Health TechHealthcare

Growing AI-fueled Healthcare is not Free from the Risk of Bias

Growing AI-fueled Healthcare is not Free from the Risk of Bias

Artificial Intelligence (AI) was previously widely utilized in hospitals throughout India for several purposes

According to a recent study by the United Nations Development Programme, electronic health records and AI-driven automation are revolutionizing healthcare in India. However, many Indians may be at a disadvantage due to the lack of adequate datasets, particularly for some populations, and the overemphasis on urban and male populations. According to the research study, which was carried out by the Aapti Institute and commissioned by the UNDP as part of the Business and Human Rights in Asia program funded by the European Union, companies that ask gig workers to display photos with masks, begin sharing location data, and measure their body temperatures during COVID-19 also put them at risk of losing privacy. Even while Artificial Intelligence (AI) was previously widely utilized in hospitals throughout India for several purposes, including image analysis and danger identification, Covid-19 has extended its use to the point that hospitals can now assess the degree of lung injury.

In many areas of healthcare, including diagnosis, therapy, automated analysis of medical test results, predictive healthcare diagnosis, automation of healthcare diagnosis with monitoring equipment, and wearable sensor-based medical devices, according to health experts, artificial intelligence has resulted in significant advancements.

The doctor-patient ratio in India is 1:1457, which means that physicians may only spend two to five minutes with each patient, increasing the likelihood of mistakes and incorrect diagnoses. According to the paper, it serves as a tool for physicians that enhances their ability to provide treatment by facilitating evidence-based decision-making and allowing them to devote more time to pressing situations.

Researchers from the Aapti Institute Aishani Rai, Vinay Narayan, and Sarayu Natarajan said in a report that structural problems with digital inaccessibility are the main cause of the dearth of data availability. This conclusion was reached after doing research and interviewing experts, and stakeholders.

For instance, just 14% of adult women in India had smartphones compared to 37% of adult males. Data for the healthcare industry is also provided through smartphone health applications that continuously track user behavior. If we depend too much on smartphone data, which is mostly collected from men with higher-than-average earnings, we risk having a distorted view of the health needs of women in general and poor women in particular.

Uninformative forecasts for minority groups result from the lack of appropriate facts on a population, leaving predictions that apply to majority populations. According to the findings, underrepresented datasets result in patients with lower socioeconomic status receiving fewer diagnostic tests and prescription drugs.

Every nation has its patterns of diseases that are frequently prominent. The cardiovascular disease strikes people in India significantly earlier than it does in middle- and high-income nations. Any AI designed to identify heart attacks would underdiagnose Indian women since doctors often diagnose heart attacks based on symptoms reported by males. In addition to capturing varied people, diversified datasets also need to include other socioeconomic factors that have an impact on health, the report stated, noting that most healthcare facilities in India are located in metropolitan regions.

The study acknowledged the enormous potential of predictive analytics to improve healthcare by lowering patient fatalities in rural areas. Healthcare analytics made for 10.81% of India’s digital healthcare industry in 2019; by 2025, it’s projected to be worth Rs 47.04 billion. As a “more complete piece of law on data protection, providing a framework of consumer rights and remedies (compensation) and penalties for a data breach,” it has been advised that the Personal Data Protection Bill, 2019, be enforced by the Center. The Electronic Health Records Standards, 2016, which outline privacy and security, must be expanded by the state in the absence of data protection law, it stated.

 

Risks for gig workers

With Covid-19, several on-demand platforms have made it mandatory for employees to disclose their body temperature to consumers, in addition to their images with masks and location information, even though the employee’s rating serves as a surveillance tool. Ratings are used to maintain strict control over employees’ duration of engagement on the app, acceptance and cancellation rates, and compliance with corporate standards since they are related to worker activities on the app (such as how many jobs they refuse and how many tasks they do). Additionally, women who work as gig workers typically do grooming and personal care tasks that necessitate them to enter private areas where their safety is at stake. It is up to you to ensure their security.

To guarantee fair working pay and conditions for gig workers, ensure that minimum wages are based on international standards, and persuade businesses to adhere to strict standards of data security in the collection to train employees on the consent form for data collection, UNDP has urged manufacturing bodies (like NASSCOM and ASSOCHAM) to issue rules of ethics that businesses must abide by.

According to the research, the 2020 Labour Codes are a positive first step toward recognizing the rights of gig workers to social security safeguards, but the government must work to make this framework operational to guarantee that the employees receive the benefits. It emphasizes that because they are categorized as “independent contractors,” employees are not entitled to social security benefits like health insurance, pension contributions, or paid maternity leave.

What's your reaction?

Excited
0
Happy
0
In Love
0
Not Sure
0
Silly
0

You may also like

Leave a reply

Your email address will not be published. Required fields are marked *

More in:Health Tech