Where Are We Now? Lumbar radiculopathy, defined as pain emanating in a radicular pattern related generally to herniated discs, is common among patients presenting with back and leg pain. A discectomy through open or less-invasive approaches, generally done as a same-day procedure or with a 1-night stay in the hospital, usually is both effective and cost effective [4-6]. Nonoperative treatment including physical therapy, targeted epidural steroid injections, and non-narcotic anti-inflammatory medications can also alleviate the symptoms of acute radiculopathy. But the SPORT trial, in its observational arm, demonstrated superiority in patient-reported outcomes and pain measures in those patients treated surgically when compared to nonoperative measures [6]. The randomized arm of that trial suffered from problems of crossover in gaining an answer relative to the intention-to-treat analysis [5]. Still, data from the SPORT trial showed that at 2 and 4 years, surgical treatment for intervertebral disc herniation and spinal stenosis were cost effective. In the current study, Glennie and colleagues [2] aim to determine whether microdiscectomy is more cost effective than nonsurgical care in patients with chronic radiculopathy, which they define as patients with 4 to 6 months of symptoms of radiculopathy due to a herniated disc. These authors had previously demonstrated that early surgery resulted in improved outcomes compared to 6 months of nonoperative care in chronic radiculopathy [1]. The present study uses incremental cost utility ratios to compare outcomes related to the cost of treatment between operative and nonoperative care for that same patient population utilizing the randomized controlled cohort from their previous trial. It is important to analyze relative outcomes and costs when talking about a common diagnosis with multiple possible interventions, because cost constraints on the medical system continue to increase. Ultimately, we may have to choose between and/or stratify interventions based on value if resources are not unlimited. Although the authors found that early surgery for chronic radiculopathy is cost effective compared with nonsurgical care, they also determined that nonsurgical care is less effective, but also less expensive. The tradeoff for the more expensive but more effective care is justified by the low estimate for incremental utility. This is an important finding as we have to decide between different options of care when resources become limited. Where Do We Need to Go? While it has been well established that microdiscectomy for disc herniation is surgically effective and cost effective when compared to nonoperative treatment, it is not clear what the best timing for surgery might be. We also do not know whether there are certain patients who take longer to get better but eventually overcome their pain with nonoperative treatment. I was a coauthor on a study [3] that previously attempted to examine the timing of surgery for intervertebral disc herniation using the SPORT trial data with both operative and nonoperative interventions before and after 6 months of symptoms. It could be very useful to examine patients with herniated discs who underwent microdiscectomy and/or nonoperative treatment at less than 6 weeks versus greater than 6 weeks and less than 3 months versus greater than 3 months of nonoperative treatment. That might help determine whether the acuity of intervention influences outcomes both in terms of patient-reported outcome measures and overall cost. Likewise, we still need to know whether the results of the current study apply to all patients with herniated discs and radiculopathy or if it depends on whether this patient population also has neurologic or motor deficits. While we know that patients with progressing motor deficits and a foot drop likely benefit from an earlier intervention, the data are not clear on this point, and how they do after surgery may depend on the timing of the operation. How Do We Get There? It is impractical and likely cost prohibitive to answer each of the unknowns with a randomized prospective study. However, large population registries with carefully collected patient-reported outcome measures and careful definitions and descriptions of the patient population may be extremely useful in answering some ongoing questions. Within large population data from registries, we could compare those who gain equal improvement from nonoperative treatments (physical therapy, epidural steroids, medications) to the nonimprovers. Additionally, cohorts of different types and lengths of nonoperative interventions could be compared (interlaminar vs. tranforaminal injections, for example). As for timing of surgery, in a large enough cohort, a comparative group analysis between interventions at 6 weeks or less could be compared to 3 months, 6 months, and longer to determine whether there is disparity in outcome and/or durability of results. The same type of analysis could be done for differing types of surgical interventions (open laminotomy/discectomy, tubular discectomy, utilization of the microscope) in terms of resource utilization and outcome. The ultimate goal is to describe and identify the patient phenotype, based on presenting characteristics that will guide whether intervention is needed and what the best intervention is. Large, carefully curated and collected registry data may well be the answer to this and many other questions related to lumbar radiculopathy and many other diagnoses. By collecting large population data even without a randomized prospective study, some trends, outcomes, and complications become apparent and useful when comparing different interventions for a similar diagnosis. The definition of the diagnosis and the in-take criteria for the patient population being observed is crucial, as well as learning whether there are certain patients who should go straight to surgery versus having injections for some period of time and/or whether the number of injections matters. This can also be decided not only through targeted randomized trials, but by using data from registry populations or prior trials, as was done in the current study [2].