Annals of Internal Medicine最新文献

Clinical impact ratings: GIM/FP/GP: [Formula: see text] Cardiology: [Formula: see text].

Background: Recombinant zoster vaccine (RZV) was preferentially recommended over live zoster vaccine (ZVL) starting in 2018.

Objective: To assess RZV effectiveness using target trial emulation, accounting for prior receipt of ZVL and immunocompetence.

Design: Analysis 1 emulated 12 sequential trials of at least 12 months' duration in which treatment assignment was assessed by real-world vaccination in the month before each trial. Individuals could participate in multiple trials for which they were eligible (for example, no prior RZV vaccination). Pooled vaccine effectiveness (VE) was estimated across the trials using weighted Fine and Gray models with robust variance estimation. Risks were estimated from the cumulative incidence function. Analysis 2 used the same methods to estimate the VE of 2 RZV doses versus 1 dose, with 10 trials beginning 60 days from the first dose.

Setting: 20% random sample of fee-for-service Medicare beneficiaries covered by Parts A, B, and D between 2007 and 2019.

Participants: Medicare beneficiaries aged 65 years or older with continuous Part D coverage, 6 months of continuous coverage before trial enrollment, no claims for herpes zoster (HZ) since 2007, and no prior RZV vaccination.

Intervention: 1 or 2 RZV doses.

Measurements: Outcomes were HZ, HZ ophthalmicus, and postherpetic neuralgia. Covariates were age, sex, race, ethnicity, prior ZVL receipt, and immunocompetence.

Results: Vaccine effectiveness against any HZ outcome was 56.1% (95% CI, 53.1% to 59.0%), with similar VE between immunocompetent (56.5% [CI, 53.2% to 59.5%]) and immunocompromised (54.2% [CI, 44.7% to 62.1%]) individuals. Individuals vaccinated with ZVL in the past 10 years benefited from RZV. A second RZV dose conferred an additional 67.9% effectiveness against any HZ outcome.

Limitation: Limited follow-up; covariate misclassification.

Conclusion: Recombinant zoster vaccine is effective in older adults, including immunocompromised adults, and 2 doses were more effective than 1. Prior ZVL recipients should be revaccinated with RZV.

Primary funding source: National Center for Advancing Translational Sciences.

Background: There is concern about widespread increases in cancer incidence rates in younger adults.

Objective: To compare international cancer incidence trends in younger adults (aged 20 to 49 years) and older adults (aged ≥50 years).

Design: Surveillance study.

Setting: Forty-two countries in Asia (n = 11), Europe (n = 22), Africa (n = 1), North and South America (n = 6), and Australasia (n = 2) with annual cancer incidence data from 2003 to 2017 in the International Agency for Research on Cancer's GLOBOCAN database.

Participants: Adults aged 20 years or older.

Measurements: Joinpoint regression to estimate the average annual percentage change (AAPC) in cancer incidence rates for 13 cancer types previously reported to be increasing in multiple countries in younger adults (leukemia and breast, endometrial, colorectal, oral, kidney, liver, pancreatic, gallbladder, prostate, stomach, esophageal, and thyroid cancer).

Results: Incidence rates increased in younger adults in most (>75%) countries between 2003 and 2017 for 6 of the 13 cancer types: thyroid cancer (median AAPC, 3.57%), breast cancer (median AAPC, 0.89%), colorectal cancer (median AAPC, 1.45%), kidney cancer (median AAPC, 2.21%), endometrial cancer (median AAPC, 1.66%), and leukemia (median AAPC, 0.78%). Incidence rates for these cancer types also increased in older adults in most countries (median AAPCs, 3% for thyroid cancer, 0.86% for breast cancer, 1.65% for kidney cancer, 1.20% for endometrial cancer, and 0.61% for leukemia). The exception was colorectal cancer, which only increased in older adults in about half the countries (median AAPC, 0.37%), and the AAPC was greater in younger than older adults in 69% of countries. For liver, oral, esophageal, and stomach cancer, rates decreased in younger adults in more than half the countries.

Limitation: Most countries with available data were high-middle-income countries, and the results might not be generalizable.

Conclusion: Cancer incidence rates increased for several cancer types in many of the countries studied; however, other than colorectal cancer, these increases occurred in both younger and older adults. These findings can help inform future research and clinical and public health guidelines.

Primary funding source: Institute of Cancer Research and National Institutes of Health Intramural Research Program.

Background: Perinatal depression can have a deleterious impact on mothers and infants.

Purpose: To evaluate psychological therapies for perinatal depression.

Data sources: 6 databases from January 2000 to March 2025.

Study selection: Randomized controlled trials (RCTs) of psychological therapies for people with depression during pregnancy and up to 1 year postpartum.

Data extraction: 6 researchers extracted study data and assessed the risk of bias and strength of evidence (SoE).

Data synthesis: Forty-four RCTs were included. Cognitive behavioral therapy (CBT; k = 25, n = 2962) was probably more effective than treatment as usual (TAU) in reducing depressive symptoms by an equivalent -1.7 points (95% CI, -2.0 to -1.3 points) on the Edinburgh Postnatal Depression Scale (EPDS; range, 0 to 30 points) (moderate SoE) and may have greater recovery rates from depressive symptoms (relative risk [RR], 1.7 [CI, 1.3 to 2.3]) (low SoE). Behavioral activation (k = 3, n = 508) may be more effective than TAU in reducing depressive symptoms by an equivalent -1.5 EPDS points (CI, -2.6 to -0.5 points) (low SoE). There may be no differences in depressive symptoms between CBT and counseling (k = 3, n = 226; EPDS, -0.5 [CI, -1.5 to 0.5]) or counseling and TAU (k = 3, n = 247; EPDS, -0.8 [CI, -2.6 to 1.0) (low SoE). Interpersonal therapy (IPT; k = 9, n = 1003) was probably more effective than TAU in reducing depressive symptoms by an equivalent -1.7 EPDS points (CI, -2.9 to -0.5 points) (moderate SoE) and may have greater recovery rates from depressive symptoms (RR, 1.2 [CI, 0.97 to 1.5]) (low SoE).

Limitations: Participants were not blinded to treatment, study variation in country, interventions, populations, or reducing SoE. Differences may not be clinically important.

Conclusion: For treatment of perinatal depression, CBT, behavioral activation, and IPT may be effective.

Primary funding source: Agency for Healthcare Research and Quality (AHRQ).

Background: Data extraction is a critical but error-prone and labor-intensive task in evidence synthesis. Unlike other artificial intelligence (AI) technologies, large language models (LLMs) do not require labeled training data for data extraction.

Objective: To compare an AI-assisted versus a traditional, human-only data extraction process.

Design: Study within reviews (SWAR) using a prospective, parallel-group comparison with blinded data adjudicators.

Setting: Workflow validation within 6 ongoing systematic reviews of interventions under real-world conditions.

Intervention: Initial data extraction using an LLM (Claude, versions 2.1, 3.0 Opus, and 3.5 Sonnet) verified by a human reviewer.

Measurements: Concordance, time on task, accuracy, sensitivity, positive predictive value, and error analysis.

Results: The 6 systematic reviews in the SWAR yielded 9341 data elements from 63 studies. Concordance between the 2 methods was 77.2% (95% CI, 76.3% to 78.0%). Compared with the reference standard, the AI-assisted approach had an accuracy of 91.0% (CI, 90.4% to 91.6%) and the human-only approach an accuracy of 89.0% (CI, 88.3% to 89.6%). Sensitivities were 89.4% (CI, 88.6% to 90.1%) and 86.5% (CI, 85.7% to 87.3%), respectively, with positive predictive values of 99.2% (CI, 99.0% to 99.4%) and 98.9% (CI, 98.6% to 99.1%). Incorrect data were extracted in 9.0% (CI, 8.4% to 9.6%) of AI-assisted cases and 11.0% (CI, 10.4% to 11.7%) of human-only cases, with corresponding proportions of major errors of 2.5% (CI, 2.2% to 2.8%) versus 2.7% (CI, 2.4% to 3.1%). Missed data items were the most frequent error type in both approaches. The AI-assisted method reduced data extraction time by a median of 41 minutes per study.

Limitations: Assessing concordance and classifying errors required subjective judgment. Consistently tracking time on task was challenging.

Conclusion: Data extraction assisted by AI may offer a viable, more efficient alternative to human-only methods.

Primary funding source: Agency for Healthcare Research and Quality and RTI International.

Hyponatremia is the most common electrolyte disorder in hospitalized patients. Hospital-associated hyponatremia includes community-acquired (e.g., hyponatremia on admission) and hospital-acquired hyponatremia. Acute-onset hyponatremia requires rapid treatment with hypertonic saline to decrease cerebral edema. In cases of chronic hyponatremia (>48 hours), the brain has time to normalize cell volume by losing solutes. However, even mild chronic cases can have adverse outcomes, such as decreased cognition, osteoporosis, increased risk for falls, and fractures. Hyponatremia is associated with increased hospital length of stay, readmissions, morbidity, and mortality. Prompt recognition and appropriate treatment are therefore necessary to improve outcomes.

Clinical impact ratings: GIM/FP/GP: [Formula: see text] Cardiology: [Formula: see text].