Page 18 of 24

PH1.11-13 | PH1.11-13 | Adverse Drug Reactions, Pharmacovigilance and Drug Interactions — SDL Guide — SDL Guide (Part 2)

Causality Assessment: WHO-UMC and Naranjo Algorithm

When a patient on drug therapy develops a new symptom or clinical finding, the prescriber must assess whether the drug is the likely cause — a process called causality assessment. This is not a binary yes/no decision but a probabilistic judgment based on several lines of evidence. Two widely used standardised tools for causality assessment are the WHO-UMC scale and the Naranjo Algorithm. Both tools ask the same fundamental questions: Is the timing of the event consistent with when the drug was started or changed? Is this event known to occur with this drug class? Did the reaction improve when the drug was stopped (dechallenge)? Did it recur when the drug was reintroduced (rechallenge)? The answers to these questions, weighted differently by each tool, produce a probability estimate that guides clinical decisions — whether to withdraw the drug permanently, switch to an alternative, or continue with monitoring.

Causality assessment also has a regulatory dimension: the WHO-UMC and Naranjo categorisations are the standard reporting formats used by pharmacovigilance centres globally. When a clinician submits an ADR report to PvPI, the causality assessment accompanies the report and determines how the signal is weighted in the national and international databases. A report with documented positive dechallenge is given greater weight in signal detection algorithms than one without. This is why performing a structured causality assessment — rather than simply recording 'possible drug reaction' — makes each individual report more valuable to the collective pharmacovigilance evidence base.

The WHO-UMC causality categories provide a framework used by national pharmacovigilance centres for classifying ADR reports:
- Certain: a plausible time relationship; consistent with known pharmacological properties; confirmed by dechallenge (improvement on stopping) and rechallenge (recurrence on reintroduction); no other explanation possible.
- Probable: plausible time relationship; consistent with pharmacology; confirmed by dechallenge; rechallenge not required or available; other explanations less likely.
- Possible: plausible time relationship; pharmacologically consistent; but other disease, drug, or chemical could equally explain the event.
- Unlikely: a time relationship that makes a drug contribution improbable; a more plausible explanation exists.
- Unclassifiable/Unassessable: insufficient data to classify.

The Naranjo Algorithm (Naranjo et al., 1981) is a standardised 10-question scoring instrument that assigns numerical points to key causality criteria. Each question is answered Yes (+1 to +2 points), No (0 to −1 points), or Don't Know (0 points). The score is summed and interpreted as:
- Score ≥ 9: Definite ADR
- Score 5–8: Probable ADR
- Score 1–4: Possible ADR
- Score ≤ 0: Doubtful ADR

The 10 Naranjo questions cover: (1) previous conclusive reports of this reaction; (2) adverse event after the drug; (3) improvement on withdrawal; (4) re-emergence on rechallenge; (5) alternative causes identified; (6) reaction recurred with placebo; (7) detected in blood at toxic level; (8) severity increased with dose increase; (9) patient had similar reaction previously; (10) confirmed by objective evidence.

Applying causality assessment to the SJS case from the hook: the patient developed blistering rash on carbamazepine (plausible time relationship: SJS typically occurs 1–4 weeks after initiation); SJS is a known ADR of carbamazepine, especially in patients with HLA-B*1502 (pharmacologically plausible); no other obvious cause; carbamazepine should be immediately discontinued (dechallenge); rechallenge would be absolutely contraindicated (life-threatening). WHO-UMC: Probable to Certain. Naranjo: typically 6–8 (Probable). Both classifications support immediate drug withdrawal and ADR reporting.

Naranjo question	Yes	No	Don't know
1. Previous conclusive reports	+1	0	0
2. ADR appeared after drug	+2	−1	0
3. ADR improved on withdrawal	+1	0	0
4. ADR recurred on rechallenge	+2	−1	0
5. Alternative causes excluded	+2	−1	0
6. ADR reappeared with placebo	−1	+1	0
7. Drug detected at toxic level	+1	0	0
8. Severity increased with dose	+1	0	0
9. Prior similar reaction	+1	0	0
10. Confirmed by objective evidence	+1	0	0

Drug Interactions: Pharmacokinetic, Pharmacodynamic, and Pharmaceutical

A drug interaction occurs when the effects of one drug are altered by the concurrent administration of another drug, food, or other substance. Drug interactions are classified by mechanism into three broad categories, each requiring a different management approach. The clinical significance of a drug interaction depends on three factors: the magnitude of the change in drug effect (a 5% change in plasma level is usually inconsequential; a 5-fold increase in a narrow-TI drug is potentially fatal), the therapeutic window of the affected drug (interactions involving warfarin, digoxin, phenytoin, cyclosporine, aminoglycosides, and lithium carry the highest clinical risk), and the reversibility of the adverse outcome (rhabdomyolysis from statin-macrolide interaction can cause permanent renal damage; in contrast, reduced antibiotic bioavailability from antacid co-administration is fully correctable by separating doses by 2 hours).

Understanding interaction mechanisms enables anticipation rather than reaction: a prescriber who knows that rifampicin is a CYP450 inducer will check for all CYP-metabolised co-medications before starting it, rather than discovering the interaction after a patient has a stroke from subtherapeutic warfarin or an unintended pregnancy from reduced oral contraceptive levels. The three mechanistic categories — pharmacokinetic, pharmacodynamic, and pharmaceutical — each generate a different pattern of clinical problems and a different management strategy:

Pharmacokinetic (PK) interactions alter the absorption, distribution, metabolism, or excretion of one drug by another, changing its plasma concentration — without directly affecting its pharmacodynamic action at the receptor:

• Absorption interactions: antacids (containing calcium, magnesium, or aluminium) chelate fluoroquinolone antibiotics and tetracyclines in the GI tract, forming insoluble complexes that are not absorbed — fluoroquinolone bioavailability can fall by 50–90%. Management: separate administration by at least 2 hours. Similarly, cholestyramine (bile acid sequestrant) adsorbs multiple drugs, reducing their absorption if given simultaneously.
• Protein binding displacement: aspirin can displace warfarin from albumin binding sites, transiently increasing the free warfarin fraction and enhancing anticoagulant effect. However, the increase in free drug is transient and counterbalanced by increased clearance of the free fraction — this interaction is less clinically significant than enzyme-based interactions but must be considered.
• Enzyme induction interactions: rifampicin is the archetypal enzyme inducer (CYP3A4, CYP2C9, P-glycoprotein). Co-administration substantially reduces plasma levels of warfarin (→ loss of anticoagulation), oral contraceptives (→ contraceptive failure), ciclosporin (→ transplant rejection), antiretrovirals, and digoxin. The interaction develops over 1–2 weeks as enzyme synthesis increases and reverses over 2–4 weeks after rifampicin is stopped. Carbamazepine and phenytoin are also important inducers.
• Enzyme inhibition interactions: erythromycin and clarithromycin inhibit CYP3A4. Co-administration with statins metabolised by CYP3A4 (simvastatin, atorvastatin at high doses) can raise statin plasma concentrations 5–10 fold, precipitating rhabdomyolysis. Simvastatin should not be co-prescribed with macrolide antibiotics. Fluconazole inhibits CYP2C9 → potentiates warfarin — INR can double or triple. Grapefruit juice inhibits CYP3A4 in the gut wall — a non-drug interaction that increases bioavailability of statins, nifedipine, and ciclosporin.

Pharmacodynamic (PD) interactions alter the pharmacological effect of a drug without changing its plasma concentration, by additive, synergistic, or antagonistic effects at the receptor or physiological level:

• Additive (beneficial): combining an ACE inhibitor + a calcium channel blocker + a diuretic for hypertension achieves blood pressure control through complementary mechanisms without duplicating toxicity.
• Synergistic (hazardous): alcohol + benzodiazepines → supra-additive CNS and respiratory depression via enhanced GABA-A conductance. Opioids + benzodiazepines → a combination now recognised to be responsible for a significant proportion of opioid overdose deaths. Combination products should carry explicit black-box warnings.
• Pharmacological antagonism (beneficial): naloxone + morphine → deliberate clinical use to reverse opioid toxicity; neostigmine reverses non-depolarising neuromuscular block by increasing ACh at the NMJ.

Pharmaceutical interactions (physical/chemical incompatibility) occur when two drugs are mixed in the same syringe or IV infusion bag and interact directly — precipitation, degradation, or inactivation before the drugs even enter the patient. Example: aminoglycosides and beta-lactam antibiotics should NEVER be mixed in the same infusion (chemical interaction degrades both); phenytoin precipitates in glucose solutions (must be given in normal saline). These interactions are preventable by pharmacy review.

Interacting drugs	Type	Mechanism	Clinical consequence	Management
Rifampicin + warfarin	PK (enzyme induction)	Rifampicin induces CYP2C9 and 3A4 → ↑ warfarin metabolism	↓ warfarin levels → ↓ anticoagulation, risk of thrombosis	Avoid if possible; if essential, increase warfarin dose; monitor INR closely; dose-reduce warfarin when rifampicin stopped
Erythromycin + simvastatin	PK (enzyme inhibition)	Erythromycin inhibits CYP3A4 → ↑ statin levels	Rhabdomyolysis risk	Avoid co-prescription; use azithromycin (not a CYP3A4 inhibitor) or a statin not metabolised by CYP3A4 (rosuvastatin)
Alcohol + benzodiazepines	PD (synergism)	Both enhance GABA-A Cl− conductance	Supra-additive CNS/respiratory depression	Avoid combination; counsel patient explicitly
Antacid + fluoroquinolone	PK (absorption)	Chelation in GI tract reduces fluoroquinolone absorption	↓ antibiotic bioavailability → treatment failure	Separate doses by ≥2 hours
Naloxone + morphine	PD (pharmacological antagonism)	Naloxone competitively displaces morphine at mu receptor	Reversal of opioid toxicity	Therapeutic — monitor for re-narcotisation
Aminoglycosides + beta-lactams (same syringe)	Pharmaceutical	Chemical degradation on mixing	↓ efficacy of both	NEVER mix in same infusion; give separately

Applying the Safety Framework: Identifying, Reporting, and Managing ADRs and Interactions

Clinical pharmacovigilance is not a passive activity — it requires the prescriber to actively apply a systematic process at every patient encounter. This section synthesises the skills from the preceding sections into practical, actionable frameworks for both ADR management and drug interaction prevention.

The distinguishing skill of a pharmacovigilance-competent clinician is not the ability to recite ADR categories from memory — it is the habit of asking, for every patient on every drug, the questions that a safety-oriented prescriber asks: Is this new symptom a drug effect? Which drug is most likely responsible? What is the causality strength? Does this need to be reported? Are there any drug interactions in this prescription that require monitoring or modification? These questions cost seconds to ask and can prevent serious, sometimes irreversible harm. They are also the questions that, when answered systematically and recorded formally, generate the pharmacovigilance data that protect future patients.

The frameworks below translate these questions into clinical action — first for ADR management, then for drug interaction prevention:

Stepwise approach to a suspected ADR:

1. Identify the adverse event: any new symptom, sign, or laboratory abnormality in a patient on drug therapy is a potential ADR until proven otherwise.
2. Assess causality: apply the temporal relationship criterion (did the symptom appear after the drug was started?), pharmacological plausibility (is this a known or biologically plausible effect of this drug?), and available WHO-UMC or Naranjo criteria.
3. Dechallenge: stop the suspected drug (when safe to do so). If the ADR resolves, this strongly supports causality.
4. Rechallenge decision: deliberate rechallenge with the suspected drug is done only when the clinical need for the drug outweighs the ADR risk, the ADR was Type A (not life-threatening), and adequate monitoring is in place. Rechallenge is absolutely contraindicated for severe Type B ADRs (anaphylaxis, SJS, agranulocytosis).
5. Report to PvPI: complete the CDSCO ADR form (downloadable at www.ipc.gov.in) and submit to the nearest Adverse Drug Reaction Monitoring Centre (AMC). Reporting does not require certainty — a 'possible' or 'probable' ADR classification is sufficient and appropriate. Reporting benefits all future patients who receive that drug.
6. Document and communicate: record the ADR in the patient's medical record, note the drug, date, reaction type, and outcome. If the patient has a serious ADR (e.g., penicillin anaphylaxis), provide a drug allergy card or alert for future prescribers.

Drug interaction management algorithm:

1. Take a complete drug history — including OTC drugs, herbal remedies, and supplements (St John's Wort, grapefruit juice). These are the most frequently overlooked interaction sources.
2. Identify potential interactions — use a drug interaction checker, cross-reference prescribing guidelines, or consult the British National Formulary (BNF) or CIMS India.
3. Classify the interaction — PK (enzyme-based, absorption-based, protein-binding) or PD (synergism, antagonism)? This determines whether dose adjustment, timing separation, or an alternative drug is the right approach.
4. Manage accordingly: (a) Avoid the combination if an equally effective, non-interacting alternative exists. (b) Adjust doses and monitor drug levels (TDM for narrow-TI drugs like warfarin, digoxin, phenytoin). (c) Separate administration timing (for absorption interactions). (d) Counsel the patient about dietary interactions (grapefruit, alcohol).
5. Document the decision — record that the interaction was identified, the decision made, and the monitoring plan in the notes.