Dataset Information

Multistage BiCross encoder for multilingual access to COVID-19 health information.

ABSTRACT: The Coronavirus (COVID-19) pandemic has led to a rapidly growing 'infodemic' of health information online. This has motivated the need for accurate semantic search and retrieval of reliable COVID-19 information across millions of documents, in multiple languages. To address this challenge, this paper proposes a novel high precision and high recall neural Multistage BiCross encoder approach. It is a sequential three-stage ranking pipeline which uses the Okapi BM25 retrieval algorithm and transformer-based bi-encoder and cross-encoder to effectively rank the documents with respect to the given query. We present experimental results from our participation in the Multilingual Information Access (MLIA) shared task on COVID-19 multilingual semantic search. The independently evaluated MLIA results validate our approach and demonstrate that it outperforms other state-of-the-art approaches according to nearly all evaluation metrics in cases of both monolingual and bilingual runs.

SUBMITTER: Singh I

PROVIDER: S-EPMC8423231 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Introduction: Maternal mortality continues to be one of the biggest challenges of the health system in Kenya. Informal settlements in Kenya have been known to have higher rates of maternal mortality and also receive maternity services of varied quality. Data assessing progress on key maternal health indicators within informal settlements are also often scarce. The COVID-19 pandemic hit Kenya in March this year and so far, the impact of the pandemic on access to maternal health has not been established. This study aims to add to the body of knowledge by investigating the effects of the COVID-19 pandemic and mitigation strategies on access to health care services in informal settlements. Methods: Qualitative methods using in-depth interviews were used to assess women's experiences of maternity care during the COVID-19 era and the impact of proposed mitigation strategies such as the lockdown and the curfew. Other aspects of the maternity experience such as women's knowledge of COVID-19, their perceived risk of infection, access to health facilities, perceived quality of care were assessed. Challenges that women facing as a result of the lockdown and curfew with respect to maternal health access and quality were also assessed. Results: Our findings illustrate that there was a high awareness of the symptoms and preventative measures for COVID-19 amongst women in informal settlements. Our findings also show that women's perception of risk to themselves was high, whereas risk to family and friends, and in their neighborhood was perceived as low. Less than half of women reported reduced access due to fear of contracting Coronavirus, Deprioritization of health services, economic constraints, and psychosocial effects were reported due to the imposed lockdown and curfew. Most respondents perceived improvements in quality of care due to short-waiting times, hygiene measures, and responsive health personnel. However, this was only reported for the outpatient services and not in-patient services. Conclusion: The most important recommendation was for the Government to provide food followed by financial support and other basic amenities. This has implications for the Government's mitigation measures that are focused on public health measures and lack social safety-net approaches for the most vulnerable communities.

Project description:An essential component of quality care for limited English proficient (LEP) patients is language access. Linguistically accessible medication instructions are particularly important, given the serious consequences of error and patient responsibility for managing often complex medication regimens on their own. Approximately 21 million people in the U.S. were LEP at the time of the 2000 census, representing a 50% increase since 1990. Little information is available on their access to comprehensible medication instructions. In an effort to address this knowledge gap, we conducted a telephone survey of 200 randomly selected NYC pharmacies. The primary focus of the survey was translation need, capacity, and practice. The majority of pharmacists reported that they had LEP patients daily (88.0%) and had the capacity to translate prescription labels (79.5%). Among pharmacies serving LEP patients on a daily basis, just 38.6% translated labels daily; 22.7% never translated. In multivariate analysis, pharmacy type (OR = 4.08, 95%CI = 1.55-10.74, independent versus chain pharmacies) and proportion of Spanish-speaking LEP persons in the pharmacy's census tract (OR = 1.09, 95%CI = 1.05-1.13 for each 1% increase in Spanish LEP population) were associated with increased label translation. Although 88.5% of the pharmacies had bilingual staff, less than half were pharmacists or pharmacy interns and thus qualified to provide medication counseling. More than 80% of the pharmacies surveyed lacked systematic methods for identifying linguistic needs and for informing patients of translation capabilities. Consistent with efforts to improve language access in other health care settings, the critical gap in language appropriate pharmacy services must be addressed to meet the needs of the nation's large and ever-growing immigrant communities. Pharmacists may require supplemental training on the need and resources for meeting the verbal and written language requirements of their LEP patients. Dispensing software with accurate translation capability and telephonic interpretation services should be utilized in pharmacies serving LEP patients. Pharmacists should post signs and make other efforts to inform patients about the language resources available to them.

Dataset Information

Multistage BiCross encoder for multilingual access to COVID-19 health information.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets