Retrospective but not rigorous.

To the Editor:—We read with much interest the article by Biki et al. regarding the effect of anesthetic technique and postoperative analgesia on the cancer recurrence rate after open radical prostatectomy. The results suggesting that epidural anesthesia/analgesia lowers the rate of recurrence are certainly intriguing; however, we are concerned with the lack of detailed information presented in various parts of this retrospective study. Not only does this diminish the quality of the publication, but it also raises questions about the validity of the results. Most areas where detailed information is omitted are located within the Materials and Methods section. The primary rationale for presenting methodology in any scientific publication is to allow the reader to determine the applicability of the study conditions to their own circumstances/practice and/or to replicate the study if desired. As such, meticulous and accurate reporting of details is essential. This may be particularly relevant for retrospective studies, as the most appropriate use of such studies is to generate hypotheses for the development of future clinical trials, the design of which will depend to a large extent on the methods used in the retrospective study. The most important example of incomplete information relates to the epidural anesthetic/analgesic. There is an almost complete lack of information regarding the intraoperative and postoperative epidural management, and, most significantly, the type and quantity of local anesthetic are mentioned nowhere. Certainly, “not all epidurals are created equal,” and knowing the type and quantity of medication administered via this route is of major relevance from both a research and clinical perspective. The authors also fail to provide data regarding the quantity of potent inhalational anesthetics or opioids actually administered in the perioperative period. Both types of agents inhibit natural killer cell activity, and may thus potentially increase the risk of cancer recurrence after surgery. Although the authors state in the Discussion section that “it is highly plausible that patients in the epidural group . . . required considerably less volatile anesthetic” and those receiving epidural anesthesia/analgesia “presumably required little opioid, whereas those given general anesthesia alone surely required considerable amounts of opioid,” they present no data to support these statements. Indeed, when the authors describe the general anesthetic as “most typically” consisting of a list of drugs, volatile anesthetics are not even included. Slightly more information is presented for opioids (fentanyl 1–2 g/kg is included in the list of “most typically” used intraoperative drugs; morphine 0.1–0.15 mg/kg is merely reported as having been “given for postoperative pain;” and the postoperative morphine patient-controlled anesthesia settings are stated for the general anesthesia–postoperative opioids group), but the quantity actually received by patients in the two groups is not reported. One further example of incomplete methodological information is not only deficient, but also inaccurate. The term “sizable minority” is used to describe the percentage of patients who received general anesthesia– postoperative opioids; however, this contradicts the actual numbers of patients in each group: 123 patients received general anesthesia–postoperative opioids whereas 102 received epidural anesthesia/analgesia. No explanation is given for this discrepancy. The above discussion leads to a more general issue: The standard of reporting expected for retrospective studies. Although some information may not be available, every attempt must be made to achieve the same standard of rigorous reporting as for clinical trials and laboratory investigations. Indeed, with the inherent drawbacks of retrospective studies, one could argue that the presentation of the information that is available should reach an even higher standard than that used for other types of scientific articles. Furthermore, if important data are not available, this calls into question whether the study should even be performed, as its validity may be suspect. As computerized recordkeeping and databases are increasingly used, it is quite possible that retrospective studies will become more and more common. To provide meaningful information, these studies should strive to achieve the same high standards expected of other scientific publications.