A Multidisciplinary Assessment of ChatGPT's Knowledge of Amyloidosis: Observational Study.

BACKGROUND: Amyloidosis, a rare multisystem condition, often requires complex, multidisciplinary care. Its low prevalence underscores the importance of efforts to ensure the availability of high-quality patient education materials for better outcomes. ChatGPT (OpenAI) is a large language model powered by artificial intelligence that offers a potential avenue for disseminating accurate, reliable, and accessible educational resources for both patients and providers. Its user-friendly interface, engaging conversational responses, and the capability for users to ask follow-up questions make it a promising future tool in delivering accurate and tailored information to patients. OBJECTIVE: We performed a multidisciplinary assessment of the accuracy, reproducibility, and readability of ChatGPT in answering questions related to amyloidosis. METHODS: In total, 98 amyloidosis questions related to cardiology, gastroenterology, and neurology were curated from medical societies, institutions, and amyloidosis Facebook support groups and inputted into ChatGPT-3.5 and ChatGPT-4. Cardiology- and gastroenterology-related responses were independently graded by a board-certified cardiologist and gastroenterologist, respectively, who specialize in amyloidosis. These 2 reviewers (RG and DCK) also graded general questions for which disagreements were resolved with discussion. Neurology-related responses were graded by a board-certified neurologist (AAH) who specializes in amyloidosis. Reviewers used the following grading scale: (1) comprehensive, (2) correct but inadequate, (3) some correct and some incorrect, and (4) completely incorrect. Questions were stratified by categories for further analysis. Reproducibility was assessed by inputting each question twice into each model. The readability of ChatGPT-4 responses was also evaluated using the Textstat library in Python (Python Software Foundation) and the Textstat readability package in R software (R Foundation for Statistical Computing). RESULTS: ChatGPT-4 (n=98) provided 93 (95%) responses with accurate information, and 82 (84%) were comprehensive. ChatGPT-3.5 (n=83) provided 74 (89%) responses with accurate information, and 66 (79%) were comprehensive. When examined by question category, ChatGTP-4 and ChatGPT-3.5 provided 53 (95%) and 48 (86%) comprehensive responses, respectively, to "general questions" (n=56). When examined by subject, ChatGPT-4 and ChatGPT-3.5 performed best in response to cardiology questions (n=12) with both models producing 10 (83%) comprehensive responses. For gastroenterology (n=15), ChatGPT-4 received comprehensive grades for 9 (60%) responses, and ChatGPT-3.5 provided 8 (53%) responses. Overall, 96 of 98 (98%) responses for ChatGPT-4 and 73 of 83 (88%) for ChatGPT-3.5 were reproducible. The readability of ChatGPT-4's responses ranged from 10th to beyond graduate US grade levels with an average of 15.5 (SD 1.9). CONCLUSIONS: Large language models are a promising tool for accurate and reliable health information for patients living with amyloidosis. However, ChatGPT's responses exceeded the American Medical Association's recommended fifth- to sixth-grade reading level. Future studies focusing on improving response accuracy and readability are warranted. Prior to widespread implementation, the technology's limitations and ethical implications must be further explored to ensure patient safety and equitable implementation.

Hospitalizations and Mortality From Myasthenia Gravis: Trends From 2 US National Datasets

BACKGROUND AND OBJECTIVES: Myasthenia gravis (MG) is a rare neuromuscular disorder where IgG antibodies damage the communication between nerves and muscles, leading to muscle weakness that can be severe and have a significant impact on patients' lives. MG exacerbations include myasthenic crisis with respiratory failure, the most serious manifestation of MG. Recent studies have found MG prevalence increasing, especially in older patients. This study examined trends in hospital admissions and in-hospital mortality for adult patients with MG and readmissions and postdischarge mortality in older (65 years or older) adults with MG. METHODS: Data from the Nationwide Inpatient Sample (NIS), an all-payer national database of hospital discharges, were used to characterize trends in hospitalizations and in-hospital mortality related to MG exacerbations and MG crisis among adult patients aged 18 years or older. The Medicare Limited Data Set, a deidentified, longitudinal research database with demographic, enrollment, and claims data was used to assess hospitalizations, length of stay (LOS), readmissions, and 30-day postdischarge mortality among fee-for-service Medicare beneficiaries aged 65 years or older. The study period was 2010-2019. Multinomial logit models and Poisson regression were used to test for significance of trends. RESULTS: Hospitalization rates for 19,715 unique adult patients and 56,822 admissions increased from 2010 to 2019 at an average annualized rate of 4.9% (MG noncrisis: 4.4%; MG crisis: 6.8%; all p < 0.001). Readmission rates were approximately 20% in each study year for both crisis and noncrisis hospitalizations; the in-hospital mortality rate averaged 1.8%. Among patients aged 65 years or older, annualized increases in hospitalizations were estimated at 5.2%, 4.2%, and 7.7% for all, noncrisis, and crisis hospitalizations, respectively (all p < 0.001). The average LOS was stable over the study period, ranging from 11.3 to 13.1 days, but was consistently longer for MG crisis admissions. Mortality among patients aged 65 years or older was higher compared with that in all patients, averaging 5.0% across each of the study years. DISCUSSION: Increasing hospitalization rates suggest a growing burden associated with MG, especially among older adults. While readmission and mortality rates have remained stable, the increasing hospitalization rates indicate that the raw numbers of readmissions-and deaths-are also increasing. Mortality rates are considerably higher in older patients hospitalized with MG.

Using a low-cost, real-time electronic immunization registry in Pakistan to demonstrate utility of data for immunization programs and evidence-based decision making to achieve SDG-3: Insights from analysis of Big Data on vaccines.

BACKGROUND: Despite the proliferation of digital interventions such as Electronic Immunization Registries (EIR), currently, there is little evidence regarding the use of EIR data to improve immunization outcomes in resource-constrained settings. To achieve the Sustainable Development Goal (SDG) of ensuring healthy lives and well-being for all ages, particularly for newborns and children under the age of 5 (goal 3b), it is essential to generate and use quality data for evidence-based decision making to overcome barriers inherent in immunization systems. In Pakistan, only 66 % of children receive all basic vaccinations, and in Sindh province, the number is even lower at 49 %. In 2012, IRD developed and piloted Zindagi Mehfooz (Safe Life; ZM) ElR, an Android-based platform that records and analyses individual-level child data in real-time. In 2017 in collaboration with Expanded Programme for Immunization (EPI) Sindh, ZM was scaled-up across the entire Sindh province and is currently being used by 2521 government vaccinators in 1539 basic health facilities, serving >48 million population. OBJECTIVE: The study aims to demonstrate how big immunization data from the ZM-EIR is being leveraged in Sindh, Pakistan for actionable decision making via three use cases (a) improving performance management of vaccinators to increase geographical coverage, (b) quantifying the impact of provincial accelerated outreach activities, and (c) examining the impact of the COVID-19 pandemic on routine immunization coverage to help devise a tailored approach for future efforts. METHODS: From October 2017 to April 2020, more than 2.9 million children and 0.9 million women have been enrolled, and more than 22 million immunization events have been recorded in the ZM-EIR. We extracted de-identified data from ZM-EIR for January 1, 2019 - April 20, 2020, period. Given the needs of each use case, monthly and daily indicators on vaccinator performance (attendance and compliance), daily immunization visits, and the number of antigens administered were calculated. Geo-coordinate data of antigen administration was extracted and displayed on geographic maps using QGIS. All generated reports were shared at fixed frequency with various stakeholders, such as partners at EPI-Sindh, for utilization in decision making and informing policy. RESULT: Our use-cases demonstrate the use of EIR data for data-driven decision making. From January - December 2019, the monthly monitoring of program indicators helped increase the vaccinator attendance from 44% to 88%, while an 85 % increase in geographical coverage was observed in a polio-endemic super high-risk union council (SHRUC) in Karachi. The analysis of daily average antigens administered during accelerated outreach efforts (AOE) as compared to routine activities showed an increase in average daily Pentavalent-3, Measles-1, and Measles-2 vaccines administered by 103%, 154%, and 180% respectively. These findings helped decide to continue the accelerated effort in high-risk areas (compared to the entire province) rather than discontinuing the activity due to high costs. During COVID-19 lockdown, the daily average number of child immunizations reduced from 16,649 to 4335 per day, a decline of 74% compared to 6 months preceding COVID-19 lockdown. ZM-EIR data is currently helping to shape the planning and implementation of critical strategies to mitigate the impact of the COVID-19 pandemic. CONCLUSION: The big data for vaccines generated through EIRs is a powerful tool to monitor immunization work-force and ensure chronically missed communities are identified and covered through targeted strategies. Geospatial data availability and analysis is changing the way EPI review meetings occur with stakeholders, taking data-driven decisions for better planning and resource allocation. In the fight against COVID-19 pandemic, as governments gradually begin to shift from containing the outbreak to strategizing a plan for sustaining the essential health services, the countries that will emerge most successful are likely the ones who can best use technology and real-time data for targeted efforts.

Engaging the private sector to increase tuberculosis case detection: an impact evaluation study.

BACKGROUND: In many countries with a high burden of tuberculosis, most patients receive treatment in the private sector. We evaluated a multifaceted case-detection strategy in Karachi, Pakistan, targeting the private sector. METHODS: A year-long communications campaign advised people with 2 weeks or more of productive cough to seek care at one of 54 private family medical clinics or a private hospital that was also a national tuberculosis programme (NTP) reporting centre. Community laypeople participated as screeners, using an interactive algorithm on mobile phones to assess patients and visitors in family-clinic waiting areas and the hospital's outpatient department. Screeners received cash incentives for case detection. Patients with suspected tuberculosis also came directly to the hospital's tuberculosis clinic (self-referrals) or were referred there (referrals). The primary outcome was the change (from 2010 to 2011) in tuberculosis notifications to the NTP in the intervention area compared with that in an adjacent control area. FINDINGS: Screeners assessed 388,196 individuals at family clinics and 81,700 at Indus Hospital's outpatient department from January-December, 2011. A total of 2416 tuberculosis cases were detected and notified via the NTP reporting centre at Indus Hospital: 603 through family clinics, 273 through the outpatient department, 1020 from self-referrals, and 520 from referrals. In the intervention area overall, tuberculosis case notification to the NTP increased two times (from 1569 to 3140 cases) from 2010 to 2011--a 2·21 times increase (95% CI 1·93-2·53) relative to the change in number of case notifications in the control area. From 2010 to 2011, pulmonary tuberculosis notifications at Indus Hospital increased by 3·77 times for adults and 7·32 times for children. INTERPRETATION: Novel approaches to tuberculosis case-finding involving the private sector and using laypeople, mobile phone software and incentives, and communication campaigns can substantially increase case notification in dense urban settings. FUNDING: TB REACH, Stop TB Partnership.