Sources Sought Notice - USING ADVANCED TECHNOLOGY IN MEDICAL REVIEW

Purpose of Notice: This is a Sources Sought notice posted for informational purposes only and will be used to assist the Centers for Medicare and Medicaid Services (CMS) in determining the availability of eligible Service-Disabled Veteran-Owned Small Business, HUB Zone, women-owned small business, and 8(a) with the technical capability necessary to successfully perform the requirements described herein. This is not a formal request for proposal (RFP) and does not commit the Centers for Medicare & Medicaid Services (CMS) to award a contract now or in the future. CMS seeks to identify and solicit feedback from capable sources regarding leading cutting-edge technology capabilities for machine learning and natural language processing software for Medicare FFS health care claims and medical records. At this time, CMS is only interested in procuring new application(s) that could be used by CMS and/or its contractors to identify improperly billed claims through non-human review of the medical record. The software shall be able to identify improper payments based on CMS coverage and documentation requirements and identify coding and medical necessity errors. CMS is seeking software for Medicare FFS's medical review programs that includes the creation of a dashboard available to CMS with real time statistics. Background CMS has historically defined program integrity very simply: - pay it right. Program Integrity focuses on paying the right amount, to legitimate providers and suppliers, for covered, reasonable and necessary services provided to eligible beneficiaries while taking aggressive actions to eliminate fraud and abuse. Program integrity requires that we protect the resources entrusted to our nation's public health care programs while also protecting the health and well-being of beneficiaries. CMS must work every day to guarantee that we are an accountable steward of Medicare and Medicaid program dollars. Accomplishing this is one of CMS top priorities. Medicare now processes and pays approximately 4 million FFS claims per day and over 900 million claims annually. To do this properly - to - pay it right - contractors must determine whether the claim represents the correct charge, for a covered, reasonable, and necessary service, provided to an eligible beneficiary, by a legitimate provider or supplier, and the item or service is properly documented in the patient record. Doing this, while still paying claims quickly, is a complex task. Increasingly, we have turned to computer edits to automatically review claims before they are paid, and to data analytics, by which we can detect patterns of overuse or other improprieties via post-pay review. But among the almost 1 billion claims processed and paid every year, fewer than 3 tenths of 1 percent receive any sort of medical record review. Put another way, 99.7 percent of all FFS Medicare claims are processed and paid within 17 days without any medical review. When they are medically reviewed, which means a human clinician checks the patient medical record to confirm compliance with Medicare FFS documentation rules, CMS sees a 5 to 1 return on investment when comparing cost to recoveries. The CMS level of medical review is less than that of private commercial payers. Private insurers aggressively use tools such as prior authorization to closely monitor whether claims meet payer guidelines, and they apply these tools and practices to as much as five percent of claims, or more than fifteen times what Medicare does. While more reviews by CMS could reduce improper payments, the need for a clinician to personally review patient charts and determine if claims meet payment requirements is very costly. The level of provider burden associated with medical review is also an important consideration. Often, improper payments occur because providers and suppliers have not followed Medicare payment, coverage, and coding guidelines or have not submitted the appropriate requested documentation. Provider/supplier error is often made with no intent to defraud Medicare or misuse its resources and CMS recognizes the importance of its role in educating providers and suppliers about Medicare's requirements. In addition to continuing our national education efforts such as hosting Open Door Forums and publishing Medicare Learning Network educational materials, we recently began focusing more on personalized education through a process called Targeted Probe and Educate (TPE), conducted by the MACs. Despite these efforts, more needs to be done to drive more review of claims through the use of cutting edge technology. As part of our initial effort to identify new technology to enhance medical review capabilities, CMS issued two Requests for Information on October 21, 2019: the Request for Information on the Future of Program Integrity, and the Request for Information on Using Advanced Technology in Program Integrity. The Requests for Information (RFIs) focused on obtaining input from stakeholders and experts on innovative methods and tools to elevate CMS program integrity efforts and on how the agency can better use emerging technologies, such as AI and machine learning, to ensure proper claims payment, reduce provider burden, and overall, conduct program integrity in a more efficient manner. The RFIs accounted for several areas, such as, Electronic Health Records and Finding Documentation Requirements, Data Analytics and Data Systems, Medicare Claim Review, and Education.[1] Each of these influence our medical review efforts. Looking forward, CMS is seeking new, innovative solutions and technologies, such as artificial intelligence tools such as machine learning and/or natural language processing, which may be more cost effective to the Medicare FFS program. Capability Statements In response to this Sources Sought Notice, CMS is seeking capability statements. Capability statements are NOT proposals and do not address price/cost. The contractor shall submit capability statements that demonstrate the capability and capacity to provide an AI and/or machine learning enterprise solution. For the purposes of responding to this notice, contractors should address their capability of the following: Provide cutting edge technology using AI - such as machine learning, natural language processing, or other tools - respondents must show the capability to read a medical record in any format (structured documents, pdf or tif, and handwritten) and to convert the medical record to structured format / useable data, when necessary. The technology shall be able to read the medical record and make recommendations regarding the accuracy and appropriateness of payment based on Medicare Fee-for-Service coverage, payment, and documentation requirements. The technology shall be able to make recommendations regarding payment, partial payment, and full or partial denial of the claim based on a software based review of the medical record. The technology shall be able to identify possible reasons for denial based on incorrect coding, insufficient documentation, and medical necessity. The technology shall be able to identify possible denial reasons for all 3 scenarios. Technology must achieve greater savings for taxpayers, allow CMS to review more claims, ensure proper claims payment, reduce provider burden, and overall, conduct medical review in a more efficient manner. The capability statement shall refer to other Federal and/or non-Federal experience of similar size and scope necessitating comparable adherence to Federal system security and privacy requirements and all medical review requirements, including statutory and regulatory requirements. The contractor shall demonstrate the solutions / systems capability in meeting all of the following requirements: Be a software and/or system with the ability to be deployed to multiple medical review contractors through licenses. Demonstrate the ability to have real time tracking data available at the individual contractor level and the Medicare FFS level. Incorporate, at a minimum, the top 10 Service Types identified by CMS. The service types could be identified in some part based on the annual improper payment rate associated with Part A, Part B and Durable Medical Equipment, Prosthetics, Orthotics and Supplies (DMEPOS) which can be found in the 2019 Medicare Fee-for-Service Supplemental Improper Payment Data Report into the technology. Demonstrate the ability to make recommendations regarding payment, partial payment, or denial based across all claim types. If the product can only edit certain claim types, please provide details. These recommendations shall occur without human review of the medical record. Communicate recommendations for payment and/or denial to the appropriate Medicare contractor. Base recommendations on Medicare FFS coverage, payment, documentation and prior authorization requirements in place on the date of service. Validate results of recommendations for payment (including recommendations for payment and recommendations for denial of payment). Identify and track reasons for payment denials, including coverage requirements not met, error categories, etc., in order to provide real-time results to providers and suppliers and generate dashboards, reports, etc., for CMS. Note: At a minimum, the contractor shall account for the categories specified in the 2019 Medicare Fee-for-Service Supplemental Improper Payment Data Report (2019 Data). Track historical interface and seamless transition of history of authenticated user access across sessions. Provide maintenance inspections and help desk support to include the development of content, establishment and refinement of operating procedures, maintenance of a comprehensive knowledgebase of previously answered questions/issues, and the tracking of operational metrics. Must be Federal Information Security Management Act (FISMA) compliant and must maintain through the life of the contract. Must have the ability to mitigate real or perceived conflicts of interest. For example, a real or perceived conflict of interest would be a product that is currently providing services to providers. This type of product could have a real or perceived conflict of interest if used to audit Medicare FFS providers. An integral part of this Sources Sought is to identify and understand what is available on the market. CMS understands that more than one product may be required to obtain the full functionality required above. Vendors expressing an interest should provide appropriate capability statements that include the following information: The name and address of the firm Dun and Bradstreet identification number Point of contact: name, title, phone, fax and e-mail; Information Technology Security Official: name, title, phone, fax and e-mail; Ability to meet requirements as listed above. Each requirement must be addressed separately. Description of the organization's experience supporting large integrations, including any multi-tiered applications with Medicare contractors and/or commercial health care plans; including but not limited to, Medicare Administrative Contractors, Recovery Audit Contractors, Medicare Advantage Plans, Medicare shared systems, etc. Provide, at a minimum, two descriptions of large-scale integration efforts to include agency/company name,, type of contract, total dollar value, and description. Description of the organizations experience deploying and field-testing similar application(s). Provide a high-level description of the architecture of the software being proposed. Ability to support CMS Expedited Lifecycle (XLC) processes or comparable system development lifecycle frameworks. Responses Requested Responses to the Sources Sought Notice are required no later than 11:30 am, EST December 21, 2020. Documentation should be sent to: tyrone.scott@cms.hhs.gov and cc:aaron.blackshire@cms.hhs.gov in the following format: Microsoft Word (or PDF) document with page size 8.5 by 11 inches. Font shall be Times New Roman Size 12 with no less than single spacing between lines. The font size for any charts will be Calibri 11. The maximum number of pages for submission is 15 pages. Please be advised that email transmitted files over 5 megabytes are not delivered during standard working hours, are only released from the CMS server after 5:00 PM EST, and may affect the timeliness of your response. Additional Information CMS also requests the following additional information: Business Information: Type of company (e.g., small business, 8(a), woman owned, veteran, etc.) as validated via the System for Award Management (SAM), the Official U.S. Government system that consolidated the capabilities of CCR/FedReg, ORCA, and EPLS.. All potential offerors must register on the SAM