The role of an artificial intelligence model in antiretroviral therapy counselling and advice for people living with HIV.

OBJECTIVES: People living with HIV may find personalized access to accurate information on antiretroviral therapy (ART) challenging given the stigma and costs potentially associated with attending physical consultations. Artificial intelligence (AI) chatbots such as ChatGPT may help to lower barriers to accessing information addressing concerns around ART initiation. However, the safety and accuracy of the information provided remains to be studied. METHODS: We instructed ChatGPT to answer questions that people living with HIV frequently ask about ART, covering i) knowledge of and access to ART; ii) ART initiation, side effects, and adherence, and iii) general sexual health practices while receiving ART. We checked the accuracy of the advice against international HIV clinical practice guidelines. RESULTS: ChatGPT answered all questions accurately and comprehensively. It recognized potentially life-threatening scenarios such as abacavir hypersensitivity reaction and gave appropriate advice. However, in certain contexts, such as specific geographic locations or for pregnant individuals, the advice lacked specificity to an individual's unique circumstances and may be inadequate. Nevertheless, ChatGPT consistently re-directed the individual to seek help from a healthcare professional to obtain targeted advice. CONCLUSIONS: ChatGPT may act as a useful adjunct in the process of ART counselling for people living with HIV. Improving access to information on and knowledge about ART may improve access and adherence to ART and outcomes for people living with HIV overall.

ChatGPT as a tool to improve access to knowledge on sexually transmitted infections.

OBJECTIVES: Specific to sexual health, individuals in need of information may be adolescents who have limited ability to formally access healthcare. These digital natives may turn to ChatGPT to address their concerns on sexually transmitted infections (STI). We sought to evaluate the veracity of ChatGPT's responses to commonly asked questions on STIs. METHODS: We instructed ChatGPT (GPT 3.5) to answer STI questions from three domains, namely, (1) general risk factors for STIs, (2) access to care and diagnosis of STIs and (3) management of STIs and postexposure prophylaxis. The responses were recorded and checked against the US Centers for Disease Control and Prevention STI Treatment Guidelines 2021. RESULTS: Overall, the responses were concise and accurate. In terms of prevention, ChatGPT could also recommend measures like safe sex practices and human papillomavirus vaccination. However, it failed to recommend HIV pre-exposure prophylaxis. When an individual expressed a symptom that could potentially represent STI (eg, dyspareunia) ChatGPT appropriately provided reassurance that other possibilities exist, but advocated for testing. In terms of treatment, ChatGPT consistently communicated the importance of partner testing and follow-up testing, but at times, failed to highlight the importance of testing for other STIs. Overall, the advice given was not tailored to the specific individual's circumstances. CONCLUSIONS: ChatGPT can provide helpful information regarding STIs, but the advice lacks specificity and requires a human physician to fine-tune. Its ubiquity may make it a useful adjunct to sexual health clinics, to improve knowledge and access to care.

Infectious etiologies of persistent and chronic diarrhea in Asian developing countries: A systematic review and meta-analysis.

Infectious causes of diarrhea contribute significantly to morbidity in Asia. We conducted a systematic review and meta-analysis of the prevalence of infectious etiologies of persistent and chronic diarrhea in Asian adults. Searches were performed on PubMed and Scopus for studies from January 1, 1970, to May 30, 2023. Sixteen studies were identified and included. The meta-analysis was conducted with the random-effects method, estimating the pooled prevalence of groups of infectious pathogens as causes of persistent and chronic diarrhea among Asian adults. The findings were highly heterogeneous and indicative of publication bias. The majority of studies were conducted on persons living with human immunodeficiency virus infection (PLHIV). The studies were predominantly from low-income and middle-income Asian countries. The most common cause was parasitic, with a pooled prevalence of 0.52 (95% confidence interval 0.28-0.65, I2 = 99%, Cochran's Q = 1027.44, P < 0.01), followed by bacterial, fungal, and viral causes, which were substantially rarer. Negative microbiological testing was also common, with a pooled prevalence for a negative test being 0.37 (95% confidence interval 0.17-0.52, I2 = 99%, Cochran's Q = 1027.44, P < 0.01). Subgroup analyses of studies conducted among PLHIV, from year 2000 and among those conducted in Southeast Asia showed a similar prevalence of parasitic causes of diarrhea. In conclusion, in Asian adults with persistent and chronic diarrhea, parasitic causes were most prevalent. However, the estimate of true prevalence is limited by significant heterogeneity among the available studies. More study in this field is required, especially examining PLHIV in the post-antiretroviral therapy era and from high-income countries.

Elevated serum lactate dehydrogenase aids prediction of mortality in Pneumocystis jirovecii pneumonia without underlying human immunodeficiency virus infection - Derivation of a clinical risk score.

Pneumocystis jirovecii pneumonia (PCP) is associated with significant mortality amongst patients without underlying human immunodeficiency virus infection (HIV). We sought to develop a risk score to predict mortality in this population. We reviewed patients with a presumed or confirmed PCP and a negative HIV test from 2006-2023. We constructed a multivariable model to identify parameters independently associated with mortality and the adjusted odds ratios were converted to weights to derive a risk score. Subsequently, we compared the performance of our score to the CURB-65 score by means of area under receiver operating characteristic curve (AUC). In total, we examined 93 patients with PCP without HIV. Mortality was 31.2%. Risk factors for mortality included older age, male sex and high serum lactate dehydrogenase levels (LDH) and C-reactive protein levels. A risk score was derived comprising age> 65 years (2 points), male sex (2 points) and LDH> 770 U/L (3 points). Our risk score (AUC 0.71, 95%CI 0.60-0.82) performed better than the CURB-65 score (AUC 0.53, 95%CI 0.41-0.66). A low-risk score of 0-1 had excellent negative predictive value for mortality (97.5%). In conclusion, a risk score comprising age, sex and LDH can predict mortality in PCP without underlying HIV and help with prognostication.

Does Prior Respiratory Viral Infection Provide Cross-Protection Against Subsequent Respiratory Viral Infections? A Systematic Review and Meta-Analysis.

The epidemiology of different respiratory viral infections is believed to be affected by prior viral infections in addition to seasonal effects. This PROSPERO-registered systematic review identified 7388 studies, of which six met our criteria to answer the question specifically. The purpose of this review was to compare the prevalence of sequential viral infections in those with previously documented positive versus negative swabs. The pooled prevalence of sequential viral infections over varying periods from 30-1000 days of follow-up was higher following a negative respiratory viral swab at 0.15 than following a positive swab at 0.08, indicating the potential protective effects of prior respiratory viral infections. However, significant heterogeneity and publication biases were noted. There is some evidence, albeit of low quality, of a possible protective effect of an initial viral infection against subsequent infections by a different virus, which is possibly due to broad, nonspecific innate immunity. Future prospective studies are needed to validate our findings.

Performance of Risk Scores in Predicting Infective Endocarditis in Patients with Staphylococcus aureus Bacteraemia in a Prospective Asian Cohort.

Background: Several risk scores have been derived to predict the risk of infective endocarditis (IE) amongst patients with Staphylococcus aureus bacteraemia (SAB), which helps to guide clinical management. Methods: We prospectively studied 634 patients admitted with SAB. The cohort was stratified into those with or without IE, and the PREDICT Day 1, Day 5 and VIRSTA scores were tabulated. Area under the receiver operating characteristic (AUC) curves were constructed to compare the performance of each score. Results: Of the 634 patients examined, 36 (5.7%) had IE. These patients were younger (51.6 ± 20.1 vs. 59.2 ± 18.0 years, p = 0.015), tended to have community acquisition of bacteraemia (41.7% vs. 17.9%, p < 0.001), and had persistent bacteraemia beyond 72 h (19.4% vs. 6.0%, p = 0.002). The VIRSTA score had the best performance in predicting IE (AUC 0.76, 95%CI 0.66-0.86) compared with PREDICT Day 1 and Day 5. A VIRSTA score of <3 had the best negative predictive value (97.5%), compared with PREDICT Day 1 (<4) and Day 5 (<2) (94.3% and 96.6%, respectively). Conclusions: Overall, the risk scores performed well in our Asian cohort. If applied, 23.5% of the cohort with a VIRSTA ≥ 3 would require TEE, and a score of <3 had an excellent negative predictive value.