We would like to thank the authors of the letters for their comments. However, we think that in the multiple regression analysis performed, adjustments were made for all confounders measured so that the difference between the groups was eliminated. A prerequisite for meaningful results from multiple analysis is to collect data on all potential confounders. Thus, the problem with observational studies is not related to the observed covariables but to all non-observed covariables for which differences between the groups could exist. In this regard we would refer here to the second part of our reply.
We reject the criticism with regard to percutaneous coronary intervention (PCI)/coronary artery bypass graft surgery (CABG). PCIs were more common in the non-rehab group (62.7% versus 52.5%). Table 2 shows a tendency towards lower mortality among patients with PCI. This means that the non-rehab group appears to be at an advantage in the initial situation. This is similar for the variable CABG.
We would like to express our thanks for addressing the issue that it is not feasible to conduct a randomized study because in Germany, every patient with severe cardiac disease has the legal right to undergo rehabilitation pursuant to the indication catalogue of the funding organizations.
Re 1) In response to the criticism of our colleagues Kowall/Stang, we have investigated how many patients actually died during the first two weeks in which cardiac rehabilitation following an acute inpatient hospital stay must be started. The analysis revealed that two patients died during this time. For the sake of completeness, we also analyzed weeks 3 and 4 (one patient died in each week). Thus, early mortality does not appear to have an impact on the results presented by us.
Re 2) Here, we have to agree. No data on Barthel index or nursing care level were collected. These have to be regarded as potential confounders that do have an impact on mortality. Because of these, the advantage of participating in cardiac rehabilitation may have been overestimated.
Re 3) The first subgroup consists of patients who did not undergo the rehabilitation program despite the approval of their rehabilitation application. Here, it is possible that the (deteriorated) health status was the reason why the patient did not participate in the rehabilitation program. Thus, the increased mortality in this group may be explained by this factor.
The second subgroup comprises patients whose rehabilitation applications were either not approved or not completely processed, for unknown reasons. Here, one can also speculate that a deteriorated health status may have played a role; thus, mortality could actually be higher in this subgroup, too.
We would like to thank the authors of the two letters. However, in our opinion the data presented in our study show that patients benefit from cardiac rehabilitation and that eligible patients in the Federal State of Saxony-Anhalt do not make enough use of this rehabilitation offering.
Prof. Dr. med. habil. Axel Schlitt, MHA
Paracelsus-Harz-Klinikikum Bad Suderode, Quedlinburg, Germany
Conflict of interest statement
Prof. Schlitt has received consultancy fees (Advisory Board) from Boehringer Ingelheim. He has received lecture fees and reimbursement of conference fees and travel expenses from Sanofi-Aventis, BMS, Pfizer, Novartis, Servier, Boehringer Ingelheim, MSD, and Bayer AG. He has received study funding (third-party funds) from GSK, Sanofi-Aventis, Mitsubishi, Endotis, Bayer AG, Boehringer Ingelheim, Novartis, Actelion, and BMS.
|1.||Schlitt A, Wischmann P, Wienke A, Hoepfner F, Noack F, Silber RE, Werdan K: Rehabilitation in patients with coronary heart disease— participation and its effect on prognosis. Dtsch Arztebl Int 2015; 112: 527–34 VOLLTEXT|
Thinking Slow About Abercrombie & Fitch: Straightening out the Judicial Confusion in the Lower CourtsSSRN Electronic Journal , 201810.2139/ssrn.3216741