In this episode of Hospital Medicine Unplugged, we unpack acute interstitial nephritis (AIN)—a frequently overlooked cause of acute kidney injury (AKI) driven largely by medications, immune reactions, and systemic diseases.

We start with epidemiology clinicians should recognize. AIN accounts for roughly 15–27% of kidney biopsies performed for AKI and about 2.8% of all kidney biopsies overall. Among biopsies done specifically for acute renal failure, AIN represents ~13.5% of cases. Drug-induced AIN dominates the landscape, responsible for 70–90% of biopsy-proven cases, and its incidence appears to be rising—particularly in older adults, where polypharmacy and underutilization of kidney biopsy can obscure the diagnosis.

Next we break down the most common causes. • Antibiotics are the leading class, responsible for ~49% of drug-induced AIN, especially penicillins, cephalosporins, rifampin, and fluoroquinolones. • Proton pump inhibitors account for ~14%, with omeprazole the single most common culprit drug. • NSAIDs (~11%) are another major contributor. Other causes include 5-aminosalicylates, diuretics, allopurinol, anticonvulsants, and chemotherapeutic agents. Emerging causes include immune checkpoint inhibitors, reflecting the expanding use of immunotherapy in oncology.

We then explore the immunologic pathophysiology. AIN is primarily driven by T-cell–mediated hypersensitivity reactions (Type IV) targeting tubular antigens or drug-related antigens processed by tubular epithelial cells. However, IgE-mediated mast cell activation (Type I hypersensitivity) may also contribute in some cases. The resulting interstitial inflammation and edema can rapidly progress to fibrosis, making early recognition and treatment critical for renal recovery.

Histologically, AIN is characterized by interstitial inflammatory infiltrates composed mainly of lymphocytes, macrophages, plasma cells, and sometimes eosinophils, along with tubulitis, interstitial edema, and tubular injury. Glomeruli are typically normal, while interstitial fibrosis and tubular atrophy signal chronicity and worse prognosis. Variants include granulomatous AIN and rare entities like IgM-positive plasma cell tubulointerstitial nephritis.

Clinically, the classic triad of fever, rash, and eosinophilia is now uncommon—present in fewer than 10–15% of patients. Instead, most patients present with nonspecific symptoms such as malaise, nausea, or asymptomatic AKI. Non-oliguric AKI is typical, often accompanied by mild proteinuria and tubular dysfunction.

Diagnosis relies on clinical suspicion, medication review, and supportive laboratory findings. Urinalysis may show sterile pyuria and white blood cell casts, which are more specific for AIN. Eosinophiluria, historically emphasized, is neither sensitive nor specific. Ultimately, kidney biopsy remains the gold standard when the diagnosis is uncertain.

We also review emerging biomarkers that may transform diagnosis. Urinary CXCL9, an interferon-γ–induced chemokine involved in lymphocyte recruitment, has shown excellent diagnostic performance with AUC values up to ~0.94 for AIN detection. Additional candidate biomarkers include urinary TNF-α, IL-9, kidney injury molecule-1 (KIM-1), and soluble C5b-9, reflecting tubular injury and immune activation.

Management begins with immediate withdrawal of the offending drug. If kidney function does not improve within 5–7 days, corticosteroid therapy is often initiated, typically prednisone ~40–60 mg daily (~0.8 mg/kg). Evidence suggests that early steroid therapy—within the first 1–2 weeks—improves renal recovery, while prolonged treatment beyond several weeks offers little additional benefit.

Finally, we discuss prognosis. About 76% of patients achieve some degree of kidney recovery within six months, with complete recovery in roughly half of steroid-treated cases. However, chronic kidney disease remains common, and long-term studies suggest up to 39% of patients may eventually develop end-stage kidney disease. Poor outcomes are associated with delayed diagnosis, prolonged drug exposure, interstitial fibrosis on biopsy, dialysis requirement, and older age.

The key takeaway: acute interstitial nephritis is a common, often medication-related cause of AKI that requires high clinical suspicion, prompt withdrawal of offending drugs, and early consideration of corticosteroids to prevent irreversible kidney damage.

In this episode of Hospital Medicine Unplugged, we break down celiac disease—from epidemiology and modern diagnostic strategies to life-threatening complications and emerging therapies beyond the gluten-free diet.

We start with epidemiology clinicians should know. The global prevalence of celiac disease is ~1.4% based on serology and ~0.7% with biopsy confirmation. Incidence rates are ~17 per 100,000 person-years in women and ~8 per 100,000 in men, with a female-to-male ratio of ~1.8. Importantly, about 70% of cases remain undiagnosed, the so-called “celiac iceberg.” Over recent decades, incidence has increased substantially, rising from <2 per 100,000 annually in the 1980s to >20 per 100,000 in many regions today.

Next we unpack genetic susceptibility and immune pathogenesis. Nearly all patients carry HLA-DQ2 or HLA-DQ8, but these genes alone are insufficient—~40% of the population carries them, yet only 1–3% develop disease, highlighting the role of environmental triggers and additional genetic factors. Gluten exposure leads to immune activation against deamidated gliadin peptides, resulting in small-intestinal inflammation, villous atrophy, and malabsorption.

We then highlight how the clinical presentation has shifted. The classic picture of malabsorption with diarrhea and weight loss is now less common in adults. Instead, non-classical presentations predominate, including iron-deficiency anemia, osteoporosis, abnormal liver enzymes, infertility, and nonspecific GI symptoms. Diarrhea still occurs in ~40–50% of patients, but many adults present with extraintestinal manifestations or even asymptomatic disease.

We also review celiac crisis, a rare but life-threatening presentation requiring hospitalization. Patients develop severe diarrhea, dehydration, electrolyte disturbances, metabolic acidosis, and profound malnutrition. Management requires intravenous fluids, electrolyte replacement, aggressive nutritional support, and sometimes corticosteroids, alongside initiation of a strict gluten-free diet, which leads to improvement in the vast majority of patients.

Diagnosis begins with serologic testing. IgA tissue transglutaminase (tTG-IgA) is the preferred initial screening test, with ~93–95% sensitivity and ~95–98% specificity, and total IgA should be measured simultaneously to detect IgA deficiency. Endomysial antibody testing has near-100% specificity and can confirm the diagnosis. In adults, upper endoscopy with small-bowel biopsy remains the diagnostic standard, demonstrating intraepithelial lymphocytosis, crypt hyperplasia, and villous atrophy.

We then discuss major complications clinicians must recognize. These include osteoporosis, infertility, neurologic complications, hyposplenism, and small-bowel adenocarcinoma. One of the most serious is enteropathy-associated T-cell lymphoma (EATL)—a rare but aggressive malignancy with very poor survival, often arising from type 2 refractory celiac disease.

Refractory celiac disease (RCD) occurs when symptoms and villous atrophy persist despite ≥12 months of strict gluten-free diet. • Type 1 RCD behaves similarly to active celiac disease and responds to immunosuppressive therapy with excellent survival. • Type 2 RCD represents a pre-lymphoma state with clonal abnormal lymphocytes, and 30–50% progress to EATL within five years.

Management still centers on the gluten-free diet, which leads to symptomatic improvement in ~70% of patients within two weeks, though histologic healing can take months and may remain incomplete in many adults.

Finally, we explore the future of therapy. While diet remains the cornerstone, multiple pharmacologic strategies are in development, including gluten-degrading enzymes, intestinal barrier modulators like larazotide, transglutaminase inhibitors, immune-modulating therapies targeting IL-15, microbiome-based therapies, and even gene-edited wheat with reduced immunogenic gluten.

The takeaway: celiac disease is common, frequently underdiagnosed, and increasingly recognized through non-classical presentations. With improved diagnostics, recognition of severe complications like refractory disease and lymphoma, and a rapidly evolving therapeutic pipeline, management of celiac disease is entering a new era beyond diet alone.

In this episode of Hospital Medicine Unplugged, we tackle polypharmacy and deprescribing—how to recognize problematic medication overload, quantify its harms, and apply structured, patient-centered strategies to safely reduce medication burden.

We begin with definitions that shape clinical practice. Polypharmacy is most commonly defined as the use of ≥5 medications, though definitions vary. Importantly, not all polypharmacy is harmful. “Appropriate polypharmacy” occurs when medications are evidence-based and optimized, while “problematic polypharmacy” arises when medications lack clear benefit or when harms outweigh benefits. Deprescribing is the systematic process of identifying and discontinuing medications whose risks exceed benefits, aligned with a patient’s goals, function, life expectancy, and preferences.

Next we review how common this problem is. Polypharmacy affects 30–40% of community-dwelling older adults, 40–50% of hospitalized older adults, and up to 90% of nursing home residents. Roughly 20–50% of older adults take at least one potentially inappropriate medication (PIM). Risk rises with multimorbidity, female sex, lower socioeconomic status, and each additional chronic disease increases the odds of polypharmacy by nearly 90%.

We then quantify the clinical consequences. • Adverse drug events occur in 20–30% of hospitalized older adults, and each additional medication increases adverse reaction risk by ~10%. • Polypharmacy is associated with higher mortality (HR ~1.2–1.7) and increased hospital admissions and readmissions. • It also increases fall risk (OR ~1.6) and contributes to hip fractures, frailty, cognitive impairment, and functional decline.

A key driver is the prescribing cascade, where a drug causes side effects that are treated with additional medications. Classic examples include: • NSAIDs → hypertension → antihypertensives • Cholinesterase inhibitors → urinary incontinence → anticholinergics • Calcium channel blockers → edema → diuretics • Antipsychotics → parkinsonism → antiparkinsonian drugs

To identify problematic medications, we review major screening tools. • 2023 AGS Beers Criteria highlights medications to avoid or use cautiously in older adults, including guidance on benzodiazepines, antipsychotics in dementia, and aspirin for primary prevention in adults ≥70. • STOPP/START version 3 includes 94 criteria for inappropriate prescriptions and 34 for underprescribing. • Additional tools include the Medication Appropriateness Index, FORTA classification, Anticholinergic Cognitive Burden scale, and Drug Burden Index.

We then walk through a practical deprescribing framework. A common 5-step protocol includes:

1. List all medications and indications

2. Assess overall risk of drug-related harm

3. Identify drugs eligible for discontinuation

4. Prioritize those with highest harm and lowest benefit

5. Implement tapering and monitor for withdrawal or recurrence

Certain medications require careful tapering to prevent withdrawal syndromes, including benzodiazepines, beta-blockers, antidepressants, corticosteroids, opioids, antiepileptics, clonidine, baclofen, and proton pump inhibitors.

We highlight high-yield deprescribing targets. • Proton pump inhibitors: up to 70% lack appropriate indication; associated with C. difficile infection, pneumonia, CKD, and fractures. • Benzodiazepines: linked to falls, delirium, and cognitive impairment, with tapering success rates 27–80%. • Antipsychotics: frequently used for dementia behaviors but carry 1.6–1.7× increased mortality risk. • Anticholinergic medications: high burden strongly linked to cognitive decline and mortality. • Sliding-scale insulin: increases hypoglycemia without improving glycemic control.

We also discuss patient and system barriers. Interestingly, 92% of older adults say they would stop at least one medication if their doctor recommended it, though many fear symptom recurrence or believe medications are necessary.

Finally, we examine solutions that work. Pharmacist-led medication reviews reduce inappropriate medications by 21–35% and lower readmissions, while clinical decision support tools in electronic health records can flag high-risk medications and prompt deprescribing conversations.

The takeaway: polypharmacy is common, harmful, and often reversible. Using structured frameworks, validated screening tools, and shared decision-making, clinicians can safely deprescribe and improve medication safety—especially for older adults with multimorbidity.