Background: The medical database MEDLINE contains only a part of all published randomized controlled (RCTs) and controlled clinical (CCTs) trials. The validity of a systematic review depends on the complete consideration of all studies pertaining to a clinical question. Therefore, a manual systematic search in medical journals is mandatory.
Methods: The last 30 years of the three peer-reviewed German-language ophthalmological journals Der Ophthalmologe, Klinische Monatsblätter für Augenheilkunde, and Spektrum der Augenheilkunde were retrospectively searched for controlled trials. All identified RCTs and CCTs were classified according to their ophthalmological subspecialty. The quality of the studies was evaluated using the CONSORT Statement. Moreover, a comparison was made of how many of the studies identified by hand searching are registered in MEDLINE and correctly classified as RCTs/CCTs.
Results: In total 984 RCTs and CCTs were identified; 62% of the trials were from the subspecialty anterior eye segment, 14% belonged to the subspecialty posterior eye segment, and 18% belonged to glaucoma. A sample size calculation was reported in less than 2% of the full-text publications and 34% reported a blinded study design. A randomization process was described in 23% of the RCTs. Most controlled trials were published in the 1990s in all three journals. From 1998 onward, there was a decrease in the number of published controlled trials in the reviewed German literature. About 63% of the studies registered in MEDLINE are correctly classified.
Conclusion: In the specialty of ophthalmology there are many important studies which provide basic information for compilation of systematic reviews and clinical guidelines. However, many of the identified clinical trials in German-language journals show deficiencies in the study quality. Only a part of the identified controlled trials are correctly classified in MEDLINE. Therefore, a retrospective registration is essential to complete the electronic database in ophthalmology.