Effective identification of previously implausible safety signals is a core component of successful pharmacovigilance. Timely, reliable, and efficient data ingestion and related processing are critical to this. The term 'black swan events' was coined by Taleb to describe events with three attributes: unpredictability, severe and widespread consequences, and retrospective bias. These rare events are not well understood at their emergence but are often rationalized in retrospect as predictable. Pharmacovigilance strives to rapidly respond to potential black swan events associated with medicine or vaccine use. Machine learning (ML) is increasingly being explored in data ingestion tasks. In contrast to rule-based automation approaches, ML can use historical data (i.e., 'training data') to effectively predict emerging data patterns and support effective data intake, processing, and organisation. At first sight, this reliance on previous data might be considered a limitation when building ML models for effective data ingestion in systems that look to focus on the identification of potential black swan events. We argue that, first, some apparent black swan events-although unexpected medically-will exhibit data attributes similar to those of other safety data and not prove algorithmically unpredictable, and, second, standard and emerging ML approaches can still be robust to such data outliers with proper awareness and consideration in ML system design and with the incorporation of specific mitigatory and support strategies. We argue that effective approaches to managing data on potential black swan events are essential for trust and outline several strategies to address data on potential black swan events during data ingestion.

Department of Medicine at NYU Grossman School of Medicine New York USA

Department of Non Communicable Disease Epidemiology London School of Hygiene and Tropical Medicine London UK

Global Safety GSK 980 Great West Road Brentford TW8 9GS Middlesex UK

R and D IT MSD Prague Czech Republic

Zobrazit více v PubMed

Lindquist M. Data quality management in pharmacovigilance. Drug Saf. 2004;27(12):857–870. doi: 10.2165/00002018-200427120-00003. PubMed DOI

Stergiopoulos S, Fehrle M, Caubel P, Tan L, Jebson L. Adverse drug reaction case safety practices in large biopharmaceutical organizations from 2007 to 2017: an industry survey. Pharmaceut Med. 2019;33(6):499–510. PubMed

Bate A, Stegmann JU. Safety of medicines and vaccines—building next generation capability. Trends Pharmacol Sci. 2021;42(12):1051–1063. doi: 10.1016/j.tips.2021.09.007. PubMed DOI

Ghosh R, Kempf D, Pufko A, Barrios Martinez LF, Davis CM, Sethi S. Automation opportunities in pharmacovigilance: an industry survey. Pharmaceut Med. 2020;34(1):7–18. PubMed

IBM. What is intelligent automation. 2021. https://www.ibm.com/cloud/learn/intelligent-automation

Bate A, Hobbiger SF. Artificial intelligence, real-world automation and the safety of medicines. Drug Saf. 2021;44(2):125–132. doi: 10.1007/s40264-020-01001-7. PubMed DOI

Lewis DJ, McCallum JF. Utilizing advanced technologies to augment pharmacovigilance systems: challenges and opportunities. Ther Innov Regul Sci. 2020;54(4):888–899. doi: 10.1007/s43441-019-00023-3. PubMed DOI PMC

Kassekert R, Easwar M, Glaser M, Ventham R, Bate A. Automation in routine use for data collection and processing for scalable faster RWE generation. Value Health. 2020 (in Press).

Glaser M, Cranfield C, Dsouza D, Duma A, Hastie K, Kassekert R, et al. Automating individual case safety report identification within scientific literature using natural language processing. Pharmacoepidemiol Drug Saf. 2021;30:118–881.

Rajkomar A, Dean J, Kohane I. Machine learning in medicine. N Engl J Med. 2019;380(14):1347–1358. doi: 10.1056/NEJMra1814259. PubMed DOI

Kumah-Crystal YA, Pirtle CJ, Whyte H, Goode ES, Anders SH, Lehmann CU. Electronic health record interactions through voice: a review. Appl Clin Inform. 2018;9(03):541–552. doi: 10.1055/s-0038-1666844. PubMed DOI PMC

Huysentruyt K, Kjoersvik O, Dobracki P, Savage E, Mishalov E, Cherry M, et al. Validating intelligent automation systems in pharmacovigilance: insights from good manufacturing practices. Drug Saf. 2021;44(3):261–272. doi: 10.1007/s40264-020-01030-2. PubMed DOI PMC

Sessa M, Khan AR, Liang D, Andersen M, Kulahci M. Artificial intelligence in pharmacoepidemiology: a systematic review. Part 1—overview of knowledge discovery techniques in artificial intelligence. Front Pharmacol. 2020;11:1028. doi: 10.3389/fphar.2020.01028. PubMed DOI PMC

Brown EG, Wood L, Wood S. The medical dictionary for regulatory activities (MedDRA) Drug Saf. 1999;20(2):109–117. doi: 10.2165/00002018-199920020-00002. PubMed DOI

Létinier L, Jouganous J, Benkebil M, Bel-Létoile A, Goehrs C, Singier A, et al. Artificial intelligence for unstructured healthcare data: application to coding of patient reporting of adverse drug reactions. Clin Pharmacol Therap. 2021;130:392. doi: 10.1002/cpt.2266. PubMed DOI PMC

Bate A, Evans SJ. Quantitative signal detection using spontaneous ADR reporting. Pharmacoepidemiol Drug Saf. 2009;18(6):427–436. doi: 10.1002/pds.1742. PubMed DOI

Taleb NN. Black swans and the domains of statistics. Am Stat. 2007;61(3):198–200. doi: 10.1198/000313007X219996. DOI

Spiegelhalter D. Risk and uncertainty communication. Annu Rev Stat Appl. 2017;4:31–60. doi: 10.1146/annurev-statistics-010814-020148. DOI

Sandman PM, Miller PM, Johnson BB, Weinstein ND. Agency communication, community outrage, and perception of risk: three simulation experiments. Risk Anal. 1993;13(6):585–598. doi: 10.1111/j.1539-6924.1993.tb01321.x. DOI

Kasperson RE, Renn O, Slovic P, Brown HS, Emel J, Goble R, et al. The social amplification of risk: a conceptual framework. Risk Anal. 1988;8(2):177–187. doi: 10.1111/j.1539-6924.1988.tb01168.x. DOI

Bekiros S, Boubaker S, Nguyen DK, Uddin GS. Black swan events and safe havens: the role of gold in globally integrated emerging markets. J Int Money Financ. 2017;73:317–334. doi: 10.1016/j.jimonfin.2017.02.010. DOI

Osterholm MT, Moore KA, Gostin LO. Public health in the age of Ebola in West Africa. JAMA Intern Med. 2015;175(1):7–8. doi: 10.1001/jamainternmed.2014.6235. PubMed DOI

Gray GL, Alles MG. Measuring a business's grit and survivability when faced with “black swan” events like the coronavirus pandemic. J Emerg Technol Acc. 2021;18(1):195–204. doi: 10.2308/JETA-2020-060. DOI

Yarovaya L, Matkovskyy R, Jalan A. The effects of a “black swan” event (COVID-19) on herding behavior in cryptocurrency markets. J Int Financ Markets Inst Money. 2021;75:101321. doi: 10.1016/j.intfin.2021.101321. DOI

Edwards IR. Causality assessment in pharmacovigilance: still a challenge. Drug Saf. 2017;40(5):365. doi: 10.1007/s40264-017-0509-2. PubMed DOI

Fan BE, Shen JY, Lim XR, Tu TM, Chang CCR, Khin HSW, et al. Cerebral venous thrombosis post BNT162b2 mRNA SARS-CoV-2 vaccination: a black swan event. Am J Hematol. 2021;96(9):E357–E361. doi: 10.1002/ajh.26272. PubMed DOI PMC

Lakshminarayanan B, Pritzel A, Blundell C. Simple and scalable predictive uncertainty estimation using deep ensembles. arXiv preprint arXiv:161201474. 2016.

Farina F, Phillips L, Richmond NJ. Intrinsic uncertainties and where to find them. arXiv preprint arXiv:210702526. 2021.

Ovadia Y, Fertig E, Ren J, Nado Z, Sculley D, Nowozin S, et al. Can you trust your model's uncertainty? Evaluating predictive uncertainty under dataset shift. arXiv preprint arXiv:190602530. 2019.

Hendrycks D, Mazeika M, Dietterich T. Deep anomaly detection with outlier exposure. arXiv preprint arXiv:181204606. 2018.

Finelli LA, Narasimhan V. Leading a digital transformation in the pharmaceutical industry: reimagining the way we work in global drug development. Clin Pharmacol Ther. 2020;108(4):756–761. doi: 10.1002/cpt.1850. PubMed DOI PMC

Blundell C, Cornebise J, Kavukcuoglu K, Wierstra D. Weight uncertainty in neural network. International Conference on Machine Learning; 2015: PMLR; 2015. p. 1613–22.

Shafaei A, Schmidt M, Little JJ. A less biased evaluation of out-of-distribution sample detectors. arXiv preprint arXiv:180904729. 2018.

Meinke A, Bitterwolf J, Hein M. Provably Robust Detection of Out-of-distribution Data (almost) for free. arXiv preprint arXiv:210604260. 2021.

Ditzler G, Roveri M, Alippi C, Polikar R. Learning in nonstationary environments: a survey. IEEE Comput Intell Mag. 2015;10(4):12–25. doi: 10.1109/MCI.2015.2471196. DOI

Finlayson SG, Subbaswamy A, Singh K, Bowers J, Kupke A, Zittrain J, et al. The clinician and dataset shift in artificial intelligence. N Engl J Med. 2021;385(3):283–286. doi: 10.1056/NEJMc2104626. PubMed DOI PMC

Chandra SR. Scalable and secure learning with limited supervision over data streams. https://utd-ir.tdl.org/bitstream/handle/10735.1/6196/ETD-5608-011-CHANDRA-8457.95.pdf?sequence=6&isAllowed=y: Texas; 2018.

Ackerman S, Farchi E, Raz O, Zalmanovici M, Dube P. Detection of data drift and outliers affecting machine learning model performance over time. arXiv preprint arXiv:201209258. 2020.

Lund R. Revenge of the white swan. Am Stat. 2007;61(3):189–192. doi: 10.1198/000313007X219374. DOI

Dietterich TG. Ensemble methods in machine learning. International workshop on multiple classifier systems; 2000: Springer; 2000. p. 1–15.

Finlayson SG, Bowers JD, Ito J, Zittrain JL, Beam AL, Kohane IS. Adversarial attacks on medical machine learning. Science. 2019;363(6433):1287–1289. doi: 10.1126/science.aaw4399. PubMed DOI PMC

Gennatas ED, Friedman JH, Ungar LH, Pirracchio R, Eaton E, Reichmann LG, et al. Expert-augmented machine learning. Proc Natl Acad Sci. 2020;117(9):4571–4577. doi: 10.1073/pnas.1906831117. PubMed DOI PMC

Madras D, Pitassi T, Zemel R. Predict responsibly: improving fairness and accuracy by learning to defer. arXiv preprint arXiv:171106664. 2017.

Mozannar H, Sontag D. Consistent estimators for learning to defer to an expert. In: International Conference on Machine Learning; 2020: PMLR; 2020. p. 7076–87.

Wabartha M, Durand A, Francois-Lavet V, Pineau J. Handling black swan events in deep learning with diversely extrapolated neural networks. Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence, IJCAI-20, International Joint Conferences on Artificial Intelligence Organization; 2020; 2020. p. 2140–7.

Kompa B, Snoek J, Beam AL. Second opinion needed: communicating uncertainty in medical machine learning. NPJ Digit Med. 2021;4(1):1–6. doi: 10.1038/s41746-020-00367-3. PubMed DOI PMC

McDermott MBA, Wang S, Marinsek N, Ranganath R, Foschini L, Ghassemi M. Reproducibility in machine learning for health research: still a ways to go. Sci Transl Med. 2021;13(586). PubMed

Molnar C. Interpretable machine learning: Lulu. com; 2020.

Danysz K, Cicirello S, Mingle E, Assuncao B, Tetarenko N, Mockute R, et al. Artificial intelligence and the future of the drug safety professional. Drug Saf. 2019;42(4):491–497. doi: 10.1007/s40264-018-0746-z. PubMed DOI PMC

Peters J, Janzing D, Schölkopf B. Elements of causal inference: foundations and learning algorithms. The MIT Press; 2017.

Markatou M, Ball R. A pattern discovery framework for adverse event evaluation and inference in spontaneous reporting systems. Stat Anal Data Mining ASA Data Sci J. 2014;7(5):352–367. doi: 10.1002/sam.11233. DOI

Olsson S, Edwards IR. Tachycardia during cisapride treatment. BMJ. 1992;305(6856):748–749. doi: 10.1136/bmj.305.6856.748-a. PubMed DOI PMC

Inman W, Kubota K. Tachycardia during cisapride treatment. BMJ. 1992;305(6860):1019. doi: 10.1136/bmj.305.6860.1019-a. PubMed DOI PMC

Layton D, Key C, Shakir SA. Prolongation of the QT interval and cardiac arrhythmias associated with cisapride: limitations of the pharmacoepidemiological studies conducted and proposals for the future. Pharmacoepidemiol Drug Saf. 2003;12(1):31–40. doi: 10.1002/pds.781. PubMed DOI

Bate A, Lindquist M, Orre R, Edwards IR, Meyboom RH. Data-mining analyses of pharmacovigilance signals in relation to relevant comparison drugs. Eur J Clin Pharmacol. 2002;58(7):483–490. doi: 10.1007/s00228-002-0484-z. PubMed DOI

Mann RD. An instructive example of a long-latency adverse drug reaction–sclerosing peritonitis due to practolol. Pharmacoepidemiol Drug Saf. 2007;16(11):1211–1216. doi: 10.1002/pds.1466. PubMed DOI

Brewer T, Colditz GA. Postmarketing surveillance and adverse drug reactions: current perspectives and future needs. JAMA. 1999;281(9):824–829. doi: 10.1001/jama.281.9.824. PubMed DOI

Kessler DA. Introducing MEDWatch. A new approach to reporting medication and device adverse effects and product problems. JAMA. 1993;269(21):2765–2768. doi: 10.1001/jama.1993.03500210065033. PubMed DOI

Bate A, Reynolds RF, Caubel P. The hope, hype and reality of Big Data for pharmacovigilance. Ther Adv Drug Saf. 2018;9(1):5–11. doi: 10.1177/2042098617736422. PubMed DOI PMC

LePendu P, Iyer SV, Bauer-Mehren A, Harpaz R, Mortensen JM, Podchiyska T, et al. Pharmacovigilance using clinical notes. Clin Pharmacol Ther. 2013;93(6):547–555. doi: 10.1038/clpt.2013.47. PubMed DOI PMC

Bhattacharya M, Snyder S, Malin M, Truffa MM, Marinic S, Engelmann R, et al. Using social media data in routine pharmacovigilance: a pilot study to identify safety signals and patient perspectives. Pharmaceut Med. 2017;31(3):167–174.

Norén GN, Orre R, Bate A, Edwards IR. Duplicate detection in adverse drug reaction surveillance. Data Min Knowl Disc. 2007;14(3):305–328. doi: 10.1007/s10618-006-0052-8. DOI

Star K, Caster O, Bate A, Edwards IR. Dose variations associated with formulations of NSAID prescriptions for children: a descriptive analysis of electronic health records in the UK. Drug Saf. 2011;34(4):307–317. doi: 10.2165/11586610-000000000-00000. PubMed DOI

Nath J. Chatbot, machine learning and artificial intelligence in pharmacovigilance: maintaining privacy, optimizing efficiency. 2018 [cited 2021 25th November]; https://chatbotsmagazine.com/chatbot-machine-learning-and-artificial-intelligence-in-pharmacovigilance-maintaining-privacy-877283e4b4b7. Accessed 11 Mar 2022.