If We Only Had the Data

Clean, Aggregated and Accessible Data Offers Many Opportunities in the World of Healthcare

What can be done with all of the healthcare data existing today? That is, what can be done if it were accessible. Healthcare data comes from many, many, sources. Its structured and unstructured. So how can an organization gain access to the amount of data that prove to be statistically significant to the problems they need to solve? Remember, the more data involved in the sample, the more accurate the level of confidence in the prediction!

"Someone has just given you access to possibly half of ALL claims data for 5, 10, or even 15 years back! Think about it! You would have every diagnosis, every procedure, every hospitalization, every test, everything from every treatment. You have a chance to revolutionize health care, save lives, and lower costs. What would you do first, second, third, even fourth?"

Would you use predictive analytics to predict the outcome of a patient and the length of time to the outcome, not to mention the cost?

Would you use prescriptive analytics to advise treatment on a patient based upon what has been successful in the past, the meds, the procedures, and practices? Could your data analysis prevent a person from moving from stage 1 to stage 2?

There's a real clamor for transparency in the cost of healthcare. Perhaps, you would develop an "estimating" model using the data to compare a patient's diagnosis with similar diagnoses to predict the price they should expect to pay? You get an estimate from your plumber, why not your Urologist?

Should you take it a step farther and provide the hospital with comparisons of their costs compared to the total healthcare marketplace? Are they competitive, and if not, what are the areas to be addressed to align them?

Perhaps, you should look at how you could develop a "real" Social Determinants of Cost (SDoH) model? The SDoH model could guide treatment for persons in challenging socio-economic situations, educationally voided conditions, or food insecure households? Did you realize the healthcare community estimates SDoH accounts for 80% of the outcomes of healthcare and the actual treatment only 20%? Doesn't that mean we should focus on SDoH?

Wait, isn't gun control a real problem in the US? Aren't there many that say Mental Health is a cause? Would you use the data to identify the signs of mental health and violence? Shouldn't we also use this for the good of the people, by finding out which treatments work for which people? Are individuals "wired" the same, and do statistics matter when it comes to mental health?

Here's a big one! Can we tie this data to the human genome mapping in an advanced manner to better identify patterns and conditions that may occur in the future for the population? If the patient has an X chromosome, persons with the same chromosome have experienced condition Y.

Oh, by the way, are vaccinations worth it? There is much discussion as to the correlation between vaccinations and autism. Is the cost worth the risk, or does the risk outweigh the cost? Has the level of autism increased as a result of protection, or is it because professionals didn't accurately diagnose autism and report it in the past?

Then, there are those evil capitalist profits that provide the funds needed to pay salaries and build infrastructure. Can we use the data to help companies better and more effectively insure their employees at the lowest possible cost? Let's build upon what we've learned from the predictive analytics mentioned earlier. For example, a patient is into the 4th of 5 stages of treatment for a specific condition. Does the company need to budget the full amount to protect the employee for the level of treatment stage 1 through 4 again next year? If we know that step 5 is primarily follow-up, with few drugs and procedures, we budget less because this stage requires fewer treatment dollars.

There's also malpractice! We see it every day on TV. A secret camera captures someone cheating on healthcare claims, duplicating or fabricating charges, and overbilling Medicare. Can we use the data to prevent malpractice? By the way, stopping malpractice results in less expense for the insurance companies and means more profits for the capitalists. Still, since the government influences pricing, doesn't the prevention of malpractice lower the cost of insurance for the proletariat? That's good, right?

Speaking of the government, all politicians say we must fix Medicare/Medicaid! Some say, "Medicare for All"! How can we use our data to reduce overcharges for Medicare/Medicaid? We hear that healthcare professionals diagnose through the process of elimination and run extra tests that cost more money, just to make sure the patient "DOESN'T" have condition "X." If our data could, within an expected level of confidence, rule out 3 of the five tests that a healthcare professional is required to run to cover all the bases and protect themselves from litigation, wouldn't that save money in healthcare treatment? If treatment is lower and more specific, that should result in lower healthcare premiums, correct?

Maybe the healthcare payers would have a better idea of how to approve or disapprove treatments that require pre-approval? We want to eliminate Type 1 errors. An example of a Type 1 error might be that a test or treatment was denied because the data indicated it was not needed. In reality, the test was necessary but not performed. That could result in a horrible outcome, even death for the patient. No one likes Type 1 errors! Still, a Type 2 error is unfortunate in its own way? Well, not as bad as death, but proceeding with the test when it wasn't necessary would undoubtedly be a waste of time for the patient, a waste of money for the payer, and a waste of resource utilization for the provider. Everyone loses, including the population, because insurance premiums are higher due to unnecessary testing. Sounds like the circular flow concept of what goes around; well, you know the rest!

Since the 1980's we've heard a clamor for "Tort Reform" to reduce the liability of healthcare providers when it comes to being sued for malpractice. The theory is that if monetary amounts from legal judgments are limited, malpractice insurance would be less expensive. Lower premiums for malpractice insurance should reduce healthcare costs. Could the data be used to support the actions of the healthcare professional or confirm the suspicions of the patient feeling there were egregious errors in treatment?

Continuing with the legal side of healthcare, could the data be used to create or supplement a healthcare law library to provide statistical evidence and support for the plaintiff's claim or a strong rebuttal for the defendant?

These ideas only scratch the surface of the discovery process, and every suggestion has a derivative that is waiting to be defined. Remember, the more data, the better the probability of accurate prediction.