Two recent articles19–20 addressed the interpretation of
nonrandomized, observation-type studies. They concluded that
quantitative results were similar to randomized, controlled studies.
However, the relatively high incidences of patients unavailable for
follow-up, and especially for follow-up physiology testing, remains a
nemesis to meaningful interpretations of many LVRS publications.
Comparing LVRS patient results to historical cohort data may also be
misleading because of clinical, pathologic, treatment, and smoking
variabilities. Some investigators have emphasized clinical and
quality-of-life follow-up,12,14,17 whereas we have
emphasized mechanisms.13 The manuscript by Flaherty et al
goes a long way toward achieving both goals.