From the Journals
Consider invasive mediastinal staging in higher risk NSCLC patients, despite guidelines
Mediastinoscopy staging for NSCLC may be cost effective for more patients than guidelines suggest.
COLORADO SPRINGS – Significant variability exists between hospitals in Washington state in their rates of invasive mediastinal staging for lung cancer, Farhood Farjah, MD, reported at the annual meeting of the Western Thoracic Surgical Association.
“We found evidence of a fivefold variation in hospital-level rates of invasive mediastinal staging not explained by chance or case mix,” according to Dr. Farjah of the University of Washington, Seattle.
Dr. Farood Farjah
“This has led to substantial concerns about quality of thoracic surgical care in the community at large,” he noted.
The Washington study is the first to show hospital-by-hospital variation in rates of invasive mediastinal staging.
Invasive mediastinal staging for lung cancer is considered important because imaging is known to have a substantial false-negative rate, and staging results have a profound impact on treatment recommendations, which can range from surgery alone to additional chemoradiation therapy.
Yet the meaning of the hospital-level huge variability in practice observed in the Washington study remains unclear.
“Our understanding of the underutilization of invasive mediastinal staging is further complicated by the fact that patterns of invasive mediastinal staging are highly variable across hospitals staffed by at least one board-certified thoracic surgeon with a noncardiac practice,” Dr. Farjah explained. “This variability could be a marker of poor-quality care. However, because the guidelines are not supported by level 1 evidence, it’s equally plausible that this variability might represent uncertainty or even disagreement with the practice guidelines – and specifically about the appropriate indication for invasive staging.”
He presented a retrospective cohort study of 406 patients whose non–small cell lung cancer was resected during July 2011–December 2013 at one of five Washington hospitals, each with at least one board-certified thoracic surgeon with a noncardiac practice on staff. The four participating community hospitals and one academic medical center were involved in a National Cancer Institute–funded, physician-led quality improvement initiative.
Overall, 66% of the 406 patients underwent any form of invasive mediastinal staging: 85% by mediastinoscopy only; 12% by mediastinoscopy plus endobronchial ultrasound-guided nodal aspiration (EBUS); 3% by EBUS only; and the remaining handful by mediastinoscopy, EBUS, and esophageal ultrasound-guided nodal aspiration. The invasive staging was performed at the time of resection in 64% of cases. A median of three nodal stations were sampled.
After statistical adjustment for random variation and between-hospital differences in clinical stage, rates of invasive staging were all over the map. While an overall mean of 66% of the lung cancer patients underwent invasive mediastinal staging, the rates at the five hospitals were 94%, 84%, 31%, 80%, and 17%.
Dr. Farjah and his coinvestigators are now conducting provider interviews and focus groups in an effort to understand what drove the participating surgeons’ wide variability in performing invasive mediastinal staging.
Discussant Jane Yanagawa, MD, of the University of California, Los Angeles, commented, “I think this is a really interesting study because, historically, lower rates of mediastinoscopy are assumed to be a reflection of low-quality care – and you suggest that might not be the case, that it might be more complicated than that.”
Dr. Yanagawa sketched one fairly common scenario that might represent a surgeon’s reasonable avoidance of guideline-recommended invasive mediastinal staging: a patient who by all preoperative imaging appears to have stage IA lung cancer and wishes to avoid the morbidity, time, and cost of needle biopsy, instead choosing to go straight to the operating room for a diagnosis by wedge resection, followed by a completion lobectomy based upon the frozen section results. Could such a pathway account for the variability seen in the Washington study?
“I think it could have,” Dr. Farjah replied. “I would say that’s probably one driver of variability.”
As for the generalizability of the findings of a five-hospital study carried out in a single state, Dr. Farjah said he thinks the results are applicable to any academic or community hospital with at least one board-certified thoracic surgeon with a noncardiac practice.
He reported having no financial conflicts of interest regarding the study.