JAMA Network Open最新文献

Importance: Continuity of care is a key aspect of high-quality primary care. Vulnerable populations often experience fragmented care. Some US Department of Veterans Affairs (VA) clinics offer primary care in patient aligned care teams (PACTS) tailored for veterans with homeless experience (VHE), termed H-PACTs.

Objectives: To test the hypothesis that primary care continuity would be higher for VHEs in H-PACTs than for VHEs in mainstream VA PACTs and to compare other service utilization patterns by primary care clinic type.

Design, setting, and participants: Retrospective observational cohort study including national survey data combined with VA electronic health records data from primary care clinics at 26 VA medical centers. Participants were VHEs who completed the national survey and had 2 or more primary care visits in the 12 months before the survey. The survey was completed between April and October 2018 and data were analyzed from April 2020 to November 2025.

Exposure: Enrollment in H-PACTs or mainstream PACTs.

Main outcomes and measures: Continuity was calculated using the usual provider of care (UPC) measure, which is the proportion of primary care visits with the most frequently seen clinician. High continuity was defined as a UPC of 0.75 or higher. Multivariable regression models examined the association of H-PACT enrollment with high continuity, and other utilization measures included mental health, specialty visits, emergency department (ED) visits, and hospitalizations.

Results: A total of 2271 VHEs in H-PACTs (2140 [94.2%] male; 932 [41.0%] Black, 1050 [46.2%] White, and 263 [11.6%] other; mean [SD] age, 58.1 [9.3]) and 1627 VHE in mainstream PACTs (1393 [85.6%] male; 674 [41.4%] Black, 740 [45.5%] White, and 192 [11.8%] other; mean [SD] age, 60.7 [12.1]) were included. Compared with those in mainstream PACTs, VHEs in H-PACTs had a higher mean (SD) UPC (0.81 [0.23] vs 0.77 [0.25]; χ21 = 21.6; P < .001) and were more likely to achieve high continuity (1483 patients [65.3%] vs 938 [57.7%]; χ22 = 25.0; P < .001). After multivariable adjustment, care in H-PACTs remained associated with high continuity (odds ratio [OR], 1.48; 95% CI, 1.33-1.66). In adjusted analyses, compared with those in mainstream PACTs, VHEs in H-PACTs had significantly more primary care visits (4.6 vs 4.0; z score = 5.28; P < .001), fewer specialty visits (6.2 vs 7.9 visits; z score = -4.66; P < .001), and were less likely to have an ED visit (OR, 0.83; 95% CI, 0.75-0.92).

Conclusions and relevance: In this study, VHEs in H-PACT clinics had higher primary care continuity with no indication of substitution of specialty or emergency visits for primary care. The H-PACT model is associated with less intensive health care delivery.

Importance: Conventional cancer statistics account only for individuals with a clinical diagnosis, overlooking potentially large numbers of undetected malignant neoplasms. Autopsy-based studies offer a unique opportunity to estimate the burden of cancer, including latent tumors.

Objective: To evaluate long-term trends and characteristics of cancers using a nationwide autopsy registry in Japan.

Design, setting, and participants: This cohort study of hospital-based autopsies over a 66-year period (1958-2023) obtained data from the Annual of the Pathological Autopsy Cases in Japan (APAC-J), a nationwide database maintained by the Japanese Society of Pathology. Latent cancers, defined as malignant neoplasms undiagnosed during life but discovered at autopsy, were analyzed based on International Statistical Classification of Diseases, Tenth Revision and International Classification of Diseases for Oncology, Third Edition. All autopsies recorded in APAC-J were included. The data were analyzed between May 7 and August 2, 2025.

Main outcomes and measures: The main outcome was the proportion of total, multiple, and latent cancers by year, age group, and sex based on trends in latent cancer detection; metastatic status of latent cancers; and the enrichment ratio in autopsy, a novel metric comparing cancer prevalence in autopsies with national mortality statistics.

Results: From 1958 to 2023, 1 486 557 autopsies were registered in APAC-J (mean age based on available age group data, 59.1 years; 62.5% male), with 55.2% including cancer diagnoses. The proportion of multiple primary cancers increased from 1.8% (420 of 22 989) in 1974 to 14.4% (957 of 6661) in 2023. The enrichment ratio in autopsy was elevated in adolescents (aged 15-19 years), young adults (aged 20-24 years), and adults aged 80 years or older. Since 1986, 36 133 latent cancers were found in 34 174 of 811 159 autopsies (4.2%). The detection rate of latent cancers increased from 1.7% (683 of 39 839) in 1986 to 7.4% (493 of 6661) in 2023. Latent prostate cancer at age 75 to 79 years was identified in 4.5% (corresponding to a 2017-2021 prevalence of 656.2 per 100 000 population), 6.9-fold higher than the clinical incidence, whereas latent thyroid cancer at age 50 to 54 years showed substantially larger excesses (0.9% of men and 1.7% of women, representing 94.5-fold and 60.7-fold higher prevalences, respectively). Overall, metastases were present in 7.3% of latent cancers (2649 of 36 133).

Conclusions and relevance: This cohort study of autopsies across Japan found a substantial reservoir of undiagnosed cancer, including some with metastatic potential. These findings highlight the persistent value of autopsy for assessing cancer burden and underscore the need to refine approaches for early detection while minimizing overdiagnosis.

Importance: Emergency departments (EDs) serve patients at high risk for overdose. There is increasing interest in peer-delivered ED interventions for substance use but little rigorous research on their effectiveness.

Objective: To examine the effectiveness of an initiative (Relay) operated by the New York City Health Department that dispatches trained peer wellness advocates (WAs) to support ED patients after a nonfatal opioid overdose.

Design, setting, and participants: This randomized clinical trial compared Relay and site-directed care (SDC) at 4 EDs in New York, New York. Adult patients presenting after opioid-involved overdose were enrolled from October 6, 2020, to June 30, 2022, with 12 months of outcome follow-up. Statistical analysis was performed from November 4, 2024, to May 6, 2025.

Intervention: ED workers (generally physicians) called the Relay hotline for patients presenting after a suspected opioid-involved overdose. WAs met patients in the ED to provide peer support and brief overdose risk reduction education. WAs attempted to contact patients for 90 days to provide ongoing support, education, and referrals using a harm reduction framework.

Main outcomes and measures: Opioid-related adverse events (any opioid-involved overdose [fatal or nonfatal] or any other substance use-related ED visit) in the 12 months after enrollment were identified using health care administrative data plus self-report.

Results: Among a total of 253 participants randomized, 127 were randomized to the Relay arm and 126 to the SDC arm. A total of 247 participants, 125 in the Relay arm and 122 in the SDC arm (190 [76.9%] men; 80 [32.4%] Black, 126 [51.0%] Hispanic or Latinx, 76 [30.8%] White, and 91 other race [36.8%]), were included in the intention-to-treat analyses. No statistically significant differences between arms were observed for the primary outcome (mean [SD] opioid-related adverse events, 3.29 [4.52] in the Relay arm and 4.10 [9.36] in the SDC arm; rate ratio, 1.02; 95% CI, 0.72-1.45; P = .90). By 12 months after enrollment, 24 participants (9.7%) had died (17 [70.8%] due to overdose). Relay participants reported high satisfaction with the ED intervention.

Conclusions and relevance: This randomized clinical trial examining the impact of an ED peer navigator intervention on subsequent opioid-related adverse events did not find significant outcome differences for Relay vs SDC participants. These findings highlight the importance of intervening to save lives in this high-risk population and suggest potential refinements to future ED peer intervention research.

Trial registration: ClinicalTrials.gov Identifier: NCT04317053.