#futureinsurance focuses on insureds' operations. The biggest competitors to #insurancecarriers will be vendors, MGAs, #captives, and the insureds themselves. @pulse_data sheds light on the future of claims & risk management.
#longtermcare facilities are sued all the time. Medically fragile people, often referred from a hospital post surgery, have an accident in rehab or in a long term stay. These accidents are frequent, severe, and routinely end up with a law suit. This is why their insurance sucks.
US Healthcare basically only treats problems. @pulse_data flips this script. They ran machine learning algorithms on urine data to find one of the most expensive diseases before it starts: chronic kidney disease. They start treatment before the patient needs dialysis.
Proactive renal care reduces patient mortality and improves the patients' experience. In addition, this radically reduces the cost of treatment. Chronic kidney disease costs healthcare providers over $100 billion annually (that number seems high - blame @forbes)
Regardless, kidney disease is expensive. We can use machine learning to see it sooner. Early treatment saves money. Everybody wins. So, how does this relate to insurance?
@pulse_data created an inference map from billing receipts to actual clinical data. They used data to find what to look for to prevent a loss - chronic kidney disease. Culling this data from other sources can lead to predictions for other types of perils.
For example, motion sensors may provide insight as to early onset of dementia as there are imperceptible changes to gait as the disease starts. Watches can track a patient's pulse rate to monitor arrhythmia and recommend a patient for cardiac monitoring before cardiac arrest.
Patient beds may have sensors in them that detect when a patient has not moved in a period of time. This can prioritize which patients need to be flipped in order to preclude pressure ulcers and assist with treatment.
Thus, risk management becomes a function of identifying the best data, running simulations, and managing care in light of the results. This changes risk management from "being careful" to an operations-focused enterprise.
Most admitted carriers lack the sophistication to underwrite to this degree of risk management. Consequently, #captiveinsurance comes into play. #captives, #MGAs, #MGUs, and similar risk financing companies can take these loss control metrics into account when underwriting risks.
These operational, data driven risk management actions will reduce claims and lower the cost of insurance. In addition, patient outcomes will improve. Meanwhile, insurance companies are paying fewer claims. Hospitals, long term care facilities, and similar players can do this.
If the healthcare providers manage their risks, then their cost of professional liability / malpractice insurance will plummet. The commercial markets won't respond adequately, and they'll be stuck overpaying for other bad actors' risks.
There is no reason why this example should be limited to healthcare. Similar simulations can be run on #workerscomp claims data and mine natural language to find leading indicators of creeping catastrophic claims. Operators are better at doing this than insurers.
Fundamentally, the operators care more about their risks than the insurers. Most risk managers are not going to bring the sophistication of big data. Most risk managers misunderstand the industry they're purportedly managing risk in. But the operators know their business.
Given that the cost of capital has never been cheaper, expect more data-driven solutions to enter the market place. These kinds of ideas are low hanging fruit. Machine learning already does this stuff. Carriers are never going to bring this to market. This is #futureinsurance
You can follow @bonedaddy03.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: