Authors: Suhana Bedi, Yutong Liu, Lucy Orr-Ewing...

IMPORTANCE

Large language models (LLMs) can assist in various health care activities, but current evaluation approaches may not adequately identify the most useful application areas.

OBJECTIVE

To summarize existing evaluations of LLMs in health care in terms of 5 components: (1) evaluation data type, (2) health care task, (3) natural language processing (NLP) and natural language understanding (NLU) tasks, (4) dimension of evaluation, and (5) medical specialty.

DATA SOURCES

A systematic search of PubMed and Web of Science was performed for studies published between January 1, 2022, and February 19, 2024.

STUDY SELECTION

Studies evaluating 1 or more LLMs in health care.

DATA EXTRACTION AND SYNTHESIS

Three independent reviewers categorized studies via keyword searches based on the data used, the health care tasks, the NLP and NLU tasks, the dimensions of evaluation, and the medical specialty.

RESULTS

Of 519 studies reviewed, published between January 1, 2022, and February 19, 2024, only 5% used real patient care data for LLM evaluation. The most common health care tasks were assessing medical knowledge such as answering medical licensing examination questions (44.5%) and making diagnoses (19.5%). Administrative tasks such as assigning billing codes (0.2%) and writing prescriptions (0.2%) were less studied. For NLP and NLU tasks, most studies focused on question answering (84.2%), while tasks such as summarization (8.9%) and conversational dialogue (3.3%) were infrequent. Almost all studies (95.4%) used accuracy as the primary dimension of evaluation; fairness, bias, and toxicity (15.8%), deployment considerations (4.6%), and calibration and uncertainty (1.2%) were infrequently measured. Finally, in terms of medical specialty area, most studies were in generic health care applications (25.6%), internal medicine (16.4%), surgery (11.4%), and ophthalmology (6.9%), with nuclear medicine (0.6%), physical medicine (0.4%), and medical genetics (0.2%) being the least represented.

CONCLUSIONS AND RELEVANCE

Existing evaluations of LLMs mostly focus on accuracy of question answering for medical examinations, without consideration of real patient care data. Dimensions such as fairness, bias, and toxicity and deployment considerations received limited attention. Future evaluations should adopt standardized applications and metrics, use clinical data, and broaden focus to include a wider range of tasks and specialties.

Topics: Natural Language Processing; Humans; Delivery of Health Care; Language; Medicine

PubMed: 39405325
DOI: 10.1001/jama.2024.21700

Review

Authors: S J Hiltabiddle

OBJECTIVE

To review factors associated with adolescent condom use for the prevention of sexually transmitted diseases (STDs).

DATA SOURCES

Thirty-six references on adolescent development, STDs, STD risk factors, and factors influencing condom use among heterosexual adolescents.

STUDY SELECTION

Twenty-two research articles addressing issues influencing condom use by heterosexual adolescents.

DATA SYNTHESIS

This review identifies unique risk factors related to adolescents and their risk for STDs, barriers to and facilitators of condom use, and suggestions for health care providers to increase condom use among adolescents.

CONCLUSIONS

Although research on condom use among adolescents has inherent difficulties, current findings, along with reported high rates of STDs in this population, indicate infrequent and inconsistent condom use. Many factors contribute to the motivation for condom use and should be assessed individually. Using the Health Belief Model as a theoretical framework, health care providers can guide the adolescent to make realistic risk assessments and identify positive ways of incorporating condoms into their sexual lives. Further research then must be conducted to test the effectiveness of this approach.

Topics: Adolescent; Adolescent Behavior; Condoms; Female; Health Behavior; Health Knowledge, Attitudes, Practice; Humans; Male; Models, Psychological; Psychology, Adolescent; Risk Factors; Sex Education; Sexual Behavior; Sexually Transmitted Diseases

PubMed: 8627404
DOI: 10.1111/j.1552-6909.1996.tb02514.x

Review

Authors: Zainul Abedeen Sobani, Shabbir Akhtar, Montasir Junaid...

Teratocarcinosarcoma is a rare, morphologically heterogeneous and highly malignant neoplasm. It is characterized by the presence of benign and malignant epithelial, mesenchymal and neural components. The carcinoma may be either squamous or adenocarcinoma and the mesenchymal component may manifest spindle, smooth, skeletal muscle, cartilage and bone features. Because of their infrequency, these lesions are often misdiagnosed, leading to management difficulties. In this case report we have shared our experience with sinonasal teratocarcinosarcoma in a 23 year old female and performed a brief review of literature.

Topics: Carcinosarcoma; Combined Modality Therapy; Diagnosis, Differential; Female; Humans; Magnetic Resonance Imaging; Nose Neoplasms; Teratoma; Young Adult

PubMed: 22755362
DOI: No ID Found

Authors: Maria Francesca Spina, Paolo Innocente Banfi, Antonello Nicolini...

Thunderstorm asthma is a rare event: in this letter we describe two cases observed during the same month of 2018 at an Italian Emergency department, assessed by the same medical team and according to the same methodology and approach. Given the infrequency of such a phenomenon and the debate around its nature, frequency, and - at times - existence, we strongly believe it is important for all specialists who observe such cases to report them, building an evidence base to expand its knowledge and understanding.

PubMed: 32983454
DOI: 10.4081/mrm.2020.679

Authors: Sa-Eeda Chippendale, Kirsten Reichmuth, Margitta Worm...

INTRODUCTION

Anaphylaxis is a severe, life-threatening generalized hypersensitivity reaction. While guidelines to reduce the morbidity, risk, and mortality of anaphylaxis are widely available, adherence to these is often suboptimal. We aimed to audit paediatric anaphylaxis at a South African tertiary allergy referral centre, comparing our data to those of the large Network of Severe Allergic Reactions (NORA) registry.

METHODS

Children treated for severe allergic reactions between January 2014 and August 2016 were identified for screening using ICD-10 coding of all admissions and discharges, pharmacy records of adrenaline autoinjector dispensing, and additional referrals from the allergy department to the study. Screened participants not meeting the inclusion criteria after preliminary questioning and/or folder review were excluded. Data were collected via a standardized questionnaire using direct interviews, and captured on a local web-based registry.

RESULTS

Of the 156 episodes analysed, >40% were graded as severe and nearly two-thirds of patients were seen for a recurrent episode. Males, younger children, and individuals of mixed-race ethnicity were more frequently affected. Skin and mucosa were most commonly involved, followed by respiratory and gastrointestinal involvement; cardiovascular and other systemic involvement occurred infrequently. Specific IgE assay was the most frequently requested test. Food-related triggers (peanut, hen's egg, fish, cashew nuts and cows' milk) predominated and decreased with age. Anaphylaxis was strongly correlated with atopic conditions. While prophylactic measures were almost universally instituted, adrenaline was rarely used, by both lay persons and healthcare professionals. Hospital admissions were infrequent, and no deaths were recorded.

CONCLUSION

Management of anaphylaxis can be improved. Specifically, the use of adrenaline prior to hospital arrival remains suboptimal. Ongoing education and training of patients, parents, teachers, and healthcare workers is identified as an area requiring intensification.

PubMed: 36185548
DOI: 10.1016/j.waojou.2022.100666

Review

Authors: N Godinho, J M Lopes, V V Augusto...

The authors describe a case of Sheehan's syndrome that followed hemorrhagic complications due to late abortion, and emphasise the infrequency of changes in plasma lipid levels in this disease. They report a particular case of hypopituitarism and point out the main concepts of pathophysiology, diagnosis, therapeutics and complications that follow this increasingly rarer condition.

Topics: Abdominal Pain; Abortion, Induced; Adult; Chronic Disease; Female; Humans; Hypopituitarism; Pregnancy; Pregnancy Trimester, Third; Uterine Hemorrhage

PubMed: 9312986
DOI: No ID Found

Review

Authors: Almudena Hernandez-Milian, Antoni Payeras-Cifre

Listeriosis is a disease caused by Listeria monocytogenes (L. monocytogenes). L. monocytogenes is bacteria that usually infects some determined inhabitants, especially high risk patients such as the elderly, immunosuppressed patients and pregnant women. However, it can also affect people who do not have these risk factors. L. monocytogenes is widespread in nature being part of the faecal flora of many mammals and it is a common foodborne source. It is acquired by humans primarily through consumption of contaminated food. Besides, between 1% and 10% of the population is a faecal carrier of L.monocytogenes. Listeriosis may occur sporadically or in outbreaks. Infection causes a spectrum of illness, ranging from febrile gastroenteritis to invasive disease, including bacteraemia, sepsis, and meningoencephalitis. This infection has a low incidence, although it is undeniably increasing, particularly due to the rise of population of over 60 years old or of under 60 years olds with a predisposing condition. The diagnosis is complicated because of its incubation period and the different clinical manifestations. Also listeriosis has a high mortality despite adequate and early treatment. The importance of bacteraemia for L. monocytogenes lies in the infrequency of this bacterium and the high mortality, even with appropriate antibiotic treatment.

Topics: Foodborne Diseases; Humans; Listeria monocytogenes; Listeriosis

PubMed: 24822197
DOI: 10.1155/2014/358051

Authors: Sajid Mahmood, Muhammad Shahbaz, Aziz Guergachi...

Association rule mining research typically focuses on positive association rules (PARs), generated from frequently occurring itemsets. However, in recent years, there has been a significant research focused on finding interesting infrequent itemsets leading to the discovery of negative association rules (NARs). The discovery of infrequent itemsets is far more difficult than their counterparts, that is, frequent itemsets. These problems include infrequent itemsets discovery and generation of accurate NARs, and their huge number as compared with positive association rules. In medical science, for example, one is interested in factors which can either adjudicate the presence of a disease or write-off of its possibility. The vivid positive symptoms are often obvious; however, negative symptoms are subtler and more difficult to recognize and diagnose. In this paper, we propose an algorithm for discovering positive and negative association rules among frequent and infrequent itemsets. We identify associations among medications, symptoms, and laboratory results using state-of-the-art data mining technology.

Topics: Algorithms; Data Mining; Databases, Factual; Information Storage and Retrieval

PubMed: 24955429
DOI: 10.1155/2014/973750