We tend to talk about “big data” as if it were just some massive scam for Facebook, Amazon, and the government to keep tabs on us and sell us stuff we don’t need.
Thought leaders in healthcare see it differently. Big data stands on the cusp of changing the way healthcare professionals innovate and provide medical care–and patients may start noticing it in lower bills and better outcomes.
So what is this scary thing called “big data?” Forget the fact that our phones and FitBits basically track our every move; it’s not only about that, for a change–though such devices are big data contributors. “Big data” simply refers to the collection and analysis of large quantities of data in an attempt to recognize patterns.
This used to be prohibitively time-consuming, but advances in machine learning have turned the tide. We can use artificial intelligence (AI) technology to recognize patterns in vast and complex data sets that human brains would take centuries to discover and unwrap.
This is not pie-in-the-sky for the far future. Big data in healthcare is expected to experience a phenomenal compound annual growth rate of 36% by 2025, with electronic health records and practical management solutions alone tipping the scales at $68 billion in value by 2024.
So what does a big data healthcare landscape look like? Here are five things you should know about big data in healthcare:
1) Big Data Assists Disease Prevention
At over $3 trillion per year in delivery expenditure, US healthcare accounts for 17.6% of gross domestic product (GDP).
This huge number could be drastically reduced if the diseases for which Americans were going to physicians could be prevented altogether.
Wearable devices have already produced a mountain of data that doctors could use in discovering diseases early, before treatment courses become costly and the prognosis poor. FitBit alone has documented:
- 90 billion hours of cardiovascular data
- 167 billion minutes of exercise
- 85 trillion (yes, with a “t”) steps taken.
Doctors can cross-reference these statistics with other data not directly related to disease, but which can help identify if a patient is at increased risk of certain conditions. Data applicable to the diagnosis and prevention of disease include:
- Family history
- Lifestyle habits (e.g. smoking, alcohol, fitness)
- Prescription history.
Imagine if doctors knew ahead of time that a certain patient was at risk of developing type-2 diabetes, a costly and debilitating condition with a long prediabetic period that often shows no symptoms?
Or catch a cancerous growth in the early stages rather than at the more advanced, more damaging or inoperable stages?
Or, what if it could sort through billions of medical records that find the five past cases, pointing to a rare disease the primary care doctor has never seen in person?
Think of the suffering reduced and the money saved by reading the data and providing early and prophylactic healthcare instead of “sick care.”
According to Forbes, big data analytics have successfully been put to use at Seattle Children’s Hospital to make diagnoses and identify treatment plans that would previously have taken weeks to make, weeks any seriously ill, hospitalized child might not have.
Parkland Hospital in Texas used big data analytics to reduce 30-day readmissions by 31% for heart patients, saving half a million dollars per year for the hospital and with an even greater impact on the patients.
2) Big Data May Reduce Human Error and Treatment Fraud
The surplus of available data eliminates much of the educated guesswork and intuition doctors used to rely on to reach diagnoses.
Human error and the vastness of human knowledge aside, fraud is another major concern driving up care costs and leading to negative patient outcomes—for example, fraudulent diagnosis and prescription fraud to obtain medications illegally.
According to Chicago-based digital marketing experts Digital Authority Partners, the best estimates place the cost of medical fraud somewhere between $80 billion and $200 billion. Taken together, human error and fraud account for 10% of the total US healthcare bill. Using machine intelligence to analyse data for irregularities that may point to fraud or human error stands to make a big dent in that lumbering $3-trillion healthcare spend.
Big data analytics deployed by the Centers for Medicare and Medicaid Services are already credited with preventing frauds that could have led to $210 million in sunk costs.
Big data also has a major role to play in reducing duplicate medical records, processing entitlement reimbursements correctly, and identifying operational inefficiencies that cost patients dearly.
3) Big Data May Drive Healthcare Innovation
Healthcare thrives on innovation. That’s how we as a species cured polio and discovered penicillin. The innovation of today will cure the diseases of tomorrow. Big data sits on the cutting edge of this race toward the future of healthcare.
According to Orthogonal, the demand for big data analytics has driven the wearables market into the field of Medical Body Area Networks (MBAN) and Remote Patient Monitoring (RPM), technologies both critical to amassing and extracting from wearable devices the kind of data from which doctors can draw meaningful and incisive conclusions. MBAN and RPM allow patient data to transfer by WiFi and mobile data directly into specialized medical databases.
Then there’s CancerLinQ, a database of doctors’ notes compiled by the National Center for Tumor Diseases in Germany, which combs records from 1 million cancer patients across over 100 clinics, seeking out patterns indicative of risk factors and cure vectors.
Big data will lead the way in the development of new diagnostic tools and new drugs.
4) Big Data May Increase Medical Equipment Reliability
Patients’ bodies aren’t the only issues concerning doctors. Sophisticated diagnostic and treatment equipment plays a key role in the success of modern medical professionals and positive patient outcomes; thus, the availability of consistent, reliable and recognizably cost-effective technology is of paramount importance.
Flawed equipment means failed or missed diagnoses and–possibly years down the line–increased healthcare costs by addressing diseases at their later stages instead of ensuring early recognition and intervention.
There’s no doubt big data plays a key and growing role in the analytics and preventative maintenance of diagnostic and treatment equipment, keeping it in peak working order–and when these machines include AI robots increasingly tasked with operating on human hearts and brains, this role couldn’t get more critical.
5) Big Data Facilitates Centralized, Cloud-Based Medical Databases
One of big data’s hallmarks is that the information must all be located in one place. For a patient who has visited multiple doctors across numerous clinics, paper records just don’t cut it.
That’s where the cloud comes in. Cloud storage of medical records saves trees, reduces hospital admission costs, and makes it easier to compile the kind of databases into which big data analytics can really sink some teeth, producing actionable inferences.
The HealthConnect cloud-based system of health records implemented by Kaiser Permanente has already claimed credit for over $1 billion in savings.
Now picture a non-proprietary system of cloud-based medical records following patients throughout their lives, whether they change doctors … change cities … change planets … wherever life takes them.
Of course, along with data storage come new data security concerns. Keeping many sensitive records in one place means security controls have a better chance to evolve in step with our ability to compile that data, also increasing the likelihood of human errors being spotted in anomalous input patterns.
Doctors, pharma and medical companies, government healthcare agencies, and patients alike are all waking up to the promise offered by big data.
If anything outweighs the paranoiac concerns of being “watched” by the devices we carry, it has to be the possibility of these devices collecting data that could one day save our lives … or at least save us a fortune in medical bills.
Data mismanagement or ignorance represents a missed opportunity and an operational inefficiency, both of which cost us all in the long run. This cost must be weighed when we consider the challenges of both safely and ethically using big data analytics in the delivery of healthcare–and weighed against the potential of the innovation of big data to reduce suffering.
What is the purpose of medicine if not to achieve that?