Chapter 20 — Key Takeaways

A scannable one-page card. For the full argument and the worked examples, see index.md.

The core claims

• Toxicology is two sciences in one name. Detection (what substance, and how much) is near the top of the validity spectrum — analytical chemistry, validated, low measurable error. Interpretation (so what: impaired? incapacitated? killed?) is a judgment that descends the spectrum the further it reaches from the instrument. The whole discipline is keeping the two apart and never letting the first vouch for the second.
• The three questions get harder as you go. What is present? (identification — strong). How much? (quantitation — strong-ish). So what? (interpretation — a human judgment built from tolerance, timing, combinations, and postmortem chemistry). Certainty drains away from the first to the third.
• Match the specimen to the question. Blood = the present (impairment, cause-of-death) but site matters after death. Urine = recent exposure over days (exposure ≠ impairment; metabolite ≠ active drug). Vitreous humor = the protected archive (isolated in the eye; resists some postmortem artifact; key for the alcohol question). Hair = weeks–months of exposure pattern, but over-read (external contamination; approximate timing).
• Screen, then confirm — and a presumptive positive alone is never proof. A sensitive immunoassay screen produces leads but is fooled by cross-reactivity; a confirmation on a different chemical principle (GC-MS, Chapter 23) identifies the actual compound. Only a confirmed result is an identification.
• Alcohol: easy to measure, hard to interpret. BAC is reliable analytical chemistry; back-calculating it to an earlier moment (retrograde extrapolation) is a chain of attackable assumptions — which limb of the absorption–elimination curve was the person on? A per se limit exists precisely because present at a level and impaired are related but not identical.
• The dose makes the poison. A finding sits in a therapeutic, toxic, or lethal range — but the ranges are population guides that overlap and shift, and tolerance can slide an individual's whole scale to the right. Present ≠ impairing ≠ fatal — three different claims at three different concentrations.
• The body does not hold still after death. Postmortem redistribution can inflate central-blood concentrations above the antemortem level; postmortem alcohol formation can manufacture ethanol. The defenses are interpretive, not instrumental: peripheral (femoral) blood, multi-specimen cross-check, and drug-specific knowledge.

The method-validity verdict (NAS 2009 / PCAST 2016)

Where they sit: confirmed identification/quantitation is with DNA at the top; interpretation of effect sits well below, because it rests on individual variables no instrument measures; an unconfirmed immunoassay "positive" is presumptive only and must not be presented as proof.

What you can honestly say on the stand

• Identification (strong): "Diazepam was confirmed by GC-MS in peripheral (femoral) blood at a concentration of [N]. That identification rests on analytical chemistry and is documented in the confirmation record."
• Interpretation (a bounded judgment): "That concentration is in the toxic/incapacitating range for the general population; whether it incapacitated this individual depends on tolerance and timing I cannot fully characterize, so I state it as strongly consistent with incapacitation, not as a certainty."
• The cold case: "A sedative was confirmed at a toxic, incapacitating concentration in peripheral blood, with a modest blood alcohol level interpreted with postmortem cautions; this strongly supports that the victim was chemically incapacitated before death. It does not identify who administered the sedative, does not establish it as the cause of death, and does not by itself prove homicide."
• What you must NOT say: that a confirmed presence proves impairment at a specific past moment; that a central-blood postmortem level equals the antemortem level; that a single back-calculated BAC is a measurement; that an unconfirmed screen is an identification; or that "drugged before death" names a perpetrator.

Key terms (one line each)

• Forensic toxicology — applying the science of poisons (drugs, alcohol, chemicals) to legal questions: identification, quantitation, interpretation.
• Postmortem toxicology — toxicology of the deceased, complicated by redistribution, neoformation, and instability that make a corpse's chemistry an unreliable mirror of life.
• Screening / confirmation — a sensitive presumptive test (immunoassay) that produces leads, then a specific definitive test (GC-MS) on a different principle that produces identifications.
• Immunoassay — an antibody-based screening test; fast and cheap but prone to cross-reactivity, hence presumptive only.
• BAC (blood alcohol concentration) — ethanol in the blood (U.S.: g/100 mL); reliably measured, but its back-calculation to an earlier moment rests on attackable assumptions.
• Metabolite — a product the body makes when transforming a substance; indicates the parent drug was present and processed, even after it is gone.
• Therapeutic / toxic / lethal range — population bands of concentration for intended effect, harm, and death; overlapping and shifted by tolerance, not per-person verdicts.

The cold-case line

The toxicology establishes that Marcus Diallo was drugged before death — a sedative at an incapacitating (toxic) level in peripheral blood, plus a modest BAC interpreted with postmortem cautions. It explains how a fit 38-year-old was overcome and is consistent with the homicide the autopsy established (Chapter 11) — but it names no perpetrator, is not the cause of death, and does not independently prove homicide. A confirmed identification (top of the spectrum) plus a bounded interpretation (a judgment). One more honest brick in the wall (Chapter 39).

The themes this chapter advanced

• The validity spectrum (Theme 2) — the chapter's spine: toxicology sits in two places at once (strong detection, judgment-level interpretation), and the discipline is to present each at its true, different strength.
• Exclusion over proof (Theme 1) — "present ≠ impairing ≠ fatal"; the toxicology corroborates and explains the homicide without proving it or naming anyone; honest verbs throughout.
• (Also advanced: cognitive bias (Theme 3) — the narrative-before-the-number problem in BAC and overdose framing, §20.4 Cognitive-Bias Watch; and the CSI effect (Theme 4) — the television fantasy that a drug level on a screen settles a case, §20.1, §20.5.)