Annals of Internal Medicine — New This Month

The month's high-yield read is the new ACP living clinical guideline on obesity pharmacotherapy — semaglutide and tirzepatide are first-line, with the older agents ranked behind them, backed by two ACP systematic reviews (below). Two randomized trials are summarized for awareness but flagged lower-yield for general IM: EXTEND (a VA-proven nurse-delivered telehealth model failed to lower HbA1c when ported to fee-for-service care) and PRACTICE (4-week prehabilitation cut complications before spinal fusion, but only in Chinese hospitals). A cluster of observational/health-services work — zoster vaccine & dementia, Stockholm3 prostate screening, CIN 2 management, diverticulitis antibiotic stewardship, dementia hospitalization spending, and the rural workforce — rounds out the month.

EndocrinologyACP GuidelineHigh yield · IM

Pharmacologic Treatments With Lifestyle Modifications in Adults With Overweight or Obesity — ACP Living Clinical Guideline (Apr 2026)

For a nonpregnant adult with obesity (or overweight + a weight-related comorbidity), which weight-management drug do you start, and in what order?

•Obesity (BMI ≥30): 1st-line semaglutide or tirzepatide (moderate-certainty) · 2nd phentermine–topiramate · 3rd liraglutide · 4th naltrexone–bupropion (all conditional, with lifestyle). •Overweight (BMI 27–30) + T2D / dyslipidemia / HTN / OSA / CVD: 1st-line semaglutide or tirzepatide · 2nd liraglutide. Phentermine–topiramate and naltrexone–bupropion not recommended here. •Both incretins are first-line because tirzepatide gives the most weight loss but semaglutide uniquely reduced mortality & MACE — the committee weighted cardiovascular benefit over weight-loss magnitude. •Cautions: phentermine–topiramate contraindicated in established CVD + teratogenic; naltrexone–bupropion suicidality black-box; consider stopping if <5% weight loss at max tolerated dose; weight regains on discontinuation — frame as chronic, often lifelong therapy.

What this changes. ACP's first guideline on obesity pharmacotherapy — codifies GLP-1 / dual-incretin agents as first-line and ranks the older drugs behind them. Editorial caveat (Wee et al.): only 6 head-to-head RCTs among 69 studies, so most cross-drug comparisons are indirect, and weight loss can cost lean/muscle mass — pair therapy with protein intake and resistance exercise.

General IM / Health ServicesEXTEND · RCTLower yield · IM

Nurse-Delivered Comprehensive Telehealth for Uncontrolled Type 2 Diabetes + Hypertension in Fee-for-Service Care

In adults with persistently uncontrolled T2D + hypertension in a fee-for-service system, does nurse-delivered comprehensive telehealth vs self-monitoring lower HbA1c?

Population: 220 adults, persistently uncontrolled T2D (HbA1c ≥8%) + hypertension; mean HbA1c 9.8%, 68% Black; 6 academic clinics.
Intervention: 12-mo nurse + pharmacist comprehensive telehealth (mobile monitoring, self-management support, medication titration).
Comparison: Device-based self-monitoring, no study telehealth.
Outcome: 12-mo HbA1c change −1.1 vs −0.7 pp; between-group −0.4 pp (95% CI −1.0 to 0.3) — NS. Only diabetes self-care improved. Suboptimal fidelity (median 9 of ≥12 encounters).

Clinical takeaway. A VA-proven comprehensive telehealth model did not translate to a fee-for-service academic system. The editorial frames it as a replication lesson — fidelity, context, and reimbursement, not just intervention design, drive real-world effect. What this changes: cautionary, no practice change — tempers porting integrated-system telehealth into FFS care.

Geriatrics / PerioperativePRACTICE · RCTLower yield · IM

Multimodal Prehabilitation for Older Adults Undergoing Spinal Fusion

In adults ≥75 undergoing elective spinal fusion, does 4-week multimodal prehabilitation added to ERAS reduce 90-day complications vs ERAS alone?

Population: 159 adults ≥75 (mean 78.7 y, ~45% frail) undergoing elective spinal fusion at 3 Chinese tertiary hospitals.
Intervention: 4-wk Vivifrail-based exercise + nutritional optimization + psychological interventions, added to ERAS.
Comparison: Standard ERAS alone.
Outcome: ≥1 complication within 90 days 74.7% vs 91.2% (RR 0.80; 95% CI 0.67–0.95; risk difference −18.0%). Shorter length of stay; slightly worse 90-day grip/gait in the prehab arm.

Clinical takeaway. Supports preoperative multimodal prehabilitation to cut complications in older spinal-fusion patients, but generalizability beyond Chinese tertiary hospitals and the resource burden are uncertain, and the ERAS Society has not yet recommended prehabilitation. What this changes: adds RCT evidence to an unrecommended area — not yet generalizable enough to mandate.

Supporting Systematic Reviews (obesity guideline package)

•Benefits & harms living SR + network meta-analysis (High yield) — 69 RCTs / 112,511 patients. Semaglutide probably reduced MACE (pooled ~0.80) and mortality; tirzepatide produced the greatest weight loss and best quality of life. Only 6 head-to-head RCTs; all agents increased discontinuations due to adverse events. •Cost-effectiveness SR (Lower yield — modeling) — 9 model-based studies, no high-certainty evidence; phentermine–topiramate and tirzepatide high value vs lifestyle, liraglutide low value.

Also this month (observational / health-services work — no PICO, no card): RZV & dementia in skilled-nursing residents — target trial emulation (n=509,926): ≥1 recombinant zoster vaccine dose linked to 24% lower relative (~6 pp absolute) 4-yr dementia risk (RR 0.76), attenuating to ~12% in the strictest negative-control analysis — residual healthy-vaccinee bias, hypothesis-generating (DOI); Stockholm3 vs PSA prostate screening, 2-yr — sensitivity for clinically significant cancer 90% vs 74% at similar specificity; no mortality/cost data yet (DOI); Immediate vs delayed treatment of CIN 2 — similar 3-yr cancer/CIN 3+ risk but far more unnecessary excisions with immediate treatment (36% vs 8%) — supports surveillance-first for lower-risk CIN 2 (DOI); Outpatient antibiotics for uncomplicated diverticulitis (VA) — prescribed in 96.6% of visits despite a decade of "selective use" guidelines (DOI); Hospital admission & spending in dementia — quasi-experimental IV: no mortality benefit, +$2547 30-day spending — some discretionary admissions may be low-value (DOI); U.S. health-care workforce by urbanicity — nonmetropolitan areas have 44% fewer patient-facing workers per capita, biggest gaps in physicians and psychologists (DOI).

Personal study digest — original paraphrases prepared for educational use; figures are drawn from the archived full text. Read the full articles at acpjournals.org (subscription required). Window: Jun 6–26, 2026; Ann Intern Med 2026, Vol. 179 (online-first). Trial registrations: EXTEND (NCT05120544), PRACTICE (NCT06140797). DOIs: 10.7326/ANNALS-25-02714 (editorial 10.7326/ANNALS-26-01972), 10.7326/ANNALS-24-03764, 10.7326/ANNALS-24-03766, 10.7326/ANNALS-26-00132 (editorial 10.7326/ANNALS-26-01922), 10.7326/ANNALS-25-05205, 10.7326/ANNALS-25-04689, 10.7326/ANNALS-25-04753, 10.7326/ANNALS-25-04053, 10.7326/ANNALS-25-05583, 10.7326/ANNALS-25-03725, 10.7326/ANNALS-26-00239.