Chapter 20: Forensic Toxicology: Drugs, Poisons, and What the Blood Reveals

"All things are poison, and nothing is without poison; the dose alone makes a thing not a poison." — Paracelsus (Philippus Aureolus Theophrastus Bombastus von Hohenheim), c. 1538 — the sixteenth-century physician whose maxim became the founding principle of toxicology.

Overview

A drug found in a body is a fact. What that fact means is an argument — and the gap between the two is where this chapter lives. The instrument that identifies a compound in a vial of blood is one of the most reliable things in all of forensic science: analytical chemistry, grounded in physics, with error rates that can be measured and stated (we will meet the instruments themselves in Chapter 23). But the question a court actually asks is never "is the drug there?" It is "was this person impaired, poisoned, or killed by it — and how do you know?" That second question drags the toxicologist out of the clean world of the instrument and into the messy biology of dose, tolerance, timing, and a corpse whose own chemistry keeps changing after death. The science of detection is strong. The science of interpretation is where the trouble — and the honest work — begins.

This is the discipline the autopsy of Chapter 11 was feeding all along. When the pathologist drew that careful set of specimens — peripheral blood from a leg vein, vitreous humor from the eye, urine, gastric contents, a piece of liver — and sent them to the laboratory, they were posing a specific question about Marcus Diallo: was he incapacitated before he died? This chapter answers it. But to answer it honestly we first have to build the toolkit: which specimen answers which question, how a cheap fast screen differs from a slow definitive confirmation, what a blood alcohol number can and cannot tell you about a moment hours earlier, why "the dose makes the poison" is the most important sentence in pharmacology, and why a body that has been dead for a day can lie to you about how much drug it held in life.

Forensic toxicology is the application of the science of poisons — drugs, alcohol, and other chemicals — and their effects on the body to legal questions. We will be honest, as the whole book insists, about where it sits. At the instrument, toxicology is near the top of the validity spectrum, in the company of DNA. At the interpretation — did this level cause this death, was this driver impaired at the time of the crash — it descends into judgment, and judgment is exactly where overstatement creeps in and where a careful cross-examiner makes a confident expert squirm. By the end you will know which of those two toxicologies you are listening to.

In this chapter, you will learn to:

• State the three questions of forensic toxicology — what, how much, so what? — and say why the third is the hard one.
• Choose the right specimen — blood, urine, vitreous humor, or hair — for the question being asked, and name what each can and cannot reveal about timing and amount.
• Tell screening from confirmation, explain why an immunoassay result must never stand alone, and trace the hand-off to instrumental confirmation (Chapter 23).
• Interpret a blood alcohol concentration (BAC) honestly — absorption, elimination, retrograde extrapolation — and separate a number from impairment.
• Apply the dose makes the poison: locate a finding in a therapeutic, toxic, or lethal range, and resist the leap from present to impairing to fatal.
• Explain postmortem redistribution and the interpretive traps of postmortem toxicology, and place the field on the validity spectrum.

Learning Paths

🔎 Investigator/CSI: Your decisions at the scene and at the autopsy set the ceiling on everything the toxicologist can do. Sections 20.2 and 20.6 are yours: which specimens get collected and from where (peripheral vs. central blood is not a detail — it can change the reported number by a multiple), how they are preserved (the right tube, the right preservative, refrigeration), and why a specimen not taken is a question that can never be answered. 🧪 Lab analyst: Weight 20.3, 20.5, and 20.6. Screening-then-confirmation is the architecture of your bench; the therapeutic/toxic/lethal framework is how you turn a number into a defensible opinion; and postmortem redistribution is the trap you must account for before you write a single interpretive sentence. ⚖️ Law/courtroom: Sections 20.4 and 20.6 are where cross-examination lives — retrograde extrapolation of BAC, the gap between a blood level and impairment, the tolerance problem, and the difference between "the drug was present" and "the drug caused the death." The toxicologist's detection is hard to attack; their interpretation is the battlefield. 👥 General reader/juror: §20.1 and §20.5 are the antidote to the television version, where a number on a screen settles the case. A drug level is the beginning of a question, not the end of one. "Present" is not "impairing," and "impairing" is not "fatal" — and learning to keep those three apart is most of what this chapter offers a juror.

20.1 The three forensic-toxicology questions

Begin, as always, with the question the evidence is being asked to answer — because in toxicology there are really three, they are asked in sequence, and they get harder as you go. A great deal of bad toxicology testimony is the result of answering the third question with the confidence that belongs only to the first.

The first question is: what is present? Is there a foreign substance in this person's body, and what is it? This is an analytical-chemistry question, and it is the one toxicology answers best. Modern instruments (Chapter 23) can identify a specific compound in a biological sample with extraordinary reliability — a molecular fingerprint, matched against reference libraries, with a measurable and low error rate. When a properly run laboratory reports that a confirmed analysis identifies, say, diazepam in a blood sample, that identification is about as solid as forensic science gets. The what is the strong end of the discipline.

The second question is: how much? What was the concentration of the substance in the blood (or other specimen)? This is quantitation, and it is harder than identification but still fundamentally a measurement, performed with calibrated instruments and quality controls. A good laboratory reports a number with an associated uncertainty. The trouble is not usually the measuring; the trouble is what the number is taken to mean, which is the third question.

The third question is: so what? What does this substance, at this concentration, in this person, mean for the legal question — was the person impaired, poisoned, incapacitated, or killed by it? This is interpretation, and it is a different kind of question entirely. It is not a measurement; it is a judgment, built from the number plus a great deal of biology the number does not contain: the person's tolerance, their size and health, the time elapsed between exposure and sampling, the interaction of multiple drugs, the route of administration, and — after death — the chemistry of the corpse itself. Two people with the identical blood concentration of the identical drug can be, respectively, comfortably medicated and dead, depending on factors the toxicologist must reason about rather than read off a dial.

   THE THREE QUESTIONS OF FORENSIC TOXICOLOGY — AND WHERE THE CERTAINTY GOES
   ─────────────────────────────────────────────────────────────────────────

   Q1  WHAT is present?        ──►  identification        STRONG
       (which substance?)            (instrument + library; low, measurable error)
                                     │
   Q2  HOW MUCH?               ──►  quantitation          STRONG-ish
       (what concentration?)         (calibrated measurement, with uncertainty)
                                     │
   Q3  SO WHAT?               ──►  INTERPRETATION         JUDGMENT
       (impaired? poisoned?           (number + tolerance + timing + biology +
        killed? — the legal Q)         postmortem chemistry → an OPINION)

   ↑ certainty is HIGHEST at Q1 and DRAINS AWAY toward Q3.
   The instrument answers Q1–Q2. A human being, reasoning under uncertainty,
   answers Q3 — and that is where testimony is won or lost.

The shape of that diagram is the shape of the whole chapter, and indeed of the whole book's argument applied to one field. The strength of toxicology is real and it is concentrated at the top: a confirmed identification is trustworthy, a careful quantitation is trustworthy. But certainty drains away as you descend toward the question the court actually cares about. The honest toxicologist knows exactly where on that ladder a given statement sits, and refuses to lend the confidence of the instrument to the judgment of the interpretation.

🔬 At the Bench A useful habit, drilled into good analysts, is to separate the sentences by their source of authority. "Diazepam is present at a concentration of X" is an instrument sentence — the analyst can defend it with calibration records and confirmation spectra. "That concentration is incapacitating for most people" is a pharmacology sentence — it rests on the published literature about the drug, and its strength depends on how well this person matches the population that literature describes. "Therefore the person could not have driven away" is an inference sentence — the weakest of the three, contingent on everything the first two leave out. When a toxicologist blurs these together into one confident pronouncement, they are laundering the certainty of the first sentence onto the third. The discipline is to keep them visibly separate, on the stand and in the report.

There is also a fourth, prior question that frames the whole enterprise and that the historical toxicologist had to solve first: where do you even look, and for what? A body can contain thousands of possible substances; you cannot test for all of them at once. The toxicologist works from the case context — the scene, the medications found, the history, the autopsy findings — to decide which specimens to take and which screens to run, then follows the leads the screens generate. This is reasonable and necessary, but notice that it is also a channel for the same bias that runs through the whole book: a toxicologist told "probable overdose, found with a needle" will order a different panel than one told "found dead in a burned cabin," and a substance no one thought to look for is a substance that will not be found. We will return to this. For now, hold the architecture: what, how much, so what — strong, strong, and hard.

🔍 Check Your Understanding 1. A laboratory reports that a confirmed analysis identifies a specific opioid in a decedent's blood at a stated concentration. Which of the three questions has it answered, and which remains open? 2. Why is it fair to say the certainty of a toxicology result "drains away" from the first question to the third? Give the source of authority for each of the three kinds of statement.

20.2 Specimens: blood, urine, vitreous, hair

Toxicology is only as good as the sample it starts from, and different samples answer different questions. The choice of specimen is not a clerical matter — it determines whether you are measuring what was circulating and active at a moment, what the body has already cleared, or what was present weeks ago. A toxicologist who confuses the questions different specimens answer will draw confident conclusions from the wrong evidence. We take the four principal specimens in turn, each for what it can and cannot establish.

Blood is the specimen that speaks to the present. Because a drug in the bloodstream is, broadly, the drug that is reaching the brain and the rest of the body, the blood concentration is the measurement most closely tied to whether a substance was active — impairing, intoxicating, or toxic — at the time the blood was drawn. This is why blood is the specimen of choice for impairment questions (the drunk or drugged driver) and for cause-of-death interpretation. But blood carries a critical complication that §20.6 will develop fully: in a living person, "blood" is reasonably uniform, but in a dead person, the concentration of a drug can differ dramatically depending on where in the body the blood was drawn. Blood from a central site (the heart, the great vessels of the chest and abdomen) can be artifactually higher than blood from a peripheral site (a femoral vein in the leg), because of a postmortem process we will name shortly. This is why the pathologist in Chapter 11 was careful to draw peripheral blood from a leg vein for Diallo's toxicology — not blood pooled in the trauma-disturbed central cavity. Where the blood came from is part of the result.

Urine speaks to the recent past, and to exposure rather than effect. The kidneys concentrate drugs and especially their breakdown products for excretion, so urine often contains substances, or their metabolites, at higher concentrations and for longer windows than blood — sometimes days after the drug itself has left the bloodstream. A metabolite is a product the body makes when it chemically transforms a substance, typically en route to eliminating it; finding a metabolite tells you the parent drug was in the body and was processed, even when the parent drug is gone. This makes urine excellent for answering "was this person exposed to this drug in roughly the last few days?" — the classic workplace or probation drug-testing question. But it makes urine poor for answering "was this person impaired right now?" A positive urine screen for, say, a cannabis metabolite can reflect use days earlier, long after any impairment has passed. Urine detects exposure; it does not measure current effect. Conflating the two — treating a positive urine result as proof of impairment at a specific moment — is one of the most common interpretive errors in the field, and one a good attorney will pounce on.

Vitreous humor is the body's protected archive. The vitreous humor is the clear, gel-like fluid inside the eyeball, and it has two forensic virtues that make it disproportionately valuable in death investigation. First, it is anatomically isolated — sealed inside the eye, relatively protected from the bacterial invasion and the redistribution that disturb blood after death — so it decomposes more slowly and resists some of the postmortem artifacts that corrupt blood concentrations. Second, because drugs reach it more slowly than they reach blood, a vitreous level can sometimes lag and thereby preserve information about an earlier state. Vitreous is particularly prized for measuring alcohol (it helps distinguish genuine antemortem drinking from postmortem alcohol production — see §20.4 and §20.6) and for chemistries like glucose. Its limitations: the volume is small, not every drug partitions into it predictably, and the relationship between a vitreous concentration and the corresponding blood concentration is not one-to-one and must be interpreted with care.

Hair is the long memory, and the most over-read of the four. As hair grows — roughly a centimeter a month — substances circulating in the blood can be incorporated into the growing shaft, so a length of hair contains a rough chronological record of exposure stretching back months. Segmental analysis (testing successive segments along the hair) can, in principle, indicate a pattern of use over time, which is why hair testing appears in chronic-exposure questions, some drug-facilitated-crime investigations, and historical-poisoning cases. But hair is treacherous and is routinely overstated. External contamination — drug residue deposited on the hair from the environment or from smoke rather than absorbed from within — can produce positives that have nothing to do with ingestion, and distinguishing external contamination from internal incorporation is a genuine, unresolved difficulty. Hair growth rates vary between people and body regions; the timing inferred from segment position is approximate; and there are documented concerns about analytical bias in some hair tests (for instance, differential binding of certain drugs to certain hair types). Hair can answer "is there evidence of exposure over a long window?" It answers "how much, and exactly when?" poorly, and "was the person impaired at a moment?" not at all.

   WHICH SPECIMEN ANSWERS WHICH QUESTION? (schematic — windows are illustrative)
   ──────────────────────────────────────────────────────────────────────────────

   SPECIMEN        DETECTION WINDOW       BEST FOR                 KEY LIMIT
   ───────         ────────────────       ───────                 ─────────
   BLOOD           hours (parent drug)    impairment NOW; cause-   site matters after death
                                          of-death interpretation  (central ≠ peripheral)
   URINE           ~days (metabolites)    recent EXPOSURE          exposure ≠ impairment;
                                                                   metabolite ≠ active drug
   VITREOUS        hours–days, PROTECTED  alcohol & chemistries    small volume; not 1:1
                   (isolated in the eye)  in death cases; resists  with blood
                                          some postmortem artifact
   HAIR            WEEKS–MONTHS           chronic / historical     external contamination;
                   (~1 cm/month)          exposure pattern         timing approximate; over-read

   RULE: match the specimen to the QUESTION. "Was the drug ever in the body?"
   and "Was the person impaired at 9 p.m.?" are answered by DIFFERENT specimens.

Read that table as a discipline, not a lookup. The single most important idea in it is that the specimen constrains the question you are entitled to answer. A hair result cannot speak to a moment; a urine result cannot speak to impairment; a central-blood result cannot be trusted as a faithful measure of the antemortem concentration without accounting for where it came from. The pathologist who collects broadly — blood from a peripheral site, vitreous, urine, gastric contents, tissue — gives the toxicologist the ability to cross-check one specimen against another, which is the surest defense against the interpretive traps of §20.6. The toxicologist who is handed only one tube, from the wrong place, inherits a question they may not be able to answer honestly.

⚖️ In the Courtroom Specimen choice is a standard and effective line of cross-examination, precisely because juries assume "a drug test is a drug test." Doctor, this was a urine result, correct? Urine detects exposure over days — it does not measure whether my client was impaired at the moment of the crash, does it? You cannot tell this jury, from urine alone, that the drug was active in his bloodstream at 9 p.m.? Or, in a death case: This concentration came from heart blood, not femoral blood — and you would agree that drug concentrations in heart blood can be artifactually elevated after death by redistribution? The honest expert concedes each point and confines the testimony to what the specimen supports. The dishonest or careless one lets the jury hear "the test was positive" as if every positive answered every question — which is exactly the overstatement this book exists to name.

20.3 Screening vs. confirmation (immunoassay → GC-MS)

No competent laboratory reports a forensic toxicology result on the strength of a single test. The architecture of the bench is a two-stage process — a fast, broad, cheap screen followed by a slow, specific, definitive confirmation — and understanding why it must be two stages, and never one, is essential to reading a toxicology report honestly.

Screening is the first stage: a rapid, inexpensive, presumptive test applied to many samples (or for many drug classes) to sort the likely-negatives from the possible-positives. A screen is designed to be sensitive — to catch nearly everything that might be present — at the deliberate cost of specificity, meaning it accepts a higher rate of false positives in exchange for rarely missing a true positive. The workhorse of toxicology screening is the immunoassay: a test that uses antibodies engineered to bind a target drug or drug class, producing a measurable signal (a color change, a fluorescence, a change in light absorbance) when the target is present. Immunoassays are fast, cheap, automatable, and ideal for triage — which is exactly why hospitals, workplaces, and crime laboratories use them as the front line.

But an immunoassay has a built-in weakness that makes it unfit to stand alone as proof: it detects a class by molecular shape, and other molecules of similar shape can trigger it. This is cross-reactivity, and it is the source of the notorious false positive. An antibody raised against amphetamine may also bind certain decongestants or other structurally similar compounds; a screen may flag "opiates" on the basis of a substance that is not the one anyone cares about. The immunoassay says, in effect, "something in this sample looks like it might be in this drug class." That is a lead, not a conclusion. Treating a positive screen as a final answer is a category error — and one that has produced wrongful job terminations, wrongful arrests, and at least the appearance of guilt where none existed.

🔬 At the Bench Think of the screen as a metal detector at an airport and the confirmation as the hand search that follows. The detector is tuned to beep at anything that might be a weapon, because the cost of missing a real one is unacceptable — so it also beeps at belt buckles, keys, and hip replacements. No one is arrested on the beep. The beep triggers a specific follow-up that determines what actually set it off. A forensic immunoassay is the beep: deliberately over-sensitive, useful precisely because it rarely misses, and meaningless as proof on its own because it cannot tell a weapon from a belt buckle. The confirmation is the hand search — and only the hand search goes in the report as an identification.

Confirmation is the second stage: a different, more specific analytical method applied to the presumptive positives to identify the actual compound (and usually to quantify it). The confirmation must rest on a different chemical principle than the screen — there is no point confirming an immunoassay with another immunoassay, since a cross-reacting substance would fool both. The gold-standard confirmatory technique in toxicology couples chromatography, which separates the components of a mixture from one another, with mass spectrometry, which identifies each separated component by its molecular fingerprint. Most often this is gas chromatography–mass spectrometry (GC-MS) — we will dismantle exactly how it works in Chapter 23, but the principle to hold now is that it does not merely react to a class; it separates the specimen into its individual constituents and then identifies each one by a pattern as distinctive, for practical purposes, as a fingerprint. Where the screen says "possibly an amphetamine-class substance," the confirmation says "methamphetamine, specifically, at this concentration." The first is a presumption; the second is an identification.

   THE TWO-STAGE ARCHITECTURE OF A TOXICOLOGY RESULT
   ──────────────────────────────────────────────────────────────

   SAMPLE
     │
     ▼
   ┌──────────────────────────┐   designed SENSITIVE (catch everything),
   │ SCREEN — IMMUNOASSAY      │   not SPECIFIC. Detects a CLASS by shape.
   │ fast • cheap • presumptive│   ⚠ CROSS-REACTIVITY → false positives.
   └──────────────────────────┘   OUTPUT = a LEAD, never a conclusion.
     │
     ├──► NEGATIVE ──► (typically reported negative; case-dependent)
     │
     ▼  PRESUMPTIVE POSITIVE
   ┌──────────────────────────┐   DIFFERENT chemical principle (so a
   │ CONFIRM — GC-MS           │   cross-reactant can't fool both).
   │ slow • specific • definitive│ Separates the mixture, then IDENTIFIES
   └──────────────────────────┘   each component by molecular fingerprint.
     │                              (full mechanism: Chapter 23)
     ▼
   IDENTIFICATION (+ quantitation)  ←— ONLY this goes in the report as an ID.

   THE RULE: a presumptive screen ALONE is never a forensic identification.
   Screen to find; confirm to identify.

The presumptive-versus-confirmatory pairing is not unique to toxicology — you met it with blood in Chapter 10 (Kastle-Meyer presumes; DNA confirms) and you will meet it again with drug chemistry in Chapter 21 and instrumentally in Chapter 23. It is one of the great organizing principles of the forensic laboratory, and it encodes a piece of hard-won humility: a fast test that is allowed to be wrong sometimes must be checked by a slow test that is not. When you read a toxicology report, the first thing to find is whether the reported result was confirmed or merely screened. A confirmed GC-MS identification is among the strongest evidence in forensic science. An unconfirmed immunoassay "positive," presented as if it were an identification, is exactly the kind of overstated, sub-validated claim the NAS 2009 and PCAST 2016 reports taught us to distrust.

⚠️ Junk-Science Alert Beware the roadside or bedside presumptive result paraded as a final answer. A color-change field test or a single immunoassay strip that "comes up positive for" a drug is a screen — sensitive by design, fooled by cross-reactants, and not a confirmation. People have been arrested, jailed, and even induced to plead guilty on the strength of cheap presumptive field tests that a confirmatory laboratory analysis later showed were wrong — the substance was soap, candy, or an over-the-counter medication. (We return to the roadside field-test problem with drug chemistry in Chapter 21.) The presence of a positive screen is not the presence of a drug. The confirmation is. If a result was never confirmed, that absence is the finding — and a court should treat the unconfirmed positive as the lead it actually is, not the proof it is dressed up to be.

🔍 Check Your Understanding 1. Why must a confirmatory test rest on a different chemical principle than the screen it confirms? What would go wrong if you "confirmed" an immunoassay with a second immunoassay? 2. A workplace screen flags "amphetamines." Before this becomes an identification, what must happen, and why is the screen alone insufficient?

20.4 Alcohol: BAC, absorption, and interpretation

Alcohol is the most common, most studied, and most litigated substance in all of forensic toxicology, and it is the perfect case study in the chapter's central lesson — that the measurement is easy and the interpretation is hard. Measuring how much alcohol is in a sample is routine analytical chemistry. Saying what that number meant for a person's state at a different time is a chain of assumptions that a good attorney can attack link by link.

The BAC — blood alcohol concentration — is the amount of ethanol in the blood, conventionally expressed in the United States as grams of alcohol per 100 milliliters of blood, written as a percentage (the familiar "0.08" of the legal driving limit means 0.08 grams per 100 mL). BAC can be measured directly from a blood sample or estimated from a breath sample (the breath alcohol that breathalyzers measure, converted to an equivalent blood concentration by an assumed ratio). The number itself, properly obtained, is reliable. What it means depends on the pharmacology of how alcohol enters, distributes through, and leaves the body — and on the inconvenient fact that the body is not static between the event of interest and the moment of the test.

Alcohol pharmacology has three phases that the interpreter must keep straight:

• Absorption. After drinking, alcohol is absorbed from the stomach and especially the small intestine into the bloodstream, and the BAC rises. The rate depends on how much and how fast a person drank, whether they had eaten (food markedly slows absorption), and individual physiology. During absorption the BAC is climbing, and it has not yet reached its peak.
• Distribution. Alcohol distributes throughout the body's water. Because people differ in body water (with body size and composition), the same amount of alcohol produces different peak concentrations in different people — a key reason a "standard drink" does not produce a standard BAC.
• Elimination. The body removes alcohol at a roughly constant rate (this is the unusual feature — most drugs are cleared as a percentage of what is present, but alcohol is cleared at an approximately fixed amount per hour, so-called zero-order kinetics). Once absorption is complete and the peak is passed, the BAC declines along a roughly linear downward slope. The elimination rate varies between individuals and with chronic drinking history, which is precisely the variability that makes back-calculation uncertain.

   THE BAC CURVE — RISE, PEAK, FALL (illustrative; NOT to scale)
   ────────────────────────────────────────────────────────────────────

   BAC
    │              ●●●  ← PEAK (absorption complete)
    │           ●●     ●●
    │         ●            ●●        ELIMINATION
    │       ●  ABSORPTION     ●●     (≈ constant rate per hour;
    │      ●  (rising; food      ●●   roughly linear decline)
    │     ●    slows it)            ●●
    │    ●                            ●●
    │   ●                                ●●
    └──●─────────────┬──────────────────────●●──────────► TIME
   drink           sample A?              sample B?

   THE TRAP: the SAME measured BAC can sit on the RISING limb or the
   FALLING limb. Where the person was on this curve at the TIME OF THE EVENT
   — not the time of the test — is what matters, and a single sample can't
   tell you which limb it came from without more information.

Here is where interpretation gets genuinely hard, and where the chapter's themes converge. The legal question is almost never "what was the BAC when we drew the blood?" It is "what was the BAC at the time of the crash" — which happened earlier, sometimes much earlier. Estimating the earlier value from the later measurement is called retrograde extrapolation (back-calculation), and it works by assuming the person was in the elimination phase and adding back the alcohol presumed eliminated in the interim (the elapsed time multiplied by an assumed elimination rate). The procedure is standard, and under the right conditions it is defensible. But it rests on assumptions that may not hold:

• Was the person actually eliminating, or still absorbing, at the time of the event? If they were still on the rising limb when the crash occurred, back-calculating as though they were eliminating overestimates the BAC at the relevant moment — potentially badly. Without knowing the person's last drink and absorption state, you may not know which limb you are on.
• What was the elimination rate? It varies substantially between individuals; using a population-average rate introduces error in either direction.
• How much time elapsed, and is it known precisely? The extrapolation scales with the assumed interval.

A careful toxicologist performs retrograde extrapolation only with explicit assumptions stated, often as a range rather than a single number, and concedes the conditions under which it fails. An overreaching one offers a confident single back-calculated value as though it were a measurement — laundering, once again, the certainty of the instrument onto the uncertainty of the inference.

🔬 Read the Evidence

text FIGURE 20.1 — "One measured number, two different stories" [constructed teaching example] THE ITEM A driver's blood, drawn at the hospital 90 minutes after a single-vehicle crash, measured at a BAC of 0.10 (0.10 g/100 mL). The prosecution asks the toxicologist to state the BAC "at the time of the crash." THE CONTEXT The only hard data are the measured 0.10 and the 90-minute interval. The driver's last drink, whether they had eaten, and their absorption state at the moment of the crash are not independently known. WHAT IT SHOWS The 0.10 measurement itself is reliable analytical chemistry. Retrograde extrapolation *assuming elimination* at a typical rate would back-calculate a HIGHER value at the time of the crash (alcohol presumed eliminated in 90 minutes is added back). WHAT IT DOESN'T It does NOT, by itself, establish which limb of the curve the driver was on at the crash. If they were still ABSORBING (e.g., drank shortly before driving, on an empty road after a quick stop), the crash-time BAC could have been LOWER than 0.10, not higher — and the extrapolation would overstate it. The single sample cannot distinguish the two. THE INFERENCE The honest output is conditional and ranged: "IF the driver was past peak and eliminating, the crash-time BAC was likely somewhat above 0.10; IF still absorbing, it may have been at or below 0.10. Resolving this requires the drinking history." Stated as a defensible range with its assumptions, not a single confident number. THE LESSON A measured BAC is a fact about the moment of the draw. A crash-time BAC is an INFERENCE that depends on where the person sat on the absorption–elimination curve — and a single sample, absent the drinking history, often cannot tell you which.

Two further interpretive points complete the alcohol picture. First, the relationship between BAC and impairment is real but not rigid. Higher BAC means greater impairment on average, and at high enough levels everyone is impaired — but tolerance varies, and the correspondence between a specific number and a specific person's functional state is statistical, not deterministic. This is exactly why the law in many places sets a per se limit (a BAC at or above which driving is illegal regardless of demonstrated impairment): the legislature, recognizing that proving individual impairment from a number is hard, simply makes the number itself the offense. That legal move is a tacit admission of the toxicological truth — present at a level and impaired are related but not identical, and the law sometimes chooses the measurable one.

Second, and crucially for the cold case, alcohol in a dead body is a special interpretive minefield, because after death bacteria can produce ethanol in the body through fermentation — postmortem alcohol formation (neoformation). A BAC measured in a decomposing body may reflect drinking before death, microbial production after death, or both, and disentangling them is a genuine forensic problem (Case Study 20.2 is built on exactly this). This is one of the great virtues of vitreous humor (§20.2): because the eye is isolated and slower to decompose, comparing the vitreous alcohol to the blood alcohol helps distinguish true antemortem drinking from postmortem production. We will return to this under postmortem traps (§20.6); flag it now as the reason a "modest BAC" in a body recovered from a fire must be interpreted with care rather than read off at face value.

🧠 Cognitive-Bias Watch Alcohol interpretation is unusually exposed to contextual bias (Theme 3; Chapter 31) because the narrative so often arrives before the number. A toxicologist told "the driver reeked of alcohol and admitted drinking" may resolve every ambiguity in the extrapolation toward a higher crash-time BAC; one told "a sober-seeming driver, possibly a medical event" may lean the other way. The same elapsed-time-and-rate arithmetic can be nudged in either direction by what the analyst expects to find. The safeguard is the book's recurring one: state the assumptions explicitly, present a range, and make clear which parts of the conclusion rest on the measurement and which rest on the assumed drinking history. A back-calculation that quietly absorbed the investigator's theory is worth less than one that exposed its own assumptions to challenge.

20.5 Drugs and poisons; the dose makes the poison

We now reach the principle that founded the entire science, stated five centuries ago and never improved upon. The physician Paracelsus wrote that the dose alone makes a thing a poison — meaning there is no such thing as a substance that is poisonous or safe in itself; there is only a quantity that is one or the other. Water in excess can kill; a lethal nerve agent in a small enough dose does nothing. Everything in toxicology flows from this, and it is the single idea that most often separates honest interpretation from the television version in which a drug is simply "found" and the case is closed.

The principle is operationalized through the idea of ranges. For a great many substances, especially drugs, the literature describes characteristic concentration bands:

• The therapeutic range is the band of blood concentrations at which a drug produces its intended effect with acceptable safety — the level a properly medicated patient would show. A finding in the therapeutic range is consistent with appropriate medical use, not with poisoning.
• The toxic range is the band of concentrations at which harmful effects appear — impairment, organ stress, dangerous physiological disturbance — above the therapeutic level but not necessarily fatal.
• The lethal range is the band of concentrations associated with death. A finding in the lethal range is consistent with a fatal intoxication — though, as we are about to see, even this is an interpretation, not an automatic verdict.

   THE DOSE MAKES THE POISON — THERAPEUTIC / TOXIC / LETHAL (schematic)
   ───────────────────────────────────────────────────────────────────────

   CONCENTRATION →   low ──────────────────────────────────────────► high

   ┌───────────────┬──────────────────┬───────────────────────────────┐
   │  (sub-        │   THERAPEUTIC     │   TOXIC          │   LETHAL    │
   │ therapeutic)  │   intended effect │   harm appears   │   death     │
   │   little/no   │   "appropriate    │   "impairing /   │   "fatal    │
   │    effect     │    medical use"   │    poisoning"    │    range"   │
   └───────────────┴──────────────────┴───────────────────────────────┘
        present  ≠  impairing  ≠  fatal     ← three DIFFERENT claims

   ⚠ THE BANDS OVERLAP and SHIFT with the person. TOLERANCE can move an
   individual's whole scale to the right: a chronic user may function at a
   concentration that is in the "lethal range" for a naive person. The ranges
   are a POPULATION guide, not a per-person verdict.

Read that diagram with the chapter's discipline. The ranges are indispensable — they are how a toxicologist turns a bare number into a meaningful statement — but they are a population guide, and three cautions keep them honest. First, present is not impairing, and impairing is not fatal: these are three different claims at three different concentrations, and sliding from one to the next without the numbers to support it is the commonest overstatement in the field. A drug merely present (in or below the therapeutic range) supports nothing more than exposure. Second, the bands overlap and are blurry, drawn from population data with wide individual variation, so a single number near a boundary rarely settles the question by itself. Third — and most important — tolerance can move the entire scale. A person who uses a drug chronically can develop tolerance such that they function normally at a concentration that would incapacitate or kill someone who had never taken it. This is why a blood opioid concentration that falls in the "lethal range" of the textbooks may be a routine maintenance level in a long-term user, and why a concentration well below that range can kill a naive person, a child, or someone whose tolerance has lapsed. The number means nothing without the person.

🔬 At the Bench The hardest interpretive problems in postmortem toxicology are combination and tolerance cases, and they share a structure. In a combined drug toxicity death, no single substance is in its own lethal range, but several central-nervous-system depressants together — an opioid, a benzodiazepine, and alcohol, say — produce a fatal synergistic effect; the toxicologist must reason about the combination, because reading each drug in isolation would miss the cause of death entirely. In a tolerance case, the toxicologist must temper the textbook ranges against the decedent's history: needle tracks, a prescription record, evidence of chronic use. In both, the number is the start of the reasoning, not the end. The discipline is to ask not "is this drug above its lethal threshold?" but "what is the most defensible interpretation of these concentrations of these substances in this person, given everything known about them?" — and to state the residual uncertainty out loud.

This brings us to the specific question the autopsy handed to this chapter, and to the framing the cold case needs. The pathologist asked whether Marcus Diallo was incapacitated before death — and incapacitation is a toxic-range concept, not a lethal-range one. A sedative (a central-nervous-system depressant of the kind used to calm or induce sleep) at a concentration well above the therapeutic range, in the incapacitating (toxic) band, would render a person unable to defend themselves or escape, without necessarily being the thing that killed them. That is a different and more specific claim than "the drug killed him," and it must be made at its true strength. Hold this distinction as we approach the Case File: the toxicology's job here is not to name the cause of death (Chapter 11 established that — blunt-force trauma, with the fire set afterward) but to establish state — whether the victim was chemically incapacitated when the fatal events occurred. "Incapacitated" is a toxic-range inference about function, defensible from a concentration in the right band interpreted against the person; it is not, and must not be inflated into, a claim that the sedative was the killer.

🔍 Check Your Understanding 1. Explain "the dose makes the poison" in your own words, and use it to refute the claim "the drug was found in his blood, so it poisoned him." 2. Why can a blood opioid concentration in the textbook "lethal range" be a survivable, even routine, level in one person and rapidly fatal in another? Name the single concept that most accounts for the difference.

20.6 Postmortem redistribution and interpretive traps

We arrive at the reason postmortem toxicology is its own treacherous specialty, distinct from the toxicology of the living. Postmortem toxicology is the analysis and interpretation of drugs and poisons in the deceased — and its defining difficulty is that the body does not hold still after death. A blood concentration measured in a corpse is not a frozen photograph of the concentration at the moment of death; it is a measurement taken after a series of postmortem processes have had time to move drugs around, generate some substances, and degrade others. The toxicologist who treats a postmortem number as if it were an antemortem one — as if the body were merely a container that preserved its contents — will be confidently, sometimes catastrophically, wrong.

The central trap has a name: postmortem redistribution (PMR). After death, the barriers and gradients that the living body maintains break down. Drugs that had concentrated in organs during life — especially the liver, lungs, and heart muscle, and especially drugs that are lipophilic (fat-loving) and have a large volume of distribution (a pharmacological measure of how extensively a drug leaves the bloodstream for the tissues) — can leak back out of those organs into the nearby blood. The result is that blood drawn from a central site (the heart, the great vessels) after death can show a drug concentration substantially higher than the concentration that was actually circulating at the moment of death — sometimes several times higher. This is not a measurement error; the instrument reports the central-blood concentration accurately. It is that the central-blood concentration is no longer representative of the antemortem level. The drug has moved.

Three practices, drilled into competent death investigation, defend against PMR, and all three were visible in the Chapter 11 autopsy:

• Sample from a peripheral site. Blood drawn from a femoral vein (in the leg), ideally tied off, is far less affected by redistribution from the central organs than heart blood. This is why the pathologist drew Diallo's toxicology blood from a leg vein — a deliberate choice to obtain the most antemortem-representative specimen available.
• Cross-check multiple specimens. Comparing peripheral blood, central blood, vitreous, and liver concentrations lets the toxicologist gauge how much redistribution has occurred and bound the antemortem value. A large central-to-peripheral ratio is itself a flag that PMR is in play.
• Interpret with the drug's known redistribution behavior. Some drugs redistribute markedly; others barely at all. The toxicologist weighs the specific drug's documented tendency before converting a postmortem concentration into a statement about the antemortem level.

PMR is the headline trap, but it is one of several, and the honest postmortem toxicologist holds the whole list in view:

• Postmortem alcohol formation (neoformation), introduced in §20.4: microbial fermentation can generate ethanol in a decomposing body, so a measured BAC may overstate or wholly fabricate antemortem drinking. The vitreous-versus-blood comparison, and markers of microbial activity, help distinguish true drinking from postmortem production. In a body recovered from a fire, with thermal and decompositional changes in play, an alcohol result demands exactly this scrutiny.
• Site and timing dependence. As above, the concentration depends on where and when the sample was taken relative to death; "the BAC was X" is incomplete without the site and the interval.
• Drug instability. Some substances degrade after death (a falling concentration that understates the antemortem level) while others are generated from precursors; the toxicologist must know which.
• The tolerance problem, carried over from §20.5: even a perfectly representative postmortem concentration means nothing without the decedent's history of use.

   POSTMORTEM REDISTRIBUTION — WHY WHERE-YOU-DRAW CHANGES THE NUMBER
   ──────────────────────────────────────────────────────────────────────

   AT DEATH                          HOURS LATER (barriers break down)
   ────────                          ────────────────────────────────
   drug concentrated in              drug LEAKS from organs into the
   LIVER / LUNG / HEART              adjacent CENTRAL blood
        │                                  │
   [central blood] ≈ true            [central blood] ↑↑ ARTIFICIALLY HIGH
   circulating level                  (no longer represents death-time level)

   [femoral/leg vein] ─────────────► LESS affected → more antemortem-
   (peripheral)                       representative  ✅ preferred specimen

   THE TRAP: report a CENTRAL-blood number as if it were the antemortem
   concentration → OVERSTATE how much drug was circulating in life.
   THE FIX: peripheral site + multi-specimen cross-check + drug-specific
   redistribution knowledge. (This is why Ch. 11 drew LEG-VEIN blood.)

Step back and place the whole field on the validity spectrum, because this is the chapter's thesis stated plainly. Forensic toxicology is not one thing on the spectrum; it sits in two very different places depending on which of the three questions (§20.1) you are asking. The identification and quantitation of a substance by confirmed instrumental analysis is near the top, with DNA — grounded in analytical chemistry, validated, with measurable error rates; this is why the book's Chapter 1 table placed instrumental toxicology among the strong methods. The interpretation of what a concentration means — impairment, incapacitation, cause of death — is a judgment that descends the spectrum the further it reaches from the instrument, because it depends on tolerance, timing, postmortem chemistry, and individual variation that no instrument measures. The same expert, in the same report, may make a top-of-spectrum statement ("methamphetamine was confirmed at this concentration in peripheral blood") and a much-lower-on-the-spectrum statement ("this level was incapacitating for this person"), and the discipline — the whole discipline of this book — is to present each at its true and different strength, never letting the certainty of the first vouch for the judgment of the second.

⚖️ In the Courtroom The toxicologist's testimony splits cleanly along the §20.1 fault line, and a skilled cross-examiner aims every question at the interpretive half. The identification is nearly unassailable: a confirmed GC-MS result, properly documented, survives cross-examination because it is analytical chemistry. So the attack moves to interpretation: Doctor, this concentration came from heart blood — you'd agree postmortem redistribution can elevate central-blood concentrations several-fold? You don't know my client's tolerance, do you? You assumed an elimination rate that varies person to person? You can confirm the drug was present — but "present" is not "impairing," and you cannot tell this jury he was impaired at the specific moment, can you? The honest expert concedes what must be conceded, confines the strong claims to identification and quantitation, and frames every interpretive statement as the reasoned, bounded judgment it is. The expert who defends an interpretive leap as though it were a measurement has stepped off the science — and a good attorney will make the jury see exactly where the step occurred.

🔍 Check Your Understanding 1. Explain postmortem redistribution and why blood drawn from the heart can mislead in a way that blood drawn from a femoral vein is less likely to. Why is this an interpretation problem and not an instrument error? 2. Toxicology sits in two places on the validity spectrum at once. Name the two, give an example statement from each, and explain what keeps the strong one from vouching for the weak one.

🗂️ The Case File

The specimens come back from the autopsy. When Marcus Diallo's body was autopsied (Chapter 11), the pathologist — having already established the case's hinge finding, that he was dead before the fire (no soot in the deep airways, low carboxyhemoglobin) — drew a careful set of toxicology specimens and sent them to the state laboratory with a specific question: was this man incapacitated before he died? The specimens were collected with postmortem toxicology's traps in mind: peripheral blood from a leg vein (to limit postmortem redistribution, §20.6), vitreous humor from the eye (the protected archive, useful for the alcohol question, §20.2 and §20.4), urine, and gastric contents. The laboratory ran the two-stage process this chapter described — a broad immunoassay screen to generate leads, then instrumental confirmation (the GC-MS work that Chapter 23 will detail) to identify and quantify what the screen flagged.

What the toxicology found. Two results matter. First, a sedative — a central-nervous-system depressant — was confirmed in the peripheral blood at a concentration well above the therapeutic range, in the incapacitating (toxic) band (§20.5). Interpreted against the femoral-site sampling and the available specimens, this is a defensible statement about state: at the time of the fatal events, Diallo was most consistent with being chemically incapacitated — sedated to the point of being unable to resist or escape. Second, the blood showed a modest BAC — a low blood alcohol concentration. Because this is a body recovered from a fire, the alcohol was interpreted with the §20.6 cautions in view: the vitreous result was used to help distinguish genuine antemortem drinking from any postmortem alcohol production, and the level is reported as modest — consistent with some drinking before death, not with heavy intoxication, and not the operative finding here.

What this adds — and only this. The toxicology answers the autopsy's question and no more. It establishes that Marcus Diallo was drugged before death — incapacitated by a sedative at a toxic, incapacitating level, with a modest amount of alcohol also present. State the strength precisely, as the chapter demands: this is a strongly-supported statement about the victim's state (incapacitation), built on a confirmed identification (top of the spectrum) and a bounded interpretation of the concentration against the person and the sampling (a judgment, honestly framed). It explains something the case badly needed explaining — how a fit 38-year-old man was overcome without the kind of extensive defensive injury one might otherwise expect: he was sedated first. That is entirely consistent with the homicide the autopsy established. But notice, with the book's discipline, what the toxicology does not say. It does not name who administered the sedative, or how, or when to the minute. It does not, by itself, prove homicide — Chapter 11's airway and trauma findings did that; the toxicology corroborates and explains, it does not independently convict. And the sedative is not being called the cause of death — the cause was blunt-force trauma; the sedative is the means by which the victim was rendered defenseless. "Drugged before death" is the honest finding. "Drugged, therefore by person X" is a step the toxicology has not earned and does not take.

Who is excluded / who remains. No one is excluded or included by the toxicology alone — it speaks to the victim's state, not to any suspect. What changes is the picture of the crime: this was not a struggle a strong man might have won, but an incapacitation followed by a killing and a staged fire. That reframing matters for everything ahead — it bears on access to the sedative (who could have obtained and administered it), and it sharpens the contrast with Cody Renner's confession, which must now be tested against a crime that required sedation as well as force. Log it in the workbook (Appendix I): victim chemically incapacitated before death (sedative, toxic/incapacitating range, peripheral blood, confirmed); modest BAC, interpreted with postmortem cautions; consistent with the established homicide; perpetrator not identified by this evidence. It is a capital mistake to theorize before one has data — and the data now say the victim never had a fighting chance, while still declining to say whose hand delivered the dose.

Conclusion

Forensic toxicology is two sciences wearing one name, and the whole of this chapter has been an argument for telling them apart. The first science — identifying a substance and measuring its concentration by confirmed instrumental analysis — is among the most reliable in all of forensic science, grounded in analytical chemistry, validated, and properly placed near DNA at the top of the validity spectrum. The second science — interpreting what a concentration means for impairment, incapacitation, or death — is a judgment that depends on tolerance, timing, route, drug combinations, and the treacherous postmortem chemistry of a body that does not hold still after death. The honest toxicologist knows which of the two they are speaking at every moment, and never lends the certainty of the instrument to the uncertainty of the inference.

We built the toolkit to see that distinction. The three questions — what, how much, so what? — strong, strong, and hard. The specimens — blood for the present, urine for recent exposure, vitreous for the protected archive, hair for the long and over-read memory — each constraining the question it is entitled to answer. The two-stage architecture — a sensitive screen that produces leads and a specific confirmation that produces identifications — and the rule that a presumptive positive alone is never proof. The alcohol case study, where measuring the BAC is easy and back-calculating it to the moment of a crash is a chain of attackable assumptions. The founding principle that the dose makes the poison, operationalized through therapeutic, toxic, and lethal ranges that are population guides, not per-person verdicts, and that tolerance can slide bodily to the right. And postmortem redistribution and its sibling traps, which are why peripheral blood, multiple specimens, and drug-specific knowledge are the price of an honest postmortem interpretation.

And the toolkit answered the autopsy's question. The specimens drawn from Marcus Diallo's body — peripheral blood, vitreous, urine — returned a confirmed sedative at an incapacitating level and a modest blood alcohol concentration: he was drugged before death. That finding does not name a killer, and it is not the cause of death; it explains the how of his being overcome, and it is fully consistent with the homicide the autopsy established. The case now knows that a strong man was first made defenseless. In the next chapter we turn from what was in the body to what was in the cabin: the forensic chemistry bench, where presumptive tests suggest and instrumental tests confirm the identity of the gasoline in the fire debris — the chemistry that begins to establish how, and with what, the fire was set.

Key Terms

• Forensic toxicology — the application of the science of poisons (drugs, alcohol, and other chemicals) and their effects on the body to legal questions, comprising identification, quantitation, and interpretation.
• Postmortem toxicology — the analysis and interpretation of drugs and poisons in the deceased, distinguished by the postmortem changes (redistribution, neoformation, instability) that make a corpse's chemistry an unreliable mirror of the antemortem state unless carefully accounted for.
• Screening — the first, presumptive stage of analysis: a fast, broad, deliberately sensitive test (typically an immunoassay) that produces leads, not identifications, and must be confirmed before it counts as proof.
• Confirmation — the second, definitive stage: a more specific method resting on a different chemical principle (typically GC-MS) that identifies the actual compound and usually quantifies it; only a confirmed result is a forensic identification.
• Immunoassay — a screening test using antibodies that bind a target drug or drug class to produce a measurable signal; fast and cheap but prone to cross-reactivity (false positives from structurally similar substances), hence presumptive only.
• BAC (blood alcohol concentration) — the amount of ethanol in the blood (in the U.S., grams per 100 mL, written as a percentage); a reliable measurement whose interpretation back to an earlier moment requires the attackable assumptions of retrograde extrapolation.
• Metabolite — a product the body makes when it chemically transforms a substance (usually toward elimination); its presence indicates the parent drug was in the body and was processed, even after the parent drug itself is gone.
• Therapeutic range — the band of blood concentrations at which a drug produces its intended effect with acceptable safety; a finding here is consistent with appropriate medical use, not poisoning.
• Toxic range — the band of concentrations above therapeutic at which harmful effects (impairment, incapacitation, organ stress) appear, without necessarily being fatal.
• Lethal range — the band of concentrations associated with death; even a finding here is an interpretation (tolerance and combinations can shift it), not an automatic verdict of fatal intoxication.

Spaced Review

1. The autopsy of Chapter 11 established that Marcus Diallo was dead before the fire and asked this chapter whether he was incapacitated first. Explain how the toxicology answer ("a sedative at an incapacitating level") corroborates and explains the homicide finding without independently proving it, and state precisely why "drugged before death" stops short of naming a perpetrator. (§20.5–20.6; Chapter 11.)
2. Distinguish screening from confirmation, and explain why a confirmatory test must rest on a different chemical principle than the screen. Connect this to the presumptive-versus-confirmatory pairing you met for blood in Chapter 10. (§20.3; Chapter 10.)
3. From Chapter 1: a witness says, "The drug was in his blood, so it impaired him." Name two distinct things wrong with that sentence, using the present ≠ impairing idea and the concept of tolerance. (§20.4–20.5; Chapter 1, §1.4 and §1.6.)
4. From Chapter 4: why is it fair to say a toxicology result is only as good as the specimen and its collection, echoing the claim that the lab's quality is the ceiling on the evidence? Give one specific collection choice from the cold-case autopsy that protected the interpretation. (§20.2, §20.6; Chapter 4, Chapter 11.)
5. Validity-spectrum question. Forensic toxicology occupies two positions on the NAS 2009 / PCAST 2016 spectrum at once. Name them, give an example statement that sits at each, and explain what specific feature of the high-spectrum statement (absent from the low-spectrum one) earns it the higher position — and why letting the first vouch for the second is the error this chapter warns against. (§20.1, §20.6; and the spectrum from Chapter 1.)