Back to Resources
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
Never paste real patient identifiers into AI tools. Even for drug information queries, avoid including patient names, MRNs, or other PHI. Frame your question generically: "a 72-year-old patient with CKD stage 3b" rather than using any identifying details. HIPAA applies to every digital tool you use.
Always verify AI-generated references. AI models frequently fabricate study names, journal citations, and statistical results. Before citing any reference from AI output in a clinical communication, confirm it exists and says what the AI claims it says. Use PubMed, Lexicomp, or your institutional database subscriptions for verification.
Ask the AI not to fabricate citations. Adding a line like "do not invent specific citations — instead indicate where I should look" reduces hallucinated references and produces more honest output. The newsletter prompt above demonstrates this approach.
Specify the output format and word limit. Drug information responses need to be concise and scannable. Setting a word limit and requesting structured formats (tables, subheadings, bullet points) ensures the output is usable in clinical communication without heavy editing.
Layer your workflow: AI first, then clinical databases. Use AI to generate the structure and narrative draft of your response, then fill in verified data from Lexicomp, Clinical Pharmacology, Micromedex, or primary literature. This hybrid approach is faster than either tool alone and more accurate than AI alone.
5 Claude Prompts for Prior Authorization
5 Claude Prompts for MTM Documentation
Claude CoWork for Pharmacists
ClaudePharmacistBeginnerPrompt Pack
5 Claude Prompts for Drug Information & Research — Pharmacists
Ready-to-use Claude prompts for clinical drug research, interaction assessments, and evidence-based drug information queries.
Why Use AI for Drug Information & Research?
Drug information queries are a daily reality for pharmacists in every practice setting. Whether a physician calls about a complex multi-drug interaction, a P&T committee needs a formulary review, or you want to summarize the evidence for an off-label use, the underlying task is the same: find the relevant clinical data, synthesize it, and communicate it clearly. Traditional DI workflow — searching Lexicomp, PubMed, and primary literature — is thorough but time-consuming, especially when the question is nuanced and the answer is scattered across multiple sources.
AI tools like Claude can accelerate the synthesis and communication steps of the drug information process. They are particularly useful for drafting structured responses, organizing complex interaction data into readable formats, and generating first-draft clinical summaries that you can then verify against authoritative references. Think of AI as a drafting assistant that gives you a strong starting point, not a replacement for your clinical database subscriptions or your own pharmacological expertise.
A critical caution: AI models can generate plausible but fabricated references, trial names, and clinical data points. This is especially dangerous in drug information work where accuracy is paramount. Every factual claim in AI-generated output — every study name, every dosing recommendation, every interaction mechanism — must be verified against a primary or tertiary source before you share it with a prescriber, a committee, or a patient. The prompts below are designed to produce well-structured drafts, but verification is always your final step.
Prompt 1: Complex Drug Interaction Assessment
Analyze a multi-drug interaction scenario and produce a structured clinical assessment.
A patient is currently taking [LIST ALL CURRENT MEDICATIONS WITH
DOSES]. The prescriber wants to add [NEW MEDICATION AND DOSE].
Perform a comprehensive drug interaction assessment. For each
clinically significant interaction identified, provide: the
interacting drug pair, the mechanism of interaction (pharmacokinetic
or pharmacodynamic), the expected clinical consequence, the severity
rating (major, moderate, minor), the level of evidence supporting
the interaction, and a specific management recommendation (dose
adjustment, monitoring parameter, timing separation, or alternative
agent). Organize the results from most to least clinically
significant. Conclude with an overall clinical recommendation
for the prescriber.Example Output:
Drug Interaction Assessment: Adding Amiodarone 200 mg daily
>
Interaction 1 (Major): Amiodarone + Warfarin
Mechanism: Pharmacokinetic — amiodarone inhibits CYP2C9 and CYP3A4, reducing warfarin metabolism.
Clinical consequence: Significant increase in INR (often 30-50% rise), elevated bleeding risk.
Evidence: Well-documented in multiple prospective studies and case series.
Management: Empirically reduce warfarin dose by 30-50% upon amiodarone initiation. Monitor INR within 3-5 days and weekly for 4-6 weeks until stable. Educate patient on bleeding signs.
>
Interaction 2 (Major): Amiodarone + Simvastatin
Mechanism: Pharmacokinetic — CYP3A4 inhibition increases simvastatin exposure.
Clinical consequence: Increased risk of rhabdomyolysis and myopathy.
Evidence: FDA-mandated labeling restriction; simvastatin doses above 20 mg daily are contraindicated with amiodarone.
Management: Reduce simvastatin to 20 mg daily or switch to a non-CYP3A4-dependent statin such as rosuvastatin or pravastatin. Monitor CK if myalgia develops.
>
Overall Recommendation: Amiodarone may be added with dose adjustments to warfarin and simvastatin. Close INR monitoring and statin dose reduction are essential before initiation.
Why it works: Asking for mechanism, severity, evidence level, and management strategy for each interaction mirrors the structure of a formal drug information response and ensures no critical detail is omitted.
Prompt 2: Formulary Exception Clinical Summary
Prepare a clinical summary to support adding a non-formulary medication or granting a formulary exception.
Write a clinical summary supporting the use of [NON-FORMULARY
MEDICATION] for [INDICATION] to present to a Pharmacy & Therapeutics
committee or formulary review board. Include: FDA-approved
indications, mechanism of action, summary of pivotal clinical trial
efficacy data (include trial names, patient populations, primary
endpoints, and NNT if available), safety profile and common adverse
effects, comparison to current formulary alternatives [LIST FORMULARY
ALTERNATIVES], unique clinical advantages of the requested agent,
estimated cost impact [PROVIDE AWP OR ACQUISITION COST IF KNOWN],
and a recommendation with supporting rationale. Format this as a
structured drug monograph summary suitable for a P&T committee
presentation. Limit to 500 words.Example Output:
Formulary Exception Summary: Evinacumab (Evkeeza) for Homozygous Familial Hypercholesterolemia
>
FDA-Approved Indication: Adjunct to other LDL-lowering therapies for homozygous familial hypercholesterolemia (HoFH) in adults and pediatric patients aged 5 years and older.
>
Mechanism of Action: Evinacumab is a monoclonal antibody targeting angiopoietin-like protein 3 (ANGPTL3), which regulates lipid metabolism independently of the LDL receptor pathway. This mechanism is critical for HoFH patients with null/null LDLR mutations who respond poorly to statins and PCSK9 inhibitors.
>
Pivotal Efficacy Data: The ELIPSE HoFH trial (N=65, double-blind RCT) demonstrated a 47.1% reduction in LDL-C from baseline versus a 1.9% increase with placebo at 24 weeks (p<0.001). NNT for achieving greater than 30% LDL-C reduction was approximately 2.
>
Safety Profile: Most common adverse effects include nasopharyngitis (16%), influenza-like illness (7%), and infusion-related reactions (6%). No hepatotoxicity signal observed in clinical trials.
>
Formulary Alternatives: Current formulary agents (rosuvastatin, ezetimibe, evolocumab) target the LDL receptor pathway and have limited efficacy in receptor-negative HoFH. Lomitapide is available but carries significant GI toxicity and hepatotoxicity risk.
>
Recommendation: Approve evinacumab with restriction to confirmed HoFH patients who have failed maximally tolerated conventional therapy. Estimated annual cost is $450,000 per patient; however, the eligible patient population at our institution is currently one individual.
Prompt 3: Off-Label Use Evidence Review
Summarize the available evidence for an off-label medication use in a concise, clinically useful format.
Summarize the clinical evidence for the off-label use of
[MEDICATION NAME] for [OFF-LABEL INDICATION, e.g., "low-dose
naltrexone for fibromyalgia" or "gabapentin for alcohol use
disorder"]. Structure the summary as follows: current FDA-approved
indications, proposed mechanism for the off-label use, summary of
published clinical evidence (differentiate between randomized
controlled trials, observational studies, and case reports),
dosing regimens used in the literature, notable safety concerns
specific to this off-label use, and whether any major clinical
guidelines or compendia (NCCN, UpToDate, AHFS DI) support this
use. End with a clinical bottom line: is the evidence sufficient
to support a trial of therapy, and under what conditions?Example Output:
Off-Label Evidence Review: Low-Dose Naltrexone for Fibromyalgia
>
FDA-Approved Indications: Opioid use disorder (50 mg daily); alcohol use disorder (50 mg daily).
>
Proposed Off-Label Mechanism: At low doses (1.5-4.5 mg), naltrexone is hypothesized to modulate microglial activation in the central nervous system, reducing pro-inflammatory cytokine release (TNF-alpha, interleukin-6) that contributes to central sensitization and widespread pain in fibromyalgia.
>
Published Clinical Evidence:
- RCTs: Two small randomized, double-blind, placebo-controlled crossover trials (N=31 and N=30) demonstrated significant reductions in pain severity (approximately 30% improvement) versus placebo. Both used 4.5 mg nightly dosing over 12-week treatment periods.
- Observational: A retrospective cohort (N=215) from a Danish pain clinic reported clinically meaningful pain reduction in 37% of patients.
- Case reports: Multiple published case series support tolerability and modest efficacy.
>
Dosing in Literature: Typically initiated at 1.5 mg nightly, titrated to 4.5 mg nightly over 2-3 weeks. Compounded capsules are required as this dose is not commercially available.
>
Safety Concerns: Generally well tolerated. Most common effects are vivid dreams and transient headache. Must not be used concurrently with opioid analgesics.
>
Guideline Support: Not currently recommended in ACR fibromyalgia guidelines. UpToDate mentions low-dose naltrexone as an emerging option with insufficient evidence for a formal recommendation.
>
Clinical Bottom Line: Evidence is promising but limited to small trials. A monitored trial of therapy is reasonable in patients who have failed first-line agents (duloxetine, pregabalin, milnacipran), provided no concurrent opioid use.
Prompt 4: Drug Information Newsletter Article
Draft a clinical pharmacy newsletter article on a topic relevant to your practice site.
Write a 400-word clinical pharmacy newsletter article about
[TOPIC, e.g., "new FDA safety communication regarding fluoroquinolone
use in uncomplicated infections" or "updated AHA guidelines on
statin therapy for primary prevention" or "emerging evidence on
GLP-1 agonists for cardiovascular risk reduction"]. The audience is
[TARGET AUDIENCE, e.g., "hospital physicians and nurses" or
"community pharmacy staff" or "long-term care facility providers"].
Include: a brief background on why this matters now, the key
clinical takeaway (what changed or what is new), practical
implications for prescribing or dispensing at our practice site,
and 2-3 action items the reader can implement immediately. Use
a professional but accessible tone — avoid excessive jargon
but do not oversimplify the clinical science. Include a "Key
References" section at the end listing 3-4 source types the
reader should consult (do not fabricate specific citations —
list reference categories like "FDA Safety Communication,
[date]" or "journal name, primary study").Example Output:
Clinical Pharmacy Update: FDA Safety Communication on Fluoroquinolone Use in Uncomplicated Infections
>
Why This Matters Now
The FDA has reinforced its strongest warnings regarding systemic fluoroquinolone use for uncomplicated urinary tract infections, acute bacterial sinusitis, and acute bronchitis. These warnings reflect accumulating evidence of serious, potentially irreversible adverse effects involving tendons, muscles, joints, nerves, and the central nervous system — collectively termed fluoroquinolone-associated disability (FQAD).
>
What Changed
Updated labeling now states that fluoroquinolones should be reserved for conditions with no alternative treatment options when used for uncomplicated infections. The boxed warning has been expanded to include aortic aneurysm and dissection risk, hypoglycemic coma, and mental health side effects including suicidal ideation, in addition to the existing tendinopathy and peripheral neuropathy warnings.
>
Practical Implications for Our Facility
For uncomplicated UTIs, first-line agents remain nitrofurantoin, trimethoprim-sulfamethoxazole, and fosfomycin. Ciprofloxacin and levofloxacin should not be prescribed empirically for uncomplicated cystitis unless culture-directed and no alternatives exist. Emergency department prescribers should be especially aware, as fluoroquinolones are frequently selected for convenience in discharge prescriptions.
>
Action Items:
1. Review your fluoroquinolone prescribing for uncomplicated UTIs this month and switch eligible patients to first-line alternatives.
2. Update any standing order sets or clinical pathways that list fluoroquinolones as first-line for uncomplicated infections.
3. Counsel patients currently on fluoroquinolones to report tendon pain, numbness/tingling, or mood changes immediately.
>
Key References: FDA Drug Safety Communication (most recent update); Infectious Diseases Society of America UTI treatment guidelines; relevant primary literature in the New England Journal of Medicine or Clinical Infectious Diseases; institutional antibiogram data.
Prompt 5: Therapeutic Alternative Comparison
Create a structured comparison of therapeutic alternatives for a specific clinical scenario.
Create a clinical comparison table and narrative for the following
therapeutic alternatives for [INDICATION, e.g., "type 2 diabetes
in a patient with established cardiovascular disease"]:
Drug A: [MEDICATION NAME AND CLASS]
Drug B: [MEDICATION NAME AND CLASS]
Drug C: [MEDICATION NAME AND CLASS]
Compare across the following parameters: mechanism of action,
efficacy for the primary indication (cite landmark trial names
if applicable), cardiovascular outcomes data, renal outcomes data,
key adverse effects, contraindications and precautions, dosing
and administration considerations, drug interactions of clinical
significance, and approximate monthly cost [INSERT COSTS IF KNOWN,
OR SAY "use institutional pricing"]. Present the comparison as a
formatted table followed by a 150-word narrative recommendation
explaining which agent you would prioritize and under what clinical
circumstances each alternative would be preferred.Example Output:
Therapeutic Comparison: SGLT2 Inhibitors vs. GLP-1 Receptor Agonists vs. DPP-4 Inhibitors for Type 2 Diabetes with Established Cardiovascular Disease
>
| Parameter | Empagliflozin (SGLT2i) | Liraglutide (GLP-1 RA) | Sitagliptin (DPP-4i) |
|---|---|---|---|
| Mechanism | Inhibits renal glucose reabsorption | Incretin mimetic, enhances insulin secretion | Inhibits DPP-4 enzyme, prolongs incretin activity |
| HbA1c Reduction | 0.7-0.8% | 1.0-1.5% | 0.6-0.8% |
| CV Outcomes Trial | EMPA-REG OUTCOME | LEADER | TECOS |
| CV Mortality Benefit | Yes (38% RRR) | Yes (22% RRR) | Neutral |
| Renal Benefit | Yes (progression of nephropathy reduced) | Modest | Neutral |
| Key Adverse Effects | Genital mycotic infections, DKA (rare) | GI (nausea, vomiting), pancreatitis (rare) | Nasopharyngitis, headache |
| Administration | Oral, once daily | Subcutaneous injection, once daily | Oral, once daily |
| Approximate Monthly Cost | $550 | $900 | $450 |
>
Narrative Recommendation: For a patient with type 2 diabetes and established cardiovascular disease, empagliflozin is the preferred first add-on to metformin based on demonstrated cardiovascular mortality reduction and renal protective effects at a moderate cost. Liraglutide is the preferred alternative if stronger glycemic lowering is needed or if the patient has contraindications to SGLT2 inhibitors (recurrent genital infections, history of DKA). Sitagliptin is appropriate only when cost, injection aversion, and tolerability are primary concerns and the patient declines agents with proven cardiovascular benefit.