Therapeutic Duplication¶

The medication-safety failure where uncoordinated prescribers each place a defensible order that lands on the same pharmacologic target, so additive exposure overruns the therapeutic window — a harm that lives in the set of orders, not in any single one.

Core Idea¶

Therapeutic duplication is the medication-safety failure mode in which two or more agents acting on the same pharmacologic target are prescribed to the same patient in parallel — without any single prescriber having intended the combination — so that the patient receives additive dose, additive side effects, or both beyond what any individual order contemplates. The defining structural character is that the pathology is not localised in any single prescription: each individual order may be clinically defensible in isolation, correctly dosed, and appropriately indicated. The defect lies in the set of concurrent orders evaluated as a whole against a shared pharmacologic target class. In clinical practice the mechanism arises through fragmented prescribing authority: a home medication established by an outpatient provider, an admitting order placed by a hospitalist unfamiliar with the home list, a cross-cover order added by a nocturnist, and a combination product (acetaminophen in a scheduled analgesic and separately in a cold remedy, two different NSAIDs on the same list) each contribute to a cumulative exposure no one intended and no single order reveals. The resultant harm profile is additive rather than synergistic in the pharmacokinetic sense: the shared target receives more drug than the therapeutic window accommodates. Within medication safety, pharmacy practice, and electronic prescribing, therapeutic duplication is the standard category name for this set-level coordination failure, and it demarcates a distinct intervention target — the medication list reviewed as an integrated set, not as a collection of individual prescriptions — calling for reconciliation at every care transition, pharmacologic-class scanning rather than interaction-pair scanning, and decision-support logic that can identify same-class co-prescribing across multiple ordering providers.

Structural Signature¶

Sig role-phrases:

the shared pharmacologic target — a single drug class or receptor that more than one agent acts upon
the parallel prescribers — multiple ordering providers (home, admitting, cross-cover) authoring agents with no coordination point and no view of each other's orders
the medication list as an integrated set — the harm-bearing object: not any single order but the full concurrent set read against its target classes
the additive exposure — cumulative dose, side effects, or both compounding on the one target beyond its therapeutic window, with each order blameless in isolation
the missing reconciliation step — the absence of any point where the set is finally read end-to-end (only present at care transitions)
the intent disambiguation — the one-bit test separating intended combination therapy (one coordinating mind) from unintended duplication (fragmented authority), since both look identical on the list
the class-scan remedy — group by shared target, count agents per class, decompose combination products into active ingredients, and detect same-class co-prescribing across providers

What It Is Not¶

Not a single bad prescription. No individual order need be wrong — each can be correctly dosed, appropriately indicated, and defensible in isolation. The harm lives in the set of concurrent orders read against a shared target class, so an audit that inspects prescriptions one at a time finds nothing and misses the overexposure entirely.
Not intended combination therapy. When one clinician deliberately stacks same-class agents for benefit, the list looks identical — two agents on one target — but that is a defensible regimen, not duplication. What makes it duplication is the absence of a coordinating mind: the stacking is an artifact of fragmented prescribing authority that no one assembled on purpose.
Not a pharmacokinetic interaction. Duplication can harm with no drug-drug interaction at all; additive action of two agents on the same target suffices to push past the therapeutic window. The question is not "do these drugs interact?" but "is more than one agent already acting on this target?" — a class-scan, not an interaction-pair check.
Not designed redundancy. Deliberate backup of a function for resilience is engineered and intended; therapeutic duplication is undesigned coincidence — overlapping orders that happen to land on one target with no one having planned the overlap. The resemblance is only that both put more than one agent on a job.
Not pharmacologic synergy. This is not two agents potentiating each other beyond the sum of their effects at the receptor or pathway level. The exposure here is straightforwardly additive on a shared target — more drug, not amplified drug — and that simple summation is what overruns the window.

Scope of Application¶

Therapeutic duplication lives across the prescribing and medication-review subfields of medication safety and pharmacy practice; its reach is within that domain, and the loose analogues in other systems (redundant security controls, overlapping subsidies) belong to the parent primes the concept instantiates — coordination failure and set-level configuration harm — not to the medical label.

Medication reconciliation at care transitions — the canonical fix-point: admission, transfer, and discharge are where the home list, admitting orders, and cross-cover additions are finally read end-to-end as one set, so same-class stacking can be detected and resolved.
E-prescribing clinical decision support — duplicate-therapy alerts fire at order entry when a new agent shares a pharmacologic class with one already on the list, the automated form of the class-scan across multiple ordering providers.
Hospital pharmacy review — the inpatient pharmacist verifying an order set against the active medication profile catches additive same-target exposure (two NSAIDs, three benzodiazepine-receptor agents) that no single prescriber assembled.
Geriatric polypharmacy audit — deprescribing and polypharmacy review in patients on a dozen-plus agents is a habitat where unintended same-class duplication is both common and high-harm, surfaced by grouping the list by target class.
Oncology regimen reconciliation — confirming that supportive-care and chemotherapy orders do not redundantly stack agents on a shared target across the multiple services co-managing a cancer patient.
Emergency and cross-cover prescribing — the setting that generates duplication through fragmented authority, where a nocturnist or cross-cover provider adds an agent without a view of the full list, and the same setting where catching it matters.
Combination-product surveillance — scanning for active ingredients hidden inside combination products (acetaminophen in both a scheduled analgesic and a cold remedy, an NSAID buried in a combination), the decomposition step that surfaces a second agent the list does not name outright.

Clarity¶

Naming therapeutic duplication makes legible a harm that order-by-order review structurally cannot see. When each prescription is correctly dosed, indicated, and defensible in isolation, an audit that inspects prescriptions one at a time finds nothing wrong — yet the patient is overexposed. The label relocates the defect from the element to the set: the harm-bearing object is the medication list read as a whole against a shared pharmacologic class, not any single line on it. That dissolves the reflex to hunt for the one bad order and tells the reviewer where the pathology actually lives, so the corrective effort points at reconciliation and class-level scanning rather than at re-litigating individual, blameless decisions.

It also sharpens the distinction medication safety most needs to keep crisp: intended combination therapy, where a single clinician deliberately stacks same-class agents for benefit, versus unintended duplication, a pure coordination defect arising from fragmented prescribing authority across providers who never saw each other's orders. The two look identical on the list — two agents on one target — and only the presence or absence of a coordinating intent separates them, which is exactly what duplication-checking must establish. The concept further distinguishes additive same-target exposure from pharmacokinetic interaction, so the practitioner asks not "do these drugs interact?" (an interaction-pair question) but "is anything else already acting on this target?" (a class-scan question) — and recognizes that the fix belongs at the care transitions and decision-support layer where the integrated set is finally read end-to-end, not at the point of any one order.

Manages Complexity¶

A long medication list is a combinatorially intimidating safety object. For a patient on a dozen or more agents — assembled across a home regimen, an admitting order set, cross-cover additions, and combination products that hide active ingredients inside other products — a reviewer who tries to certify the list as safe by brute force faces a quadratic sprawl: every drug checked against every other for pharmacokinetic interaction, plus dose-summing across overlapping orders, plus indication-mapping to confirm each is still warranted. The number of pairings grows with the square of the list length, and most of that work is spent on pairs that share no target and can never compound, while the genuinely harmful arrangement — additive load on one shared target, contributed by orders no single prescriber assembled — is exactly the thing a pairwise interaction scan is not built to surface, because each order is individually correct and only the set is wrong.

Therapeutic duplication compresses that sprawl by changing the unit of analysis from the pair to the pharmacologic class. The relevant question is no longer "do any two of these drugs interact?" but "for each target class, is more than one agent already acting on it?" — and that single reframing collapses an all-pairs search into a class-by-class scan: group the list by shared pharmacologic target, and read off, class by class, whether the count of agents on that target exceeds one. The analyst tracks one regularity — same-target co-occurrence — across a handful of classes rather than re-deriving harm for each of the many drug pairs, and the structural object that bears the harm (the medication list read as an integrated set against its target classes) is precisely the object the scan inspects, so the search effort lands where the pathology actually lives instead of being spread thinly over blameless individual orders.

The branch structure that follows from a positive hit is correspondingly compact. When a class shows more than one agent, exactly one further parameter resolves the outcome: was the stacking intended — a single clinician deliberately combining same-class agents for benefit — or unintended, a coordination defect arising from fragmented prescribing authority? Intended combination therapy and unintended duplication present identically on the list (two agents, one target) and are separated only by the presence or absence of a coordinating intent, so that one binary decides whether the finding is a defensible regimen or a harm to correct. And the remedy menu is fixed by where the integrated set is finally read end-to-end: reconciliation at every care transition, class-level scanning rather than interaction-pair scanning, and decision-support logic that detects same-class co-prescribing across multiple ordering providers. The reviewer thus reasons from a class-grouped count and a single intent flag straight to the verdict and the intervention point — turning a quadratic, pair-by-pair audit of a blameless list into a linear class-scan with a one-bit disambiguation and a transition-layer fix.

Abstract Reasoning¶

Therapeutic duplication licenses a set-level diagnostic that order-by-order reasoning structurally cannot reach: the analyst infers a harm from the configuration of orders even when every individual order is correctly dosed, indicated, and defensible. The reasoning runs FROM "three benzodiazepine-receptor agents appear on one list, each blameless in isolation" TO "the shared target is receiving additive dose beyond its therapeutic window, and the patient is overexposed though no single prescription is wrong." The harm-bearing object is relocated from the element to the set — the medication list read as a whole against its target classes — so the analyst stops hunting for the one bad order (there is none) and reasons about the list as an integrated whole. This is the move that lets a reviewer predict overexposure and incipient toxicity (the delirium on night two) from a list whose every line passes an isolated audit.

The operative interventionist reframing changes the unit of analysis from the pair to the pharmacologic class. Instead of asking "do any two of these drugs interact?" — an all-pairs question whose effort is spent mostly on targets that can never compound — the analyst asks, per class, "is more than one agent already acting on this target?" The procedure is a class-scan: group the list by shared pharmacologic target and read off, class by class, whether the agent count exceeds one, treating combination products as decomposed into their active ingredients so a hidden second acetaminophen or NSAID is surfaced. So the reasoner attacks the list by target class rather than by drug pair, and predicts that the genuinely harmful arrangement — additive load on one shared target assembled by uncoordinated prescribers — will appear in the class-grouped count precisely where a pairwise interaction scan would miss it.

The decisive boundary-drawing move is the intended-versus-unintended disambiguation, a one-bit test applied to every positive hit. Intended combination therapy (a single clinician deliberately stacking same-class agents for benefit) and unintended duplication (a coordination defect arising from fragmented prescribing authority) present identically on the list — two agents, one target — and are separated only by the presence or absence of a coordinating intent. The analyst therefore does not infer harm from same-target co-occurrence alone but reasons that the finding is a defensible regimen or a defect depending on whether a single mind assembled it, and treats establishing that intent as the load-bearing step of duplication-checking. The concept further fixes where the fix belongs: because the defect lives in the set and is invisible at any single order, the analyst predicts that no point-of-order check can catch it and that the integrated set is only ever read end-to-end at care transitions — so the corrective lever is reconciliation at every transition and decision-support logic that scans same-class co-prescribing across multiple ordering providers, not the re-litigation of individual, blameless decisions. And the analyst keeps a clean line against pharmacokinetic interaction: duplication can harm with no interaction at all, because additive same-target action suffices, so the reasoning is about cumulative exposure on a class, not about a pairwise drug-drug effect.

Knowledge Transfer¶

Within medicine the concept transfers as mechanism, with full diagnostic and interventional fidelity, across the breadth of prescribing practice. It moves across services — hospital medicine, geriatric polypharmacy review, oncology regimen reconciliation, emergency and cross-cover prescribing, ambulatory care — and across drug classes — two ACE inhibitors, three benzodiazepine-receptor agents, an NSAID stacked on an NSAID hidden in a combination product, acetaminophen appearing in both a scheduled order and an as-needed cold remedy. In every one of these settings the same machinery carries intact: the harm-bearing object is the medication list read as an integrated set against its pharmacologic target classes; the diagnostic is the class-scan (group by shared target, count agents per target, decompose combination products into active ingredients); the load-bearing disambiguation is intended-versus-unintended (a single coordinating mind versus fragmented prescribing authority); and the remedy family is fixed — reconciliation at every care transition, class-level rather than interaction-pair scanning, and decision-support logic that detects same-class co-prescribing across multiple ordering providers. The vocabulary survives unchanged across these subfields because the substrate is constant: a shared pharmacologic target, parallel uncoordinated prescribers, and a list that is only ever read end-to-end at a transition. The transfer is mechanistic, not analogical, because pharmacologic class and additive same-target exposure are literal everywhere within prescribing.

Beyond medicine the honest verdict is that the named concept does not transfer — importing "therapeutic duplication" onto non-medical systems is metaphor, and the underlying regularity that does recur is better carried by the more general patterns the concept instantiates than by the medical label. The shape that genuinely travels is parallel, uncoordinated interventions compounding on a shared target: redundant security controls firing on one asset, overlapping subsidies or multiple funders backing the same line of work, defense-in-depth interlocks stacking past their intended margin. But that shape, stripped of its medical content, is just a coordination failure on a set — and it is already owned by the parent primes (set-level configuration harm, coordination failure, double-counting / redundancy miscalibration), not by anything distinctive that "therapeutic duplication" adds. The home-bound cargo is exactly what makes the medical concept precise and what fails to survive extraction: the pharmacologic-class grouping as the unit of analysis, the medication-reconciliation protocol as the canonical fix, the combination-product decomposition step that surfaces a hidden second agent, the e-prescribing decision-support alert keyed to same-class co-prescribing. A firewall rule set has no pharmacologic class to scan and no reconciliation visit at which the set is read; an overlapping-subsidy problem has no receptor and no combination product hiding an active ingredient. So the cross-domain lesson should carry the parent prime — look at the configuration, not the element; uncoordinated parallel actions on one target compound — and not "therapeutic duplication," whose force comes entirely from the medication-safety apparatus bolted to that thin structural core. As the seed records, the concept decomposes cleanly into existing primes plus medical apparatus with no distinctive structure left over, which is precisely why it earns the domain-specific label rather than the prime one; this is the boundary made explicit in Structural Core vs. Domain Accent.

Examples¶

Canonical¶

The canonical case is acetaminophen (paracetamol) duplication. A patient may take a scheduled acetaminophen for pain, an opioid–acetaminophen combination tablet (such as hydrocodone/acetaminophen) for breakthrough pain, and an over-the-counter cold remedy that also contains acetaminophen — each product used as directed, none obviously wrong. Summed, the daily acetaminophen can exceed the ~4 g hepatotoxic threshold, and acetaminophen overdose is a leading cause of acute liver failure in the United States. Precisely because the harm lives in the hidden overlap, the FDA in 2011 asked manufacturers to cap acetaminophen at 325 mg per tablet in prescription combination products and mandated a boxed warning, an intervention aimed at the set-level exposure rather than any single order.

Mapped back: Acetaminophen is the shared pharmacologic target reached three ways; the analgesic, the combination tablet, and the cold remedy are authored or bought without a coordinating view — the essence of the medication list as an integrated set going unread. The cumulative dose crossing 4 g is the additive exposure overrunning the therapeutic window, and the FDA's per-tablet cap plus the requirement to name the ingredient are the combination-product decomposition remedy institutionalized.

Applied / In Practice¶

Medication reconciliation at hospital admission is the working deployment of the class-scan. The Joint Commission's National Patient Safety Goal on reconciling medication information requires that, at each care transition, a clinician compare the patient's home medications against newly ordered ones and resolve discrepancies. In practice this routinely catches duplications such as a patient admitted on home lisinopril whose admitting team, working from an incomplete list, adds ramipril — two ACE inhibitors acting on the same target, risking additive hypotension, hyperkalemia, and renal injury — or an NSAID ordered on top of one the patient already takes. Reconciliation is the one point where the whole list is read end-to-end.

Mapped back: The outpatient and admitting clinicians are the parallel prescribers with no shared view; lisinopril-plus-ramipril is a same-class hit surfaced only when the medication list as an integrated set is finally read. Reconciliation supplies exactly the missing reconciliation step, and confirming that no clinician intended the ACE-inhibitor pair is the intent disambiguation that separates a defect from deliberate combination therapy — the class-scan remedy operating at the transition layer where the concept says the fix belongs.

Structural Tensions¶

T1: A mechanical class-scan versus an intent judgment the system cannot supply (the load-bearing bit is the missing one). The concept's compression is genuine: grouping the list by target and counting agents per class turns a quadratic pair audit into a linear scan. But the whole verdict then hinges on one further bit — was the same-class stacking intended combination therapy or unintended duplication? — and that bit is exactly what a fragmented prescribing system does not record. Intent lives in the head of whichever clinician did or did not coordinate, is distributed across providers who never saw each other's orders, and is rarely documented on the list where the scan operates. So the easy, automatable step (detection) is bolted to the hard, under-supported step (disambiguation), and a scan that cannot read intent produces same-class hits that are as often defensible regimens as defects — the false positives that, at the decision-support layer, train clinicians to override the alerts wholesale. Diagnostic: Can the intent behind this same-class pairing actually be established from the record, or is the harm/benefit verdict resting on a coordinating intent the fragmented system never captured?

T2: Pharmacologic class as the unit versus class as a leaky proxy for shared effect. Reframing from "do any two drugs interact?" to "is more than one agent on this target?" is the concept's decisive move — but pharmacologic class is only a proxy for the thing that actually harms, additive action on one physiological effect, and the proxy leaks both ways. It over-fires when two agents nominally share a class yet do not simply add (a partial agonist alongside a full one, different tissue selectivities, one sub-therapeutic), flagging safe combinations as duplication. And it under-fires when genuinely dangerous compounding crosses different classes converging on one effect — several drugs from unrelated classes all lowering blood pressure, or QT-prolongers of different mechanisms — which a same-class scan never groups together. The clean class-based unit is more targeted than pairwise scanning but inherits its own blind spots wherever "same class" and "same effect" diverge. Diagnostic: Do the flagged agents actually add on one effect (true duplication), or does the harm live in a cross-class convergence the same-class grouping cannot see?

T3: Additive-by-definition versus interaction-amplified reality (the pure case that undercounts). The concept draws a clean line: duplication harms by additive same-target exposure, no pharmacokinetic interaction required, so the question is cumulative load, not a drug-drug effect. That purity is analytically useful — it explains how a list of individually correct orders overexposes. But real overexposure frequently combines additive same-target load with an interaction (one agent inhibiting the other's clearance, or a metabolic route saturating), so the total exposure exceeds simple summation. Treating duplication as strictly additive, and interaction as a separate category, risks undercounting exactly the compound cases where the two mechanisms stack, and can license a false reassurance that "these don't interact" when additive load plus a partial interaction is what crosses the window. Diagnostic: Is the exposure here truly the arithmetic sum of same-target doses, or is an interaction amplifying the stack beyond additivity — so the "additive-only" framing understates the risk?

T4: Set-level, no-blame diagnosis versus diffused ownership of the fix. Relocating the harm from the element to the set is the concept's core insight and a humane one: it stops the reflex to hunt for the one bad order, since each is defensible, and points effort at reconciliation rather than at re-litigating blameless decisions. But declaring every individual order blameless and the harm a property of "the set" can diffuse responsibility so far that no one owns the correction — the fragmented authority that assembled the duplication is the same fragmentation that leaves the fix ownerless, with "the system" blamed and no clinician accountable for reading the list end-to-end. The framing that correctly avoids scapegoating an individual order can, unchecked, dissolve the accountability that would get the duplication resolved. Diagnostic: Does the set-level framing here still assign someone concrete ownership of the reconciliation, or has "the harm is in the set" become a reason no individual is answerable for fixing it?

T5: Autonomy versus reduction (a medication-safety category or the coordination-failure primes). Therapeutic duplication is a precise, standard clinical concept with real apparatus — the pharmacologic-class grouping as the unit of analysis, the medication-reconciliation protocol, the combination-product decomposition step, the same-class e-prescribing alert — and within prescribing it transfers as full mechanism across services and drug classes. But the entry itself notes it decomposes cleanly into existing primes plus medical apparatus with no distinctive structure left over: strip the pharmacology and what remains — uncoordinated parallel interventions compounding on a shared target; look at the configuration, not the element — is owned by parents (set-level configuration harm, coordination failure, double-counting/redundancy miscalibration). Redundant security controls or overlapping subsidies have no receptor, no class to scan, no reconciliation visit. Diagnostic: Resolve toward the parent primes (coordination failure, set-level configuration harm) when the compounding is in a non-medical system; toward the named concept only where pharmacologic class, additive same-target exposure, and reconciliation are literally present.

Structural–Framed Character¶

Therapeutic duplication sits at the framed-leaning position on the structural–framed spectrum: a safety-failure category constituted by the practice of prescribing, resting on a thin but portable coordination-failure skeleton — indeed the entry itself notes it decomposes cleanly into existing primes plus medical apparatus with no distinctive structure left over. The criteria mostly point framed. Its evaluative weight is high: the concept names a harm and a failure mode — additive exposure overrunning the therapeutic window — so to call a medication set therapeutically duplicated is to convict its configuration, a verdict of defect rather than a neutral description. It is strongly human-practice-bound: although the additive pharmacology on a receptor is a physical fact, "duplication" as the entry defines it is a coordination failure — the load-bearing distinction between defect and defensible regimen is the presence or absence of a coordinating mind across fragmented prescribing authority — so the phenomenon is constituted by the practice of multiple providers prescribing without a shared view, and dissolves the moment one imagines a single coordinated prescriber; nothing here is a mind-independent regularity the way isostasy is. Its institutional origin is real: the concept is a standard medication-safety category with its own apparatus — reconciliation mandated at care transitions (Joint Commission), e-prescribing duplicate-therapy alerts, combination-product surveillance — artifacts of clinical safety institutions and regulatory practice. On vocab_travels it is domain-pinned: pharmacologic class, therapeutic window, reconciliation, combination-product decomposition have no referents off the prescribing substrate. And on import_vs_recognize it patterns as recognition within medicine and as outright metaphor beyond it.

The one structural-looking feature is the portable skeleton the entry isolates: uncoordinated parallel interventions compounding on a shared target — look at the configuration of the set, not the individual element — which is coordination_failure plus set-level configuration harm (and double-counting / redundancy miscalibration). That skeleton is genuinely substrate-spanning, recurring in redundant security controls firing on one asset and overlapping subsidies backing one line of work. But it does not lift therapeutic duplication off the framed side, because the compounding-on-a-shared-target structure is precisely what the entry instantiates from those umbrella primes, not what makes "therapeutic duplication" itself travel: the cross-domain reach belongs to the general coordination-failure / set-level-harm pattern, while the entry's distinctive content — the pharmacologic-class grouping as the unit of analysis, medication reconciliation as the canonical fix, combination-product decomposition, the same-class e-prescribing alert — is domain accent that stays home, so much so that a firewall rule set has no class to scan and no reconciliation visit at which the set is read. Its character: an evaluatively charged, prescribing-practice-constituted medication-safety failure mode, structural only in the coordination-failure-on-a-shared-target skeleton it instantiates from its umbrella and equips with the pharmacologic-class and reconciliation apparatus of medication safety.

Structural Core vs. Domain Accent¶

This section decides why therapeutic duplication is a domain-specific abstraction and not a prime, and it carries the case for its domain-specificity in the same move.

What is skeletal (could lift toward a cross-domain prime). Strip the pharmacology and a thin relational structure survives: multiple uncoordinated agents each act on one shared target, so their contributions add up past a tolerance no single actor intended — a harm that lives in the configuration of the set, not in any one element. The portable pieces are abstract — several parallel interventions authored without a coordination point, one shared target they all land on, additive compounding, and a defect readable only from the whole set. That skeleton is coordination_failure (parallel actors with no shared view) plus set-level configuration harm and double-counting / redundancy miscalibration. It is genuinely substrate-spanning — redundant security controls firing on one asset, overlapping subsidies backing one line of work, defense-in-depth interlocks stacking past their margin all instance it — which is exactly why it is the core therapeutic duplication instantiates, not what makes the entry the particular thing it is. Indeed the entry itself notes it decomposes cleanly into existing primes plus medical apparatus with no distinctive structure left over.

What is domain-bound. Almost everything that makes the concept therapeutic duplication in particular is medication-safety furniture that does not survive extraction. The pharmacologic-class grouping as the unit of analysis; the therapeutic window the additive exposure overruns; the medication-reconciliation protocol at care transitions as the canonical fix; the combination-product decomposition step that surfaces a hidden second active ingredient (acetaminophen buried in both a scheduled analgesic and a cold remedy); and the e-prescribing duplicate-therapy alert keyed to same-class co-prescribing are the worked vocabulary, instruments, and empirical cases of one clinical discipline. The decisive test is what the machinery has to grip on off-substrate: a firewall rule set has no pharmacologic class to scan and no reconciliation visit at which the set is read end-to-end; an overlapping-subsidy problem has no receptor and no combination product hiding an active ingredient. And the load-bearing distinction between a defect and a defensible regimen — the presence or absence of a coordinating mind across fragmented prescribing authority — is itself a fact about clinical practice, so the phenomenon dissolves the moment one imagines a single coordinated prescriber.

Why this does not clear the prime bar. A prime is a relational structure whose vocabulary travels and whose cross-domain transfer is recognition of the same mechanism, not analogy. Therapeutic duplication's transfer is bimodal. Within medicine it travels as full mechanism — across hospital medicine, geriatric polypharmacy, oncology reconciliation, emergency and cross-cover prescribing, and every drug class (two ACE inhibitors, three benzodiazepine-receptor agents, stacked NSAIDs, hidden acetaminophen), the same harm-bearing set, the same class-scan, the same intent disambiguation, and the same reconciliation remedy carry unchanged, because pharmacologic class and additive same-target exposure are literal everywhere in prescribing (recognition). Beyond medicine, importing "therapeutic duplication" onto non-medical systems is outright metaphor — a redundant-controls or overlapping-subsidy problem borrows the shape and renames every component. And when the bare structural lesson is wanted cross-domain — look at the configuration, not the element; uncoordinated parallel actions on one target compound — it is already carried, in more general form, by the parents the entry instantiates: coordination_failure and set-level configuration harm (with double-counting / redundancy miscalibration alongside). The cross-domain reach belongs to those parents; "therapeutic duplication," as named, carries the pharmacologic-class unit, the reconciliation protocol, and the combination-product decomposition as medication-safety accent that should stay home.

Relationships to Other Abstractions¶

Current abstraction Therapeutic Duplication Domain-specific

Parents (3) — more general patterns this builds on

Therapeutic Duplication is a kind of Medication Error Domain-specific

Therapeutic Duplication is a Medication Error specialized to an unintended concurrent same-target overlap created across fragmented orders.
Therapeutic Duplication presupposes, typical Therapeutic Window Prime

Harm-relevant Therapeutic Duplication typically presupposes a Therapeutic Window whose upper margin the uncoordinated additive exposure approaches or crosses.
Therapeutic Duplication is a decomposition of Systemic Fragmentation Prime

Removing medication vocabulary leaves siloed local actors making defensible decisions whose missing cross-boundary integration produces duplicated global action and set-level harm.

Children (1) — more specific cases that build on this

Polypharmacy Domain-specific is part of, conditional Therapeutic Duplication

Polypharmacy conditionally contains Therapeutic Duplication when fragmented additions place multiple uncoordinated agents on one pharmacological target.

Hierarchy paths (62) — routes to 17 parentless roots

Therapeutic Duplication → Medication Error → Medical Error → Swiss Cheese Model (Layered Defense with Aligning Holes) → Conjunctive Path Activation → Causality → Dependency

Show alternative paths (61)

Not to Be Confused With¶

Drug-drug interaction. A pharmacokinetic or pharmacodynamic effect in which one agent alters another's action (inhibiting its clearance, potentiating its effect). Therapeutic duplication can harm with no interaction at all — additive same-target dosing suffices to overrun the therapeutic window. The check differs: interaction is an all-pairs "do these two drugs interact?" scan; duplication is a class-scan "is more than one agent already on this target?". Tell: does the harm require the drugs to influence each other's pharmacology (interaction), or does it arise from plain additive load on one shared target (duplication)?
Intended combination therapy (rational polypharmacy). A single clinician deliberately stacking same-class agents for benefit — an appropriate regimen that looks identical on the list (two agents, one target). What separates it from duplication is the presence of a coordinating mind; duplication is the artifact of fragmented authority no one assembled on purpose. Tell: did one prescriber knowingly combine the agents for a clinical reason (combination therapy), or did the overlap accrete across uncoordinated providers with no shared view (duplication)?
Polypharmacy. The general condition of a patient taking many medications concurrently. Duplication is a specific set-level defect that can occur within polypharmacy — same-target overlap — but most polypharmacy is not duplication (many drugs on many different targets). Part-vs-whole. Tell: is the concern simply the number of medications (polypharmacy), or specifically that more than one lands on one shared pharmacologic target without coordination (duplication)?
Prescribing cascade. A distinct fragmented-prescribing failure in which one drug's side effect is misread as a new condition and treated with a second drug, growing the regimen. Both stem from uncoordinated prescribing, but a cascade adds agents on different targets to chase an iatrogenic symptom, whereas duplication stacks agents on the same target. Tell: was a new drug added to treat a prior drug's adverse effect (prescribing cascade), or do multiple agents redundantly act on one shared target (duplication)?
Pharmacologic synergy / potentiation. Two agents amplifying each other beyond the sum of their individual effects at the receptor or pathway. Duplication is straightforwardly additive — more drug, not amplified drug — and that simple summation is what crosses the window. Tell: is the combined effect greater than the arithmetic sum of the parts (synergy), or just the parts added together on one target (duplication)?
Coordination failure / set-level configuration harm (the parent primes it instantiates). The substrate-neutral skeleton — uncoordinated parallel interventions compounding on a shared target, a harm readable only from the whole set, not the element — carried by coordination_failure and set-level configuration harm (with double-counting / redundancy miscalibration alongside). Not confusable peers but the umbrella; redundant security controls and overlapping subsidies are its non-medical instances. Tell: outside prescribing, the portable lesson is these parents — treated more fully elsewhere — while pharmacologic class, additive same-target exposure, and reconciliation are therapeutic duplication's home-bound accent.

Neighborhood in Abstraction Space¶

Therapeutic Duplication sits in a moderately populated region (48^th percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.

Family — Pharmacokinetics & Drug Response (19 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-07-12