Performance of the Veterans Choice Program for Improving Access to Colonoscopy at a Tertiary VA Facility
Introduction: The Veterans Choice Program (VCP) was designed to provide a pathway for veterans to access health care in the community if wait times at the US Department of Veterans Affairs (VA) were > 30 days. However, the performance of this program, in terms of timeliness, quality assurance, and overall utilization by veterans for colonoscopy is not well studied.
Methods: We reviewed records of veterans at VA Pittsburgh Healthcare System (VAPHS) who underwent VCP colonoscopy from June 2015 through March 2017. We compared the number of days from the scheduling encounter to the first available colonoscopy at VAPHS to the actual colonoscopy through the VCP. Additionally, we examined the availability of procedure and pathology results, documentation of quality metrics, and if clear follow-up recommendations were present in community care records. We then separately examined VCP utilization in a representative sample (5% margin of error, 95% CI) of all colonoscopy referrals through the VCP.
Results: During the study period 3,855 veterans were eligi ble for colonoscopy via the VCP, and 190 colonoscopies were performed through VCP. Records were absent for 29 exams (15.3%). There was no statistically significant difference for the number of days from a veteran’s initially scheduled first-available colonoscopy at VAPHS when compared with the actual VCP colonoscopy (median 2 days earlier, P = .62). Pathology results were absent in 14 of 118 (11.9%) patient records, and follow-up recommendations were absent in 29 of 161 (18%) cases. Documentation of colonoscopy quality metrics were deficient in 27% to 70% of procedure reports. In a utilization sample of 350 veterans, only 26 (7.4%) veterans referred for colonoscopy had documented VCP colonoscopies, and 231 (66%) had a VAPHS colonoscopy. The median actual wait time for colonoscopy was 61 days for VAPHS and 66 days through VCP ( P = .15).
Conclusions: Colonoscopies referred through the VCP were not performed sooner in aggregate compared with the first available colonoscopy at VAPHS, although there was wide variability in wait times. We recommend additional mechanisms be put into place when outsourcing to community care: Ensure seamless and require prompt transfer of records back to the VA, require reporting of quality metrics standard at the VA for community care colonoscopies, and establish clinically meaningful wait-time thresholds for referral into the community, rather than static ones.
Limitations
There are several limitations to our analysis. First, all data were extracted via chart review by one author; therefore, some scanned procedure or pathology reports or pre- and postprocedure records may have been missed. Second, these data are representative of a single VA medical center and may not reflect trends nationwide. Third, there are many factors that can influence veteran decision making regarding when and where colonoscopy procedures are performed, which could be related to the VCP community care referral process or independent of this. Finally, colonoscopies performed through the VCP are grouped and may not reflect variability in the performance of community practices that veterans were referred to though the VCP.
Adenoma detection rates (ADR) were not included in the assessment for 2 reasons. First, there was an insufficient number of screening colonoscopies to use for the ADR calculation. Second, a composite non-VA ADR of multiple community endoscopists in different practices would likely be inaccurate and not clinically meaningful. Of note, the VAPHS does calculate and maintain ADR information as a practice for its endoscopists.
Conclusions
Our findings are particularly important as the VA expands access to care in the community through the VA Mission Act, which replaces the VCP but continues to include a static wait time threshold of 28 days for referral to community-based care.10 Especially for colonoscopies with the indication of screening or surveillance, wait times > 28 days are likely not clinically significant. Additionally, this study demonstrates that there also are delays in access to colonoscopy by community-based care providers, and potentially reflects widespread colonoscopy access issues that are not unique to the VA.
Our findings are similar to other published results and reports and raise similar concerns about the pitfalls of veteran referral into the community, including (1) similar wait times for the community and the VA; (2) the risk of fragmented care; (3) unevenquality of care; and (4) low overall utilization of VCP for colonoscopy.11 We agree with the GAO’s recommendations, which include establishing clinically meaningful wait time thresholds, systemic monitoring of the timeliness of care, and additional mechanisms for seamless transfer of complete records of care into the VA system. If a referral is placed for community-based care, this should come with an expectation that the care will be offered and can be delivered sooner than would be possible at the VA. We additionally recommend that standards for reporting quality metrics, including ADR, also be required of community colonoscopy providers contracted to provide care for veterans through the VA Mission Act. Importantly, we recommend that data for comparative wait times and quality metrics for VA and the community should be publicly available for veterans so that they may make more informed choices about where they receive health care.
Acknowledgments
The authors thank Kaneen Allen, PhD, for her administrative assistance and guidance.
