalence of Anaemia

There was a significant discrepancy in the prevalence of anaemia reported by the three methods:

• Fully Automated Analyser (Reference): 28% (28/100 participants).
• Urit Meter: 60% (60/100 participants).
• Mission Meter: 64% (64/100 participants).
• Observation: Both portable meters overestimated the prevalence of anaemia by more than double the reference method.

B. Diagnostic Accuracy (vs. Reference Analyser)

• Sensitivity: Both Urit and Mission meters demonstrated 100% sensitivity. They correctly identified all participants who were anaemic according to the reference method (no false negatives).
• Specificity: Both meters showed low specificity:
  • Urit: 55.6%
  • Mission: 50.0%
• False Positives: The meters generated a high number of false positives (32 for Urit, 36 for Mission), classifying non-anaemic individuals as anaemic.

C. Agreement and Bias (Bland–Altman Analysis)

• Negative Bias: Both meters consistently underestimated hemoglobin values compared to the analyser.
  • Urit Mean Bias: –1.665 g/dL
  • Mission Mean Bias: –1.55 g/dL
• Limits of Agreement (LOA): The LOAs were wide, indicating poor agreement at the individual level.
  • Urit LOA: –4.443 to +1.123 g/dL.
  • Mission LOA: –4.50 to +1.40 g/dL.
  • Implication: Individual readings could differ by up to ±4 g/dL, making the devices unreliable for precise individual diagnosis near critical cut-off points.

D. Demographic Trends

• Gender: Females had significantly higher anaemia prevalence than males across all methods (e.g., 73.2% vs 27.6% by Urit; 36.6% vs 6.9% by analyser).
• Age: Prevalence varied by age group depending on the device, but generally remained higher in reproductive-age groups.

5. Significance and Conclusion

• Clinical Utility: The study concludes that while portable Hb meters are excellent for screening due to their high sensitivity (ensuring no true anaemic cases are missed), they are not suitable as standalone diagnostic tools for confirming anaemia or determining prevalence in epidemiological surveys due to low specificity and high false-positive rates.
• Public Health Implication: Relying solely on POC meters may lead to the overestimation of anaemia burden, potentially causing inefficient allocation of resources (e.g., unnecessary iron supplementation).
• Recommendation: In resource-limited settings, POC meters should be used for initial screening, but positive results should be confirmed by a fully automated analyser or venous sample analysis to ensure accurate diagnosis and prevent misclassification. The study underscores the need for strict validation of POC devices against venous samples in local contexts.