Object: The aim of this study was to evaluate the quality of clinical studies appearing in neurosurgical journals during three decades.
Methods: Clinical studies published in 1982, 1992, and 2002 in Neurosurgery, Journal of Neurosurgery (JNS), and British Journal of Neurosurgery (BJN) were evaluated. The feature study types were categorized as follows: Type 1, prospective randomized controlled trials (PRCTs); Type 2, other prospective studies; Type 3, retrospective case-controlled and cross-sectional studies; and Type 4, case reports. Among a total of 786 articles, the following study types were identified: Type 1, eight (1%); Type 2, 46 (6%); Type 3, 81 (10%); and Type 4, 651 (83%). Overall, the proportion of study types did not differ among journals. Between 1982 and 2002, increases in the proportion of study Types 1 (0 compared with 2%, respectively), 2 (4 compared with 10%, respectively), and 3 (5 compared with 13%, respectively) and decreases in the proportion of study Type 4 (92 compared with 75%, respectively; p < 0.001) were apparent only in JNS and Neurosurgery (p < 0.01). Between 1982 and 2002, the median number of patients (two compared with 14, respectively) and the mean number of authors per study (3.4 compared with 4.8, respectively) increased (p < 0.001). The JNS had a greater mean number of authors per study than the other journals (p < 0.001).
Conclusions: During the three decades evaluated, case reports remained the predominant study design in these prominent neurosurgical journals, notwithstanding the modest increases in study Types 1 and 2 and the associated escalations in author and patient numbers in the same period.