Development and evaluation of RapTAT: A machine learning system for concept mapping of phrases from medical narratives

Rapid, automated determination of the mapping of free text phrases to pre-defined concepts could assist in the annotation of clinical notes and increase the speed of natural language processing systems. The aim of this study was to design and evaluate a token-order-specific naïve Bayes-based machine learning system (RapTAT) to predict associations between phrases and concepts. Performance was assessed using a reference standard generated from 2860 VA discharge summaries containing 567,520 phrases that had been mapped to 12,056 distinct Systematized Nomenclature of Medicine – Clinical Terms (SNOMED CT) concepts by the MCVS natural language processing system. It was also assessed on the manually annotated, 2010 i2b2 challenge data. Performance was established with regard to precision, recall, and F-measure for each of the concepts within the VA documents using bootstrapping. Within that corpus, concepts identified by MCVS were broadly distributed throughout SNOMED CT, and the token-order-specific language model achieved better performance based on precision, recall, and F-measure (0.95 ± 0.15, 0.96 ± 0.16, and 0.95 ± 0.16, respectively; mean ± SD) than the bag-of-words based, naïve Bayes model (0.64 ± 0.45, 0.61 ± 0.46, and 0.60 ± 0.45, respectively) that has previously been used for concept mapping. Precision, recall, and F-measure on the i2b2 test set were 92.9%, 85.9%, and 89.2% respectively, using the token-order-specific model. RapTAT required just 7.2 ms to map all phrases within a single discharge summary, and mapping rate did not decrease as the number of processed documents increased. The high performance attained by the tool in terms of both accuracy and speed was encouraging, and the mapping rate should be sufficient to support near-real-time, interactive annotation of medical narratives. These results demonstrate the feasibility of rapidly and accurately mapping phrases to a wide range of medical concepts based on a token-order-specific naïve Bayes model and machine learning.     

Improving knowledge and trust in vaccines: A survey-based assessment of the potential of the European Union Clinical Trial Regulation No 536/2014 plain language summary to increase health literacy

The European Clinical Trial Regulation No 536/2014 is the first mandate for a non-technical, publicly disclosed, plain language summary (PLS) of clinical trial results. This easy-to-understand summary has the potential to inform the public about clinical trial results and thereby improve health literacy in vaccines.To investigate the utility of the PLS, we undertook 2 online surveys (July/October 2020) in the United Kingdom, the United States and India. Participants were selected by quota sampling to ensure representation of gender, age and parental status. Those lacking interest in vaccine clinical research were excluded. In survey 1, participants were questioned about their interest in and expectations of vaccine trial results. In survey 2, the perceptions of participants to a range of written communication styles used in publicly available PLSs were evaluated.A total of 66 (13%) and 122 (29%) individuals were excluded solely due to lack of interest in vaccine clinical research in surveys 1 and 2, respectively; 450 respondents (150/country) completed survey 1 and 300 (100/country) completed survey 2. In survey 1, there was a correlation (p < 0.01) between claimed knowledge of and trust in vaccines. Healthcare professionals were the most trusted source for vaccine information, while vaccine companies were ranked relatively low. In survey 2, infographic PLS formats were considered easiest to understand, most engaging and the strongest communicators. Emphasizing the main points of the infographics in the text did not improve comprehension or recall. Most respondents (86%) indicated that they would like to see this type of communication in the future.Overall, this research suggests that the PLS, by optimizing content and format, has a potential to increase health literacy, and thereby, as part of a wider integrated communication strategy, build vaccine knowledge and confidence.     

Speech,language and communication: a public health issue across the lifecourse

Speech and Language skill is an essential component of normal child development. Speech, language and communication needs are widespread and are an important determinant of outcomes over the life course. This article discusses the evidence for this relationship and what can be done to improve speech, language and communication capacity for children today.     

Story comprehension in pre-readers: understanding goal structure and generating inferences when a story has competing goals

ABSTRACT

The purpose of this study was to explore pre-readers’ comprehension of a story with competing character goals. Fifty-eight children in three age groups (2½–3 years; 3–4 years; and 4–5 years) were read a story in which the protagonist’s goal was unstated and conflicted with the goal of the secondary character. Understanding of explicit story content was tested through a joint story retell, during which children were prompted to supply 10 story elements. Children in the two older groups supplied more than half of the elements; those in the youngest group obtained a significantly lower score, supplying an average of two elements. Responses to questions suggested that a majority of participants in each age group had accurately inferred the protagonist’s goal. However, few children demonstrated that they had linked actions with goal, an indication that they had difficulty grasping the goal structure of a story with competing goals.     

Promoting and Advancing Health Literacy at NYU Langone Health

NYU Langone Health is one of the nation’s preeminent academic medical centers. Their mission is to serve, teach, and discover. This is achieved through an integrated academic culture devoted to excellence in patient care, education, and research. NYU Langone promotes and advances health literacy through Patient and Family Education Services of the Patient Experience Department. Since the institution provides health care services to dozens of neighborhoods in New York City and beyond, health literacy, as well as cultural sensitivity, are top priorities.     

Commentaries: the Turn of the Scrooge: One Yank's Perspective on Accent Reduction

This study uses a qualitative approach to examine common techniques used by speech-language pathologists (SLPs) to manage fatigue in communication intervention following traumatic brain injury (TBI). Thirteen SLPs from New Zealand hospital and rehabilitation services completed semi-structured interviews in which they discussed their current management strategies. A number of management-related themes recurred throughout the interviews, highlighting the basis of fatigue-management-focused (FMF) communication intervention. The four fatigue-related themes were: intervention structure, client and family strategies, monitoring by both the client and therapist, and lifestyle and daily activities. From these four themes, a model of current SLP practice was developed for clients with TBI. This model will provide the basis for future studies evaluating the effectiveness of FMF communication intervention.     

Qualitative research within the Deaf community in Northern Ireland: A multilingual approach

In this methodology article, the authors illustrate how they conducted multilingual qualitative research in an exploration of the barriers that Deaf people in Northern Ireland face when attempting to access the system of justice. The authors’ research practices are informed, to the extent possible, by the principles of community-based participatory research (CBPR). They explore the challenges of conducting research in American Sign Language (ASL), British Sign Language (BSL), and Irish Sign Language (ISL), and spoken English, facilitated by sign language interpreters fluent in BSL and ISL. Centering the research on the lived experiences of Deaf people who navigate the system of justice, the authors implemented CBPR-informed research methods, which ultimately led to sustained discussion and joint action by the authors and members of the Northern Ireland Deaf community aimed at the removal of barriers that Deaf people face when interacting with the justice system. By writing about their methodological approach in Northern Ireland, the authors wish to be transparent about their work in the hope that other researchers can replicate their successes and avoid the limitations of conducting this work in partnership with members of the Deaf community in other countries.     

How do deaf speakers talk about time?

The grammatical and lexical aspects of temporal reference in the expressive language of prelingually deaf speakers in comparison with normally hearing age-matched speakers were examined. Significant differences emerged between groups in the grammaticalization of time. The deaf subjects used fewer verb inflections to mark time and made many errors with inflections. In addition, a significant triple interaction was found between group × verb typology × verb inflection, which suggested that the deaf were using the verb inflections to code the aspectual features of completion and duration inherent in verbs. Lexical aspects of time reference were found to be similar for both groups. The results are discussed with reference to the normal development of linguistic expressions of time in children acquiring language, and the implications for language therapy with the deaf are considered.