Eleven days versus nine. That is the gap in hospital stay after stomach cancer surgery between patients who arrived with a documented diagnosis of depression or anxiety and those who did not, in a new analysis of 1,502 American surgical patients. Two days sounds small. Stretched across a recovery that already runs long, and paired with a complication rate that ran roughly half again as high, it points at something the operating room cannot fix.
Kennedy Jensen and colleagues at the University of Utah, working with collaborators at Oregon Health & Science University and Florida State, pulled their patients from MarketScan, a large commercial claims database covering hundreds of U.S. insurance plans. They looked for adults who had a gastric adenocarcinoma diagnosis and then underwent a total or subtotal gastrectomy, meaning surgeons removed all or part of the stomach, between 2011 and 2021. Everyone in the group had at least six months of continuous insurance coverage before diagnosis and six months after surgery, so the team could see what happened on either side of the operation.
Of those 1,502 patients, 221 (15 percent) had a coded diagnosis of anxiety, depression, or both: 168 with anxiety, 88 with depression. That figure is worth pausing on, because it almost certainly undercounts. Claims data only records what a clinician wrote down and billed for. Someone struggling quietly, or never asked, does not appear.
The two groups were similar in most respects. Sex, region, and whether patients lived in a city or a rural area all looked comparable. Patients with depression or anxiety were somewhat younger at the time of surgery, a median of 59 against 63, and more of them had a total gastrectomy rather than a partial one, 46 percent against 34 percent. The authors say they cannot explain that surgical difference from claims data alone.
What the numbers showed
The team ran regression models adjusting for age, sex, rurality, and which operation the patient had. After that adjustment, a depression or anxiety diagnosis independently predicted a longer stay (an incident rate ratio of 1.26, meaning about a quarter longer). It also predicted more complications within 30 days, 13 percent against 8 percent, and the same gap held at 90 days. Most of that difference came from one specific problem: venous thromboembolism, or dangerous blood clots. Twenty-one patients in the smaller group had one, against 75 in the much larger comparison group.
Costs followed the same pattern. The median bill for the surgical admission itself came to $50,262 for patients with depression or anxiety and $42,380 for those without. By 90 days, counting readmissions and emergency visits, the gap had widened to $56,983 against $43,707. These are total paid amounts from settled claims, insurer payments plus what patients owed in deductibles and copays, not the inflated sticker prices hospitals list.
One finding cut against expectation. Readmissions and emergency department visits did not differ between the groups at all, at either 30 or 90 days. The authors offer two possible readings: complications may have been handled during the original hospitalization, or claims data may simply miss some post-discharge visits.
What the study cannot tell you
The authors are direct about the limits, and a reader should be too. This is a look backward at billing records, not an experiment. MarketScan does not record tumor stage, how complex the operation was, or whether patients received chemotherapy before or after surgery. Any of those could be doing work the model attributes to mental health. The team also raises the possibility of reverse causation: more advanced, frightening disease may itself produce more anxiety and depression, in which case the psychiatric diagnosis is partly a marker of a sicker patient rather than a cause of a rougher recovery. Direction of causality, they write, is impossible to establish here.
The database also excludes uninsured patients and those on Medicaid, and it does not record race or ethnicity, so the team could not adjust for either. These are commercially insured Americans, mostly through employer plans.
Why it matters
Stomach cancer is usually treated with more than surgery. Many patients receive chemotherapy before the operation and more afterward, and finishing that full course matters for long-term survival. A rough recovery can delay or derail it. The authors point to published work showing that complications after gastrectomy are linked to worse survival partly because patients never get their adjuvant chemotherapy started or completed. So an extra two days and a higher clot rate are not just a hospital inconvenience. They sit upstream of the treatment that does the curing.
What makes depression and anxiety interesting here, as opposed to age or tumor stage, is that they can potentially be changed. Because many gastric cancer patients get chemotherapy before surgery, there is a window of weeks or months in which someone could screen for these conditions and treat them. The authors cite meta-analyses suggesting psychological prehabilitation can shorten hospital stays and ease pain, and earlier work finding that treating depression in cancer patients lowered overall costs.
The research team is emphatic on one point, stating it twice in the paper: none of this should be read as a reason to withhold curative surgery from someone with a psychiatric diagnosis. The finding is a prompt to offer more support, not less treatment. Whether that support actually closes the gap is a question this study cannot answer, and the authors call for prospective trials to find out.