Introduction

Pharmacovigilance (PV) is the “science and activities relating to the detection, assessment, understanding, and prevention of adverse effects or any other possible drug-related problems” (WHO, 2015). PV practices for most cases depend on analysing clinical trials, biomedical writing, observational examinations, Electronic Health Records (EHRs), social media and Spontaneous Reporting (SR). Where Health Care Professionals (HCPs), producers or patients send suspected Adverse Drug Reactions (ADRs) to a national PV centre (Harpaz et al., 2014). ADRs marked by scholars as one of the significant cause of morbidity and mortality around the world, In western countries 1% to 35% of the total hospital admissions caused by the ADRs effects(Alexopoulou et al., 2008).

Study by U.S Food and Drug Administration (FDA) content that annual death count that range from 44,000 to 98,000 due to medication errors, 7000 deaths occurred due to ADRs (U.S. Department of Health and Human Services, 2018). In 2015 a total of 253,017 serious adverse event reports received by FDA out of those 44,693 deaths were associated with ADRs (FAERS, 2015). In South Africa, 2.9% of the medical admissions death were contributed by ADRs, in which 56 of 357 deaths (16%) were ADR cases (Mouton et al., 2015). Hence post-market drug surveillance is significant  n recognising potential Adverse Reactions(ARs). The existing system of post-market  surveillance can be slow and under-proficient because 94% of ADRs cases are under-reported (Hazell & Shakir, 2006).

Problem Statement

According to the Tanzania Medicines And Medical Devices Authority(TMDA) pharmacovigilance guideline, all healthcare practitioner that interact with patients and consumer of the medicinal products are responsible for reporting ADRs (Mugoyela, Robert, & Masota, 2018). The goal is to detect adverse reaction as early as possible, especially severe, unknown and infrequent reactions so to monitor them within the population.

However, as it was analysed in figure 1, Healthcare workers sometimes do not report ADRs due to complacency, insecurity, diffidence, indifference, ignorance, fear of medico-legal consequences and lack of time to complete the form diagnosis (Biagi et al., 2013). Recently, many hospitals have introduced the Electronic Medical Records (EMRs). Some of these systems include a Clinical Data Warehouse (CDW) for the secondary use of the clinical data, which includes data relating to drug safety (Coloma et al., 2013;  arpaz et al., 2013). These sorts of information are commonly gathered routinely during administrative processes and clinical practice by various healthcare professionals (Trifirò, Sultana, & Bate, 2018).

Despite the availability of electronic healthcare data, there is no consensus on the best methods of identifying adverse reactions from these data sources (Yom-Tov & Gabrilovich, 2013). EMR holds the promise about active monitoring of ADR, Harpaz et al. (2013), suggest that extracting clinical narratives from EMR can lead to a significantly large improvement in adverse events detection.

This study proposes the real-time NLP framework to auto-extract ADR cases from clinical health records. Despite several efforts done by previous scholars, there are still challenges which need to be addressed. This study is a step taken to improve the pharmacovigilance by automatically signal the presences of adverse events in real-time.

Objectives

The main objective of the study is to improve the Pharmacovigilance system by proposing the NLP Framework for automating the extraction of ADR cases from the EMRs. Specific Objectives:

  1. Analyse Electronic Medical Records in UDOM hospitals and collect data sets
  2. Develop the real-time NLP Framework for improving reporting of ADR cases from EMRs
  3. To demonstrate and evaluate the developed framework in UDOM hospitals.

References

Alexopoulou, A., Dourakis, S. P., Mantzoukis, D., Pitsariotis, T., Kandyli, A., Deutsch, M., & Archimandritis, A. J. (2008). Adverse drug reactions as a cause of hospital admissions: A 6-month experience in a single center in Greece. European Journal of Internal Medicine, 19(7), 505–510. https://doi.org/10.1016/j.ejim.2007.06.030

Biagi, C., Montanaro, N., Buccellato, E., Roberto, G., Vaccheri, A., & Motola, D. (2013). Underreporting in pharmacovigilance: An intervention for Italian GPs (Emilia-Romagna region). European Journal of Clinical Pharmacology, 69(2), 237–244. https://doi.org/10.1007/s00228-012-1321-7

Coloma, P. M., Avillach, P., Salvo, F., Schuemie, M. J., Ferrajolo, C., Pariente, A., … Trifirò, G. (2013). A reference standard for evaluation of methods for drug safety signal detection using electronic healthcare record databases. Drug Safety, 36(1), 13–23. https://doi.org/10.1007/s40264-012-0002-x

FAERS. (2015). FDA Adverse Event Reporting System  FAERS) – FAERS Reporting by Patient Outcomes by Year. Retrieved April 8, 2019, from https://www.fda.gov/drugs/guidancecomplianceregulatoryinformation/surveillance/ adversedrugeffects/ucm070461.htm

Harpaz, R., Callahan, A., Tamang, S., Low, Y., Odgers, D., Finlayson,  ., … Shah, N. H. (2014). Text Mining for Adverse Drug Events: the Promise, Challenges, and State of the Art. Drug Safety, 37(10), 777–790. https://doi.org/10.1007/s40264-014-0218-z Harpaz, R., Vilar, S.,  uMouchel, W., Salmasian, H., Haerian, K., Shah, N. H., … Friedman, C. (2013). Combing signals from spontaneous reports and electronic health records for detection of adverse drug reactions. Journal of the American Medical Informatics Association, 20(3), 413–419. https://doi.org/10.1136/amiajnl-2012- 00930

Hazell, L., & Shakir, S. A. W. (2006). Under-reporting of adverse drug reactions. British Medical Journal (Clinical Research Ed.), 290(6478), 1355.  https://doi.org/10.1136/bmj.290.6478.1355

Johnson, A. E. W., Stone, D. J., Celi, L. A., & Pollard, T. J. (2018). The MIMIC Code Repository: Enabling reproducibility in critical care research. Journal of the American Medical Informatics Association, 25(1), 32–39. https://doi.org/10.1093/jamia/ocx084

Mouton, J. P., Mehta, U., Parrish, A. G., Wilson, D. P. K., Stewart, A., Njuguna, C. W., … Cohen, K. (2015). Mortality from adverse drug reactions in adult medical inpatients at four hospitals in South Africa: A cross-sectional survey. British Journal of Clinical Pharmacology, 80(4), 818–826. https://doi.org/10.1111/bcp.12567

Mugoyela, V., Robert, R., & Masota, N. (2018). Investigation of Factors Affecting Preparedness of Reporting Adverse Drug Reactions among Nurses in Public and Private Hospitals in Dar Es Salaam, Tanzania. Pharmacology & Pharmacy, 09(01), 38–51. https://doi.org/10.4236/pp.2018.91004

Trifirò, G., Sultana, J., & Bate, A. (2018). From Big Data to Smart Data for Pharmacovigilance: The Role of Healthcare Databases and Other Emerging Sources. Drug Safety, 41(2), 143–149. https://doi.org/10.1007/s40264-017-0592-4

U.S. Department of Health and Human Services. (2018). Drug Interactions; Labeling – Preventable Adverse Drug Reactions: A Focus on Drug Interactions. Retrieved April 11, 2019, from https://www.fda.gov/drugs/developmentapprovalprocess/developmentresources/druginteractionslabeling/ ucm110632.htm

WHO. (2015). WHO | Pharmacovigilance. Retrieved April 8, 2019, from WHO website: http://www.who.int/medicines/areas/quality_safety/safety_efficacy/pharmvigi/en/

Yom-Tov, E., & Gabrilovich, E. (2013). Postmarket drug surveillance without trial costs: Discovery of adverse drug
reactions through large-scale analysis of web search queries. Journal of Medical Internet Research, 15(6), 1–12. https://
doi.org/10.2196/jmir.26144