Dataset for: Social dilemma in the excess use of antimicrobials incurring antimicrobial resistance
Data files
Dec 09, 2022 version files 962.90 KB
-
Data_S1.csv
114.18 KB
-
Data_S2.csv
114.18 KB
-
Data_S3.csv
115.03 KB
-
Data_S4.csv
115.05 KB
-
Data_S5.csv
124.44 KB
-
Data_S6.csv
64.04 KB
-
Data_S7.csv
61.81 KB
-
Data_S8.csv
125.01 KB
-
Data_S9.csv
124.37 KB
-
README.md
4.78 KB
Abstract
This is the dataset for the study of "Social dilemma in the excess use of antimicrobials incurring antimicrobial resistance".
The emergence of antimicrobial resistance (AMR) caused by the excess use of antimicrobials has come to be recognized as a global threat to public health. There is a ‘tragedy of the commons’ type social dilemma behind this excessive use of antimicrobials, which should be recognized by all stakeholders. To address this global threat, we thus surveyed eight countries/areas to determine whether people recognize this dilemma and showed that although more than half of the population pays little, if any, attention to it, almost 20% recognize this social dilemma, and 15–30% of those have a positive attitude toward solving that dilemma. We suspect that increasing individual awareness of this social dilemma contributes to decreasing the frequency of AMR emergencies.
Methods
We designed a questionnaire to observe a social dilemma in the excess use of antimicrobials incurring antimicrobial resistance by placing two types of imaginary artificial-intelligence (AI) physicians who perform medical practice from either an individual or societal perspective.
We assume two AI medical diagnosis systems: “Individual precedence AI” (abbreviated Individual-AI) and “World precedence AI” (abbreviated World-AI). Both AIs diagnose and prescribe medicine automatically. The Individual-AI system diagnoses patients and prescribes medicine to prevent infections based on an individual perspective, including all prophylactic prescriptions against rare accidental infections (not yet present and unlikely to occur). It does not consider the global risk of AMR in the decision. The World-AI system, instead, takes into account the global mortality rate of AMR, aiming to reduce the total number of all AMR-related deaths. Because of this, this AI system does not prescribe antimicrobials against rare and not-yet-present infections.
This questionnaire design allows us to observe the social dilemma. For example, it shows a typical social dilemma caused by preferring the use of Individual-AI for diagnosing oneself but preferring the use of World-AI for diagnosing strangers.
The survey entitled “Survey on Medical Advancement” was administered to 8 countries/areas. The survey was conducted 4 times.
For the two surveys in Japan, an internet survey company, Cross Marketing Inc. (https://www.cross-m.co.jp/en/), created the questionnaire webpages based on our study design. The company also collected the data. As of April 2020, Cross Marketing Inc. has 4.79 million people in an active panel (survey participants who registered in advance). Here, the definition of an active panel is a survey respondent who has been active within the last year. For the panels, the questionnaire and response column were displayed on the website through which the respondents could complete and submit their responses. We extracted 500 submissions for each gender and each age group by random sampling from all samples collected during the survey periods. The surveys in the 7 countries/areas (i.e., the United States, the United Kingdom, Sweden, Taiwan, Australia, Brazil, and Russia) are conducted by Cint (https://www.cint.com/). Cint is the world’s largest consumer network for digital survey-based research. The headquarters of the company is in Sweden. Cint maintains a survey platform that contained more than 100 million consumer monitors in over 80 countries as of May 2020. For surveys in the US, UK, Sweden, Taiwan, Australia, Brazil, and Russia, Cint Japan (https://jp.cint.com/), which is the Japanese distributor of Cint, created translated questionnaire webpages based on our study design. The company also collected the data. We extracted at least 500 (US, UK, SWE, BRA, RUS) or 250 (TWN, AUS) submissions for each gender (male and female) and each age group (20 s, 30 s, 40 s, 50 s, and 60 s) by random sampling from all samples collected between survey periods.
Note that both companies eliminated inconsistent or apathetic respondents. For example, respondents with inconsistent responses (e.g., the registered age of the respondent differed from the reported age at the time of the survey.) were eliminated before reaching the authors. In addition, respondents with significantly short response times (i.e., shorter than 1 min) were eliminated because they may not have read the questions carefully.