European Urology

Editorials

Outcomes research becomes more and more important for urologists in the clinic as well as in private practice. At first glance, one may overlook the importance of this field, since many of us are not able to spontaneously differentiate and judge the different fields of outcomes research. Karakiewicz et al. [1] have done so in their review of the methodology of outcomes research. Very different fields such as classic epidemiology, the cost-efficiency relationship of treatment, quality of life research, prediction tools (TNM, nomograms), the issue of cancer screening, quality of surgical procedures and legal problems of patient informed consent (e.g., before PSA testing) are summarized under “Outcomes research”. This information makes it clear that every clinician is faced with this expanding field in his day-to-day practice. The authors manage to explain the different scientific levels of outcomes research (macro, meso and micro level) and give examples of each level.

I would like to comment on two issues that have been brought up by this article.

The first issue is surgeon volume in radical prostatectomy. Already we are faced with different suggestions of minimum numbers of RP to be performed per surgeon and per year [2], [3], and [4]. The authors add the cutoff of 40 RPs per year to this discussion. Most of the data have been generated in the United States and, upon closer inspection, many of these publications suffer from methodologic problems [5], and [6] such as (1) Most papers refer to Medicare data, which are insurance data of a patient population that is very much different from the usual population that undergoes RP today. (2) The events analyzed occurred only if a patient revisited the doctor for a problem. There was no prospective evaluation of complications and outcomes. Instead, events have been registered only if the insurance company had to pay for them. (3) Most of the studies have concentrated on only side-effects of RP such as incontinence and strictures [7], and [8]. Data on oncologic outcome became available only recently [9]. In conclusion, there certainly is a relation between frequency of RP and oncologic as well as functional outcome. However, the data presented so far do not allow for a scientifically correct cutoff below which the single surgeon will no longer be reimbursed.

The second issue is nomograms. The authors nicely point out how important nomograms will become in the near future. Because there are several treatment options for localized prostate cancer, not only the patient has difficulties; the treating physician also runs into problems with counselling. The decision-making is much harder, and the role of the patient in this process gets more and more important in terms of patient satisfaction after treatment. In 2004, the Finish published a trial [10] in which they were able to show that patients do not always follow the recommendation of their physician even if they are correctly informed. As a consequence, patient satisfaction with treatment is much higher if the patient could choose his treatment in the beginning, even if the choice differs from the physician's recommendation. The recommendation for RP against brachytherapy or watchful waiting is not an easy one, and nomograms are of great help in this decision. However, the fact that the nomograms had to be validated several times demonstrates that the raw data that were analyzed for the nomograms were far from perfect. The biopsy Gleason or clinical stage in the Partin tables is as an example. Moreover, most of the nomogram data have been generated in a very special patient cohort that often is very different from the cohort in whom the nomogram is used. Again, this should not argue against nomograms. However, they have to be improved to obtain a more and more precise prediction of a patient's outcome after treatment.

The two examples elucidate why outcomes research is so important. The article by Karakiewicz et al. describes the different levels of outcomes research, which is a good start to a better understanding of the methodology of papers that will be published in the near future.