Randomized Controlled Trials in Neurosurgery

• Main Author: Gorayeb, Rodrigo Panico
• Publication Date: 2020
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: http://hdl.handle.net/10451/52492
• topic: Domínio/Área Científica::Ciências Médicas::Medicina Básica

Background: Randomized controlled trials (RCTs) are the gold standard studies to evaluate the efficacy of therapeutic interventions. However, performing RCTs in surgical disciplines, including neurosurgery, faces several specific challenges such as recruitment difficulties, surgical selection bias, inclusion of an appropriate control group, definition of clinically relevant outcomes and perceived lack of equipoise. Therefore, little is known about the number of published RCTs in neurosurgery, how they are performed and reported and their scientific quality. This, in turn, raises questions regarding the sources of knowledge that currently support clinical decision-making by neurosurgeons. Therefore, the present investigation aims to explore the current scenario of RCTs in neurosurgery and sources of knowledge used for clinical decision-making. Hypothesis and significance: Given the previously reported difficulties inherent to design and perform RCTs in neurosurgery, it is hypothesized that RCTs in this field are low in number and of suboptimal quality, leading neurosurgeons to use other sources of knowledge for clinical decision-making. The current research aims to better understand the current scenario and ultimately contribute for better decision-making in this field. Aims and methodological approach: This research comprised four aims. Aim 1. To investigate the attitudes of Brazilian neurosurgeons in terms of access and use of different types of scientific information for clinical decision-making, by means of a cross-sectional survey that was sent to all members of the Sociedade Brasileira de Neurocirurgia. The answers were analyzed descriptively and comparisons were made among subgroups of practice place and time. Aim 2. To evaluate how easy it is to have access to published RCTs in neurosurgery using several electronic search strategies applied to commonly used databases. This was achieved by carrying out open searches on PubMed, the Cochrane Library and the Centre for Reviews and Dissemination, an advanced search on PubMed based on clinical entity-related keywords, and hand-searches on the reference list of the identified RCTs. The outcomes were the sensitivity and specificity calculated for the open keyword searches on PubMed, the Cochrane Library and the CRD database and for the Cochrane’s HSSS, based on the total number of the identified RCTs. Aim 3. To investigate the scientific quality of RCTs in neurosurgery through a systematic review performed in all identified randomized trials with two or more comparative groups and at least one neurosurgical intervention. Study design and other methodological aspects were analyzed. The quality of included RCTs was assessed using the Cochrane Risk of Bias tool. Aim 4. To investigate how feasible it is to use sham controls in neurosurgery RCTs by analyzing the frequency, type and indication of used sham interventions. Results: Aim 1: The survey’s response rate was 32%. Among the respondents, 53% had more than 10 years experience, 67% worked in public hospitals, 34% performed spine and 30% brain tumor surgeries. Therapeutic decisions were based mostly on internship learning (54%) and personal professional experience (52%). Most common information sources were scientific abstracts (53%) and the internet (47%). 89% believed Evidence-Based Medicine to be relevant, 93% believed protocols or guidelines were necessary and 74% subscribed to a medical journal. Nonetheless only 43% had protocols implemented in their services, 93% highly valued a surgeonś personal experience and 63% showed little familiarity with the interpretation of scientific concepts in the literature. 83% were willing to try an innovative treatment alternative if it showed improvement of the outcomes and reduction of severe complications. Aim 2: A total of 1102 RCTs identified through combined search strategies. PubMed open search yielded 4660 articles, among which 365 were RCTs (sensitivity: 33.1%; specificity: 7.8%). Cochrane open search yielded 621 among which 36 were RCTs (sensitivity: 3.2%; specificity:5.8%) and CRD open search returned 78 articles, among which 4 were RCTs (sensitivity: 0.4% sensitivity; specificity: 5.1%). The Cochrane HSSS retrieved 10702 results, among which 340 were RCTs (sensitivity: 30.9%; specificity: 3.2%). Aim 3: RCTs in neurosurgery were found to lack quality, as most lacked information on study design (93.6%), randomization method (59.9%), blinding (59.8%), and data analysis (76.3%). Although the overall risk of bias decreased over time, 25.5% of the RCTs published between 2010 and 2013 lacked a clear risk of bias classification due to insufficient critical information. The methodological aspects more frequently classified with high risk of bias were “blinding of participants and personnel” (21.2%) and “incomplete outcome information” (28.8%). Aim 4: From the 1102 identified RCTs, 82 (7,4%) used sham interventions. The most common indication was pain treatment (67,1%), followed by the treatment of movement disorders and other clinical problems (18,3%) and brain injuries (12,2%). The most used sham interventions were saline injections, both in the spine (31,7%) and peripheral nerves (10,9%), followed by cranial surgery (26,8%), spine surgery (15,8%) and insertion of probes or catheters for sham lesions (14,6%). Most RCTs using sham interventions were double-blinded (76,5%), whereas 9,9% were single-blinded, and 13,6% did not report the type of blinding. Discussion: Although RCTs constitute the highest level of evidence, in the field of neurosurgery they are currently not the main source of knowledge underlying therapeutic decision-making, as most neurosurgeons, particularly those with over 10 years of experience, prefer to rely on their own expertise. Additionally, RCTs in neurosurgery are not easily accessible through commonly used search strategies applied to electronic databases, which is mostly due to poor quality of reporting and indexing. Moreover, published RCTs in neurosurgery lack quality in terms of experimental methodology, and they are poorly reported, with missing information on several critical design aspects. Although the inclusion of sham procedures in neurosurgical RCTs is feasible, very few include such procedures, which limits the clinical relevance of the estimated effects. Our results highlight an urgent need to improve the methodological quality, reporting and indexing of RCTs in neurosurgery, which may be achieved through the implementation of reporting guidelines, adequate training and rigorous peer-review.