PHOENIX, ARIZ. – Cost and quality are not always synonymous, particularly when it comes to complex surgical procedures such as esophagectomy.

A review of records on more than 6,700 patients who underwent esophagectomy during a 4-year period showed that factors such as patient age, severity of illness, and hospital/surgeon volume can have a major effect on resource utilization and costs, said Dr. Daniel E. Abbott, assistant professor of surgery at the University of Cincinnati.

"There are certainly actionable risk factors for poor outcomes, such as mortality, and increased resource utilization, including dollar costs, the opportunity costs of increased length of stay, readmission, and rehabilitation and skilled nursing facilities," he reported at the annual Society of Surgical Oncology Cancer Symposium.

"I would argue that careful patient selection can have profound influences on cost-effectiveness. I think that as our health care is evolving and our outcomes are increasingly scrutinized, there will be increasing pressure to have better outcomes at lower costs," he added.

Dr. Abbott and his colleagues examined clinical variables in the cases of 6,737 esophagectomy patients treated from 2009 through 2012 in the University Healthsystems Consortium (UHC), an organization comprising 120 university hospitals and 299 affiliates.

They evaluated patient characteristics such as age and race, severity of illness index (1-4), esophagectomy type, and center and surgeon volume, and evaluated the effects of these variables on clinical outcomes that contribute to resource utilization, including deaths, readmissions, length of stay (LOS), and discharge disposition.

They found that the median LOS for all patients was 10 days (interquartile range, 8-17 days), but for patients over age 70, the median LOS was 11 days (P less than .01 vs. patients 70 and under).

Older patients did not have significantly higher readmission rates, but of the 4.2% of all patients who died in hospital, those over age 70 had more than twice the death rate of younger patients (7.0% vs. 3.2%, P less than .01).

Older patients were also significantly more likely to be discharged to a skilled nursing or rehabilitation facility than were younger patients (31.9% vs. 10.6%; P less than .01).

Total median cost per patient was $25,952, but again, older patients accounted for more of the expenses, at a median of $27,628 vs. $25,841 for those 70 and under (P less than .01).

In a multivariate analysis, patients over 70 had a more than twofold increase in risk of death (odds ratio, 2.12; P less than .01). Other factors significantly associated with greater risk for death were greater severity of illness (OR, 14.0; P less than .01) and black race vs. other races (OR, 1.88; P less than .01).

Factors associated with more frequent readmissions included greater severity of illness (OR, 1.33; P less than .01), and black race (OR, 1.34; P = .01), while patients of high-volume surgeons were less likely to need readmission (OR, 0.87; P = .04).

Lengths of stay were greater among patients over age 70, with every year over 70 translating into a 6% greater LOS (OR, 1.06; P =.03); older patients had a 16% increase in LOS for every year over 70 (OR, 1.16; P less than .01).

Similarly, each increase in severity of illness index score above 2 was associated with a 31% increase in LOS (OR, 1.31) and a 475% increase in intensive care unit (ICU) days (OR, 4.75; P less than .01 for both LOS and ICU days).

Black patients had a 22% increase in LOS vs. other races (OR, 1.22; P less than .01) and a 31% relative increase in ICU days (OR, 1.31; P = .01).

Because age and severity of illness were both strong predictors for mortality, readmission, and other perioperative outcomes, the authors conducted a further analysis combining the two variables, and found that for every 5 years of age, there were significant increases in the risk for death among patients with a greater severity of illness, compared with low severity.

Dr. Abbott noted that some of the odds ratios for older, sicker patients were "ridiculously high," probably because of the smaller sample sizes.

In a multivariate analysis of cost, factors associated with higher costs were age (OR, 1.14; P less than .01), greater severity of illness (OR, 2.14; P less than .01), and black race (OR, 1.15; P less than .01).

And in an analysis of cumulative resource use, the authors found what Dr. Abbott called a "snowball effect," in that hospitals with the lowest total costs discharged the majority of patients home, and that as costs increased, hospitals were less likely to discharge patients home and more likely to discharge them either to home health services or to extended care. In addition, as costs rose, the percentage of patients who died in hospital also rose.

He acknowledged that because the study used administrative data from university hospitals, it was skewed toward high-volume centers, and that the database did not include survival data, information about longitudinal resource use, or procedure-specific complications.

The authors did not disclose the funding source of the study. Dr. Abbott reported having no financial disclosures.

Patient selection can keep the lid on esophagectomy costs