Journal of General Internal Medicine最新文献

Background: ChatGPT has quickly gained popularity as a source of online health information (OHI). However, it is unclear how having a usual source of primary care (USPC) is related to OHI-seeking.

Objective: Explore how having a USPC and other characteristics thought to affect access-to-care influence the use of ChatGPT and other OHI forms.

Design: Cross-sectional national survey.

Participants: Adult members of ResearchMatch, a non-profit affiliate of the National Institutes of Health, between June and August 2023.

Main measures: The survey evaluated demographics, health characteristics, and OHI-seeking behaviors, including ChatGPT usage. OHI sources were categorized as "passive" (Google, Wikipedia, WebMD) and "interactive" (forums, Q&A sites, ChatGPT). Descriptive statistics, t-tests, and chi-square tests compared users by USPC status. Multiple logistic regression estimated adjusted effects on ChatGPT use.

Key results: Of 21,499 adults invited to participate in the survey, 2406 (11.2%) responded. Among respondents, 56% reported having a USPC. Those with a USPC, compared to those without, were older, spoke English as their primary language, had higher income, and had more formal education (all p<.001). Participants with a USPC were more likely to use passive OHI (OR 2.46, 95% CI 1.55-3.90, p<.001) and less likely to use interactive OHI (OR 0.73, 95% CI 0.60-0.89, p=.002) or ChatGPT (OR 0.56, 95% CI 0.44-0.71, p<.001). Age over 50 (OR 0.11, 95% CI 0.06-0.20, p<.001), non-White race (OR 0.51, 95% CI 0.38-0.70, p<.001), very good or better health (OR 0.71, 95% CI 0.55-0.92, p=.009), and college education (OR 0.61, 95% CI 0.39-0.97, p=.035) were inversely related to ChatGPT use.

Conclusions: In this national survey of patients participating in a clinical research matching service, those with regular primary care access relied less on ChatGPT, suggesting that a personal primary care relationship may attenuate the need or motivation to use AI-derived OHI.

Background: Criminal legal involvement (CLI) has been associated with increased burden of disease, worse health outcomes, and high healthcare utilization. The health needs of women with CLI are often overlooked despite the rising proportion of women in the US legal system.

Objective: Examine differences by binary gender in the prevalence of disease and healthcare utilization among individuals with lifetime CLI.

Design/setting: Cross-sectional nationally representative multivariate logistic regression and negative binomial regression.

Participants: Adult respondents with lifetime CLI (N = 37,279) with response rate of 66-69%.

Measures: The primary independent variable was binary gender (woman/man). Outcomes included medical conditions, substance use disorders, mental illness, and healthcare utilization. Covariates included sociodemographic confounders.

Results: Women comprised 29.1% of respondents with lifetime CLI and reported 0.83 times the odds of chronic heart disease compared to men (95%CI 0.74, 0.92; p = 0.001), but 1.86 times the odds of COPD (95%CI 1.63, 2.13; p < 0.001), 1.78 times the odds of asthma (95%CI 1.63, 1.93; p < 0.001), and 1.30 times the odds of cancer (95%CI 1.08, 1.53; p = 0.005). While women were less likely to have a substance use disorder (OR 0.84; 95%CI 0.78, 0.90), they were more likely to currently smoke (OR 1.29, 95%CI 1.20, 1.28; p < 0.001) and to have any mental illness (OR 2.45; 95%CI 2.26, 2.63; p < 0.001). Women reported increased rates of all forms of healthcare utilization compared to men after adjustments.

Conclusions: Women with lifetime CLI experience a different profile of diseases compared to men with increased prevalence of any mental illness and chronic disease, especially respiratory diseases. They also had higher rates of healthcare utilization. Additional research should focus on interventions tailored to the unique needs of this population.

Background: Patient knowledge regarding why they are prescribed their medications is paramount for the delivery of quality and patient-centered cardiometabolic care.

Objective: To generate national estimates of the proportion of patients who do not know why they are taking their anti-hypertensive, anti-diabetes, or anti-lipid medications.

Design: Cross-sectional study using data from the National Health and Nutrition Examination Survey from January 2013 through March 2020.

Participants: Non-institutionalized adults in the United States.

Main measures: Medication use was ascertained by respondent self-report and pill bottle review of their anti-hypertensive, anti-diabetes, or anti-lipid medications. The primary outcome was the proportion of respondents with a knowledge deficiency about why they were taking their respective medications defined by respondents reporting (1) they did not know why or (2) an incorrect indication for why they were taking each medication. We also report knowledge deficiency prevalences stratified by respondent characteristics and use multivariable logistic regression analysis to assess for relationship between respondent characteristics and odds of having a knowledge deficiency.

Key results: 4.9% (95% confidence interval [CI] 4.0-5.6%, n=391) of respondents on cardiometabolic medications had ≥ 1 knowledge deficiency, which corresponds to approximately 3,152,243 US adults. This was driven most strongly by knowledge deficiencies in anti-hypertensive medication use (4.5%, 95% CI 3.7-5.3%). There were large variations in knowledge deficiency prevalence by respondent characteristics. Most notably, 12.9% (95% CI 9.8-16.0%) of adults ≥ 80 years and 12.2% (95% CI 8.6-16.0%) of adults taking ≥ 10 medication had a knowledge deficiency. Of older adults with polypharmacy (≥ 5 prescriptions), 11.2% (95% CI 8.9-13.5%) > 65 years and 15.8% (95% CI 11.5-20.2%) > 80 years had a knowledge deficiency.

Conclusions: Most adults report a plausible reason for their use of their cardiometabolic medications. However, the prevalence of having a knowledge deficiency varied greatly by respondent demographic and health characteristics.

Purpose: Uncontrolled hypertension is a significant US health problem, despite existing effective treatments. This study assessed the impact of variations in patterns of blood pressure data on physician perceptions of hypertension control using different forms of data visualization.

Method: Physicians (N = 57) reviewed eight brief vignettes describing a fictitious patient; each vignette included a graph of the patient's blood pressure data. We examined how variations in mean systolic blood pressure (SBP), blood pressure standard deviation (SD), and form of visualization (e.g., line graph with raw values or smoothed values only) affected judgments about hypertension control and need for medication change.

Results: Smoothing successfully reduced visual noise for the physicians. For controlled hypertension, physician judgments were more consistent with clinical guidelines when using the smoothed graph compared with the raw data graph. Judgments about hypertension control with the smoothed graph were similar to judgments made using the raw data graph for cases of uncontrolled hypertension.

Conclusion: Data visualization can direct physicians to attend to more clinically meaningful information, thereby improving their judgments of hypertension control.

Background: Real-time or synchronous telemedicine can be a valuable adjunctive strategy for chronic disease management, but few studies have assessed its impact on hypertension control among safety-net populations.

Objective: To evaluate whether telemedicine is associated with blood pressure (BP) control.

Design: Retrospective cohort study. Mixed-effects logistic regression models clustered by the patient estimated associations between telemedicine and BP control after adjusting for patient factors and neighborhood context.

Participants: Patients seeking care at an urban, multisite community health center with hypertension and ≥1 BP measurement between 2020 and 2022 (3663 patient-year observations across 2086 unique patients).

Main measures: The primary outcome was BP control defined as a binary variable. We used the Centers for Medicare & Medicaid Services' Controlling High Blood Pressure quality measure criteria of systolic blood pressure < 140 mmHg and diastolic blood pressure < 90 mmHg in the most recent recording in the measurement year.

Key results: Among the 2086 patients with hypertension in our sample, there were 1257 (60.3%) Latinx and 425 (20.4%) Black patients. Over 90% lived in a neighborhood of high deprivation as categorized by the social deprivation index. Telemedicine visits, compared to none, were not associated with blood pressure control (1-2 telemedicine visits aOR, 1.05 [95% CI, 0.86-1.28]; ≥3 telemedicine visits aOR, 0.86 [95% CI, 0.68-1.09]). One in-person visit per year, compared to 2-3, was associated with lower odds of BP control (aOR, 0.72 [95% CI, 0.55-0.94]). Black patients, compared to Latinx patients, were less likely of having BP control (aOR, 0.64 [95% CI, 0.48-0.87]).

Conclusions: In this community health center cohort of patients with hypertension, telemedicine did not compromise blood pressure control. Persistent racial disparities in blood pressure control underscore the need for equity-centered strategies for hypertension management in safety-net primary care settings.

Background: The diagnosis section in hospital discharge summaries is critical for continuity of care and patient safety, yet it varies widely in quality, format, and content due to a lack of standards.

Objective: This study aims to develop a cross-specialty standard for the structure and content of the diagnosis section, based on the preferences of German physicians. The study examines physicians' satisfaction with the diagnosis section, their rating of its importance, and their preferences for its specific elements, comparing perspectives between inpatient and outpatient physicians.

Design, participants, approach: This mixed-methods study integrated a scoping review, focus group discussion, and a nationwide survey of 602 physicians (317 outpatient primary care and 285 inpatient physicians; 4.1% response rate), most trained in internal medicine. Quantitative analyses evaluated physician satisfaction and preferences, while qualitative feedback provided deeper insights regarding preferred content and format.

Key results: Although 95.7% of physicians considered the diagnosis section crucial for follow-up care, only 36.9% were satisfied with its current content and format. 91.2% supported standardizing the diagnosis section, identifying 18 content elements to be included for every current treatment diagnosis. Strong consensus (> 95.0% agreement) was reached for "name of the diagnosis," "severity/stage/classification/TNM," "localization/extent/pattern of involvement," "course e.g., acute, chronic, recurrent," "expression," "complications," "date of initial diagnosis," and "etiology/cause." 86.4% preferred separating current and chronic/prior diagnoses with headings. Outpatient physicians were more likely than inpatient physicians to rate "ICD-10 codes" as mandatory (46.2% vs. 14.8%, p < 0.001) and to consider "recommendations for further procedures" (76.6% vs. 63.6%, p < 0.001) and "follow-up appointments" (77.3% vs. 63.5%, p < 0.001) as necessary. Additionally, a list of practical recommendations for clinicians to better document diagnoses was derived.

Conclusions: This study proposes a cross-specialty standard for the diagnosis section based on physician preferences for a clearly structured format and 18 key content elements.

Background: Hypertension onset is linked to sleep, but the precise sleep duration affecting it remains unclear. Our goal is to pinpoint the sleep duration impacting hypertension incidence, offering valuable insights for prevention and management.

Methods: We conducted a systematic search in PubMed, Embase, Web of Science, CNKI, Cqvip, and Wanfang Database, up to May 30, 2023, focusing on cohort studies examining the association between nighttime sleep duration and hypertension risk in adults aged 18 and above. Two authors independently performed data extraction, quality assessment, and synthesis based on predefined criteria. A random-effects model was used to estimate pooled effect sizes with 95% confidence intervals (CIs). Heterogeneity was quantified using the I² statistic, with potential sources explored through subgroup and sensitivity analyses to validate the robustness of the results.

Results: Out of the 173,734 participants included in the meta-analysis, 41,528 eventually developed hypertension. The analysis revealed a correlation between short sleep duration and increased risk of hypertension: 1.07 (95% CI 1.00-1.14) for those sleeping ≤ 7 h, 1.04 (95% CI 1.02-1.07) for 6-7 h, and 1.17 (95% CI 1.06-1.28) for < 6 h. For women, with sleep duration 6-7 and < 6 h, the pooled risk of hypertension incidence was 1.07 (1.02-1.12) and 1.12 (1.06-1.19). In individuals under 60 years of age, an elevated risk of hypertension was observed with sleep durations of less than 6 h and between 6 and 7 h, with pooled risks of 1.24 (95% CI 1.10-1.39) and 1.05 (95% CI 1.00-1.11), respectively.

Conclusions: Hypertension is significantly correlated with sleep duration under 7 h, especially in women and those under 60, highlighting the importance of sleep management in hypertension prevention and treatment strategies.

Trial registration: PROSPERO: CRD42022345513.