Abstract: See Article, p 666 Less than one-third of all research abstracts presented at the Society for Obstetric Anesthesia and Perinatology annual meetings are ultimately published.1 Researchers have an ethical duty to make the results of their research publicly available,2 which applies to studies with positive, negative, or inconclusive findings. However, the results of negative or inconclusive studies are often not published.3 This might, in part, be related to study design issues, such as inadequate statistical power, which can result in negative outcomes, or it might be related to negative studies often being viewed as incapable of changing practice. In this issue of Anesthesia & Analgesia, Sharawi et al4 report the results of a randomized, triple-blind, noninferiority study, which compared 2 local anesthetic solutions used to extend epidural labor analgesia to anesthesia for cesarean delivery. Historically, 3% chloroprocaine has been considered the drug of choice for urgent cesarean delivery due to its onset time, as well as ease of preparation. However, a recent Bayesian network meta-analysis demonstrated that the onset of surgical anesthesia is likely to be faster for 2% lidocaine with bicarbonate in comparison to 3% chloroprocaine.5 No previous studies have directly compared the onset times for a mixture of 2% lidocaine, with sodium bicarbonate, fentanyl, and epinephrine to 3% chloroprocaine for extension of epidural analgesia for cesarean delivery. The clinical question is important for practicing obstetric anesthesiologists; however, difficult to study using a randomized controlled trial methodology, as it would be challenging to recruit and randomize women undergoing unplanned emergent cesarean deliveries, then assess outcomes in a timely manner without compromising clinical care. Two elements of the study by Sharawi et al4 are worthy of attention: the clinical model used to simulate the conversion of epidural analgesia to epidural anesthesia and the choice of a noninferiority study design. Patients who were enrolled in the study were scheduled for elective cesarean delivery. One hour before the scheduled surgery, combined spinal-epidural analgesia was initiated. The only medication injected intrathecally was 150 µg of morphine, as epidural chloroprocaine can interfere with morphine analgesia postoperatively.6 Epidural analgesia was initiated using 0.0625% bupivacaine with 2 µg/mL fentanyl until a T10 dermatomal level was achieved to pinprick. Once a block to a T10 level was achieved, maintenance of labor analgesia was initiated using a continuous epidural infusion of the same solution. When the epidural extension was required, 3 mL of the study drug was injected as a test dose, and subsequently, the remaining 17 mL of study solution was injected over 2 minutes. The primary outcome of the study was time taken to achieve a loss of touch sensation bilaterally to the T7 dermatomal level. This was assessed every minute following the start of epidural extension anesthesia. To our knowledge, no other model exists for replicating the conversion from epidural analgesia to anesthesia for cesarean delivery. By using the presurgical admission time to initiate and establish labor analgesia, investigators can replicate the emergency scenario in a more controlled setting. This can allow researchers to interrogate different techniques used to initiate analgesia (eg, epidural versus combined spinal-epidural versus dural puncture epidural) or different neuraxial dosing regimens (eg, programmed intermittent epidural bolus versus continuous infusion, weak versus stronger concentrations of local anesthetics, different volumes of local anesthetics administered, or even ropivacaine versus bupivacaine) on cesarean delivery anesthesia. This model also has the advantage that women should not require additional interventions, earlier hospital admissions, or longer fasting durations before their scheduled surgery. This model could also potentially be used to identify predictors for the development of inadequate intraoperative anesthesia, or failed anesthesia, following epidural extension in more controlled conditions than in an intrapartum cesarean delivery scenario. While this model does not allow for understanding the interplay between labor and epidural conversion, it is innovative and may help facilitate future obstetric anesthesia studies that have not previously been possible or easy to execute. The second novel aspect of the study for obstetric anesthesiologists is the choice of a noninferiority study design to compare the 2 solutions. In the absence of any available direct comparison, and anecdotal beliefs of both agents being fastest, the goal of the study was to verify that the mixture of 2% lidocaine, with sodium bicarbonate, fentanyl, and epinephrine had a similar onset time to 3% chloroprocaine. A noninferiority design was the appropriate choice of study design as these studies evaluate whether an alternative treatment is no worse than the standard treatment by a prespecified amount (ie, noninferiority margin).7 A potential mistake would have been to have used a superiority study design, not found a statistical difference between the 2 solutions, and then conclude that the 2 solutions have the same onset time. Five potential outcomes for noninferiority studies exist: noninferior and superior, noninferior, noninferior and inferior, inconclusive, and inferior.8 When using noninferiority studies, the choice of the prespecified noninferiority margin is of critical importance. The United States Food and Drug Administration (FDA) guidance states that noninferiority margins should be within 50% of a minimally important clinical difference.9 However, in this study,4 3 minutes was chosen as the prespecified noninferiority margin because a clinically significant reduction in epidural anesthesia onset time does not exist, and it takes approximately 3 minutes to mix the lidocaine, fentanyl, bicarbonate, and epinephrine.10 Had a higher noninferiority margin been selected (ie, 4 minutes), this study would have proven noninferiority. The Sharawi et al4 study was inconclusive, as their confidence interval included the noninferiority margin, and the observed difference in mean onset times also crossed 0. This suggests that the study was underpowered. Ironically, had a superiority study design been used, the 2% lidocaine mixture (with sodium bicarbonate, fentanyl, and epinephrine) would likely have demonstrated a faster onset time than 3% chloroprocaine. So readers may ask, why does an inconclusive study warrant publication and an accompanying editorial? The results of this study will not help anesthesiologists choose between 2% lidocaine, with sodium bicarbonate, fentanyl, and epinephrine or 3% chloroprocaine when extending labor epidural analgesia for cesarean delivery anesthesia. However, it does suggest important avenues for future research, which will help guide clinical decision-making. Future superiority studies should focus on comparing these 2 epidural extension solutions, using similar methodology, to guide clinicians as to which solution to utilize for intrapartum cesarean delivery. Furthermore, it highlights the use of an important study design, which should be considered when designing obstetric anesthesia studies. In the era of drug shortages, we are often faced with changing our drug or dosing strategies for the initiation of epidural analgesia or postcesarean delivery analgesia. Noninferiority study design could and should be used to help answer clinical questions, as certain medications we routinely use become unavailable. It is our role as clinicians to be able to understand and interpret the outcomes of these studies, and the role of reviewers and editors to look beyond superiority. DISCLOSURES Name: Paloma Toledo, MD, MPH. Contribution: This author helped draft the article and approved the final version of the article. Name: Preet Mohinder Singh, MD. Contribution: This author helped draft the article and approved the final version of the article. Name: Pervez Sultan, MBChB, FRCA, MD (Res). Contribution: This author helped draft the article and approved the final version of the article. This manuscript was handled by: Honorio T. Benzon, MD.