Reasons Why You Should Be Managing Your Own Health Data

Holding on to Your Data Is Not Owning It

Originally published by Technology Hits on Medium

Photo by Dan Nelson on Unsplash

With growing emphasis on value-based healthcare reimbursement, we are amidst finding ourselves more than ever challenged by the parallel escalation of corporate interest in “patient health information.” Health data security and the monetary value in quality data are the drivers of such tension.

Despite all the rationing in releasing the pressure through patient engagement, whether by providing the patients access to their records or creating interoperable digital data systems, there is more to saving patients’ sovereign state than simply easing accessibility issues.

Of course, Increased patient demands for more detailed medical data require a more sophisticated way to store files. That will help enhance patient confidence. But that is not even close to satisfying the sovereign patient hood. Surely it will lead to fewer medical errors, improve data access for doctors and patients, speed up medical information access during doctor visits, and make billing easier. It even enables creating a patient history after the visit streamlining the continuity of care. Nevertheless, it won’t guarantee doctors, more so the patients, that sophisticated secure data management will not abuse their data.

Abuse of Personal Healthcare Data doled under the need for optimal Patient Care.

The urgency stimulates big Data in healthcare to solve, reduce administrative workloads, and increase stakeholders’ profit. It also solves global problems of humanity, such as forecasting epidemics and combating existing diseases. But, given the means of data collection and utility of those data, it is becoming evident that the balance of the profit from data mining and the process to make it useful is not fairly distributed among the stakeholders. For example, data collected as a social and personal determinant of health without the patient’s consent with full transparency has been widely used in the past to upsurge insurance premiums. That profits the insurance company and the data collector more than it helps the patient.

Today, emerging companies charge a variable fee to clinicians looking to connect with their patients. Those fees are collected and dispersed to patients based on how many projects in which they’ve participated. For example, the person who’s executed five surveys sat for three meetings, and participated in a handful of focus groups will get a more significant profit share than someone who’s user-tested one product. Connecting patients with the medical industry, too, is a lucrative market. These companies want to make something that wouldn’t benefit a small group of people while patients are doing the work. Because that’s the way the power dynamic has traditionally been in healthcare. They accomplish that by ration patients take ownership of their data.

It is widely accepted that individuals should control their data and have the right to make decisions about admittance to their personal information. Some even proclaim patient’s ownership of their data based on their direct participation in its cohort. Data provided by individuals are a form of labor that powers Artificial Intelligence (AI).

Clinical Data is co-constructed via a collaborative process implicating the patient, the clinician, and many other professionals within the health system. The primary clinician extracts, interpret, methods, classifies and describes the patient’s medical history. Other health experts formulate radiology images, test blood samples, and conduct gene sequencing. In countries with public health systems, the state pays to collect and store this data. Everyone in the patient’s care, including the patient, has a stake in healthcare data. Nonetheless, that is not the case today.

The current trend on Medical Data Management

The current mission of data management is around making it portable, interoperable, and private. Acknowledging that the health data is valuable, the primary stream data industry focuses on extracting or mining the data to deep learn of Commoditization.

The right to data portability enables individuals to procure and reuse personal data for their motives across different services. It also allows them to move, copy or transfer personal data quickly from one information technology (IT) setting to another safely and securely, without affecting the usability. However, making health information portable today is limited to clusters of centralized interoperable systems.

Data interoperability is relevant amid increasing public dependence on digital information. Data interoperability deals with the potential of systems and actions that establish, exchange, and consume data to have clear, shared goals for that data's scopes, context, and meaning.

One-off slants to Data carry concealed costs felt by people and organizations influenced by such data — most people dearth agency when it comes to their information. The value of the insights gained from such data is restricted because the real potential of such datasets is unclear. Even the prevailing interoperable systems are not interoperable enough to satisfy the appropriate agency to the patients or physicians.

Despite all the corporate rhetoric around making personal data more secure and safe, today's data privacy is not even close to perfect. Most people are still concerned, confused, and feeling a lack of control over their personal information. Majorities believe their data is less secure now than before. Today, data collection presents more risks than benefits, and it is impossible to go through daily life without being traced.

Most large institutions hold the ownership and daunting task of overseeing massive and ever-changing data abundances. They reserve to themselves the job of leveraging data as a strategic investment. The economic implication of digital services and data mining practice within the GDP (Gross Domestic Product) calculation, as most, if not all countries currently ignore the value generated by free services using digital data, and often massive amounts thereof.

Data Mining amid increasing Corporate interest in Patient Information and its contribution to Deep Learning (DL)

Data Mining is primarily borrowed in several applications. It is mainly utilized in understanding consumer research marketing, product analysis, demand, and supply analysis. Data Mining provides companies with a competitive edge in the business climate. A new data mining concept, business intelligence, is becoming used widely by leading corporate houses to stay ahead of their competitors. Business Intelligence (BI) helps in furnishing the latest data and is used for competition analysis, market research, economic trends, consumer behavior, industry research, geographical information analysis, etc. Therefore, data and data mining are valuable digital commodities. However, to the irony, most of the mined data are indiscriminately collected freely from the consumers.

In the recent decade, deep learning (DL) and neural networks have received much attention in media stories. Deep learning has been broadcasted as the subsequent era of computing.

McKinsey estimates trillions of dollars of impacts globally from deep learning. That Specifically predicts a 1% — 9% increase in revenues for companies that deploy Deep Learning.

Data Portability in the age of Information Technology

Since Data portability” is a characteristic of information technology that lets a user carry their data, Similarly, it cannot stave off increasing centralized database. Portability works together with transparency. Furthermore, if a portion of data is easily accessible or portable, the rest is secret or not transparent. The data integrity is an incomplete picture of the patient’s relationship with a service. Conversely, suppose a patient can find out all the company's information but has no way to take it and interact with it, i.e., not portable. In that case, they are denied entrance to understand further and analyze that information.

Centralized data takes back the power away from individuals by restricting portability and transparency. For instance, instead of reaching out to companies to repossess their data, companies should come to them for their data. Patients in the latter scenario control their information, hence controlling other people or companies' access to it.

Advances in modern data tokenization have made it possible to live with a new form of privacy-aware processes at scale, in the cloud, and with massively reduced risk. However, strict new privacy laws and related enforcement bodies can bring incentives to business responsibility.

Today, no forward-thinking business can tolerate the risk of data compromise, lawsuit, spying, and defamation from human blunder or direct attack, especially those well into their modern hybrid voyage. Industries riding the new crest of powerful yet emerging technologies utilizing centralized data schemes to strive and grow lays data into entirely new risk states.

The exponential rise of digital data availability and the requisite to process it in business and scientific fields have made it compulsory for users to analyze and mine helpful knowledge. Traditionally, the centralized data warehousing model has made it easier to gather all data and run an algorithm upon them.

Standard Deep Learning Models require centralizing of the training data on one machine or in a data center. For example, when an electronic health record start-up wants to develop a model to understand its physician propensity to diagnose a disease, it runs the models on the data collected from patient records and physician clinical judgments. Such data may comprise the time disbursed on patient care and notes written to drive a given diagnosis. Typically, 1000’s of data points are collected on every user over a period. Such data are conceded and sent over to a centralized data center or machines for computer analysis.

Federated Learning is the new agent of Data Monopoly

Recently, a new scheme has been considered for models trained from user interaction with mobile devices called Federated Learning.

Federated learning allots the Deep learning process over to the edge by enabling mobile devices to collaboratively learn a shared model using the device’s training data and keeping them. It dissociates the need for doing deep learning with the necessity to store the data in the cloud. Federated learning is dominated by tech giants such as Google, Amazon, and Microsoft, who offer cloud-based artificial intelligence (AI) solutions and APIs to their subscribers. This model provides users with little control over the utility of AI products and their data collected from their devices, locations, etc. That opens a new door to the monopolization of only a few strong players.

The role of Decentralization in the Modern Data Management

In the blockchain or information technology, decentralization pertains to the shift of control and decision-making from a centralized entity, be it individual, organization, or group, to a distributed network of individual users. Decentralized networks strive to reduce the need for the level of trust amongst their participants. It, therefore, daunts their ability to exert authority or control over one another in ways that degrade the functionality of the network.

Decentralization matters by dispersing the management of and access to resources in an application; more outstanding and fairer service can be achieved. Of course, decentralization typically has some tradeoffs, including lower transaction speed, but such tradeoffs are worth the improved stability and service levels they produce in the long run.

Using a decentralized identity protocol, individuals can flawlessly exchange their information without fear of authority changes or systemic disruptions.

Entities on the Internet can disappear without prior notice and fast. So not depending on a centralized authority to store information on behalf of individuals empowers users to transport their identity assets more readily. Such portability under a decentralized state also functions in tandem with an identity’s perseverance.

Blockchain’s decentralized attribute permits unhitched supply chain management systems to interoperate secure manner without excessive investment expenses. Because of the overwhelming want for a supply chain revolution, leveraging these characteristics ensures that blockchain can be helpful and effective in the real world in real-time.

Decentralized architectures are becoming the way to protect one’s privacy. Although, despite the benefits they provide for data sovereignty, decentralized architectures need to account for specific issues that otherwise might ultimately impinge upon users’ privacy. Decentralization can preserve data confidentiality; however, it may still be vulnerable to metadata analysis by corporations. We can quickly address this with conventional security protocols. That is why the hybrid models of centralized and decentralized systems are emerging to enhance blockchain technology’s effectiveness and utility.

As mentioned earlier, the emerging DL market model is dominated by tech giants. It offers users little control over the usage of deep learning products and their data collected from their devices, locations enhancing the monopolization of elite players. Ultimately, it would edge the participation of smaller businesses or even larger enterprises in profound learning innovation and lack of interoperability and interpretability of decisions driven by those systems.

With DL's emergence, we see the beginning of a decentralized deep learning market, born at the intersection of on-device AI, blockchain, edge computing, or the Internet of Things (IoT). The latter is being challenged today through the introduction of Federated learning.

So, Who Should Own Medical Records?

Patients should have their Rights because they are the bedrock of their medical records. Having access to complete medical records will empower them to make informed decisions, choose physicians, services and carry their data anywhere at any time in a secure manner with confidence. It has also been found that sharing doctor notes with patients impacted their outcomes positively.

Other stakeholders like laboratories should have the right over the data that they examined. Thus, Every stakeholder should also reward them for the value they contributed, leaving them free to use it for research and other purposes, maintaining all security and privacy measures.

We should share medical records’ benefits based on the contribution value if any other arrangement or model is not used. We can bring a change to the ever-challenging healthcare industry by involving, engaging, and educating all stakeholders.

By giving proper ownership and reducing liability for corporate entities, we ensure information security, lower costs, increase medical care quality, address citizen’s expectations, and promote efficient practices. Ideally, We can achieve that through Portability, Interoperability, Transparency utilizing a decentralized ledger system.

#Patientinformation #bigdata #healthinformation #informationsecurity #healthcare

0 views0 comments