Accordion Health makes high-precision granular predictions of healthcare costs and outcomes. We help people save money and thrive by identifying paths to low-cost, high-quality outcomes. We work with providers, plans, and employers to target the actions they can take to achieve these outcomes.
With over a decade of data analytics experience in the healthcare domain, Accordion Health has developed many innovative technologies to help better understand and utilize healthcare data. Our team has integrated datasets from many different sources into our technologies. Using our patent-pending forecast engine, we can view all potential medical paths of individuals. We’ve created a temporal analytics engine which uses our forecast engine and multi-source integration to answer various questions related to MA plans. Our risk adjustment module, which operates over almost any kind of data format, combines with our temporal analytics engine and allows plans to make informed decisions for the future. Finally, our advanced machine learning models, combined with health records, redefine physician assessment and matching metrics by assessing both the short-term and long-term differences in physicians.
The healthcare landscape is rapidly evolving. Payment models are shifting from Fee For Service to Value-based Care models. Every year, hundreds of different performance metrics are added/removed and measured by governments and non-profit organizations. More importantly, patient population is drastically changing. Plans now need to look ahead and prepare for the changes.
Combining our multi-source integration and forecast engine, Accordion Health answers various questions on time-varying quantities, such as plans' Star Ratings, enrollment sizes, reimbursement rates, and competitors' performance measures.
Insight is limited by its data; incomplete data only provides you with incomplete insights. After 100 years of regulation and fragmentation in the healthcare industry, there is no single healthcare dataset that captures a comprehensive picture of patients.
Accordion Health addresses this issue by integrating multiple datasets from payers, employers, providers, and even the datasets with different aggregation levels such as state-level, county-level, and ZIP code-level. We bring socio-economic datasets from Census, surveillance survey datasets from Centers for Disease Control and Prevention (CDC), and public/private datasets from social media.
Prediction problems in healthcare are much more than applying machine learning toolkits to preprocessed datasets.
Thousands of different medical events can happen for a patient. Patients may enter or exit the system at any time (e.g. travel, moving, etc.), giving discontinuities in the data. Sometimes, doctors and nurses may put wrong information in the database. Most of Electronic Health Records are unstructured. Just preprocessing the data and defining scientific problems require deep domain expertise in both data analytics and healthcare industry.
Accordion Health brings over a decade of data analytics experience in the healthcare domain, and our experience and insights are coded into our patent-pending forecast engine. The forecast engine automatically groups and tracks individuals with similar demographics and medical histories, showing the paths for all potential medical events.
Payment models are shifting from Fee For Service to "risk"-adjusted capitated rates. Plans need to know how much they will be paid and spend in the next year to manage their finances. Accountable Care Organizations, who are considering the transition to Medicare Advantage plans, also want to know how much they will have in their budget to operate the business.
Accordion Health has implemented various risk adjust models and their data preprocessing modules. Our risk adjustment module operates over almost any kind of data formats including 837, CCLF, any other commerical standards. Combined with our temporal analytics engine, plans can see where their risk scores are heading, where the gap is happening, and how to close the gap.
Every doctor is different. Some doctors specialize in specific procedures for young adults, while some others are better at treating cases with chronic conditions. Existing physician performance metrics not only fail to account for differences in patients but also suffer from short-term vision, analyzing outcomes only 1-2 weeks after a medical treatment.
Accordion Health's cutting-edge machine learning models combined with electronic health records will redefine physician assessment and matching metrics. The algorithms quantitatively assess both the short-term and long-term differences in physicians from a cost, outcome (readmission rate / complication rate), and utilization perspective for each individual patient.
Accordion Health provides many solutions for problems faced by Medicare Advantage Plans, ACOS, employers, and consumers. We provide an interactive Star Rating dashboard, combined with our data analytics consulting services, to help MA Plans make changes to improve their services. For MA Plans and ACOS, we supply a wide range of risk-adjustment related services. Employers can save money on health benefits by providing a tool, our Chestnut mobile app, to their employees that allows them to take preventive action with little effort. Our public facing product, Pistachio, allows consumers to easily find the perfect set of healthcare plans tailored to their needs and preferences. Through the use of our other mobile app, Almond, consumers will be equipped with actionable insights on how to plan for post-operation events.
Every year the Centers for Medicare and Medicaid Services (CMS) collects data from Medicare Advantage (MA) plans. Each different quality and performance metric (in total more than 40, picked from hundreds of different measures) is assigned a score (1 to 5 stars, with 5 being the best). These stars are then aggregated with weights to generate a single, overall star rating from 1 star (poor performance) to 5 stars (excellent performance) for each MA plan. Now why is this important? Because if you get four or five stars, CMS gives you "bonuses!”
We can't stress enough that this billion dollar bonus system is not easy to understand; yes, the system is esoteric. Accordion Health compiled multiple years of different quality measurements and integrated them with multiple socio-economic, regional, and risk-adjustment datasets. Try our ASTRID (Accordion STar Rating Interactive Dashboard). ASTRID, combined with our in-depth data analytics consulting services, helps you see this complex bonus system clearer. Find out the metrics and markets your business needs to focus on in the next year.
Payment models are shifting from fee-for-service to risk-adjusted capitation. In the capitated payment world, payments are arranged proportional to the number of enrollees. Wait, what happens if my plan has more sick patients than the others? Don't forget about the term "risk-adjusted.” Using a statistical model to estimate potential "risk,” the payments to plans with a greater amount of sick patients are adjusted to be higher than the payments to plans with healthier people. The theory is simple, but the reality isn't. Risk is calculated from a certain "statistical" model. Let's recall the famous quote from statistician George E. P. Box: "all models are wrong, but some are useful.” In many cases, the risk-adjusted payments are significantly off from what you will eventually pay.
Accordion Health implemented various existing risk models, analyzed their performance and corner cases, and developed a cutting-edge risk model that estimates the true risk as close as possible. We provide a wide spectrum of services ranging from calculating multiple risk scores for MA plans and ACOs, and provide our accurate custom risk scores for payers to manage their finances.
Healthcare costs are rising and the burden is on self-funded employers. Good health benefits will attract top quality talents, but they are expensive for employers and are only getting worse every year. Should employers change their Third Party Administrators? Should employers find new insurance brokers? Should employers re-negotiate the contracts with provider groups?
What if Accordion Health has a much simpler and faster solution for your problems? The secret sauce is in our data-driven patient-doctor matching algorithm. Did you know that just by choosing a better personalized doctor, you can get a higher quality procedure but pay less? Every doctor is different. Some doctors specialize in specific procedures for young adults, while some others are better at treating cases with chronic conditions. Choose a better doctor with Chestnut (mobile app): the easiest way to save money and get healthier.
Navigating within the US healthcare system can be extremely frustrating for consumers. We understand that between choosing a health plan, doctor, and nursing facility there are so many decisions to make, but not enough data around. Don't trust the ratings and reviews on the internet. Patient satisfaction is hardly related to the effectiveness of care for the most part.
Pistachio: Accordion Health has compiled hundreds quality metrics about Medicare Advantage plans from reliable public/non-profit agencies. Our dynamic question algorithm asks about which services you want to be covered, which quality metrics you care for, and which chronic conditions you have, then recommends a set of perfect healthcare plans for your needs. Check out pistachio.gethealthnuts.com for you and your parents.
Almond: Accordion Health has analyzed millions of different medical pathways. Depending on your medical history, chronic conditions, age, and gender, your complications after a surgery can be drastically different. Backed by our patent-pending forecast engine, Almond provides a real-time forecast on post-operation events. You can prevent a bad event happening if you can predict it. Check out our Almond app (coming soon).