scholarly journals Patient and Consumer Safety Risks When Using Conversational Assistants for Medical Information: An Observational Study of Siri, Alexa, and Google Assistant

10.2196/11510 ◽  
2018 ◽  
Vol 20 (9) ◽  
pp. e11510 ◽  
Author(s):  
Timothy W Bickmore ◽  
Ha Trinh ◽  
Stefan Olafsson ◽  
Teresa K O'Leary ◽  
Reza Asadi ◽  
...  
Author(s):  
Timothy W. Bickmore ◽  
Ha Trinh ◽  
Stefan Olafsson ◽  
Teresa K O'Leary ◽  
Reza Asadi ◽  
...  

BACKGROUND Conversational assistants, such as Siri, Alexa, and Google Assistant, are ubiquitous and are beginning to be used as portals for medical services. However, the potential safety issues of using conversational assistants for medical information by patients and consumers are not understood. OBJECTIVE To determine the prevalence and nature of the harm that could result from patients or consumers using conversational assistants for medical information. METHODS Participants were given medical problems to pose to Siri, Alexa, or Google Assistant, and asked to determine an action to take based on information from the system. Assignment of tasks and systems were randomized across participants, and participants queried the conversational assistants in their own words, making as many attempts as needed until they either reported an action to take or gave up. Participant-reported actions for each medical task were rated for patient harm using an Agency for Healthcare Research and Quality harm scale. RESULTS Fifty-four subjects completed the study with a mean age of 42 years (SD 18). Twenty-nine (54%) were female, 31 (57%) Caucasian, and 26 (50%) were college educated. Only 8 (15%) reported using a conversational assistant regularly, while 22 (41%) had never used one, and 24 (44%) had tried one “a few times.“ Forty-four (82%) used computers regularly. Subjects were only able to complete 168 (43%) of their 394 tasks. Of these, 49 (29%) reported actions that could have resulted in some degree of patient harm, including 27 (16%) that could have resulted in death. CONCLUSIONS Reliance on conversational assistants for actionable medical information represents a safety risk for patients and consumers. Patients should be cautioned to not use these technologies for answers to medical questions they intend to act on without further consultation from a health care provider.


2021 ◽  
Author(s):  
Timothy W Bickmore ◽  
Stefán Ólafsson ◽  
Teresa K O'Leary

BACKGROUND Prior studies have demonstrated the safety risks when patients and consumers use conversational assistants such as Apple’s Siri and Amazon’s Alexa for obtaining medical information. OBJECTIVE The aim of this study is to evaluate two approaches to reducing the likelihood that patients or consumers will act on the potentially harmful medical information they receive from conversational assistants. METHODS Participants were given medical problems to pose to conversational assistants that had been previously demonstrated to result in potentially harmful recommendations. Each conversational assistant’s response was randomly varied to include either a correct or incorrect paraphrase of the query or a disclaimer message—or not—telling the participants that they should not act on the advice without first talking to a physician. The participants were then asked what actions they would take based on their interaction, along with the likelihood of taking the action. The reported actions were recorded and analyzed, and the participants were interviewed at the end of each interaction. RESULTS A total of 32 participants completed the study, each interacting with 4 conversational assistants. The participants were on average aged 42.44 (SD 14.08) years, 53% (17/32) were women, and 66% (21/32) were college educated. Those participants who heard a correct paraphrase of their query were significantly more likely to state that they would follow the medical advice provided by the conversational assistant (<i>χ</i><sup>2</sup><sub>1</sub>=3.1; <i>P</i>=.04). Those participants who heard a disclaimer message were significantly more likely to say that they would contact a physician or health professional before acting on the medical advice received (<i>χ</i><sup>2</sup><sub>1</sub>=43.5; <i>P</i>=.001). CONCLUSIONS Designers of conversational systems should consider incorporating both disclaimers and feedback on query understanding in response to user queries for medical advice. Unconstrained natural language input should not be used in systems designed specifically to provide medical advice. CLINICALTRIAL


10.2196/30704 ◽  
2021 ◽  
Vol 23 (11) ◽  
pp. e30704
Author(s):  
Timothy W Bickmore ◽  
Stefán Ólafsson ◽  
Teresa K O'Leary

Background Prior studies have demonstrated the safety risks when patients and consumers use conversational assistants such as Apple’s Siri and Amazon’s Alexa for obtaining medical information. Objective The aim of this study is to evaluate two approaches to reducing the likelihood that patients or consumers will act on the potentially harmful medical information they receive from conversational assistants. Methods Participants were given medical problems to pose to conversational assistants that had been previously demonstrated to result in potentially harmful recommendations. Each conversational assistant’s response was randomly varied to include either a correct or incorrect paraphrase of the query or a disclaimer message—or not—telling the participants that they should not act on the advice without first talking to a physician. The participants were then asked what actions they would take based on their interaction, along with the likelihood of taking the action. The reported actions were recorded and analyzed, and the participants were interviewed at the end of each interaction. Results A total of 32 participants completed the study, each interacting with 4 conversational assistants. The participants were on average aged 42.44 (SD 14.08) years, 53% (17/32) were women, and 66% (21/32) were college educated. Those participants who heard a correct paraphrase of their query were significantly more likely to state that they would follow the medical advice provided by the conversational assistant (χ21=3.1; P=.04). Those participants who heard a disclaimer message were significantly more likely to say that they would contact a physician or health professional before acting on the medical advice received (χ21=43.5; P=.001). Conclusions Designers of conversational systems should consider incorporating both disclaimers and feedback on query understanding in response to user queries for medical advice. Unconstrained natural language input should not be used in systems designed specifically to provide medical advice.


1991 ◽  
Vol 54 (12) ◽  
pp. 913-916 ◽  
Author(s):  
JOHN P. ERICKSON ◽  
PHYLLIS JENKINS

Salmonella spp. and Listeria monocytogenes strains were inoculated into four commercial mayonnaise products: sandwich spread, real mayonnaise, reduced calorie mayonnaise dressing, and cholesterol-free reduced calorie mayonnaise dressing. Products represented a broad cross-section of aqueous phase acetic acid, salt, sucrose, and other compositional factors. Results showed that Salmonella spp. inactivation rates were unaffected by formula composition. The organism was rapidly inactivated, decreasing ≥8 log10 CFU/g in ≤72 h, in each of the four products. L. monocytogenes inactivation rates were directly correlated with aqueous phase acetic acid concentrations as follows: sandwich spread ≥ real mayonnaise &gt; cholesterol-free reduced calorie mayonnaise dressing &gt; reduced calorie mayonnaise dressing. L. monocytogenes inactivation rate in sandwich spread and real mayonnaise was similar to Salmonella spp. The reduced calorie mayonnaise dressings showed gradual, incremental population declines. L. monocytogenes decreased 3 and 5 log10 CFU/g in 72 h in reduced calorie and cholesterol-free reduced calorie mayonnaise dressings, respectively. The higher anti-listerial activity in the cholesterol free formulation was attributed to egg white lysozyme. This study documented that commercial mayonnaise, including reduced calorie mayonnaise dressing varieties, represent negligible consumer safety risks.


Author(s):  
David Vogel

This chapter analyzes European and American policies toward a range of consumer safety risks; including drugs, children's products, and cosmetics. It shows how European and American risk regulations have converged, though the dynamics through which this occurred differed substantially. Pharmaceutical regulation constitutes the most important exception to the broader pattern of increased transatlantic regulatory policy divergence. What makes this area of regulatory policy distinctive is that its political salience increased in the United States but not in Europe. Pharmaceutical regulation also represents an important exception to the dominant pattern of transatlantic regulatory policy diffusion. In this case, European regulatory policies did affect those of the United States, first by highlighting the transatlantic drug lag, and more recently by American decisions to adopt some European practices to expedite drug approvals.


Author(s):  
Hiroto Ishiki ◽  
Jun Hamano ◽  
Hiroka Nagaoka ◽  
Yoshinobu Matsuda ◽  
Akihiro Tokoro ◽  
...  

Background: Neuroleptics are commonly used in palliative care settings. However, adverse events of neuroleptics, known as extrapyramidal symptoms (EPSs), might be overlooked in clinical practice. We conducted this study to explore the prevalence of EPSs in palliative care setting. Methods: This multicenter, prospective, observational study included patients who 1) were referred to either a specialist palliative care team or a palliative care unit, 2) had a diagnosis of cancer, and 3) were ≥20 years of age. We investigated the prevalence of EPSs and medications used. The primary outcome was the overall Drug-Induced Extrapyramidal Symptom Scale (DIEPSS) score. Results: Between November 2015 and October 2016, 149 patients from 5 centers in Japan were enrolled. The median age was 67 years (range: 21–88 years) and the study population included 81 men (54.4%). The cancer types included lung (55 patients, 36.9%), upper gastrointestinal tract (5, 3.3%), hepatobiliary (19, 12.8%); breast (12, 8.1%); head and neck (10, 6.7%), gynecologic (10, 6.7%), genitourinary (10, 6.7%), and others (28, 18.8%). The median Karnofsky performance status was 60 (20–100). Most patients (86.6%) did not experience delirium. Thirty-nine (26.2%) patients received one or more EPS-inducing medications. EPSs occurred in 4 (2.7%) patients with a cutoff score of 5 points for 5 parkinsonism items in DIEPSS. Conclusion: A lower frequency (<3%) of patients than expected in this population had EPSs. Therefore, we concluded that an interventional study is not feasible. However, medications that cause EPSs are often used in palliative care; therefore, a longitudinal study is warranted. Trial registration: This study was registered in the University Hospital Medical Information Network Clinical Trials Registry (UMIN-CTR) (UMIN000019810) on 16, November, 2015.


2021 ◽  
Author(s):  
Marank de Steenwinkel ◽  
Juanita A Haagsma ◽  
Esther C.M. van Berkel ◽  
Lotte Rozema ◽  
Pleunie P.M. Rood ◽  
...  

Abstract Background Patient satisfaction is an important indicator of emergency care quality and has been associated with information dispensation at the Emergency Department (ED). Optimal information dispensation could improve patient experience and expectations. Knowing what kind of information patients want to receive and the preferred way of information dispensation are essential to optimize information delivery at the ED. The purpose of this cross-sectional observational study was to evaluate patient satisfaction concerning information dispensation (including general, medical and practical information), the need for additional information and preferences with regards to the way of information dispensation at the ED of a teaching hospital in the Netherlands. Results423 patients (patients ≥18 years with Glasgow Coma Scale 15) were enrolled (response rate 79%). The median patient satisfaction score concerning the overall information dispensation at the ED was 7.5 on a rating scale 0-10. Shorter length of ED stay was associated with higher patient satisfaction in multivariate analysis (P<0.001). The majority of respondents was satisfied regarding medical (N=328; 78%) and general information (N =233; 55%). Patients were less satisfied regarding practical information (N =180; 43%). Respondents who indicated that they received general, medical and practical information were significantly more often satisfied compared to patients who did not receive this information (P<0.001). Two thirds (N=260; 62%) requested more general information. Half of the respondents (N=202; 48%) requested more practical information and a third (N=152; 36%) requested more medical information. The preferred way for receiving information was orally (N=189; 44.7%) or by leaflets (N=108; 25.5%).ConclusionThe majority of respondents were satisfied concerning information dispensation at the ED, especially regarding medical information. Respondents requested more general and practical information and preferred to receive the information orally or by leaflets.


10.2196/16768 ◽  
2020 ◽  
Vol 22 (4) ◽  
pp. e16768 ◽  
Author(s):  
Yan Li ◽  
Shan Ye ◽  
Yidong Zhou ◽  
Feng Mao ◽  
Hailing Guo ◽  
...  

Background The internet allows patients to easily look for health information. However, how Chinese patients with breast cancer use the internet has rarely been investigated, and there is a scarcity of information about the influence of internet use on survival. Objective This observational study aimed to investigate the details of online medical information searching by Chinese patients with breast cancer and to determine whether internet use has any survival benefits. Methods Patients who were diagnosed with invasive breast cancer at Peking Union Medical College Hospital between January 2014 and December 2015 were enrolled. We obtained information on their internet-searching behavior and gathered data from the patients’ medical and follow-up records. The associations between internet use and other clinic-pathological factors were analyzed. A Cox proportional-hazards model and the Kaplan-Meier method were used for disease-free survival (DFS) analyses. Results A total of 973 patients with invasive breast cancer who underwent definitive surgery took part in the study. Among them, 477 cases (49.0%) performed web-based breast cancer information searching before the initial treatment. A multivariate logistic regression analysis suggested that web-based breast cancer information searching was significantly associated with younger age (odds ratio [OR] 0.95, 95% CI 0.94-0.97, P<.001), higher education level (OR 1.37, 95% CI 1.01–1.86, P=.04), and breast conserving surgery (OR 1.35, 95% CI 1.04-1.77, P=.03). Baidu (73.4%, 350/477) and WeChat (66.7%, 318/477) were the two most popular online information sources for breast cancer; however, only 44.9% (214/477) felt satisfied with the online information. In contrast to the nonweb searching group, the web-using patients who were satisfied with online information showed significantly improved DFS (hazard ratio 0.26; 95% CI 0.08-0.88, P=.03). Conclusions The patients who were most likely to search the internet for breast cancer information were younger and well-educated, and they were more likely to have breast conserving therapy. Web-using patients who were satisfied with the internet information showed significantly improved DFS. Patients should browse credible websites offering accurate and updated information, and website developers should provide high-quality and easy-to-understand information to better meet the needs of patients with breast cancer.


1992 ◽  
Vol 25 (2) ◽  
pp. 181-200 ◽  
Author(s):  
Diana E. Forsythe ◽  
Bruce G. Buchanan ◽  
Jerome A. Osheroff ◽  
Randolph A. Miller

Sign in / Sign up

Export Citation Format

Share Document