I recently had the opportunity to attend the annual Surveillance, Epidemiology and End Results (SEER) Managers and Principal Investigators Meeting with my HDRP colleagues Lynne Harlan and Joan Warren; DCCPS Director Bob Croyle; and many colleagues from the Surveillance Research Program.
The Kentucky Cancer Registry and the NCI-designated Markey Cancer Center at the University of Kentucky hosted the meeting. The Registry is well-integrated with the Cancer Center’s Cancer Prevention and Control Program, which has led to recent novel collaborations between basic and population scientists. These connections are highly valued by Dr. Mark Evers, Cancer Center Director, who spoke at the meeting.
At the meeting I was reminded of the complexity of gathering high quality data. Data collection continues to rest primarily with hospital-based registrars and central registry staff. Changes in health care delivery have and will continue to make their work more difficult. For example, an increasing number of cancers are diagnosed and/or treated in outpatient care settings, which are generally not legally compelled to give registrars or registries access to medical records. Oncology providers and practices face time pressures that make it difficult to respond to requests for additional information sent by registries. The increasing number of biomarkers used in treatment decision-making, along with how many labs are involved in biomarker testing, have made it difficult to gather some of the data most of interest in epidemiology and outcomes research. These challenges are unlikely to diminish in the future.
Led by Lynne Penberthy, a team of NCI staff are carefully evaluating the possibility of using various sources of electronic data to augment treatment data and address some of the aforementioned challenges. For example, they are conducting a small pilot study with IMS Health to see if their extensive data on physician prescribing can augment SEER treatment data, particularly oral therapies. There are ongoing discussions about potential data linkages with national laboratory companies and vendors who manage claims for oncology practices. Lynne and her staff are also completing several important studies of data quality, from which they will develop ongoing quality assurance activities to be conducted by the registries and NCI. Although not related to data capture, several of the registries are taking steps to create a “virtual pooled registry” that would simplify researcher access to data from multiple registries.
HDRP staff consults on many of these projects. We recognize that high quality cancer registry, claims, and other data are vital for health care delivery research. Indeed, the SEER-Medicare data linkage has resulted in over 1,000 manuscripts, some large proportion of which probably would not exist without that resource. Identifying high quality data and facilitating their availability to the research community is an important component of expanding health care delivery research at NCI. Additional information on these efforts will be available later in 2015.
What new data sources would allow you to address gaps in the evidence base for cancer care? What are those gaps?