Hospital and Forensic Alcohol Results: Check the Specimens Before Explaining the Difference

A difference between a hospital alcohol result and a forensic laboratory result requires checking exactly which specimens were tested. A 2015 comparison of 39 cases found strong overall correlation, yet included one large discrepancy even though both laboratories tested serum in that case. The authors suspected that different samples had been analyzed; they did not prove a sample switch.

The study illustrates why a group correlation cannot establish the identity, comparability or accuracy of every individual specimen. It also separates the date of laboratory testing from the time blood was collected.

The specimens were collected while the patients were alive

Saitman and colleagues retrospectively compared hospital ethanol reports with testing at the San Diego County Medical Examiner’s Office. The specimens were collected at hospital admission before death. The forensic laboratory’s later involvement did not make the admission samples postmortem blood.

The reviewed casework included 280 hospital cases, of which 59 screened positive for ethanol. Thirty-nine had sufficient appropriate material for the comparison. This was a selected medical-examiner case series, not a representative sample of living DUI suspects.

All hospital results were from serum. At the forensic laboratory, 35 specimens were whole blood and four were serum. Thus, most comparisons changed both the laboratory and the specimen matrix.

The methods also differed

The hospital results came from clinical testing, generally using an alcohol-dehydrogenase enzymatic assay. The forensic laboratory used headspace gas chromatography with flame-ionization detection, including separate screening and quantification columns and duplicate analyses.

The laboratory received different collection-tube types and stored specimens at 4°C after arrival. The maximum reported delay between hospital and forensic testing was ten days.

These were analyses of material from the admission collection set. A later testing date is not evidence that blood was drawn from the patient at that later time. The two reports cannot therefore be used as though they measured the patient’s ongoing metabolism between laboratory testing dates.

A strong overall correlation included a conspicuous outlier

The reported coefficient of determination was R² = 0.942 across all 39 cases. Removing case 11 increased it to 0.981. Both figures describe how the paired concentrations varied across the dataset; neither guarantees agreement in every case.

For case 11, the hospital serum result was 0.340 g/dL. The forensic quantification was 0.195 g/dL, also in serum. Because both results involved serum, a serum-to-whole-blood water-content explanation does not resolve that particular discrepancy.

The investigators considered analysis of different samples the most likely explanation. Their limitations section nevertheless described a sample switch as a possibility. The paper does not document a proven identification error, establish which result was correct, or provide a general frequency of sample misidentification.

The reported average difference is not a universal correction

The authors reported that forensic results were generally lower than hospital results, with an average difference described as 14.1%. Differences were particularly noticeable in the lower concentration range.

That group finding cannot isolate one cause. Most pairs compared hospital serum with forensic whole blood; methods and calibration differed; hospital collection procedures varied; and testing was separated in time. The study was not a controlled experiment changing one variable while holding every other condition fixed.

It therefore does not justify subtracting 14.1% from every hospital result, assigning every discrepancy to enzymatic interference, or attributing every lower forensic result to specimen deterioration. A proposed explanation needs support in the actual specimen and laboratory records.

What a case review should establish

Match accession numbers and specimen identifiers across both reports. Confirm whether the laboratories tested aliquots of the same specimen, different tubes from the same draw, or samples collected at different times. Identify serum, plasma or whole blood separately, and retain each result’s original units.

Next, distinguish collection, receipt, storage and analysis times. Review the methods, calibrators, quality-control results and documented handling. Where sample identity remains uncertain, that uncertainty should remain explicit rather than being replaced by an assumed matrix conversion.

The case series did not measure driving performance or validate an impairment threshold. Nor does its suspected outlier establish that another person’s hospital sample was switched.

The site’s 2017 hospital-serum uncertainty analysis concerns uncertainty estimated within one laboratory’s analytical process. This 39-case comparison addresses a different issue: interpreting disagreement across laboratories when specimen matrix, identity, methods and handling must first be reconciled.

Primary source

Saitman A, Estrada J, Fitzgerald RL, McIntyre IM. Comparative analysis of hospital and forensic laboratory ethanol concentrations: A 15 month investigation of antemortem specimens. Journal of Forensic and Legal Medicine. 2015;33:23–27. https://doi.org/10.1016/j.jflm.2015.03.012.