Cancer Treatment & Research | Moffitt



Clinical Trial Matching and Personalized Medicine SolutionRequest for Proposal Vendor Conference CallRFP 20-10-SSPDate:Tuesday, April 21, 2020Time:10:30 AMLocation:Conference CallAttendees:Moffitt Cancer Center: Lori Perks, Desiree Hanson, Dr. Kevin Hicks, Dr. Rachel Howard, Phillip Reisman, Erica Royster, Mattias Van Looveren2bPrecise Health: Kelly Miller Kelly.Miller@, Phillip Lajoie, Erin Scales, Eyal Greenberg BC Platforms: Jessica Johnson jessica.johnson@, Liisa Koivukoski, Jessica Johnson, Greg Andreola, Mike Bochnaik, Michael Saylor DEEP 6 AI: Spencer Mitchell spencer@deep6.ai, Eric Gildenhuys, Nour Malki Deep Lens: Simon Arkell simon@deeplens.ai, Dave Billiter, Mark Vance, Natalie Beeler, TJ Bowen GenomOncology: Anshu Vipparla anshu@, Teresa Fletcher, Frank McMullin, AJ Sura, Farhan Quraishi, James Cole IBM Watson Health: Briana Saucier bsaucier@us.Innoplexus Holdings: Aseem Sharma aseem.sharma@ Inteliquet: Robert Robbins rrobbins@, Kirk Junker, Jennifer Mutz, John Pappa, Ranesh Phillips InterSystems: Duke Garvin Duke.Garvin@, Matt Nee, Michael Rsenblum, Randy Pallotta, Nicholai Mitchko, Qi Li, Gary Monger LifeOmic: Briana Saucier bsaucier@us. NC3 Planners: William Newkirk william.newkirk@ VieCure: Mark Lori mark.lori@, Dr. Fred Ashbury, Michael Power XpertDox, LLC: Sameer Ather sameer.ather@, Bhupesh Panwar Not Present: Linguamatics, Public Sector Performance Associates (Monipatry Sanchez msanchez@), The HCI GroupIntroductions: Moffitt Cancer Center introductions. Roll call from each participating vendor taken.Summary: Moffitt Cancer Center is a comprehensive cancer center, designated by the National Cancer Institute, with facilities in Tampa, Florida. With facilities located on USF's campus, Moffitt Cancer Center operates health care, research and screening facilities, including a 206-bed acute care hospital, ambulatory clinic, research center and screening center. Moffitt was incorporated in 1984 and received its original 501(c)(3) tax exemption determination in 1985. The Moffitt system has grown to include three (3) not-for-profit subsidiaries of the Institute, including the H. Lee Moffitt Cancer Center and Research Institute Hospital, Inc. (the “Hospital”), the H. Lee Moffitt Cancer Center and Research Institute Foundation, Inc. (the “Foundation”), and the H. Lee Moffitt Cancer Center and Research Institute Lifetime Cancer Screening Center, Inc. (d/b/a “Moffitt Medical Group”), as well two (2) for-profit subsidiaries, the Moffitt Genetics Corporation (d/b/a “M2Gen”) and the Moffitt Technologies Corporation (“MTC”).Moffitt Cancer Center is seeking a solution that allows clinicians and staff to match patients to potential clinical trials and enables Moffitt’s Personalized Medicine Clinical Service workflows.Questions and answers regarding RFP: Vendor: XpertDoxQ. How do you find matching trials for a single patient among all those actively recruiting, and how do we identify all possible candidate patients for a specific trial. Which approach is Moffitt interested in? A: Both are of value.Vendor: VieCureQ. Are you looking for a solution not just to support trials and precision oncology, but also management of patients who are not currently on trial?A. Yes. For example, there could be patients utilizing the Personalized Medicine Clinical Service who are not necessarily enrolled on a clinical trial. These are also in scope.Vendor: InterSystems Q. Would Moffitt prefer this solution to be delivered as a service? A. That is the preference as this will facilitate easier integration, but this is not necessarily a strict requirement.Q. How many internal Moffitt users of the solution can we expect?A. Moffitt team will get a more specific answer to this question and provide response within one business day.Q. Is there an opportunity to submit additional questions and receive answers in time to incorporate into the RFP?A. Typically all questions will be addressed on this call, but if any additional questions arise, please forward to RFP email address before the close of business today (4/21) and answers will be distributed with the call minutes no later than tomorrow (4/22).Q. What data sources will feed into the solution?A. The two primary data sources feeding into the solution will be Moffitt’s Electronic Health Record (Cerner) and Clinical Trials Management System (OnCore). Ability to receive data from other systems is valuable as others may be needed in the future.Q. How much data needs to be managed each year? Should we expect to pull data from non-Moffitt sources? A. Yes, the ability to connect to external data sources (e.g. , personalized medicine / pharmacogenomics knowledge bases, and others) would be valuable.Q. Are patients who have data at an ambulatory site in scope?A. No, we’re not planning to connect directly to other hospital/community sites. Outside records from these sites would come to Moffitt during new patient intake and would be stored (mostly in PDF format) in our EHR. These would then be ingested by the solution.Q. Approximately how many clinical trials will be managed by the solution annually?A. At present we have 300 therapeutic trials open to accrual at this Center.Q. What downstream applications will the solution be integrating with?A. While we are interested in a user interface and reporting tools, this depends on the features of the solution. Whether we would need any downstream applications and what they would be is not pre-specified.Vendor: InnoplexusQ. Would patients to be matched to clinical trials only be coming from Moffitt systems, or will we also be looking at patient data from other public data sets?A. We will be focusing on trial matching for patients within Moffitt systems only.Vendor: IBM Watson HealthQ. Clarification of the name of Moffitt’s current Clinical Trials Management System.A. OnCore.Vendor: GenomOncologyQ. Would you prefer an on-premise solution or a Cloud solution?A. Cloud would be preferred, but is not critical. If you do not have Cloud capabilities, there may be potential to host the solution in an Azure instance.Q. Regarding section 5.5 of the RFP, are there any other end user groups we need to target training to in addition to the clinical trial coordinators and the personal medicine team members?A. These are the primary end users and the majority of training should be targeted to these groups.Q. What is the preferred contract length?A. This typically depends on the vendor.Vendor: Deep LensQ. What is the current solution you already have in place?A. At present, most work within the scope of this RFP is completed manually. Trial coordinators extract relevant information for trials matching manually and pass on directly to physicians, and the personalized medicine team have a home grown database they’ve been using for several years which includes mostly manually-entered demographic, clinical and molecular information on each patient.Q. Is this system going to be used for all clinical trials or just oncology?A. All Moffitt trials are oncology.Vendor: Deep 6 AIQ. Can you provide a little more about molecular testing at Moffitt? For example, do you primarily do testing in-house or is it sent out? Do the majority of these tests go to the same lab? What is the system used to store the data?A. In some cases, testing is completed in-house on our own hardware (although resulting data may still be sent out for analysis), and in others the samples are sent out to one of several different reference labs. Most data come back in PDF format, some in computable forms (XML, JSON, BAM, JSON, etc.). Typically, data is returned in batches. The ideal solution will be able to consume data in any of these formats.Q. When patients come from other systems and you scan their records in, what then happens to that information? Is the PDF entered electronically, summarized in a patient note for example?A. In some cases, a limited amount of this information is manually entered in the EMR (some specific variants may entered in a discrete field, for example), but were trying to steer away from that error-prone solution. We would prefer to ingest data from these documents electronically using tools such as optical character recognition or natural language processing going forward.Q. Do you envision the trial matching solution will integrate with the Pathways system?A. Down the road this would be desirable, but this is not critical for the current RFP.Vendor: BC PlatformsQ. Could you provide additional clarification on RR.5, generating and exporting personalized medicine consults in different file types? A. We currently have a homegrown MySQL database that we manually annotate with clinical variants before performing some quality checks. The information in the database can auto-populate a report (Microsoft Word document) for an individual patient, presenting the relevant variants in the preferred format for the physician to receive. When the report is delivered to the primary oncologist, this allows them to rapidly review the report content without needing to open multiple documents including the original reports, unless needed for validation. We also have an “intelligence” database which permits automatic linkage between a gene and variant with a pre-defined descriptive writeup.Vendor: 2bPrecise Health Q. Can you walk us through the clinical trials workflow today, specifically recruitment and onboarding? A. The process applied to both existing and new patients seen in various different clinics. Typically, an investigator on the trial will refer the patient to the Trials Office if they believe they are eligible for a specific study. A list of studies currently open to accrual (manual process managed by supervisor or coordinator within respective teams) is sent out approximately biweekly to the physicians.Q. Are you looking for a solution that will provide clinical decision support in terms of alerting a physician or coordinator to a patient’s potential eligibility?A. Yes.Q. Regarding requirement R3, what exactly is meant by “creating a curated knowledge base”?A. At present, the personalized medicine team list all clinical trials a patient is eligible for in their report based on genetic markers. This trial eligibility is currently automated from a list of active trials, but the list is not always up to date. With only a small personalized medicine staff and over 300 frequently changing trials available at the institution, it’s not efficient to manually maintain such a trial list.Q. Can it be assumed that there would be local support in curation of such a knowledge base?A. Yes, an active role of the personalized medicine team in this process would be preferred.Q. Several of the requirements refer to precision medicine. When we think precision medicine, we think more of point-of-care clinical support as the clinician works with the patient in real-time. You mentioned PharmGKB - can we assume availability of pharmacogenomic information for corresponding germline/somatic results at point of care is core to the requirements here?A. Yes.Q. Regarding requirement R.8 and “commonly used eligibility criteria” – do we have a clear definition of what these commonly used eligibility criteria are?A. A specific list does not exist, but it would be possible to narrow down what was most useful. This would take some time to develop.Q. You mentioned PDFs from reference labs, how are these received? Electronically? Scanned? Faxed? A. In some cases they are faxed. We typically try to get a “clean” pdf from the reference lab where we can, downloaded from a portal or transferred via FTP or email rather than scanned in by hand.Q. While negotiating with reference labs to get these clean versions of the PDF files, are you also pursuing greater consistency in their provision of the machine-readable formats of these files (XML, JSON, etc.)?A. Yes, this is a high priority. Some reference labs have an API or are sharing files by FTP, we are working to ensure maximum consistency and accuracy in these files. While not all reference labs currently have that capability, we are working on getting the cleanest possible version of the PDF in the interim.Q. Are you interested in finding internal trials only or nationwide trials?A. For clinical trials, the focus is primarily on internal trials. However, it would be beneficial for the patients if we could also match to those nationwide.A. For personalized medicine, nationwide trials are also important.Follow-up questions and questions received via emailQ. You mentioned the ultimate desire to have unstructured data including PDFs parsed and accessible via NLP, is this a future goal or part of the RFP?A. Extracting this information from molecular results from reference labs is definitely within scope for the RFP. Parsing of next generation sequencing and/or single gene sequencing data is a priority. The ability to capture all information from external clinical PDFs (potentially hundreds of pages per patient) is a long-term goal, but certain elements (e.g. diagnoses and histology from outside path reports) are highly desirable. Q. Regarding Requirement R.9, are you expecting the solution to extract eligibility criteria? A. Yes.Q. Are these eligibility criteria already coded in Oncore?A. No, not in all cases. These are typically being uploaded as a PDF, so are not discrete.Q. Is there a requirement for a patient-facing component of the solution? Should there also be a mechanism whereby it can either directly or indirectly facilitate patient education, etc.? A. A patient facing component is not critical to the RFP. It would be valuable to present data that should be surfaced to a patient in a manner that can easily integrate with our primary patient-facing systems, if possible.Q. Is pushing analysis back to the trial management system the desired end target? A. Yes, would like the ability to push data back to the clinical trial management system.Q. Do you have any restrictions on joint submissions? Also, if organizations submitted jointly, will you also accept individual proposals from these companies?A. There are no restrictions on joint submissions. Yes, the individual proposals will also be considered. Q. Would Moffitt accept a partial bid that proposes a solution focused to “enable Moffitt’s Personalized Medicine Clinical Service Workflows?”A. If the Clinical Trial Matching requirements cannot be met, please consider partnering with vendor(s).Q. Are there any restrictions on the use of secondary data?A. Secondary data sources will be considered on a case-by-case basis. Q. What information from Oncore do you want to be integrated with the Clinical Trial Matching Tool and vice versa, what information would you look for a Clinical Trials Matching tool to feed back into Oncore?A. Examples – Patient-level: Trial history, current trial status, various consent metrics;Trial-level: Eligibility checklist, consent forms, order templates;Any newly created data could be fed back to Oncore. Q. Approximately how many users do you anticipate (or a band; such as 20-30 users)?A. Once fully implemented, between 100 and 500 concurrent users. Q. Given that this RFP is for both precision medicine and clinical trials matching, we could envision a scenario where CTM and PM each have a distinct ‘best in breed’ company and solution. Is there a precedence or preference for Moffitt between a) selecting two companies that are each best in breed versus b) selecting one solution?A. In this case, we prefer a partnership between two companies offering “best in breed” solutions. Q. R.1 states: Ability to derive discrete values from internal and external structured and unstructured documents: Which structured clinical parameters are available for a patient and what is the extent of capability for the solution to be able to match patients to trials based on patient’s unstructured clinical information (free-text fields in EHR data)?A. Structured clinical data is available for basic demographics, cancer characteristics, scheduling, treatment, labs, vitals, and other information discretely captured in the EHR. The solution should able to match patients to trials based on free text and scanned documents related to the patient’s current and past medical history.Q. Requirement R.4 states: “Ability to interface with Moffitt’s clinical and research systems and enterprise data warehouse.”, please specify what is expected, (For Example: what is the expected/assumed to be part of the offered project (fixed price) in terms of clinical data capture from Cerner?) Please define what is meant by “Ability” in this requirement context?A. Many data elements will available discretely from our own source systems. However, discretization of a subset of critical data elements from unstructured text or documents in the EHR (e.g. diagnoses and histology from outside path reports, or molecular test results) is highly desirable.Q. Refers to “R.14 Ability to extract and store discrete molecular results data (somatic and germline) from reference laboratory reports.” Could you list the main reference lab reports with their formats from which data shall be extracted to structural format (if not such already)?A. Multiple reference labs are used. Most data come back in PDF format, some in computable forms (XML, JSON, BAM, JSON, etc.).Q. Are the reference lab reports giving results in in variant level (f ex 7:g.55249071C>T p.Thr790Met ENST00000275493.2:c.2369C>T) or in the level of biomarkers (for example “patient is-> “ HER2-positive”) or both?A. BothQ. In which format does Oncore system describe molecular (variant or biomarker) eligibility criteria?A. Unstructured text.Q. What are the clinical trial databases outside Moffitt (the precision med side) from which matching should be done?A. is the primary source. Q. What does this requirement, “R.2 Ability to interface with and incorporate reference datasets, standardized ontologies, and curated databases.” exactly mean? Could you specify / give examples of use cases here?A. ICD-10 codes, PharmGKB, reference genome, etc.Q. Refers to above, what exactly is meant by “R.3 Ability to create, store, and use a curated knowledge base.” Could you describe the use case or desired workflow specifics in writing?A. Manual curation of variants based on the latest published literature. Q. Requirement R.8 states “Ability to curate patient-level data pertaining to commonly used eligibility criteria for clinical trials.” and R.9 states “Ability to derive discrete variables from Moffitt’s Clinical Trials Management System (CTMS) pertaining to eligibility criteria for particular trials.” There was no commonly used eligibility criteria provided, but is converting the eligibility criteria from current pdf documents (per trial) to searchable format part of the RFP?A. Yes. Q. R.15 Ability to interface with reference laboratories via Application Programming Interface (API). Is the assumption correct that actual integration would be done after the initial project? (not part of first implementation)A. Most likely. Q. RR.5 Ability to generate and export personalized medicine consults in text file and document formats (.doc/.docx, .odt, .pdf, .rtf, .txt). Can Moffitt provide examples on how you currently generate the personalized medicine consult summary document?A. The Word document is auto populated based on predefined elements from the “intelligence” database. Q. RR.6 Ability to review key performance indicators including total actionable genetic results generated, clinical recommendations made, results of clinical recommendations, outcomes, drug prescribing, toxicity, and more. Please specify where the data comes from (refers to earlier question on pulling data from Cerner), from which these KPI should be calculated and any specifics on the actual KPI calculation method?A. With the examples above in mind, much of the data required to assess whether these KPIs have been met will be available discretely in the EHR. This includes physician orders, administered medication data and diagnoses by visit, as well as survival outcomes. Detailed treatment response or recurrence information is typically unstructured but would be highly valuable.Action Items:Email RFP@: One (1) person from each company shall send the names of the participants who were on the call today from your company. The minutes will be posted online with the RFP. Bid packages are due May 5, 2020 by 2:00 PM EST to RFP@. Please pay close attention to the diversity questions and response requirements set forth in the RFP. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download