Correspondence
In Reply
;
As Prof Mrozckowski correctly explains, DRG and OPS (Operationen- und Prozedurenschlüssel) codes are generally intended for billing for service rendered. They may bias the results of evaluations and are potentially inferior to data from clinical databases. However, it should be pointed out that one of the main outcome measures of our study (in-hospital mortality) is unlikely to be subject to relevant biases. Furthermore, the German medical services of health insurers, (Medizinischer Dienst der Krankenkassen, MdK) check in minute detail the cases billed to health insurers/sickness funds, especially those in which complications occurred and were invoiced (for example, the duration of ventilation).
Furthermore, publications from the US that evaluated administrative data from a database established over 20 years (ACS-NSQIP) showed sufficient concordance of both systems, with billing data—as expected—inferior to registry data (1).
In Germany, no universally clinical database exists that is of a comparable size and extent to the US ACS-NSQIP database. The DGAV-StudoQ registry, for example, as a clinical database in Germany could certainly in future enable the evaluation of individual specific questions and thereby help to set out relevant questions and quality indicators (2, 3). The selective documentation of such registries may, however, be a problem.
Our response to PD Dr Ronellenfitsch: in the meta-analysis, appendectomy is named as the long acknowledged gold standard for any form of appendicitis, and an analysis of alternatives for the different forms of appendicitis is discussed. We think that for many vital indications, surgical procedures are without alternative or superior to other measures.
As Prof Holzheimer comments, the case fatality rate in the German DRG data is higher than the Swedish registry by a factor of 10, which raises the question of the type of documentation and interpretation of the data. The unbeatable advantage in the analysis of billing data is that all patients receiving care are included, without any selection process. For the evaluation of general complications (seroma, wound infection, urinary tract infection, nerve damage, etc) the DRG data do not seem appropriate because they are coded only generally. This needs to be taken into account when comparing the review cited by Prof Holzheimer, in which general complications are described, with our data (only severe complications). No causal associations can be generated from the retrospective studies; the causes will have to be the subject of further studies.
We wish to respond to the methodological comments of Prof Mroczkowski and Drs Nimptsch and Krautz as follows:
For each billed case, the precise hours of ventilation are documented in the DRG code, and we evaluated these. The threshold of 48 hours is a commonly used value in the literature. We cannot draw any conclusions about the temporal sequence of procedures and complications in a treatment case. Accordingly, we evaluated the published complications—as is usual practice in the literature of administrative data (4)—under the assumption that these are likely to develop as events after surgery (for example, peritonitis after gastro-esophageal surgery) (5).
A general challenge in analyzing billing data lies in the fact that ICD/OPS codes are subject to continuous change. The codes for Lichtenstein procedures valid since 2016 (5–530.33 and 5–530.34) were included in the interrogation of the data (2009–2015), but did not yield any evaluation result (because—as Drs Nimptsch and Krautz commented—they were not yet firmly embedded in the OPS) and were therefore not included in the final data evaluation. In our published study (6), we considered minimally invasive inguinal herniotomy (laparoscopic transperitoneal closure [5–530.31] and endoscopic total extraperitoneal closure surgery [5–530.32]). Furthermore we checked the raw data and the published data once again in great detail and uncovered two errors: the codes for thyroidectomy (5–063) and 4/5 gastrectomy (5–436) are analyzed and included in the publication, but not reproduced in eTable 1 (2). A relevant erratum was published. We thank your correspondents for their constructive criticism.
DOI: 10.3238/arztebl.2020.0363
Dr. med. Philip Baum
PD Dr. med. Armin Wiegering
Klinik und Poliklinik für Allgemein-, Viszeral-,
Transplantations-, Gefäß- und Kinderchirurgie
Universitätsklinik Würzburg, wiegering_a@ukw.de
Conflict of interest statement
The authors declare that no conflict of interest exists.
| 1. | Lawson EH, Louie R, Zingmond DS, et al.: A comparison of clinical registry versus administrative claims data for reporting of 30-day surgical complications. Ann Surg 2012; 256: 973–81 CrossRef MEDLINE |
| 2. | Jurowich C, Lichthardt S, Matthes N, et al.: Effects of anastomotic technique on early postoperative outcome in open right-sided hemicolectomy. BJS Open 2019; 3: 203–9 CrossRef MEDLINE PubMed Central |
| 3. | Wiegering, A, Wellner U, Seyfried F, et al.: [MTL30 as surrogate parameter for quality of surgically treated diseases: Establishment based on the StuDoQ register of the German Society for General and Visceral Surgery]. Chirurg 2017; 88: 977–82 CrossRef MEDLINE |
| 4. | Panis Y, Maggiori L, Caranhac G, Bretagnol F, Vicaut E: Mortality after colorectal cancer surgery: a French survey of more than 84,000 patients. Ann Surg 2011; 254: 738–43 CrossRef MEDLINE |
| 5. | Nimptsch U, Haist T, Krautz C, Grützmann R, Mansky T, Lorenz D: Hospital volume, in-hospital mortality, and failure to rescue in esophageal surgery—an analysis of German hospital discharge data. Dtsch Arztebl Int 2018; 115: 793–800 VOLLTEXT |
| 6. | Baum P, Diers J, Lichthardt S, et al.: Mortality and complications following visceral surgery—a nationwide analysis based on the diagnostic categories used in German hospital invoicing data. Dtsch Arztebl Int 2019; 116: 739–46 VOLLTEXT |
