Thinking in Systems to Solve Payers Problems
Opportunity a Key Priority
The healthcare industry is plagued by over-utilization and over expenditures. Many payer organizations today want to improve their payment systems to appropriately pay providers and avoid instances of improper payments or cases of fraud, waste, or abuse. Up to now, many payers utilize legacy systems with additional attached methodologies to review claims based on rules or policies. Opportunities are where big volumes of low amount claims are scrutinized. Using AI is an avenue, but not all AI approaches are solving the problem, here is where Mobilizing Computable Biomedical Knowledge (MCBK) artifacts comes into play to save the day.
Why use Mobilized Computable Biomedical Knowledge in claim reviews?
Naviquis utilizes Computable Biomedical Knowledge as part of its analytical tools and focused process to evaluate claims using knowledge about human health, conditions affecting human health, patients history, situations pertaining to best clinical practices and guidelines that are explicit and transformed into machine-executable processing code through a series of expert coordinated systems and algorithms, and therefore can be represented and reasoned upon using logic, formal standards, and mathematical and statistical models and other computable approaches. The use of this methodology ensures that absolutely all claims are truly revised for all its merits, including monitoring providers, procedures, patient’s needs and coding rules and regulations requirements, but most importantly, clinical best practices and paid based on appropriate parameters.
The adoption of computable biomedical knowledge will be essential to the realization and precision in detection of claims for fraud, waste, and abuse and next generation clinical analytics is key in the success of high functioning payer organizations. The technique allows to monitor all aspects of the transactions and review health systems, patients, and providers. Discerning based on the knowledge input, which claims are to be paid and which claims should be denied.
When innovation merges with creativity
Several domains are necessary to implement such tool as MCBK: data science, medical knowledge, best practices, data modeling and analytics, and computer coding and programming. The digital tools created from this knowledge are structured in a way that claims can be parsed, analyzed, and evaluated by the tool in record speed, following all aspects of medical and dental knowledge.
Although the field of MCBK is still new, we have taken its principles to advance technologically in our models and algorithms to check claims appropriateness for payment. The incorporation of recommendations, computable clinical guidelines, resources, and predictive models surpass what simple Artificial Intelligence algorithms can do. What makes our methodology a winner is the use of digital objects based on the different clinical domains’ contents.
The results show that on average 3 to 6% of claimed amounts; which have already gone through the payer’s processing systems are flagged red for anomalies. This means there is an opportunity for savings on the payers’ side, not to mention the opportunity to educate providers in the proper billing practices. Since the MCBK uses biomedical an dental knowledge and other attributes, identifying issues is accompany by its applicable best clinical practices, rules, and protocols to be communicated to offending.
When innovation connects with creativity results are impressive
Evolving technology available allows us to create disruptions like the utilization of MCBKs for claim reviews. As many say: “They best way to predict the future is to create it.” We have reimagined fundamental capabilities of Data Science and connected it with algorithms and model development using MCBK’s to create a robust system. The plain old analysis and informatics create models to estimate the occurrence of deviations from the norm, mostly statistical based using the standard deviation from the median. Using other parameters like best practices, patient history, rules and regulations and forensics a model using MCBK surpasses plain Artificial Intelligence. AI models are as good as the data they are trained with; sometimes such data have gross biases based on the type and location of its population. Other times, data used to train is only a small percentage of total cases, therefore not representing the complete picture.
Utilizing the MCBK and multi-vector approach although already known in data science, nobody had thought of combining these approaches. There is where our innovation comes from.
Moreover, these objectives go in synch with the quintuple aim of healthcare: to lower costs, ensure provider’s wellbeing,
optimize patient’s experience, and ensure that there is equity and inclusion, not to mention improved population health. We have created high performance models and algorithms with this approach that benefit payors.
“Innovation is seeing what everybody has seen and thinking what nobody has thought” Nobel Laureate – Albert Gyorgyi
Not surprisingly, high performers have both fewer opportunities and a higher success rate than middle and low performers. Being able to focus on the healthcare industry with correct models prioritizes the right opportunities and success factors which make our research and work worthwhile.
When asking top executives at the major insurance companies around the world; what are the main reasons you keep overpaying on healthcare claims. The main answer is: Because of lack of bandwidth and resources. Mostly using legacy systems and homegrown payment software, adding additional checks and balances to the mix becomes prohibitive and many times, there is lack of alignment between the decision makers wishes and what the team can produce in a reasonable amount of time.
Our technology enables knowledge in the healthcare domains to be represented in computable form and expressed in sophisticated models to be attached to existing systems and produce outstanding results and savings for the insurance industry. We are unleashing the power of information technology, data science and MCBK to deliver useful and actionable insights that are readily available to integrate with legacy systems. Adding these intelligent structures and integrating to payors’ databases, legacy systems, enterprise systems or ERP creates enormous synergies an efficiency. All together these integrations move the organization from descriptive analytics to prescriptive analytics and may be the difference between success and struggles.
The aim of prescriptive analytics is to recognize what is going on, figure out what is most probable of happening and make decisions to steer the outcomes in the most favorable performance possible.
The difference between AI and MCBK
Pure Artificial Intelligence algorithms use machine learning disciplines and tools which develop, study, and use models and procedures to “learn” and perform certain tasks without having implicitly programmed those. AI works just by looking at examples and trained data, whether supervised or not. The system finds its own correlations and approximations. In certain form, these methodologies are a “black box”. Even the best data scientist cannot explain 100% where the results come from an AI. These issues raise questions about accountability, trust, and precision. Who is responsible if the AI is biased?
Meanwhile, using MCBK, we explicitly start from the medical, biomedical, or dental knowledge, usually a clinical narrative, protocols, or best practices for a particular procedure. From such narratives, policies and human readable knowledge, we create a pseudo code, which in turn becomes the basis for the algorithms and models which are later clearly programmed by our technical team. There is no longer a “black box”. We know exactly what the model is supposed to do and can explain where the results come from. Everything is backed up by documentation and reasons why a procedure or protocol should be performed or not. Likewise, whether it should be paid or not and even whether it is abusive or over-utilized.
We consider the use of MCBK for developing algorithms a better approach than plain AI because it is a robust approach which looks at many angles and resembles closely real-life situations without incurring in guesses or in convoluted mathematical approximations and statistical predictions. It is a trusted methodology and an ethical use of information technology. Our methodology helps building corporate capabilities to the next level, transforming the payers into a truly data driven organization without the risks involved.
コメント