Artificial intelligence (AI) has the potential to enhance clinical decision-making, including in infectious diseases. By improving antimicrobial resistance prediction and optimizing antibiotic prescriptions, these technologies may support treatment strategies and address critical gaps in healthcare. This study evaluates the effectiveness of AI in guiding appropriate antibiotic prescriptions for infectious diseases through a systematic literature review.

We conducted a systematic review of studies evaluating AI (machine learning or large language models) used for guidance on prescribing appropriate antibiotics in infectious disease cases. Searches were performed in PubMed, CINAHL, Embase, Scopus, Web of Science, and Google Scholar for articles published up to October 25, 2024. Inclusion criteria focused on studies assessing the performance of AI in clinical practice, with outcomes related to antimicrobial management and decision-making.

Seventeen studies used machine learning as part of clinical decision support systems (CDSS). They improved prediction of antimicrobial resistance and optimized antimicrobial use. Six studies focused on large language models to guide antimicrobial therapy; they had higher prescribing error rates, patient safety risks, and needed precise prompts to ensure accurate responses.

AI, particularly machine learning integrated into CDSS, holds promise in enhancing clinical decision-making and improving antimicrobial management. However, large language models currently lack the reliability required for complex clinical applications. The indispensable role of infectious disease specialists remains critical for ensuring accurate, personalized, and safe treatment strategies. Rigorous validation and regular updates are essential before the successful integration of AI into clinical practice.

Patients with stroke while hospitalized experience important delays in symptom recognition. This study aims to describe the overall management of an in-hospital stroke population and how it compares with an out-of-hospital community-onset stroke population.

In this retrospective observational study, we included consecutive patients with in-hospital and out-of-hospital strokes (both ischemic and hemorrhagic) over a period of one year treated at a comprehensive stroke center. Demographic and clinical data were extracted, and patient groups were compared with regard to stroke treatment time metrics.

A total of 362 patients diagnosed with acute stroke were included, of whom 38 (10.5%) had in-hospital and 324 (89.5%) had out-of-hospital strokes. The median delay to stroke recognition (time between the last time seen well and first time seen symptomatic) was significantly longer in in-hospital compared to out-of-hospital strokes (77.5 [0–334.8] vs. 0 [0–138.5] min, p = 0.04). The median time interval from stroke code activation to the arrival of the stroke team at the bedside was significantly shorter in in-hospital versus out-of-hospital cases (10 [6–15] vs. 15 [8–24.8] min, p = 0.01). In-hospital strokes were less likely to receive thrombolysis (12.8% vs. 45.4%, p < 0.01) with significantly higher mortality (18.2% versus 2.6%, p < 0.01) and longer overall median hospital stay (3 [1–7] vs. 12 days [7–23], p < 0.01) compared to out-of-hospital strokes.

This study showed significant delays in stroke symptom recognition and stroke code activation for in-hospital stroke patients despite comparable overall stroke time metrics. Development of in-hospital stroke protocols and systematic staff training on stroke symptom recognition should be implemented to improve care for hospitalized patients.

Antimicrobials are frequently used for palliation during end-of-life care, but adverse effects, such as antimicrobial resistance, are a concern. Shared decision-making is beneficial in end-of-life care conversations to help align antimicrobial-prescribing with patient preferences. However, there is limited data regarding optimal incorporation of antimicrobial-prescribing discussions into shared decision-making conversations. We explored healthcare provider, patient, and support caregiver (eg, family member/friend) perceptions of barriers and facilitators to discussing antimicrobial-prescribing during the end-of-life period.

Qualitative study.

Healthcare providers; palliative care/hospice care patients/caregivers.

We conducted semi-structured interviews on shared attitudes/beliefs about antimicrobial-prescribing during end-of-life patient care at one acute-care and one long-term-care facility. Interviews were analyzed for thematic content.

Fifteen providers and 13 patients/caregivers completed interviews. Providers recognized the potential benefit of leveraging shared decision-making to guide antimicrobial-prescribing decisions. Barriers included limited face-to-face time with the patient and uncertainty of end-of-life prognosis. Patients/caregivers cited trust, comprehension, and feeling heard as important characteristics which act as facilitators in fostering effective shared decision-making around antimicrobial use. Communication in which providers ensure patients are involved in shared decision-making discussions could be increased to ensure patients and their providers develop a mutually agreeable care plan.

Shared decision-making is a practice that can guide antimicrobial-prescribing decisions during end-of-life care, thus potentially minimizing antimicrobial-related adverse effects. Our findings highlight opportunities for increased shared decision-making around antimicrobial use during end-of-life care. Interventions designed to address the identified barriers to shared decision-making have the potential to improve antimicrobial-prescribing practices at end-of-life.

We performed a systematic literature review and meta-analysis on the effectiveness of coronavirus disease 2019 (COVID-19) vaccination against post-COVID conditions (long COVID) in the pediatric population.

Systematic literature review/meta-analysis.

We searched PubMed, Cumulative Index to Nursing and Allied Health Literature (CINAHL), EMBASE, Cochrane Central Register of Controlled Trials, Scopus, and Web of Science from December 1, 2019, to August 14, 2023, for studies evaluating the COVID-19 vaccine effectiveness against post-COVID conditions among vaccinated individuals < 21 years old who received at least 1 dose of COVID-19 vaccine. A post-COVID condition was defined as any symptom that was present 4 or more weeks after COVID-19 infection. We calculated the pooled diagnostic odds ratio (DOR) (95% CI) for post-COVID conditions between vaccinated and unvaccinated individuals.

Eight studies with 23,995 individuals evaluated the effect of vaccination on post-COVID conditions, of which 5 observational studies were included in the meta-analysis. The prevalence of children who did not receive COVID-19 vaccines ranged from 65% to 97%. The pooled prevalence of post-COVID conditions was 21.3% among those unvaccinated and 20.3% among those vaccinated at least once. The pooled DOR for post-COVID conditions among individuals vaccinated with at least 1 dose and those vaccinated with 2 doses were 1.07 (95% CI, 0.77–1.49) and 0.82 (95% CI, 0.63–1.08), respectively.

A significant proportion of children and adolescents were unvaccinated, and the prevalence of post-COVID conditions was higher than reported in adults. While vaccination did not appear protective, conclusions were limited by the lack of randomized trials and selection bias inherent in observational studies.

We performed a systematic literature review and meta-analysis on the effectiveness of coronavirus disease 2019 (COVID-19) vaccination against post-COVID conditions (long COVID) among fully vaccinated individuals.

Systematic literature review/meta-analysis.

We searched PubMed, Cumulative Index to Nursing and Allied Health, EMBASE, Cochrane Central Register of Controlled Trials, Scopus, and Web of Science from December 1, 2019, to June 2, 2023, for studies evaluating the COVID-19 vaccine effectiveness (VE) against post-COVID conditions among fully vaccinated individuals who received two doses of COVID-19 vaccine. A post-COVID condition was defined as any symptom that was present four or more weeks after COVID-19 infection. We calculated the pooled diagnostic odds ratio (DOR) (95% confidence interval) for post-COVID conditions between fully vaccinated and unvaccinated individuals. Vaccine effectiveness was estimated as 100% x (1-DOR).

Thirty-two studies with 775,931 individuals evaluated the effect of vaccination on post-COVID conditions, of which, twenty-four studies were included in the meta-analysis. The pooled DOR for post-COVID conditions among fully vaccinated individuals was 0.680 (95% CI: 0.523–0.885) with an estimated VE of 32.0% (11.5%–47.7%). Vaccine effectiveness was 36.9% (23.1%–48.2%) among those who received two doses of COVID-19 vaccine before COVID-19 infection and 68.7% (64.7%–72.2%) among those who received three doses before COVID-19 infection. The stratified analysis demonstrated no protection against post-COVID conditions among those who received COVID-19 vaccination after COVID-19 infection.

Receiving a complete COVID-19 vaccination prior to contracting the virus resulted in a significant reduction in post-COVID conditions throughout the study period, including during the Omicron era. Vaccine effectiveness demonstrated an increase when supplementary doses were administered.

To evaluate the frequency of antibiotic prescribing for common infections via telemedicine compared to face-to-face visits.

Systematic literature review and meta-analysis.

We searched PubMed, CINAHL, Embase (Elsevier platform) and Cochrane CENTRAL to identify studies comparing frequency of antibiotic prescribing via telemedicine and face-to-face visits without restrictions by publish dates or language used. We conducted meta-analyses of 5 infections: sinusitis, pharyngitis, otitis media, upper respiratory infection (URI) and urinary tract infection (UTI). Random-effect models were used to obtain pooled odds ratios (ORs). Heterogeneity was evaluated with I2 estimation and the Cochran Q statistic test.

Among 3,106 studies screened, 23 studies (1 randomized control study, 22 observational studies) were included in the systematic literature review. Most of the studies (21 of 23) were conducted in the United States. Studies were substantially heterogenous, but stratified analyses revealed that providers prescribed antibiotics more frequently via telemedicine for otitis media (pooled odds ratio [OR], 1.26; 95% confidence interval [CI], 1.04–1.52; I2 = 31%) and pharyngitis (pooled OR, 1.16; 95% CI, 1.01–1.33; I2 = 0%). We detected no significant difference in the frequencies of antibiotic prescribing for sinusitis (pooled OR, 0.86; 95% CI, 0.70–1.06; I2 = 91%), URI (pooled OR, 1.18; 95% CI, 0.59–2.39; I2 = 100%), or UTI (pooled OR, 2.57; 95% CI, 0.88–7.46; I2 = 91%).

Telemedicine visits for otitis media and pharyngitis were associated with higher rates of antibiotic prescribing. The interpretation of these findings requires caution due to substantial heterogeneity among available studies. Large-scale, well-designed studies with comprehensive assessment of antibiotic prescribing for common outpatient infections comparing telemedicine and face-to-face visits are needed to validate our findings.

Suicide rates are high in borderline personality disorder (BPD) where interpersonal problems trigger intense affective dysregulation and impulses to act on suicidal thoughts. To date, however, no study has examined how interpersonal stressors contribute to momentary within-person links among affect and impulsivity with suicidal ideation (SI), and how those links vary over time in people's daily lives.

A total of 153 individuals diagnosed with BPD and 52 healthy controls completed a 21-day ecological momentary assessment protocol. Of these 153 individuals with BPD, 105 had a history of suicide attempts. Multilevel structural equation modeling was used to examine dynamic links among interpersonal perceptions, affect, state impulsivity, and suicidal intent.

Aggregated across interactions, lower perceived warmth in others was associated with SI. This direct relationship, however, did not extend to momentary within-person associations. Instead, interpersonal conflicts were linked to SI indirectly via greater negative affect and lower positive affect. While a robust within-person link between interpersonal perceptions and impulsivity emerged, impulsivity did not account for the relationship between interpersonal perceptions and SI.

This intensive longitudinal study illustrates momentary interpersonal signatures of an emerging suicidal crisis. Among people with BPD at high risk for suicide, interpersonal triggers initiate a cascade of affective dysregulation, which in turn gives rise to SI.

We performed a systematic literature review and meta-analysis measuring the burden of antibiotic use during end-of-life (EOL) care.

We searched PubMed, CINAHL (EBSCO platform), and Embase (Elsevier platform), through July 2019 for studies with the following inclusion criteria in the initial analysis: antibiotic use in the EOL care patients (advanced dementia, cancer, organ failure, frailty or multi-morbidity). If the number of patients in palliative care consultation (PCC) was available, antibiotic use data were pooled to compare the proportion of patients who received antibiotics under PCC compared to those not receiving PCC. Random-effect models were used to obtain pooled mean differences, and heterogeneity was assessed using the I2 value.

Overall, 72 studies met the inclusion criteria and were included in the final review: 22 EOL studies included only patients with cancer; 17 studies included only patients with advanced dementia; and 33 studies included “mixed populations” of EOL patients. Although few studies reported antibiotic using standard metrics (eg, days of therapy), 48 of 72 studies (66.7%) reported antibiotic use in >50% of all patients. When the 3 studies that evaluated antibiotic use in PCC were pooled together, patients under PCC was more likely to receive antibiotics compared to patients not under PCC (pooled odds ratio, 1.73; 95% CI, 1.02–2.93).

Future studies are needed to evaluate the benefits and harms of using antibiotics for patients during EOL care in diverse patient populations.

To evaluate the effectiveness of chlorhexidine (CHG) dressings to prevent catheter-related bloodstream infections (CRBSIs).

Systematic review and meta-analysis.

We searched PubMed, CINAHL, EMBASE, and ClinicalTrials.gov for studies (randomized controlled and quasi-experimental trials) with the following criteria: patients with short- or long-term catheters; CHG dressings were used in the intervention group and nonantimicrobial dressings in the control group; CRBSI was an outcome. Random-effects models were used to obtain pooled risk ratios (pRRs). Heterogeneity was evaluated using the I2 test and the Cochran Q statistic.

In total, 20 studies (18 randomized controlled trials; 15,590 catheters) without evidence of publication bias and mainly performed in intensive care units (ICUs) were included. CHG dressings significantly reduced CRBSIs (pRR, 0.71; 95% CI, 0.58–0.87), independent of the CHG dressing type used. Benefits were limited to adults with short-term central venous catheters (CVCs), including onco-hematological patients. For long-term CVCs, CHG dressings decreased exit-site/tunnel infections (pRR, 0.37; 95% CI, 0.22–0.64). Contact dermatitis was associated with CHG dressing use (pRR, 5.16; 95% CI, 2.09–12.70); especially in neonates and pediatric populations in whom severe reactions occurred. Also, 2 studies evaluated and did not find CHG-acquired resistance.

CHG dressings prevent CRBSIs in adults with short-term CVCs, including patients with an onco-hematological disease. CHG dressings might reduce exit-site and tunnel infections in long-term CVCs. In neonates and pediatric populations, proof of CHG dressing effectiveness is lacking and there is an increased risk of serious adverse events. Future studies should investigate CHG effectiveness in non-ICU settings and monitor for CHG resistance.

Individuals diagnosed with borderline personality disorder (BPD) often describe their lives as stressful and unpredictable. However, it is unclear whether the adversity faced by those with BPD is a product of stress reactivity or stress generation. Here, we examined the dynamic, prospective associations between BPD and stressful life events over 3 years. Given the heterogeneity present in BPD, we sought to understand which empirically derived dimensions of this heterogeneous disorder explain stress reactivity v. stress generation.

Participants included 355 individuals diagnosed with BPD and followed longitudinally at three annual assessments. Auto-regressive cross-lagged panel models were used to examine prospective associations between stressful life events and three latent dimensions implicated in BPD: negative affect, disinhibition, and antagonism.

Antagonism and disinhibition, but not negative affect, prospectively predicted dependent stressful life events (events the individual may have some role in). Evidence for decompensation under stress was more tenuous, with independent stressful life events (those presumably outside the individual's control) predicting increases in negative affect.

Our longitudinal study of a well-characterized clinical sample found more evidence for stress generation than for stress-induced decompensation in BPD. Stress generation in BPD is driven by externalizing dimensions: antagonism and disinhibition. These results highlight the utility of empirically derived dimensions for parsing heterogeneity present in BPD, leading to improvements in diagnostic evaluation, clinical prediction, and individualized approaches to treatment planning.