JAMA Network Open最新文献

Importance: Widespread naloxone access is a key policy response to the opioid crisis. Naloxone availability is typically estimated from pharmacy sales, which exclude naloxone provided by community organizations, hospitals, and clinics, or sold over-the-counter.

Objective: To estimate naloxone knowledge, carrying, purchase, and use among US adults.

Design, setting, and participants: This survey study included noninstitutionalized adults aged 18 years and older from a national sample and a sample self-reporting opioid dependence. Respondents answered online questions between June 7 and June 29, 2024, about naloxone knowledge, carrying, purchase, and use.

Exposures: Opioid misuse, risk of overdose, risk of overdose by person known to respondent.

Main outcomes and measures: Naloxone knowledge, prevalence of naloxone carrying, purchase, and administration.

Results: The survey included 1515 individuals from a national sample (median [IQR] age, 45 [33-58] years; 770 women [50.8%]; 215 Black [14.2%], 1087 White [71.8%]) and 512 who self-reported opioid dependence. In the national sample, 50 respondents (3.3%) reported opioid dependence, yielding 562 respondents reporting opioid dependence (median [IQR] age, 41 [35-48] years; 404 female [70.2%]; 17 Black [3.0%], 494 White [87.9%]). Overall, 1164 respondents-700 (46.2%) in the national sample and 500 (89.0%) reporting opioid dependence-had heard of naloxone and correctly identified its purpose. One hundred sixty participants (10.6%) in the national sample and 340 participants (60.5%) in the sample reporting opioid dependence reported carrying naloxone. Among those reporting they were "very likely to overdose," 22 respondents (31.0%) in the national sample and 31 (73.8%) with opioid dependence reported carrying naloxone. Among those who know someone very likely to overdose, 43 participants (25.4%) in the national sample and 190 participants (70.1%) reporting opioid dependence reported carrying naloxone. Among those who ever carried naloxone, 108 (42.4%) in the national sample and 97 (22.6%) reporting opioid dependence had ever purchased naloxone. Overall, 128 respondents (8.4%) in the national sample and 267 respondents (47.5%) reporting opioid dependence reported administering naloxone to someone else while 93 (6.1%) in the national sample and 221 (39.3%) reporting opioid dependence reported being administered naloxone.

Conclusions and relevance: In this survey study of naloxone, most respondents reporting opioid dependence correctly identified naloxone's purpose and carried it. Most naloxone carried was not purchased, suggesting a need for new ways, including rapid online surveys, to monitor naloxone possession.

Importance: Flooding is a major environmental hazard, with events increasing in intensity and frequency in the context of climate change. Floods cause significant health and economic impacts, particularly among vulnerable populations, including older adults. However, comprehensive analyses of the health consequences of flooding remain limited.

Objective: To evaluate the morbidity and health care costs among Medicare beneficiaries associated with flood exposure in the US.

Design, setting, and participants: This retrospective cohort study analyzed emergency department (ED) use and unplanned hospitalization among Medicare beneficiaries 65 years or older living in zip code tabulation areas (ZCTAs) that were exposed to large-scale flood events from January 1, 2008, to December 31, 2017. This analysis was conducted from April 3 to December 15, 2023.

Exposure: The primary exposure was the presence of a flood as recorded in the Multisourced Flood Inventories, a spatially distributed flood database.

Main outcomes and measures: A conditional fixed-effects regression approach was used to explore the incidence of all-case and cause-specific ED visits and hospitalizations before and after floods. The primary outcomes measured were the incident rate ratios (IRRs) and associated 95% CIs. Attributable risk percentages and estimated attributable excess visits were calculated. Stratified analyses were performed for evaluation of effect modification. Health care costs associated with these events were measured and standardized to 2017 US dollars.

Results: Among 11 801 527 Medicare beneficiaries 65 years or older (mean [SD] age, 74.4 [7.6] years; 56.3% female), the rate of all-cause ED visits and hospital admissions increased by 4.8% (IRR, 1.05; 95% CI, 1.04-1.05) and 7.4% (IRR, 1.07; 95% CI, 1.07-1.08) after flood exposure, respectively. The mean ZCTA-level cost was $3230 (95% CI, $3198-$3261) per ED visit and $11 310 (95% CI, $11 252-$11 367) for hospitalizations. The national costs to the Medicare system were estimated to be $69 275 429 (95% CI, $63 010 840-$76 315 210) for ED visits and $191 409 579 (95% CI, $172 782 870-$206 181 300) for hospitalizations. Stratified analyses highlighted greater impacts for certain demographic groups, including adults older than 85 years, and specific seasonal patterns.

Conclusions and relevance: In this cohort study of Medicare beneficiaries 65 years or older, flood exposure was associated with increased health care use and costs, underscoring the need for targeted public health strategies and improved disaster preparedness, especially for older adults. These findings contribute to a more comprehensive understanding of the health-related costs of flooding and can be used to inform future climate change resilience and health care planning.

Importance: The KEYNOTE-A18 trial demonstrated that adding concurrent and adjuvant pembrolizumab to chemoradiotherapy and brachytherapy significantly improved survival in patients with newly diagnosed, locally advanced cervical cancer. However, considering the annual global incidence of 660 000 cases of cervical cancer, including 13 820 in the US in 2024, incorporating this regimen into the standard of care could have substantial health care economic implications for both patients and the health care system.

Objective: To determine the cost-effectiveness of adding pembrolizumab to the first-line treatment of newly diagnosed, locally advanced cervical cancer.

Design, setting, and participants: This economic evaluation created a Markov model simulating 50-year outcomes to evaluate cost-effectiveness from the payer perspective for patients receiving either pembrolizumab or placebo in addition to chemoradiotherapy plus brachytherapy. Probabilities, including disease progression, survival, and treatment-related toxic effects, were derived from KEYNOTE-A18 clinical trial data in patients with newly diagnosed, locally advanced cervical cancer. Costs and health utilities were obtained from published literature; 1-way, 3-way, and probabilistic sensitivity analyses were used to assess model uncertainty. Data analyses were conducted from April to November 2024.

Exposure: Pembrolizumab.

Main outcomes and measures: Costs, measured in 2024 US dollars, and effectiveness, measured in quality-adjusted life-years (QALYs) were used to calculate an incremental cost-effectiveness ratio (ICER). A willingness-to-pay threshold of $100 000 per QALY was chosen, below which pembrolizumab would be considered cost-effective.

Results: KEYNOTE-A18 enrolled 1060 patients (529 in pembrolizumab group, 531 in placebo group). The median age was 50 years. Pembrolizumab increased costs by $257 000 and effectiveness by 1.40 QALYs, yielding an incremental cost-effectiveness ratio of $183 400 per QALY. The addition of pembrolizumab became cost-effective if its monthly cost was decreased from $16 990 to $9190 (a 45.6% reduction) or its maximum duration of 24 months was decreased to 10 months. The model was insensitive to assumptions about treatment-related toxic effects, progression-free survival, and overall survival. Probabilistic sensitivity analysis indicated that at a willingness-to-pay threshold of $100 000 per QALY, the addition of pembrolizumab was cost-effective 37.3% of the time.

Conclusions and relevance: In this economic evaluation of adding concurrent and adjuvant pembrolizumab to first-line treatment of newly diagnosed, locally advanced cervical cancer, this regimen was not cost-effective at current prices despite data demonstrating improved survival with this regimen.

Importance: US veterans have a higher risk of suicide than the general civilian population. Research suggests that COVID-19 infection is associated with increased risk of suicide attempts or other forms of self-directed violence (SDV) among veterans.

Objective: To identify subgroups of veterans with high risk of post-COVID-19 suicide attempts or SDV.

Design, setting, and participants: This is a retrospective cohort study conducted using data from the Veteran Health Administration (VHA). Participants included VHA enrollees with a first case of COVID-19 between May 1, 2021, and April 30, 2022, residing in the 50 states or Washington, DC.

Exposure: COVID-19 infection.

Main outcomes and measures: The main outcome was a suicide attempt or SDV 12 months after COVID-19 infection. Latent class analysis was used to identify subgroups. Outcome rates and 95% CIs per 10 000 veterans were calculated. Multinomial regressions were used to model outcome risk and marginal risk ratios with 99.5% CIs to compare outcome risk across latent classes.

Results: The cohort included 285 235 veterans with COVID-19 and was predominantly male (248 118 veterans [87.0%]) and younger than 65 years (171 636 veterans [60.2%]). Chronic pain (152 788 veterans [53.6%]), depression (98 093 veterans [34.4%]), and posttraumatic stress disorder (79 462 veterans [27.9%]) diagnoses were common. The 12-month outcome rate was 73.8 events per 10 000 (95% CI, 70.7-77.0 events per 10 000). Two latent classes with high rates of suicide attempt or SDV were identified. The first high-risk subgroup (46 693 veterans [16.4%]) was older (34 472 veterans [73.8%] aged ≥65 years) and had a high prevalence of physical conditions (43 329 veterans [92.8%] had hypertension, and 36 824 veterans [78.9%] had chronic pain); the 12-month outcome rate was 103.7 events per 10 000 (95% CI, 94.7-113.3 events per 10 000). The second high-risk subgroup (82 309 veterans [28.9%]) was generally younger (68 822 veterans [83.6%] aged <65 years) with a lower prevalence of physical conditions but high prevalence of mental health conditions (61 367 veterans [74.6%] had depression, and 50 073 veterans [60.8%] had posttraumatic stress disorder); the 12-month outcome rate was 162.9 events per 10 000 (95% CI, 154.5-171.8 events per 10 000), and compared with the lowest risk subgroup, the 12-month risk of suicide attempts or SDV was 14 times higher in this subgroup (risk ratio, 14.23; 99.5% CI, 10.22-19.80).

Conclusions and relevance: In this cohort study of veterans with COVID-19, 2 veteran subgroups with distinct health profiles had high rates of suicide attempts and SDV, suggesting that different groups may require different approaches to suicide prevention after COVID-19.

Importance: Evidence-based, patient-centered treatment for alcohol use disorder (AUD) can include pharmacotherapy with naltrexone, acamprosate, or disulfiram; however, these medications are rarely used. Medicaid managed care plans (MCPs) manage health services for nearly 80% of Medicaid enrollees and are the largest payer for addiction treatment services. Little is known about Medicaid MCP policies for AUD medications.

Objectives: To describe Medicaid MCPs' coverage and management of acamprosate, naltrexone, and disulfiram for AUD and examine associations of plan characteristics and state policies with medication coverage.

Design, setting, and participants: In this cross-sectional study, a content analysis was performed of 2021 insurance benefit data for 241 comprehensive Medicaid MCPs in states using Medicaid managed care, as well as secondary sources. Data were analyzed from May to August 2024.

Main outcomes and measures: Medicaid MCP-reported medication coverage and utilization management requirements (eg, prior authorization, quantity limit requirements) for acamprosate, disulfiram, and oral and injectable naltrexone together and for each medication separately. Independent variables included plan characteristics (profit status, market share) and the state policy environment in which plans are embedded (Section 1115 substance use disorder waiver, state-defined preferred drug list). Regressions examined associations of plan characteristics and state policies with medication coverage.

Results: In this cross-sectional content analysis of 241 comprehensive Medicaid MCPs in 2021, 217 (90.0%) covered at least 1 medication for AUD: 132 (54.7%) covered acamprosate, 203 (84.2%) covered oral naltrexone, 175 (72.6%) covered injectable naltrexone, 152 (63.0%) covered disulfiram, and 103 (42.7%) covered all 4 medications. Prior authorization and quantity limits were rarely applied, except for injectable naltrexone, for which 75 plans (42.8%) imposed at least 1 of these utilization management requirements.

Conclusions and relevance: This study suggests that efforts to expand AUD medication prescribing may be limited by gaps in health insurance coverage. Medicaid MCPs and states can support AUD medication utilization by covering these medications without applying utilization management strategies.