Explaining Variations in Lung Cancer Resection Mortality Across Hospitals

Differences in failure-to-rescue rates may explain a wide variation in mortality in patients undergoing lung cancer resection at different hospitals, according to a new study.

Despite improvements in overall mortality rates with lung resection in recent years, there remain significant differences between hospitals in perioperative outcomes. “While the opportunity for quality improvement is clear, the optimal approach to achieve this goal is uncertain because the mechanisms underlying high morbidity and mortality rates have not been well defined,” wrote study authors led by Tyler R. Grenda, MD, of the University of Michigan in Ann Arbor.

The new study was a retrospective analysis of outcomes across hospitals. The researchers ranked 1,279 hospitals that conduct resection for lung cancer, and analyzed differences between 18 low-mortality hospitals (LMHs) and 25 high-mortality hospitals (HMHs). The study covered a total of 645 lung resections (441 in LMHs and 204 in HMHs), and the results were published online ahead of print in JAMA Surgery.

There were 7 perioperative deaths in the LMH group (1.6% unadjusted mortality rate), compared with 22 deaths in the HMH group (10.8%, P < .001).

However, those who underwent lung resection at HMHs presented with greater illness severity, more patients had poorer functional status than at LMHs, they were more likely to have more than two comorbid conditions (ischemic heart disease and diabetes mellitus in particular), and HMHs performed a significantly higher percentage of emergency procedures (2.5% vs 0.2%; P = .006).

After adjusting for risk, the difference in mortality was smaller but still significant, at 1.8% for LMHs and 8.1% for HMHs (P < .001).

The authors found differences in operative approach between the groups. Patients at HMHs were slightly more likely to undergo an open thoracotomy than at LMHs (adjusted odds ratio [OR], 3.15 [95% confidence interval (CI), 1.00–9.93]). Patients were less likely to undergo a thoracoscopic lung resection at HMHs than at LMHs (adjusted OR, 0.31 [95% CI, 0.10–0.99]).

There were no significant differences between the hospitals with regard to overall complication rate, at 23.3% in HMHs vs 15.6% in LMHs, for an adjusted OR of 1.79 (95% CI, 0.99–3.21). Rates of prophylactic antibiotic administration were similar between the groups, though HMHs were more likely to continue giving the antibiotics more than 24 hours after the procedure.

There was a significant difference with regard to failure to rescue for any complication; failure to rescue was more likely to occur at HMHs, with an adjusted OR of 6.55 (95% CI, 1.44–29.88). The authors highlighted this difference as a potential reason for the variation in mortality rates.

“Our findings suggest that variations in mortality rates between LMHs and HMHs may be driven by practices related to identification and effective management of complications rather than processes targeted at prevention of adverse events,” they concluded.