Development of a Retrospective and Prospective Multidisciplinary Pancreas Database Joseph Herman, MD, MSc Barbara Biedrzycki, PhD, CRNP Amol Narang, MD Radiation and Medical Oncology The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins
Why is Multidisciplinary Care Important?
Multi D Care Benefits Less confusion for patients/families Improved communication between staff Inter disciplinary teaching One stop shop Improved outcomes Less errors? Cost effective?
Rational for Pancreatic MDC Pancreatic cancer is highly lethal Time is of the essence Correct staging is necessary for correct therapy Optimal care for pancreatic cancer involves numerous specialties Standard Care my not be best care for any given individual Tailored therapies Relatively rare so that, in general, no one community physician is an expert on pancreatic cancer Diagnosis can often be paralyzing and overwhelming
Standard Scenario PCA Diagnosis??Liver Met Jaundice Surgeon PPCT Borderline Resectable Radiation Oncologist Medical Oncologist Therapy Starts 1 week 1 week 1 week 2 weeks 1 week 6 weeks!!
PMDC Scenario Jaundice PMDC PPCT/Expert Review Pathology Review Medical Oncology Radiation Oncology Surgeon Pain Medicine Nutrition Clinical Trial Assessment Consensus on Optimal Treatment 1 week Days
Pancreatic MDC Website
Pancreatic MDC: Patient Schedule 7 9 AM: Necessary imaging and lab studies obtained 9 10 AM: Patients given overview of support services (10 15 min briefings by nutrition, nursing, social work and National Familial Pancreas Tumor Registry) 10 12 PM: Patients seen by fellows, residents, NPs and PAs for a complete history and physical exam 12:30 2 PM: Multi D team meets
Pancreatic MDC: Case Review Present Cases using outline Review Pathology Review Images CT/PET/MRI/ EUS Discuss Case and reach consensus See patients and discuss options Enroll in trials/studies Dictate note and cc to referring physicians
Patients and Methods Pancreatic Multidisciplinary Clinic 203 pts (November 2006 October 2007) Data Collection Patient demographics Clinicopathologic factors Outside vs. MDC findings / recommendations
Geographic Distribution of Referrals International: 4
Results: Initial Cross Sectional Imaging Initial Assessment Percent of Patients (%) Resectable 46 Locally advanced / unresectable 35 Metastasis 18 Locally advanced / unresectable + metastasis 1 N = 174
MDC Review: Cross Sectional Imaging Change in Clinical Stage 38 out of 174 (22%) Resectable Locally advanced/ Unresectable No Metastasis Metastatic Disease Suspicious Mass No Lesion Locally advanced/ Unresectable Resectable N = 3 N = 26 N = 4 N = 5
MDC Review: Pathology Change in Diagnosis 7 out of 203 (3%) Neuroendocrine tumor (n=2) Breast metastasis (n=1) Gastrointestinal stromal tumor (n=1) Gallbladder cancer (n=1) Benign inflammatory process (n=1) Serous cystadenoma (n=1)
Conclusions Pancreatic MDC is an efficient means to assess patients with presumed pancreatic cancer MDC format facilitates consensus recommendations and less confusion regarding the therapeutic plan A single day MDC may help to improve patient education, permit greater interaction with support staff (e.g. social work, nutrition, etc.) and decrease patient anxiety Pancreatic MDC provides an important expert opinion for many patients that may lead to dramatic changes in their care
PMDC Coordination Identify appropriate patients Educate on different levels Optimize treatment options Follow up for outcomes Enhance patient satisfaction
Original Data Management System Microsoft Excel Spreadsheet Simple Weekly log Track numbers
Excel Variables Demographics Diagnosis Imaging Disciplines
Original Data Management System Microsoft Excel Spreadsheet Validity 0 or 1 Versionitis Integrity
Better Data Management System Needed PMDC and Technology Patients Clinicians Researchers Administrators
Work to date Database design based on relevant clinical and research needs of multidisciplinary team Temporal organization Preclinic, clinic, post clinic Information type Pathology, radiology, recommendations, treatment, etc. Construction of prospective and retrospective databases with attention to data compatibility
Barriers in Database Construction Compatibility between multiple databases to ensure appropriate analysis Data abstraction Real time modification and testing of forms
Barriers of Combining Retrospective and Prospective data Recall bias Error of data entry with retrospective data: the prospective data is reviewed again at the pancreatic mdc consensus conference Ability to link both datasets for analysis Need to get updated data for retrospective patients Integration with biomarker data
Other Challenges Working with data programmers to develop forms that are complete, concise, but efficient and realistic for clinicians to enter data IRB approval for both research and clinical use Ability to export data for clinical notes Knowing when to say when, collect what you really need
Acknowledgements JHU: Dan Ford, Diana Gumas, Dorothy Damron, Susan Booker Harris Corporation: Raju Prasannappa, Seth Puckett, Christian Hertel