|
About us
|
|
| | Background |
|
| | Physmark has been providing software for healthcare for over three decades. Its Medicomp software introduced in 1986 was the first software for the management of Independent Physician Associations that accepted capitated monthly payments for delivery of care. PayerSoft was the next iteration of software for Integrated delivery systems (IDS). Over 50 of the largest IDSs used PayerSoft in managing their claims processing and payment systems. In 2002 a newer version of PayerSoft was used by a self-insured organization, as an audit tool to monitor the administrative services provided by a large healthplan. Payments for these services were on a per member, per month basis. |
|
| | The very first audit revealed fraud and abuse in the form of overpayments of $4 million, because over 3,000 patients had their names entered twice in the antiquated system. At the urging of the Executive Director of that organization we undertook in 2004 a systematic study of costs in healthcare, with the goal of finding new ways to lower them, without sacrificing the quality of care. |
|
| | In 2009 the work was complete and we applied for a patent. In 2014 we received a patent and that is what underlies our software, CareMaps. |
|
| | Our Approach |
|
| | Economists when managing an economy start with a forecasting model, to lay a fixed 12 month path and then monitor variations to study causes and find tweaks to correct the flaws. We decided to follow this approach but we ran into some problems. |
|
| | First, we found that in healthcare, forecasting or predictive analytics, as it is called, was based on statistics. Data from a period were used to identify correlations and formulate regression relations. Then the regression equations were extrapolated to future periods, even though that was mathematically unjustified. We decided to adopt a dynamic model, where time dependence was explicit. |
|
| | Second, this was not easy since we had really no metrics for health of populations and for changes that place in them. Also, we had to take a stab at describing the mathematics of time evolution of populations. We used analogies from weather prediction to modify a 225 year old equation to arrive at a solution. |
|
| | We found that the 12-month forecasts had less than 5% error in predicting the numbers of people developing various chronic conditions and cancer and also the costs they incur for treatments. |
|
| | Just as the economists moved away from correlations to econometric and mathematical models, we demonstrate how modeling can be introduced in healthcare as well. |
|
| | As described elsewhere, forecasting also makes CareMaps a powerful digital forensicTM tool to detect fraud and abuse. This can find use in Medicare and private insurance managed care programs. |
|
| | Premiums are a puzzle wrapped in a mystery and CareMaps is able to interpret them as costs of treatment. As we will explain, the only way to reduce costs in healthcare is to lower premiums. |
|
