Hematocrit Interference Correction Algorithms

Hematocrit Interference Correction Algorithms

Hematocrit is the share of your blood volume taken up by red blood cells, and it changes what a test strip sees. Without correction, a strip reads too low when hematocrit is high and too high when it is low, because red cells slow the movement of glucose and other chemicals toward the electrode [1]. Correction algorithms measure something during the test that tracks hematocrit, such as the sample's electrical resistance or how fast a chemical crosses the strip, and then adjust the glucose result [1][2].

Why red blood cells skew a strip reading

A meter tests whole blood, but it is calibrated to report a plasma-equivalent number: the value a lab would get from the liquid part of the blood [1]. LifeScan researchers describe 42% hematocrit as the usual design point. Above it, readings drift low compared with the lab. Below it, they drift high [1].

The main cause is crowding. Red cells are barriers that glucose and the mediator (the chemical that carries electrons from the enzyme to the electrode) have to travel around, which lengthens their path and lowers the current the meter measures [1]. Thicker blood adds to the effect: the same paper notes that blood's relative viscosity roughly doubles, from about 4 to about 8, as hematocrit rises from 40% to 60% [1]. A 2011 lab study from Germany lists other explanations researchers have proposed, including red cells crowding out plasma so less of it reaches the reaction zone [2].

How large the error gets

Here's what surprised us. In a 2013 lab study of 18 meters, only 6 kept the gap between their readings at the highest and lowest hematocrit levels under the study's 10% limit, and the worst spread was 68% [3]. Where interference showed up, low hematocrit (below 35%) produced readings that were too high [3]. Two years earlier, the same research group had found 8 of 12 home-use meters failing the same 10% test [2].

Hematocrit isn't fixed, which is why this matters outside a lab. The 2013 authors cite a healthy reference range of about 30% to 50%, widening to 10% to 70% among hospital patients [3]. Smoking, prolonged exercise, high altitude, pregnancy, heavy menstrual bleeding and kidney disease are among the things they list that move it [3].

Three ways meters correct for it

LifeScan's researchers sort the fixes into two families. Compensating methods measure the sample and correct the math; design methods build the strip so red cells interfere less [1]. The sources we read describe three working approaches.

Measuring the blood's electrical properties

Red blood cells resist electric current more than plasma does, so more red cells mean more resistance between two fixed electrodes [1]. Meters probe this with alternating current (AC), which avoids the chemical reactions a steady current would cause at the electrodes. What they measure is called impedance, and the meter turns it into a correction factor for the glucose signal [1]. Our AC impedance and multi-pulse technology article goes further into the electronics.

The FDA's clearance summary for the Roche Accu-Chek Guide (2016) says only that "the sample and the environmental conditions are evaluated using AC and DC signals" [4]. Roche Diagnostics Operations was granted a patent in August 2008, US 7,407,811 B2, that claims applying an AC signal and a DC (direct current) signal to a fluid sample and combining both responses to report a concentration [5]. Its background section names hematocrit as one of the confounders the invention was meant to handle [5]. A patent shows what was claimed, not which meter uses it.

Reading the current over time (multi-pulse)

LifeScan's own researchers published the most detailed account we found. The OneTouch Verio strip sets a gold electrode facing a palladium electrode, with the glucose chemistry only on the palladium side [1]. For the first 4 seconds of the 5-second test, the meter holds the gold electrode at +300 millivolts and records current from mediator that has diffused across the gap. How fast that happens depends on hematocrit and viscosity [1].

For the final second the meter reverses the voltage, then combines the two currents as a ratio in its algorithm to cancel most of the hematocrit effect [1]. In the authors' testing at 19% to 61% hematocrit, all 2,700 results fell within 15 mg/dL or 20% of the lab value, the ISO limits the paper used [1]. FDA's 2011 clearance summary for the original OneTouch Verio system lists a 20% to 60% hematocrit claim [6]. As of October 2026, FDA's device identifier database lists that original meter as out of commercial distribution since December 2025, while OneTouch Verio Flex and Verio Reflect meters and OneTouch Verio strips are still distributed [7].

The 2011 German study tested a related idea it called dynamic electrochemistry, which applies a complex voltage pattern and analyzes the shape of the response [1]. That meter's spread was 8%, inside the 10% limit [2].

A dedicated hematocrit electrode, or a strip that keeps cells out

Nova Biomedical's StatStrip, a hospital meter, adds an electrode that measures hematocrit directly and feeds it to a correction algorithm [3]. It showed less than 2% bias across hematocrit levels in 2011 [2]. In 2013, Nova's home-use Nova Max Plus and Nova Max Link, built on the same strip technology, stayed stable at 4% and 7% [3].

The design route changes the strip itself, adding structures that keep red cells away from the sensing chemistry [1]. Abbott's FreeStyle Lite strips use coulometry, which counts the total charge produced over the reaction rather than sampling the current at one moment [8]. We cover the difference in coulometry vs. amperometry. In the 2013 study, the FreeStyle Freedom Lite narrowly missed the 10% limit at 11% [3].

What a meter's hematocrit range means

Each FDA 510(k) summary we read lists the hematocrit range the maker tested. In Abbott's 2010 FreeStyle Lite summary, samples at seven hematocrit levels were compared with the same glucose level at 40% hematocrit and with a lab analyzer, and all met a 15% acceptance limit [8]. The claimed ranges:

• Accu-Chek Guide (cleared 2016): 10% to 65% [4]
• FreeStyle Lite and FreeStyle Freedom Lite (cleared 2010): 15% to 65% [8]
• Original OneTouch Verio system (cleared 2011, no longer distributed): 20% to 60% [6][7]
• True Metrix (cleared 2014): 20% to 70% [9]

As of October 2026, the FDA recommends that all True Metrix users, including store-brand versions, switch to another way of testing when possible, because of a Class I recall over its E-5 error code [10].

A wider claimed range isn't a ranking, since each maker chose its own test levels. Someone whose hematocrit runs outside the usual range, during dialysis or pregnancy for example, has a good reason to ask about it, and which meter fits is a decision for their care team.

What we still can't see is most of the math. Of the sources we read, only LifeScan's 2012 paper laid out its correction formula [1]; the Accu-Chek Guide's FDA summary stops at "AC and DC signals" [4].