From time to time, you may encounter a DWI or vehicular assault/homicide prosecution where the State is attempting to introduce a defendant’s blood alcohol level obtained from a hospital lab rather than the State Police or county forensic laboratory.

Traditionally, law enforcement obtains a blood sample for forensic analysis by responding to the hospital where the suspect is being treated. A hospital phlebotomist is directed to draw blood and the samples are eventually transported by the investigating police department to the NJSP Police Forensic Lab or a county lab to determine blood alcohol content (“BAC”).

In some instances, however, law enforcement may have been wholly occupied with roadside care taking activities as a result of a serious accident. There may not have been the time or resources for police to notice evidence of alcohol consumption at the scene. In other cases, as a result of an accident, law enforcement may not initially have reasonable suspicion or probable cause to justify a consent search or a warrant for a forensic blood draw. In all these scenarios, there may be a long delay before law enforcement is aware of alcohol involvement. When that awareness ultimately occurs, it may be too late to obtain a forensic blood draw because the defendant has been released from the hospital. In those instances, the State may be forced to rely on the hospital laboratory BAC results.

Is there a difference between this hospital BAC result and a forensic lab BAC result? If so, does it matter? More appropriately for this article–are hospital lab BAC results forensically reliable for a DWI or criminal prosecution?

What is “Hospital” Blood?

State and county forensic labs analyze whole blood. Whole blood is different than hospital blood. Therefore, different analytical techniques are used to analyze whole blood and hospital blood for the presence of ethanol.

First, what is whole blood? It is the native blood as it exists in your bloodstream.

It contains all of the traditional constituent components of blood without extraction of any of those components.

It is the only permissible expression of BAC for a per se DWI offense in New Jersey.

Hospitals do not analyze whole blood for ethanol. They analyze only a part of whole blood called serum. Serum is the fluid that remains from a blood sample after extraction and/or distillation of blood cells, platelets, and fibrins (clotting agents). After extraction of the aforementioned components, the remaining serum will have a yellow or amber hue. It is this serum that a hospital lab tests for the presence of ethanol and a myriad of other compounds.

Why do hospitals use serum instead of whole blood for their lab analysis? The answer is simple. Hospitals are charged with diagnosing and treating patients—often in emergency situations. Hospitals need to analyze blood quickly, determine what is in a patient’s system, diagnose, and treat appropriately. Therefore, hospitals need a blood test method, which is quick and simple to conduct. They need a test that requires minimal sample preparation and can determine the presence of multiple compounds at the same time. Given these objectives and parameters, hospitals choose to perform enzymatic immunoassays known as EIA tests. As we will see, scientific testing that is “fast” and “easy” usually does not satisfy the more rigorous demands of a forensic analysis.

How is Hospital EIA Testing Performed?

Serum EIA testing for the presence of ethanol and other compounds is a multi-step process. First, whole blood is drawn from the patient and taken to the hospital lab. Trichloroacetic acid (“TCA”) is added. The TCA drastically lowers the blood pH causing it to de-proteinize. The blood sample is then subjected to a vortex and centrifuge so that the solid proteins — such as blood cells that have been stripped by the TCA — can be extracted to the bottom of the test tube.

The result is a multi-layered, multi-colored sample. The top layer, called the supernatant, is the serum with its characteristic amber hue. This serum is what the hospital will analyze using its EIA process.

Serum is analyzed using an EIA analyzer, such as the ubiquitous “Siemens Dimension” clinical chemistry system. The Siemens analyzer is a spectrophotometer — a fancy scientific term for a highly sensitive, adjustable flashlight. The EIA analyzer will cast a beam of light at a selected wavelength through the “clearish” serum. At the other end of the analyzer is an absorbance meter. Beer’s Law of Optics tells us that specific compounds will absorb light at characteristic, measurable rates at certain wavelengths. Most hospital labs program their EIA analyzers to emit light in the 340nm wavelength. This is a popular wavelength because it allows the detection of substances that are critical for a hospital diagnosis, including glucose, BUN, and electrolytes.

In the 340nm wavelength, the Siemens analyzer also confirms the presence and quantity of a co-enzyme called NADH.

Why do we care about NADH? When we ingest ethanol, our body immediately recognizes it as a toxin and begins to metabolize it. Our body uses an enzyme called alcohol dehydrogenase (“ADH”) and a co-enzyme called NAD to begin the metabolic process. NADH is a by-product of this chemical metabolic reaction. The theory behind EIA ethanol testing is that if we know how much NADH there is, we can mathematically calculate how much ethanol there must have been in the metabolic process to produce the NADH by-product.

The EIA process is quick. With the minimal sample preparation set forth above, a hospital can have serum results within approximately 20 minutes. Although EIA instruments are widely accepted for clinical applications, they do have significant limitations.

The Lack of Forensic Reliability of a Hospital BAC Result

While EIA testing achieves clinical and diagnostic objectives, it is not forensically reliable for a DWI or criminal prosecution where a blood alcohol level needs to be proven beyond a reasonable doubt. The reasons for its forensic unreliability can be organized into four categories.

Lack of Specificity

Specificity” is a cardinal characteristic of any scientific testing. A forensic scientific test must be specific enough to measure what it is designed to measure, specifically, and to the exclusion of everything else in the universe. In a DWI or criminal prosecution involving blood alcohol levels, a forensic scientific test must be able to determine the presence of ethanol to the exclusion of every other substance. Hospital EIA testing cannot do this.

Here’s why. EIA tests quantify the presence of NADH. While we know NADH is the by-product of the ethanol metabolic process, it is also the by-product of other metabolic processes. This means that EIA testing is cross-reactive: EIA tests identify the presence of NADH regardless of what underlying substance may have created its presence. This can often lead to a false-positive result. A false positive is the worst scientific result. It means that ethanol may be reported when, in fact, none was present. It can also mean that an incorrect level of ethanol is being reported.

So, what else creates NADH that can cause this false positive? Lactate does.

The elimination chemistry associated with the lactate metabolic process will generate the same NADH in our blood that ethanol generates when it is being metabolized. Why do we care about lactate? Frequently, patients are brought into emergency rooms suffering a trauma, which can include shock and bleeding. When muscles are even briefly deprived of blood oxygen due to bleeding (hypoperfusion) or when our blood pH changes due to bleeding (acidosis), our body reacts by immediately producing lactate. This lactate must be metabolized. In the process, NADH is generated. This NADH is present in our bloodstream when the hospital extracts blood for an EIA analysis. The EIA testing cannot differentiate between the NADH generated during ethanol metabolism versus NADH generated during lactate metabolism. EIA testing identifies and quantifies NADH, but it cannot tell us its source. The EIA testing may be identifying the presence of NADH generated solely from shock, bleeding, and lactate metabolism. In essence, the EIA test can suggest the presence of ethanol when there is none. EIA testing can likewise suggest an ethanol level wholly inconsistent with the true level.

In addition to its endogenous presence in our body, lactate can also be introduced intravenously as part of an IV “Ringer’s Lactate Solution.” Lactate is often infused in a saline IV to act as an acidosis buffer –preventing drastic swings in blood pH during trauma and shock. This Ringer’s Lactate Solution will trigger a lactate metabolic process that produces NADH exactly like ethanol metabolism. Once again, the EIA test is only measuring the presence of NADH through light absorption and cannot distinguish the source of the NADH. Lactate can also be generated in the blood sample through the introduction of the pre-test TCA de-proteinizing solution previously mentioned. All these factors can lead to a grossly false-positive result for the presence of ethanol.

In addition to the above sources of lactate, isopropyl alcohol (rubbing alcohol) will also react with blood to produce NADH. If isopropyl alcohol is used by hospital staff to cleanse a venipuncture site prior to a blood draw, it can contaminate the sample and create NADH. In fact, Siemens specifically lists lactate and isopropyl alcohol as “interfering agents” in EIA testing.

Hemolysis can also produce lactate. Hemolysis is the rupturing of red blood cells. Red blood cells contain lactate dehydrogenase (“LDH”). This lactate dehydrogenase will trigger a metabolic process that produces NADH, just like ethanol. Hemolysis can occur when blood samples are subjected to shaking or improper phlebotomy. Improper phlebotomy includes improper needle gauge selection and/or drawing blood directly from or above an IV line. Hemodiluted blood will contain higher serum concentrations of LDH and, therefore, higher concentrations of NADH.

Here’s the bottom line: an EIA test can accurately measure the presence of NADH. It cannot determine whether the NADH was generated by ethanol, lactate, TCA, hemolysis, or some other interfering process. The high likelihood of false positives and the lack of specificity make EIA testing unsuitable for forensic purposes.

Lack of Scientific Consensus on a “Conversion Rate”

As I mentioned, hospitals will test serum, not whole blood. Serum is mostly water. Ethanol has an affinity for water since it is water-soluble. Therefore, there will be a higher concentration of ethanol in serum than in whole blood. This is problematic in a criminal prosecution since only a whole blood BAC expression is permitted.

The serum ethanol level must, therefore, be “converted” to a whole blood ethanol expression. Here’s the problem: the exact conversion rate between serum ethanol levels and whole blood ethanol levels is impossible to accurately predict.

Why? Because the amount of water in our blood is scientifically impossible to predict. The amount of water in our blood determines the conversion ratio. The amount of water in our blood is based on our hematocrit or packed cell volume. The more red blood cells (solids) we have the less volume of water in our blood.

Hematocrit between individuals can vary widely. Hematocrit in one individual can vary widely during the course of a day. Without knowing the precise hematocrit of someone’s blood, it is impossible to apply a linear conversion factor to translate a serum ethanol level to a whole blood ethanol level.

Various “conversion” studies have been performed in the last 30 years. The results vary widely, just like hematocrit levels. Some studies show a conversion range of 0.98-1.09. Other studies show a conversion rate as high as 1.59 (translation: ethanol serum is 59% higher than whole blood ethanol). If you were to attempt to include all of these studies with a data coverage factor of approximately 99%, you would be left with a 71% swing in a high-low conversion rate.

A 71% swing in the scientific world is akin to a wild guess. From the State’s perspective, it could be argued that even using the most “defense favorable” conversion factor reflects a defendant’s BAC above the per se limit when dealing with high BAC levels. However, the scientific point is lost when we devolve into a debate on the most favorable or least favorable conversion factor. The point is that there is no scientific consensus on any conversion factor. Therefore, an ethanol expression by the hospital lab is useless for forensic purposes.

Indirect Testing Method

EIA testing is also forensically unsuitable because it is an indirect testing method. Indirect testing methods are scientifically unsuitable when a robust direct method is available. Science frowns upon indirect testing methods because the likelihood of error and bias increases to unacceptable levels — particularly for forensic purposes.

What is an indirect test? If I introduced you to my dog Lucy and asked you to weigh her, you would have several options. One option would be to place her on a scale. That would be a direct measurement of Lucy’s weight. A second method would be to weigh yourself, step off the scale, then step back on the scale with Lucy in your arms. You could subtract one number from the other to “back in” to Lucy’s weight. This is an indirect method in its most rudimentary form.

When reasons for the absence of a forensic blood draw and analysis are due to dilatory investigative behavior, then an indirect method should not be tolerated as the “next best thing” — particularly since a defendant’s liberty may be at stake.

“Best Practice” Concerns

Previously, I have stressed the importance of best practices when science is given a forensic application. There are a myriad of reasons why hospital EIA testing is not an example of best forensic scientific practices.

First, there are human factors. The NJSP requires that blood alcohol results be obtained in conjunction with “quality control” samples. Quality control samples assure the proficiency of the analytical device. NJSP quality control samples generally must be within 1.5% of their mean or ± 0.002 of their target value. Those are extremely tight parameters. In the healthcare world, laboratory employees are also proficiency tested. Generally, the nationwide average only requires accuracy within 25% of the target value!

This is a huge difference from the standard under which forensic analysts must operate. In addition, states like Pennsylvania only require an overall “80% pass rate” for EIA lab employee proficiency. Bottom line: a hospital lab analyst can be wrong with 20% of their samples and, in the remaining “correct” analyses, they can miss the target by up to 25%! This may be acceptable for clinical diagnostic purposes, but it is not the precision within which our forensic labs operate, nor is it the precision mandated by proof beyond a reasonable doubt.

Another human factor: the simple act of drawing arterial rather than venous blood can change an ethanol result by up to 40%. Hospital staff is often pressed in an emergency to obtain blood from whatever port is currently available. While well meaning, this erroneous action can have a tremendous impact on the accuracy and reliability of the BAC result.

There are equipment issues, as well. Hospital labs are notorious for improper and infrequent calibration of their EIA testing devices. There are hospital calibration protocols, but they’re not often followed. This is not particularly comforting when we know that the calibration process is how a lab “teaches” the device how to recognize results. Re-use of consumables used in the testing process such as test tubes, specimen prep cups, TCA containers, and pipette tips, without a standard operating procedure that will ensure cleanliness can also wreak havoc on results.

In addition, hospitals usually perform a single test to determine the presence of compounds, including ethanol. This is less rigorous than NJSP forensic lab testing, which requires both duplicate and replicate analysis of a defendant’s blood sample for the presence of ethanol.

Finally, it is often difficult for hospitals to prove appropriate traceability of the reagents used in the EIA testing process. This is critical. If reagents that are used as a catalyst in the process or as gauges to determine results are not properly constituted, then the results derived from the use of those reagents are scientifically unreliable.

What To Do

If you are involved in a case where a hospital has provided a BAC expression, your discovery request should, at a minimum, seek the operation manual for the EIA device used, the cartridge package insert, the sample preparation procedures, an actual test printout, quality control printouts, QC package inserts, and calibration data and policies reflecting the frequency of the calibration of the device at issue. Often, these materials will bear out significant problems.

EMT, paramedic, and emergency room records should be reviewed to determine whether augmented saline was administered pre-blood draw. Those records may also describe the phlebotomy method and almost certainly the venipuncture site.

It is also important to review the reported lactate levels in the blood panel. Hospitals often test for lactate levels as a marker for shock. In the absence of reported lactate levels, myoglobin (Mb) levels should be reviewed. Myoglobin can be a marker for the presence of lactate. Mb levels ≤85ng are considered in the normal range

What The Experts and EIA Manufacturers Think

The primary manufacturer of EIA analytical devices, Siemens Industries, indicates in its cartridge package inserts and its operations manual that EIA testing is an in vitro diagnostic test that may be used only in the diagnosis and treatment of ethyl alcohol issues. This is essentially a non-forensic use disclaimer. Other EIA device manufacturers go as far in their operations manuals to indicate that the results of EIA testing are not to be used for forensic purposes. The well-known medical treatise Garriott’s Medicolegal Aspects of Alcohol concurs: “In the forensic laboratory, biochemical methods [EIA methods] are not usually utilized for determining blood alcohol concentration due to their lack of specificity…for forensic purposes, enzyme methods must be confirmed by an alternative technique.”

If that’s not enough, the godfather of blood alcohol testing over the last 40 years, Dr. Kurt Dubowski, has plainly stated in his writings that “the results of blood alcohol analysis on serum or plasma specimens should not be converted to whole blood concentration.”

While hospital blood testing has a critical role in our treatment and well-being, it has no place in the courtroom.