E153: Having Established Primary Care: Impact on Postoperative Outcomes among Patients Undergoing Surgery for Gastrointestinal Cancer

Regular primary care (PC) visits are essential to monitor wellness, enable proactive screening, and manage pre-existing comorbidities. In turn, patients without adequate access to PC may be at risk of adverse outcomes. We sought to characterize the impact of established PC visits on post-operative outcomes following surgical resection of gastrointestinal cancers.  

Methods:

Patients diagnosed with esophagus, gastric, hepatobiliary, pancreas, and colorectal cancers between 2018-2021 were identified using Medicare Standard Analytic Files. Individuals with PC encounters within a year before the surgery were identified by Berenson-Eggers Type of Service and specialty codes, excluding encounters for immunizations. Pre-existing comorbidities were identified using 10^th edition International Classification of Diseases codes. A postoperative textbook outcome (TO) was defined as the composite absence of complications, no prolonged length of stay, no readmission within 90 days, and no mortality.

Results:

Among 115,654 cancer patients (esophageal: n=1,854, 1.6%; gastric: n=4,690, 4.1%; hepatobiliary: n=6,873, 5.9%; pancreatic: n=8,912, 7.7%; colorectal: n=93,325, 90.7%), 78,443 (67.8%) had at least one PC visit within the year prior to surgery. Among patients without significant comorbidities, individuals with established PC were more likely to achieve TO (OR 1.31 [95%CI 1.27-1.36]) and had lower odds of a post-operative complication (OR 0.87 [0.83-0.90]), extended length of stay (OR 0.90 [0.86-0.94]), and 90-day mortality (OR 0.89 [0.83-0.95]) (Figure). Among patients with chronic medical conditions, the effect of preoperative PC visits was less pronounced: TO (OR 0.96 [0.92-0.99]), complication (OR 1.05 [1.00-1.10]), extended length of stay (OR 1.04 [0.98-1.09]), re-admitted within 90-days of discharge (OR 1.12 [1.06-1.17], and 90-day mortality (OR 1.02 [0.95-1.09]). The number of visits with PC was inversely related to odds of TO. Specifically, while a lower number of PC visits (< 5 times/year) was associated with an increased odds of TO (OR 1.09 [1.05-1.13]), multiple PC visits ( >10 times/year) was associated with a lower likelihood of TO (OR 0.87 [0.80-0.94]).

Conclusions:

Two in 3 patients with gastrointestinal cancer had established PC within the year prior to resection. Among patients without major comorbidities, established PC was associated with a higher chance of TO. PC could not mitigate the risk of worse post-operative outcomes among patients with high medical comorbidities, or individuals who needed to see their PC >10 times/year.